2

Table of Contents
Preface

Introduction

A History of the Behavioral Therapies

William T. O'Donohue, Deborah A. Henderson, Steven C.

Hayes, Jane E. Fisher, and Linda J. Hayes

University of Nevada, Reno

Chapter 1

The Importance of Case Studies to Methodology of

Science

Thomas Nickles

University of Nevada, Reno

Chapter 2

Joseph Wolpe: Challenger and Champion for Behavior

Therapy

Roger Poppen

Southern Illinois University

Chapter 3

3
B. F. Skinner’s Contribution to Therapeutic Change: An
Agency-less, Contingency Analysis

Julie S. Vargas

West Virginia University

Chapter 4

Jacob Robert Kantor (1888-1984): Pioneer in the

Development of Naturalistic Foundations for Behavior
Therapy

Paul T. Mountjoy

Western Michigan University

Chapter 5

Child Behavior Therapy: Early History

Sidney W. Bijou

University of Nevada, Reno

Chapter 6

Studies in Behavior Therapy and Behavior Research

Therapy:

June 1953-1965

Ogden R. Lindsley

4
University of Kansas and Behavior Research Company

Chapter 7

A Brief Personal Account of CT (Conditioning Therapy),

BT (Behavior Therapy) and CBT (Cognitive-Behavior
Therapy): Spanning Three Continents

Arnold A. Lazarus

Rutgers University and the Center for Multimodal

Psychological Services

Chapter 8

Swimming Against the Mainstream: The Early Years in

Chilly Waters

Albert Bandura

Stanford University

Chapter 9

The Rise of Cognitive Behavior Therapy

Albert Ellis

Albert Ellis Institute for Rational Emotive Behavior Therapy

Chapter 10

5
From Psychodynamic to Behavior Therapy: Paradigm
Shift and Personal Perspectives

Cyril M. Franks

Rutgers University

Chapter 11

Cognitive Behavior Therapy: The Oxymoron of the

Century

Leonard Krasner

Stanford University

Chapter 12

The Development of Behavioral Medicine

W. Stewart Agras

Stanford University School of Medicine

Chapter 13

Toward a Cumulative Science of Persons: Past, Present,

and Prospects

Walter Mischel

Columbia University

Chapter 14

6
A Small Matter of Proof

Donald M. Baer

University of Kansas

Chapter 15

Do Good, Take Data

Todd R. Risley

University of Alaska

Chapter 16

Application of Operant Conditioning Procedures to the

Behavior Problems of an Autistic Child: A 25-Year
Follow-Up and the Development of the Teaching Family
Model

Montrose M. Wolf

University of Kansas

Chapter 17

The Active Unconscious, Symptom Substitution, and

Other Things That Went ‘Bump’ in the Night

Gordon L. Paul

University of Houston

7
Chapter 18

Values and Constructionism in Clinical Assessment: Some

Historical and Personal Perspectives on Behavior Therapy

Gerald C. Davison

University of Southern California

8
Dedication
This book is dedicated to the loving memory of Janet Bijou, a
true friend of behavior therapy, and whose intelligence and
kindness reflect the humanitarian spirit guiding us all.

9
Distributed in Canada by Raincoast Books

Copyright © 2001 Context Press

Context Press is an imprint of New Harbinger Publications,

Inc.

5674 Shattuck Avenue

Oakland, CA 94609

www.newharbinger.com

All Rights Reserved

Epub ISBN:9781608825882

The Library of Congress has cataloged the print edition as:

A history of the behavioral therapies : founders’ personal

theories /

edited by William T. O’Donohue ... [et al.].

p. cm.

“Based on a conference held at the University of Nevada,

Reno, in

June, 1999"—Pref.

Includes bibliographical references.

10
ISBN 1-878978-40-3

1. Behavior therapy—History—Congresses. I. O’Donohue,

William T.

RC489.B4 H55 2001

616.89’142’09—dc21

2001047081

11
Preface
This book is based on a conference held at the University of
Nevada, Reno in June, 1999. The editors organized this
conference in order to provide an opportunity for us and
others to better understand the development of the behavioral
therapies and to capture a part of the historical record before it
was lost forever. Most importantly, we wanted to honor the
founders of the behavioral therapies and to watch as this
group of approximately 20 individuals interacted — most of
them as old friends, but also perhaps for the last time, at least
in this large of a group.

The chapters the arise from this conference show the unique
properties of these unique individuals. While participants had
an outline of topics to address, each chapter reflects the topics
that the presenters felt best revealed their intellectual history
and the context and content of their contribution. The
difference between chapters in toen and approach could not
realistically be eliminated without muffling the very voices
we wanted to hear. In the end, we made the conscious
decision to give these leaders of the field the freedom to tell
their story in their own way.

We want to thank these individuals for taking time from their

busy lives to make this conference a huge success. Although
many of these founders are certainly of retirement age — all
are clearly “flunking” retirement. Old friends met once again,
sometimes after years without contact. Others met for the first
time. There was much catching up and many great stories

12
were told. We would also like to thank them for sharing the
stories of their lives and their work. These stories were often
quite moving, and sometimes very funny. What these
individuals showed us is that the story of the development of
behavior therapy is not a dry story of purely intellectual
commitments and technical developments. Rather it is clearly
a story of deeply held values, caring, compassion, conflict,
fate, and, at times, personal tragedy. This book is dedicated to
these (and other) founders of our discipline.

We also dedicate this book to the spouses and families of

these founders. Many individuals brought family members
and it was both impressive and touching. It was clear that in
many cases there was a real partnership in which spouses not
only supported but also clearly substantively contributed to
the success of their partners.

We also want to comment on one key aspect of the

conference. Many of the founders expressed their deep debt
and gratitude to two central figures — Sidney Bijou and
Albert Bandura. These two individuals often worked quietly
and behind the scenes helping others find jobs, training key
students, and in general generously giving help and
encouragement to others. The special gratitude and honor
extended to these two individuals should be noted.

We also want to thank Professor Leo Reyna for presenting a

paper at the conference. Leo is one of the great teachers in our
profession; for example, his influence on Joseph Wolpe was
critical to the development of behavior therapy.
Unfortunately, due to serious health problems he was not able
to write up his fascinating paper as a chapter for this volume.

13
We would also like to thank supporters of this conference.
Dean Robert Mead and Vice President Ken Hunter were
particularly helpful. Their generous help made this conference
possible.

Finally, we also want to thank Tuna Townsend for his

generous assistance on many aspects of this project.

William T. O'Donohue

Deborah A. Henderson

Steven C. Hayes

Jane E. Fisher

Linda J. Hayes

14
Participants in the Nevada Conference
on the History of the Behavior
Therapies

Top row (from left to right): Ogden Lindsley, Gordon Paul,

Don Baer, Bill O’Donohue, Mont Wolf, Paul Mountjoy,
Deborah Henderson

Second row (from left to right): Cyril Franks, Jane Fisher,

Todd Risley, Sid Bijou, Walter Mischel, Stewart Agras, Steve
Hayes

15
Bottom row (from left to right): Roger Poppen, Jerry
Davison, Julie Vargas, Leo Reyna, Len Krasner, Albert
Bandura, Linda Hayes

16
Introduction

A History of the Behavioral

Therapies
William T. O'Donohue, Deborah A. Henderson, Steven C.
Hayes, Jane E. Fisher, Linda J. Hayes

University of Nevada, Reno

What is behavior therapy?

But what history are we trying to capture in this volume?

What is behavior therapy, or what was it, at least in its
beginnings? It might prove useful at least for a general
orientation to provide a few definitions of behavior therapy,
particularly those that were offered around the time of its
formation:

“Behavior therapy derives its impetus from experimental

psychology and is essentially an attempt to apply the findings
and methods of this discipline to disorders of human
behavior” (Rachman, 1963, p. 3).

“The attempt to alter human behavior and emotion in a

beneficial manner according to the laws of modern learning
theory” (Eysenck, 1964, p. 1).

“Treatment deducible from the sociopsychological model that

aims to alter a person’s behavior directly through application

17
of general psychological principles” (Krasner & Ullmann,
1965, p. 244).

“Behavior therapy, or conditioning therapy, is the use of

experimentally established principles of learning for the
purpose of changing maladaptive behavior” (Wolpe, 1969, p.
VII).

Bandura (1969) placed the principles of behavior

modification within the “conceptual framework of social
learning…By requiring clear specification of treatment
conditions and objective assessment of outcomes the social
learning approach…contains a self-corrective feature that
distinguishes it from change enterprises in which
interventions remain ill-defined and their psychological
effects are seldom objectively evaluated” (p. v).

There seems to be some consensus, at least among these

various definitions, that behavior therapy is an orientation to
understanding and ameliorating human suffering, through
behavior change, that is influenced by principles derived from
experimental psychology, particularly learning research.
Behavior therapists have traditionally emphasized outcome
research (and still do) and more recently some of behavior
therapy’s subdivisions stress the importance of cognitive
variables. We can see that these definitions, although
containing some similarities, also suggest controversies. What
qualifies as a psychological principle? Exactly what
psychological principles are the most important and useful?
Which learning theory? What is the proper relationship
between a psychological principle and a therapy technique?
What evidence should be most persuasive? What are the
relative advantages and disadvantages of various research

18
methodologies (e.g., single subject vs. group designs)? What
constitutes a legitimate problem for the behavior therapist?
What constitutes improvement and how is this best to be
measured? How important are cognitive factors, and how are
they to be accounted for within a behavioral paradigm?

The various answers to these sorts of questions have led to a

great deal of variegation in behavior therapy. In fact, Kazdin
(1978) stated that in behavior therapy’s second decade,

“By now behavior modification is so variegated in its

conceptualization of behavior, research methods and
techniques that no unifying schema or set of assumptions
about behavior can incorporate all the extant techniques.
Many of the theoretical postions expressed within behavior
modification represent opposing views about the nature of
human motivation, the mechanisms that influence behavior
and the relative influence of such factors, and the most
suitable focus of treatment for a given problem” (p. 374).

Is this diversity bad? Does it mean there is now no such

coherent entity or activity as “behavior therapy”? We answer
both of these in the negative for two main reasons. First, clear
and simple definitions expressing essential properties are
often difficult to provide, as phenomena are often quite
complex, variegated and even “fuzzy.” Wittgenstein (1958)
famously points this out:

Consider for example the proceedings that we call “games.” I

mean board-games, card-games, ball-games. Olympic games,
and so on. What is common to them all — Don’t say: “There
must be something common, or they would not be called
“games’” — but look and see whether there is anything

19
common to all — For if you look at them you will not see
something that is common to all, but similarities,
relationships, and a whole series of them at that…Are they all
“amusing”? Compare chess with naught and crosses. Or is
there always winning and losing, or competition between
players? Think of patience. In ball games there is winning and
losing; but when a child throws his ball at the wall and
catches it again, this feature has disappeared. Look at the
parts played by skill and luck; and at the difference between
skill in chess and skill in tennis. Think now of games like
ring-aring-aroses; here is the element of amusement, but how
many other characteristic features have disappeared! And we
can go through the many, many other groups of games in the
same way; can see how similarities crop up and disappear.
And the result of this examination is; we see a complicated
network of similarities overlapping and criss-crossing;
sometimes overall similarities, sometimes similarities of
detail (p. 31-32).

We believe the streams and strands of behavior therapy are

best captured by

Wittgenstein’s notion of complicated networks of similarities

and dissimilarities.

Second, Kuhn (1979) has suggested that as a science develops

smaller sets of scientists begin to work on more specialized
problems and the particular nature of these problems often
call for idiosyncratic solutions. Thus, the more mature science
then develops a micro-paradigmatic structure where small
groups of scientists share similar problem solving exemplars
and assumptions that have complicated networks of

20
similarities but the field as a whole does not share one general
paradigm.

Some of this diversity can be seen in the various theories

associated with behavior therapy (O’Donohue & Krasner,
1995). From its early reliance on reinforcement, punishment,
and reciprocal inhibition we can now see behavior therapists
relying on the matching law, implosion theory, learned alarms
theory, self-efficacy theory, attribution theory, information
processing theories, relapse prevention, etc. Some of these
theories were developed, as Kuhn described, in response to
more particular problems of concern to particular behavior
therapists. However, there clearly was diversity from the very
beginning of behavior therapy: a key point of difference was
the “Iowa school” more influenced by Hull and Mower’s
learning accounts; and the “Harvard school” more influenced
by Skinner and operant psychology. Behaviorism is also one
source of influence on many but not all of the early behavior
therapists. Behaviorism is often thought to be monolithic but
actually there are a variety of behaviorisms (O’Donohue &
Kitchener, 1999). In the formation of behavior therapy these
unique behaviorisms had unique influences. Skinner certainly
was one of the most influential figures in the development of
behavior therapy (he and Wolpe would vie for the title of the
most influential). Unfortunately many of the more recent
behavior therapists have much less understanding of
behaviorism, experimental methodology and learning
research than did the first generation behavior therapists. The
interested reader is directed to O’Donohue and Ferguson
(2001) for a brief introduction to Skinner.

Why Study the History of Behavior Therapy?

21
Why should a history of the behavioral therapies be
attempted? Coleman (1988 p. 3-4) provides an interesting set
of reasons for studying history:

A first claim is that history is therapy. To be unaware of past

influences allows them to have unchecked effect upon our
present-day research program.

A second claim sees history as a cautionary tale. This is the

point made in the aphorism attributed to George Santayana, to
the effect that those who are ignorant of the past are doomed
to repeat its errors.

A third claim is that the study of history provides a more

perfect understanding of the present. This claim, advanced by
Julian Jaynes, postulates understanding rather than
error-avoidance as the primary concern, and holds that the
only way to understand the present is to study the past history
upon which the present depends.

A fourth claim sees history as a basis for the unification of

psychology. The history of psychology presumably supplies a
“more generalized knowledge” that counterbalances
specialization and facilitates appreciation of the unity of
historically related specializations.

A fifth proposal regards history as a form of travel. Reading

in the history of psychology enlarges one’s horizon of
possibilities, and this may provide a liberating influence from
the limitations of one’s particular time and place and
specialty.

22
Sixth, studying history is like getting an immunization shot.
Knowledge of the great pendulum swings that have occurred
in psychology and of the fads and orthodoxies that have come
and gone can reduce the persuasive power of new fads.

A seventh claim is that history teaches appreciation. The

study of historical background of psychology makes one
aware of the cumulative nature of scientific work in
psychology.

An eighth claim regards history as a treasure-hunt. A few

have claimed that the study of history occasionally results in
the recovery of material, that is, a neglected idea or theory,
which subsequently proves very useful in the present.

Ninth, history is a crystal ball. Erwin Esper, an early

behaviorist who recorded a history of psychology in his later
life, proposed that a knowledge of historical roots could
contribute to efforts to predict future developments.

A tenth claim sees history as a silver bullet. Knowledge of

how a present-day controversy began and how it gradually
came to be formulated may be of great and unexpected
assistance in resolving that controversy.

Eleventh, history is a lesson in sobriety. According to

Crutchfiled and Krech, “knowledge of the history of one’s
science teaches the scientist humility and tolerance of
opposing views”

A twelfth and final claim is that history is a mark of the

educated person. A knowledge of history provides a
familiarity with events, concepts, and cultural landmarks that

23
are common coin in literary and conversational context of the
larger world of ideas: even the specialist might wish to
become better acquainted with such material.

Beyond these general reasons we think there are several

reasons for studying the founding of behavior therapy. The
most obvious answer is that this history can be of interest in
its own right. Many are interested in family genealogy not
because momentous ancestors are found, but because of a
more basic curiosity regarding simply knowing what has
happened in the past — particularly how this chain has lead to
where we are now. A history can also attempt to answer key
questions that fundamentally are historical: What is the
“founding event”? (interestingly in what follows we will see
different accounts of who first coined the label “behavior
therapy”). What were the major sources of impetus in the
creation of behavior therapy? What changes/developments
occurred after the founding; what factors were responsible for
these? Finally, all sorts of more specific historical questions
can be framed; for example, what interactions did Skinner and
Wolpe have and what were these like?

Studying history reminds us of how far we’ve come — how

different things were then. When behavior therapy first came
on the scene, psychologists were largely just “testers,”
permitted to do little therapy. Large state mental hospitals
warehoused schizophrenics, who were often out of control, as
this predated the discovery of effective anti-psychotics.
Therapy was controlled by physicians and particularly by
psychoanalysts. Today the scene is so different that new
behavior therapists are in danger of having little appreciation
of how much things have changed.

24
Origins often interest us and they often relate to other
interesting and sometimes watershed events. In this history of
the behavioral therapies we see the influence of the Great
Depression; World War II; American ascendancy in the 20th
century; the Cold War and its arms race; the rise of
technology; the economic prosperity of the latter part of the
century at least in the West; as well as the rise and spread of
the research university.

Knowing where we came from also allows us to assess

whether we have lost something; whether we have drifted
away from some of the strengths that allowed us to be where
we are today. We are now producing our fourth generation of
behavior therapists; and many of the third and fourth
generation have had little to no direct contact with the
founders of our field. A valuable lesson may be that some of
what made behavior therapy successful has been lost across
the generations. We will discuss some of these possible
“object lessons” below.

We also write this history because we think it is worthy to

capture what we take to be a unique event in the history of
psychotherapy. The famous historian of science Thomas
Kuhn (1963) has stated that a key development of a science is
when a scientific revolution occurs that allows a field to
emerge from a pre-paradigmatic state to a state of normal,
paradigmatic science. At the heart of the paradigm is a
successful puzzle solution that serves as an exemplar for other
scientists to emulate to solve further problems. We suggest
that the learning based interventions that were developed,
implemented and most importantly tested by the first
generation behavior therapist represented the first time
psychotherapy emerged from a pre-paradigmatic state to a

25
state that is closer to what Kuhn calls normal science. Clinical
problems such as enuresis, chronic skills deficits of the
developmentally disabled, and certain anxiety problems were
demonstrated to be remediated, for the first time in history.
Before this psychotherapy was based on less sophisticated
epistemologies. However, this is not to say that before
behavior therapists came on the scene no one was taking a
scientific approach. Clearly this was not the case. The
Minnesota school was taking sophisticated approaches to test
development and validation. The Rogerians, and others, were
doing outcome research. And in fact some of the key
publications of our discipline were short on experimental
rigor. Many were case studies demonstrating the possibilities
of efficacy. Some were single subject designs with no
replications. Some were simply persuasive manifestos.
Although we do not have space here it would be interesting to
trace developments in the quality of the evidence during the
development of behavior therapy. This project could also
attempt to examine if philosophies of science can faithfully
capture these developments. Were behavior therapists
attempting to falsify their commitments, as Popper would
claim constitutes good science? Were they extending
problem-solving exemplars, as Kuhn would suggest? Were
they opportunistic Dadaists, permissively following
Feyerabend’s recommendation, “Anything goes”? Were they
none of these?

Another reason to study the history of behavior therapy is that

at times individuals make arguments from history. These
arguments are generally of the form:

1. In its origins behavior therapy had properties a, b, c

…

26
 2. Properties a, b, c … accounted for the success of
behavior therapy.
3. Now behavior therapy has lost or reduced properties
a. b, c, …
4. Therefore if behavior therapists want to increase their
success, they ought to reorient to properties a, b, c,…
Obviously the soundness of this kind of argument depends
upon the accuracies of the historical claims contained in the
first premise. Glib histories can be useful to someone wanting
to make certain points. Glib, superficial histories are plentiful
and the reader ought to be on guard against believing these.
Smith (1986) has nicely shown that glib histories of
behaviorism falsely associated it with logical positivism
instead of characterizing it as associated with indigenous
psychological epistemologies.

Thus behavior therapists at times might want to assume a

meta-scientific perspective to attempt to understand questions
such as: How can behavior therapy make better progress?
What are the important features of behavior therapy that
account for its growth? What factors are hindering its growth?
How has it changed over time? These are important questions.
Our point is simply that in order to answer these questions
adequately a careful, accurate understanding of the historical
record in needed. Otherwise the clever rhetorician can attempt
to make historical claims that serve his or her particular
interests.

Finally, we also record history to give due recognition to

those responsible for events. This, also, is a purpose of the
present volume. We believe that the individuals in the
chapters that follow deserve to be honored. They worked

27
hard, often as a small minority fighting a powerful and
entrenched majority, because they would not accept as
satisfactory what they saw. What they saw was much human
suffering that was not being effectively treated by the
standard interventions of the day. And what they also saw
was that the alleged evidential basis for claims surrounding
these interventions was problematic. They had a vision that
things could be better and they were willing to pay the price
to attempt to make things better. Their efforts, we believe, are
responsible for bettering the lives of many individuals —
most importantly the clients who experience more relief from
their pain; but these efforts also paved the way for better
opportunities and careers for the second and third generation
behavior therapists, as journals were launched; professional
associations were formed; and inroads to hiring this strange
new breed of therapists and researchers were made. To
paraphrase Isaac Newton, we see farther because we stand on
their shoulders.

Historiography
Behavior therapists tend to be explicitly concerned with
method. Writing a history calls for a historical method.
Historiography is the study of the method of writing history.
As such it concerns itself with a meta-question: How should
history be properly studied? Any history attempts to fairly
capture part of the story knowing that the story also
legitimately could be told in other ways. Because of the
complexity of the tale that the historian is attempting to tell,
the would-be historian makes hard choices. A key choice is
what method is to be used to capture and explicate a slice of

28
the historical record. The historian is constrained by a lack of
knowledge. Much is lost in the mists of the past.

Historical Methods

Externalist histories attempt to specify the larger forces (that

is, outside of the phenomena to be explained) that influenced
the topic of the historical account, in this case behavior
therapy. These forces could be economic, technological,
political, social, or ideological. Historicist accounts are a
variation of externalist accounts. Historicism is the view that
history should be told from the point of view of inexorable
historical forces. This view would attempt to identify the
forces that would have made the rise of behavior therapy an
historical inevitability. Marxist historical accounts utilize
historicism and externalist methods as these attempt to
identify the economic and technological forces that give rise
to historical events.

Internalist histories, on the other hand, take a more

microanalytic view, attempting to explain the historical
stream by more immediate influences. In these accounts,
personal interactions and other more day-to-day personal
events comprise the majority of the account. A “great
person” account, a type of internalist account, attempts to tell
history through the influence of extraordinary individuals.
Individual acts of genius are the main constituents of this
account.

Whiggish accounts of history are often seen as fundamentally

flawed historical accounts. In these accounts the past is seen
from the lens of what is now judged to be acceptable or true.
Those events that can be seen as contributing to the present

29
are seen as good or important. Those that hindered or are seen
as irrelevant to what is presently seen as correct are then
depicted as bad or unimportant. Whiggish accounts are
sometimes referred to as “presentist” accounts.

Cognitive accounts of history attempt to tell history through

its propositional content. A cognitivist account would proceed
along the lines of key beliefs and arguments. An example
would be: In 1952 Eysneck argued that the eventual basis for
the superiority of extant psychotherapies over placebo or
spontaneous remission was problematic. This claim leads
others to believe that new therapies needed to be developed
and that these new therapies needed to be more adequately
tested.

In this volume we have attempted to tell a history of the

behavioral therapies through the autobiographies of some of
the key players (and in the case of a few particularly
important individuals who have died — Skinner, Wolpe and
Kantor, through canonical biographies). This
autobiographical approach generally has some externalist
qualities but is mainly an internalist approach. It is inherently
a “great person” approach. This autobiographical approach
has many advantages, particularly in that it allows the major
actors to tell their own stories directly. There is no historian
making errors that are inevitable when a third party, who
often was not present, attempts to reconstruct events.
Importantly, this approach is not only a “read” of the
historical record; it actually goes a long way toward helping
to create the historical record. With some limitations briefly
described below, canonical histories are written by obtaining
the statements of the key actors. Finally, the autobiographical
account is potentially a very personal and hence potentially

30
more “alive” account. However, there are at least two major
limitations to the autobiographical approach that also need to
be recognized: 1) the biases introduced by allowing
individuals to tell their own stories; and 2) errors introduced
by the particular sample of individuals chosen.

To be explicit regarding the second possible source of error, a

different, more complete account would have been given if
three additional sets of autobiographies were obtained. First,
autobiographies of some of the basic researchers whose
research set the stage for the behavioral therapies. These
would include Pavlov, Schenov, Bechterev, Watson,
Thorndike, Guthrie, Tolman, Hull, Spence, and Mowrer,
among others. Kazdin’s (1978) excellent History of Behavior
Modification: Experimental Foundations of Contemporary
Research nicely documents these influences. The reader is
highly recommended to read this text to gain a fuller
historical account of this stage of the development of the
behavior therapies. Second, autobiographies of individuals
who engaged in developing proto-behavior therapies —
usually practical but fairly isolated applications of
conditioning principles — in the first part of the 20th century.
These individuals include people such as John Watson,
Rosalie Rayner, Mary Cover Jones, Jules Masserman, O.
Hobart Mowrer, Andrew Salter, John Dollard, Neal Miller,
among others. These individuals often were innovative and
produced an exemplar — perhaps of a treatment technique,
perhaps of a conceptual framework in which techniques could
be developed — that shed light on a path that was followed in
the last half of the 20th century. Finally, due to logistical
limitations as well as conflicting schedules the set of
autobiographies and biographies of first generation behavior
therapists is not exhaustive. The stories of Ayllon, Azrin,

31
Becker, Brady, Eysenck, Ferster, Geer, Gewirtz, Goldiamond,
Greenspoon, Kalish, Kanfer, Keller, Lan, London, Lovaas,
Malott, Marks, Michael, Patterson, Rachman, Salzinger,
Sidman, Staats, Stolz, Ullmann, Yates, among others are not
told in this volume, but if these were a fuller account would
have been made. Another history of the second generation of
behavior therapists would also be key to further understand
the history of behavior therapy. It should be notes that we
largely tell the story of the rise of behavior therapy in the
United States, although it also had important roots in Great
Britain and South Africa. Part of the South African story is
told here in the biography of Wolpe and the autobiography of
Lazarus. However, the British story is admittedly slighted,
mainly due to the logistics of travel. Finally, a more complete
account of the history of the behavioral therapies would track
the influences noted by Krasner and Ullmann (1965):

• The concept of behaviorism in experimental

psychology (e.g., J. R. Kantor, 1924, 1963).
• The instrumental (operant) conditioning concepts of
Thorndike (1931) and Skinner (1938).
• The technique of reciprocal inhibition as developed
by Wolpe (1958)
• The studies of the group of investigators at Maudsely
Hospital in London under the direction of H.J.
Eysneck (1960, 1964).
• The investigations (from the 1920s through the
1940’s) applying conditioning concepts to human
behavior problems in the United States (e.g., Mowrer
& Mowrer, 1938; Watson & Rayner, 1920).
• Interpretations of psychoanalysis in learning theory
terms (e.g., Dollard & Miller, 1950), enhancing

32
 learning theory as a respectable base for clinical
work.
• Classical conditioning as the basis for explaining and
changing normal and deviant behavior (Pavlov, 1928)
interactionism, social psychology and sociology.
• Theoretical concepts and research studies of social
role learning and research in developmental and child
psychology which emphasized vicarious learning and
modeling (Bandura, 1970; Jones, 1924).
• Social influence studies of demand characteristics,
experimenter bias, hypnosis, and placebo (Frank,
1961).
• An environmental social learning model as an
alternative to a disease model of human behavior
(Bandura, 1969; Ullmann & Krasner, 1965).
• Dissatisfaction with psychotherapy and the
psychoanalytic model.
• The development of the clinical psychologist as
scientist-practitioner.
• A group of psychiatrists emphasizing human
interaction (e.g., Adolph Meyer, 1948; Harry Stack
Sullivan, 1953).
• A utopian stream emphasizing the planning of social
environments to elicit and maintain the best of man’s
behavior (e.g., Skinner’s 1976 Walden Two).
We chose first generation behavior therapists as we wanted to
tell the story of the beginning of behavior therapy. We
defined first generation as individuals we found critical to the
development of behavior therapy in the 1950’s and early
1960’s. We chose these particular individuals as they have
been the most influential. Others were excluded either due to

33
their early deaths (e.g., Hans Eysneck, Charles Ferster), space
limitations, or to their personal decision not to participate.

The structure of most chapters follows suggestions given to

the authors by the editors. We asked the authors to give a
brief intellectual and personal biography. We wanted to trace
their personal histories so we could see the route each took to
become an early player in behavior therapy. We then asked
them to present an intellectual case study of one of their key
works. We wanted to do this as we thought this focus would
provide us with specificity regarding the sources of influence
and the consequences of one publication. We wanted this
story to have both bandwidth and fidelity. Finally, we asked
each of these individuals to take a look at the current status of
behavior therapy and provide useful object lessons from their
perspectives. What should we pay more attention to? What is
going right? What is going wrong? We thought that these
originators might have some wisdom that ought to be
recorded and noticed by current and future generations.

Object Lessons

What might emerge as important lessons from the past? We

think there are at least three deserving comment: 1) what
struck us as editors was the scholarship of the originators of
behavior therapy. They knew experimental psychology well.
They knew learning research and theory well. In fact many of
them might be reasonably thought of as learning researchers
who were utilizing humans with problems as their subjects.
They also were often scholars in general psychology, and read
in the original language forerunners like Pavlov. They also
were intellectuals and read widely in fields like philosophy,
sociology, political science and the like. For example, the

34
scholarship contained in Ullmann and Krasner’s (1969) early
writings has largely been unmatched. Currently we see a more
narrow technological focus in behavior therapy. The larger
intellectual perspective is largely lost and the behavioral
therapies are not enriched by the broader ideas and
developments in psychology and other fields.

The second object lesson, in our view, is that the role of

process in behavior therapy has been lost. Today most
research in behavior therapy is outcome research; horse race
studies examine what condition (placebo, no treatment, active
treatment) wins. Active treatment is often a packaged manual.
Many current behavior therapists see this as sound and
satisfactory. Admittedly it certainly can seem so, particularly
when compared to the many completely nonscientific
approaches to psychotherapy. However we conjecture that
there has been a shift. In the early days of behavior therapy
process research and outcome research were more frequently
intertwined. These early behavior therapists were testing the
effects of psychological processes (such as satiation,
extinction, etc.) on clinical problems. Now behavior therapists
test the effects of complex eclectic, often atheoretically
derived packages (many with no clear relation to a known
psychological process or principle).

The third object lesson is that we have lost the iconoclastic

value of the first generation: they were questioners; they were
radical; they had the courage to fight the power structure.
They were rule breakers. It is important to remember that they
were breaking the rules and they were not assured that this
would lead to success. As the discipline has matured it has
come to be more conventional, less revolutionary. Part of the
reason for this is that some of our values have been

35
assimilated into the power structure; it is also apparent that
we have made compromises. We no longer see radical
critiques of the medical model; we often use the DSM system;
we design treatment packages and research protocols that will
get funded by federal agencies; we eclectically adopt
assessment or treatment methods from other schools; our
ethical challenges to suboptimal treatment are less frequent
and more muted; and our goals have become more modest —
we seek no longer to change institutions or society but rather
to make a living. The boldness, zeal and courage to question
radically and to profess one’s convictions even when
unpopular have largely given way to a discipline which is
more conservative and conventional. Part of a mature
discipline is to be derivative — to do normal science; but in
this we also have lost some of the creativity of the early years.
Where are our revolutionaries that fight against the
problematic compromises or who see the radical possibilities
within our set of commitments and remind us that we are to
go in a vastly different direction? Should we be reminded of
the ideals of the founders and look again whether our
comittments to evidential standards could be improved;
whether our extrapolations from basic psychology could be
better; whether we seek to properly intertwine process and
outcome research; whether behavior therapy is profiting from
corrupt larger intellectual developments?

It is our hope that future generations of behavior therapists

keep alive the radical critiques of accepted ways of treating
changeworthy behavior as well as the commitment to finding
ways to improve human lives through a commitment to
scientific methods.

36
References
Bandura, A. (1969). Principles of behavior modification. New
York: Holt, Rinehart & Winston.

Bandura, A. (1970). Social learning theory. Morristown, NJ:

General Learning Press.

Coleman, S. R. (1988, May 28). What does the history of

psychology have on sale today? Paper presented at
symposium on history, philosophy and behavior analysis, at
the 14th annual convention of the Association for Behavior
Analysis. Philadelphia, PA.

Dollard, J. & Miller, N. (1950). Personality and

psychotherapy: an analysis in terms of learning, thinking and
culture. New York: McGraw-Hill.

Eysenck, H. J. (1960). Behaviour therapy and the neuroses;

readings in modern methods of treatment derived from
learning theory. New York: Pergamon Press.

Eysenck, H. J. (Ed.) (1964). Experiments in behaviour

therapy; readings in modern methods of treatment of mental
disorders derived from learning theory. New York: Pergamon
Press.

Eysenck, H. J. (1952). The effects of psychotherapy: An

evaluation. Journal of Consulting Psychology, 16, 319-324.

Frank, J. D. (1961). Persuasion and healing. Baltimore, MD:

Johns Hopkins University Press.

37
Jones, M. C. (1924). The elimination of children’s fears.
Journal of Experimental Psychology, 7, 382-390.

Kantor, J. R. (1924). Principles of psychology. New York:

Knopf.

Kantor, J. R. (1963). The scientific evolution of psychology.

Chicago: Principia.

Kazdin, A. E. (1978). History of behavior modification:

Experimental foundations of contemporary research.
Baltimore: University Park Press.

Krasner, L. & Ullmann, L. P. (Eds.). (1965). Case studies in

behavior modification. New York: Holt, Rinehart and
Winston.

Kuhn, T. S. (1963). The structure of scientific revolutions.

Chicago: University of Chicago Press.

Kuhn, T. S. (1979). The essential tension. Chicago:

University of Chicago Press.

Meyer, A. (1948). The concept of wholes. New York:

McGraw-Hill.

Mowrer, O. H. & Mowrer, W. M. (1938). Enuresis: A method

for its study and treatment. American Journal of
Orthopsychiatry, 8, 436-459.

O’Donohue W. & Ferguson, K. (2001). The psychology of

B.F. Skinner. Thousand Oaks: Sage.

38
O’Donohue, W. & Kitchener, R. (1999). Handbook of
behaviorism. San Diego: Academic Press.

O’Donohue, W. & Krasner, L. (1995). Theories of behavior

therapy: exploring behavior change. Washington, DC: APA
Books.

Pavlov, I. P. (1928). Lectures on conditioned reflexes (W. H.

Gantt, Trans.). New York: International Publishers.

Rachman, S. (1963). Introduction to behavior therapy.

Behaviour Research and Therapy, 1, 3-15.

Skinner, B. F. (1938). The behavior of organisms. New York:

Appleton-Century.

Skinner, B. F. (1976). Walden two. New York: Macmillan.

Smith, L.D. (1986). Behaviorism and logical positivism.

Stanford: Stanford University Press.

Sullivan, H.S. (1953). The interpersonal theory of psychiatry.

New York: Norton.

Thorndike, E. L. (1931). The fundamentals of learning. New

York: Bureau of Publications, Teachers College.

Ullmann, L.P. & Krasner, L. (1969). A psychological

approach to abnormal behavior. Englewood Cliffs, NJ:
Prentice-Hall.

Watson, J. B. & Rayner, R. (1920). Conditioned emotional

reactions. Journal of Experimental Psychology, 3(1), 1-14.

39
Wittgenstein, L. (1958). Philosophical investigations. New
York: Macmillan.

Wolpe, J. (1958). Psychotherapy by reciprocal inhibition.

Stanford, CA: Stanford University Press.

Wolpe, J. (1969). The practice of behavior therapy. New

York: Pergamon Press.

40
Chapter 1

The Importance of Case

Studies to Methodology of
Science
Thomas Nickles1

University of Nevada, Reno

1. Introduction: Historical Studies of

Recent Science
Until the 1960’s or so, history of science was concerned
almost entirely with old science, for example, the scientific
revolution from Copernicus to Newton, the chemical
revolution of Lavoisier and Priestley, and the experimental
revolution of Wundt and company in psychology. Indeed, the
very phrases ‘history of present science’ and (even) ‘history
of recent science’ sounded oxymoronic. To be sure,
philosophers of physics were interested in the early
20th-century revolutions in relativity theory and quantum
theory, but historians found it difficult to see the 20th-century
work of living scientists as history, both because it was not
really past and because it did not readily lend itself to good
story, that is, a semi-literary production accessible to the
historians’ usual audience — the educated general reader. For
the technical details were formidable, and, Einstein and Marie

41
Curie aside, it was increasingly difficult to understand
scientific breakthroughs as the achievement of great,
Romantic geniuses working in isolation, against scientific,
religious, or political orthodoxy.

Since the 1960’s, the situation has changed radically. Today

the field of “science studies” is flourishing. Science studies,
which is also called Social Studies of Science (the name of its
leading journal), emerged in Britain, in the 1970’s, in the
form of radical sociology of science, and quickly spread to
the Continent and to America. This movement attacked
standard philosophy of science as too a priori and dogmatic,
and it attacked internalist history of science for remaining too
much in the old history of ideas genre, in which ideas were
often treated as if they develop in a disembodied manner
according to their own internal logic and timeless standards of
evaluation. Today the field of science and technology studies
encompasses not only the sociology of science but also the
anthropology, social history, political science, and economics
of science and technology. In its generic sense, it includes
old-style history and philosophy of science and even
psychology of science, but most science studies scholars still
hold philosophy, internalist history, and individual
psychology in suspicion. A few years ago two prominent
scholars (Bruno Latour & Steve Woolgar, 1986) only half
facetiously suggested a ten-year moratorium on psychology
of science in order to allow genuinely social accounts of
cognition to flourish. Many philosophers, meanwhile,
continue to appreciate internalist history and also individual
psychological treatments of scientific work, be those
treatments cognitivist, behavioral, or (occasionally) Freudian.

42
But beyond warning that individual biography, including
psychobiography of any kind, is considered old fashioned in
some quarters, my purpose today is not to belabor these
differences among the science studies disciplines. The point
of bringing up science studies is that its emphasis is on
20th-century science, especially recent science and
technology and science policy as they have been transformed
by World War II and postwar developments. The historian of
technology Derek Price (1963) marked this difference as a
distinction between “little science” and “big science.”

Even historians have now moved into the 20th century. At

George Washington University, for example, there is a Center
for History of Recent Science, headed by Horace Judson. In
his spring 1999 newsletter, Judson, who works on the history
of recent biology and biotechnology, notes that until recently
writing about recent and current science was considered a
journalistic activity rather than one of serious scholarship
(Judson, 1999).

Yet understanding recent history of science is obviously

important, for several reasons. One is that, insofar as the
character of 20th-century science has changed, it is dangerous
to draw one’s methodological lessons from the classical,
founding cases alone. And, after all, professional psychology
is little more than a century old and developing rapidly.
Another sort of reason is that, along with the development of
market economies and bourgeois life, the rise of modern
science and technology has surely done more than anything
else to shape modern culture. These changes have become
increasingly rapid and pervasive in the 20th century. A fourth
point relates to the scale of scientific activity. Price noted that
the amount of scientific investigation had increased

43
exponentially from the time of the scientific revolution until it
began leveling off in the 1970’s. It is sobering to think that
about 80% of the scientists who have ever lived are still alive
today. There is therefore some urgency in studying the history
of recent science before the participants die.

Now it may seem that recent and contemporary history is easy

to do. After all, one still has most of the principals available
for direct interrogation (or to do the job themselves). The
challenging old intellectual mysteries about what scientist X
and Y were really doing — and what led them to their
problems and then to their epochal solutions — now seem to
vanish, for we can simply ask them!

However, things are not so easy. As Judson says, journalistic

accounts, with their anecdotal evidence and apocryphal
stories, are not to be confused with serious, professional
historical scholarship. Moreover, modern communications
technology often makes it more, rather than less, difficult to
trace the emergence of new ideas and practices. Today’s
scientists tend not to leave extensive paper trails, as did the
likes of Newton, Faraday, and Darwin — in the form of early
drafts of papers, frequent, detailed letters to colleagues, lab
notebooks, and diaries of personal reflections. Rather,
scientists now communicate by telephone, e-mail, and
personal contact at conferences and workshops. Circulated
preprints and grant proposals rather than the published record
are often where the action is — things that often have an
evanescent existence. When scientists retire, they often throw
away their personal files.

Furthermore, the idea that we can simply ask living scientists

what they meant, and what motivated them, turns out to be

44
naïve, an instance of the intentional fallacy and the whig
fallacy, both of which I shall explain later.

In his book, The Structure of Scientific Revolutions, originally

published in 1962 and expanded in 1970, Thomas Kuhn
called attention to these difficulties and remarked that
“History, if viewed as a repository for more than anecdote or
chronology, could produce a decisive transformation in the
image of science by which we are now possessed.” Here was
a call to do serious case studies, and of recent science as well
as old science.

Some philosophers as well as historians took up the call.

Kuhn himself had shown the methodological relevance of
historical cases by deftly employing them to discredit Karl
Popper’s and the logical positivists’ accounts of scientific
research. The new generation of historically-oriented
philosophers promoted the idea that philosophical accounts of
science, including scientific method, should be tested against
the data of history, with historical cases serving as data
points, as it were. To that degree, in other words, philosophy,
or at least philosophy of science, became an empirical rather
than a purely a priori subject. Today science studies is
thoroughly empirical in studying every facet of scientific
work. But it was the work of people such as Kuhn, Paul
Feyerabend, Stephen Toulmin, and Imre Lakatos in the
1960’s and early 1970’s that introduced this new kind of
philosophy of science based on historical case studies. This, I
shall contend, was an important step in weaning philosophy
of science from the traditional dogma that good science
should be constrained by a foundational epistemological
program.

45
A second reason for taking case-studies seriously derives
from a more radical suggestion, one that goes to the heart of
learning theory as applied to scientists themselves, that is, to
the very idea of scientific method. The basic idea of this
deflationary conception of method is that method itself is
better construed as a set of exemplary cases than as a set of
rules. This suggestion is at least implicit in The Structure of
Scientific Revolutions. In the next two sections, I proceed to
contrast two approaches to scientific method in this light.

2. Two Approaches to Scientific

Method, Rule-Based and Case-Based
The old philosophy of science associated with logical
positivism was, in most versions, a rule-based methodology
of science. There were rules for testing hypotheses and,
occasionally, even rules for generating empirical hypotheses.
Moreover, the positivists required that any abstract,
theoretical language be linked to observation language by
means of so-called correspondence rules, where the link
should be as close to explicit definition as possible —
although in fact this was rarely possible. This last sort of rule
brings out the empiricism as well as the logic of logical
empiricism. For this methodology of science was a
conservative epistemological program. One could do good
science only by being a good, conservative-empiricist
philosopher. The positivists, of course, recognized that
old-time foundational epistemology of the kind sought by
Descartes is impossible: we cannot achieve certainty in
empirical science. But they nevertheless insisted that every
result, indeed, every legitimate term of art, be empirically

46
well-founded before proceeding to the next step. For them the
source of all meaning was the data language, and the source
of all justification was the empirical data themselves. The
result was a one-dimensional, indeed, one-directional
conception of justification, a cumulative account of scientific
development, both doctrinal and conceptual, and an
empiricist, “building-block” theory of language learning and
concept formation, whereby each theoretical term could be
learned only in terms of less theoretical language and that
language in terms of still less, and so on back to the
observation language.

A story from the letters of the composer Arnold Schönberg

illustrates this idea. A blind man was discoursing with
someone who happened to use the term ‘white’, which for the
blind man was tantamount to a theoretical term.

“White?,” he asked.

“You know, the color of a swan,” came the reply.

“Swan?”

“Yes, the large bird with the long, curved neck.”

“Curved?”

“Here, like the shape of my arm as I am holding it now.” The

blind man felt his friend’s arm.

“Oh!” he exclaimed. “Now I know what white means!”

47
Compare the now-dated behaviorist joke: Two behaviorists
meet on the street. One greets the other: “You feel fine. How
do I feel?”

Some strands of behaviorism were broadly compatible with

logical empiricism, although the connection has been greatly
exaggerated (Smith 1986). To be sure, behavioristic
psychologists have, by and large, been conservative
empiricists when it comes to concept formation and theory
formation; however, they (and especially Skinnerian behavior
analysts) have largely ignored or explicitly repudiated the
linguistic analyses of the positivists. In general they are closer
to the Darwinian, naturalistic epistemologies of the American
pragmatists. This is most obvious in Skinner’s case, given his
emphasis on selection by consequences as an extension of
Darwinian methodology; but Edward C. Tolman and Clark
Hull owed a debt to Darwin as well (Smith, 1986).

Now as everyone knows, no science has been more concerned

with method than psychology has. And method is standardly
conceived as a set of rules.

The picture of science that we get from Kuhn is quite

different. On Kuhn’s view, insofar as you are deeply
concerned with method, you are in deep trouble! The
emergence of rules signals a science in crisis. Normal
scientists working under a paradigm don’t need to appeal to
methodological rules (he says), because their work is directly
modeled on exemplary achievements that define the field of
inquiry in the first place. Given his severe criticism of the
positivists, Popperians, and the entire rules tradition of
concept learning and problem solving in general, we can say
that Kuhn, the most famous scientific methodologist of our

48
century, was actually an anti-methodologist. Although his
one-time colleague, Feyerabend, was notorious for his book,
Against Method (Feyerabend, 1975), Kuhn himself was
against method in his own way.

And yet Kuhn also has a positive message. We can read him
as propounding the view that the sciences operate by means
of a very different sort of method than the standard,
rule-based one. This is not only an alternative method: it is
also an alternative learning theory (including an account of
concept formation), indeed, an alternative treatment of what it
is to engage in rational inquiry. Like Ludwig Wittgenstein
before him and the cognitive psychologist Eleanor Rosch
(e.g., Rosch, 1973) and her associates after him, Kuhn
rejected the view that a concept is defined in all-or-nothing
fashion in terms of a set of necessary and sufficient conditions
or rules. Rather, concepts are based on a resemblance or
similarity relation, and hence graded.

This similarity relation is not entirely natural. It itself is in

large-part culturally learned, by paying attention to what that
culture takes to be exemplary cases or prototypes. Kuhn
termed such cases “exemplars,” and exemplar became the
primary meaning of ‘paradigm’ in the expanded, 1970 edition
of Structure of Scientific Revolutions. On Kuhn’s view, a
certain amount of cultural relativity is unavoidable, since we
never have direct, unmediated experience of the Kantian
Thing-in-Itself. Different linguistic communities lump
different things together as similar. Things that are similar in
the folk physics or folk psychology of Aristotle or of the
person in the street, may be quite distinct in modern scientific
physics or psychology; and vice versa.

49
What this amounts to, I would claim, is a rhetorical turn in
our conception of human cognition, of learning and inquiry, a
turn away from logic and toward rhetoric. For rhetoric is
concerned with simile, metaphor, analogy, and such tropes
rather than with sets of logically necessary and sufficient
conditions. Insofar as Kuhn is right, logical rules, where they
function at all in the process of inquiry, are derivative from
exemplars. In my next section I will sketch how this view
extends to a case-based as against a rule-based conception of
scientific inquiry.

First, however, I want to factor in Kuhn’s strong contention

that science is a problem-solving activity. Inquiry, he insisted,
is erotetic, that is, question-centered or puzzle-centered, rather
than theory-centered. In itself this idea was not new. One can
find it quite explicitly stated in Ernst Mach, Charles Peirce,
Popper, and also in artificial intelligence work by Herbert
Simon and his colleagues from the mid-1950’s on (e.g.,
Newell & Simon, 1972). But it was Kuhn who made the most
convincing case that scientific work is a problem-solving
rather than a directly truth-seeking activity and that scientific
knowledge in manifested in practical skill more in than the
construction of aesthetically pleasing world-views (see
Rouse, 1987, chap. 2; Nickles, 1988). In other words, Kuhn
not only promoted what I would call a pragmatic account of
scientific inquiry, in contrast to the epistemological
foundationism still present in modern empiricism, but he also
shifted the emphasis from knowing-that to knowing-how.
Previous philosophical accounts of all stripes had tended to
reduce knowing how to do something to knowing that a set of
propositions is true, or that a set of rules suffices for
accomplishing a task, a conception largely retained in

50
Chomskian linguistic theory and in standard artificial
intelligence.

A related element of this pragmatic, problem-solving

construal of scientific inquiry is that what I call heuristic
appraisal or heuristic pursuit becomes at least as important as
epistemic appraisal. When one reads the positivists and other
philosophers of science, even Popper, one finds most of the
emphasis placed on so-called confirmation theory or
corroboration, which attempts to answer the question, To
what extent can this scientific theoretical claim be justified,
and on what logical and empirical basis? This is part and
parcel of the epistemic tradition coming down from Descartes
and embracing both traditional rationalism and empiricism.
Conservatives in both camps want to avoid at all costs the
Type II error of accepting a false hypothesis. An extreme case
in the empiricist tradition is W. K. Clifford, the late
19th-century British mathematician and physicist. Clifford
once wrote: “It is wrong always, everywhere, and for
everyone, to believe anything upon insufficient evidence.”
This is the empiricist counterpart to Descartes’s refusing to
take the next step until he had achieved absolute certainty.

The American pragmatists sharply rejected this Cartesian

conception of science. This sort of epistemological purity,
said Peirce, would block the road to inquiry — and that was
his own chief constraint on scientific investigation: “Do not
block the road to inquiry.” In his celebrated article, “The Will
to Believe,” the psychologist and pragmatist philosopher
William James (1897) also rejected this whole conservative
ethos of science. In terms of our more recent terminology, his
point was that Clifford was so afraid of committing a Type II
error that he was likely to commit Type I errors of rejecting

51
true hypotheses, by refusing even to entertain them seriously
because they went far beyond currently available evidence.
Nothing ventured, nothing gained. Science is inherently a
risky business. In managing risk, whether in everyday life,
business life, or science, we need to consider the utilities, the
possible payoffs, as well as the epistemic probabilities.
Accordingly, James stressed the importance of the fertility of
a hypothesis. Don’t worry so much about where it came from
but look to where it might lead.

This is heuristic appraisal, in contrast to epistemic appraisal.

Philosophers have long emphasized that science is a
long-term, self-correcting enterprise, not one likely to arrive
immediately, if ever, at the absolute truth about the universe.
Yet they have paid remarkably little attention to the
importance of heuristic appraisal. But surely scientists in the
trenches are as much concerned with fertility as with truth. A
fertile idea or technique is one that can guide you toward
interesting problems, help you write good grant proposals,
and so on. Your conviction that a thesis is true does not
automatically make it fertile, does not suggest that you can do
anything with it. Truth does not equate to fertility. Better to
have a fertile hypothesis that you know is false than a sterile
one that you believe true! One scientific counterpart to
writer’s block may be to believe something without seeing
how to do anything with it.

Those who emphasize epistemic appraisal tend to have a

backward-looking perspective. They ask the question: How
strongly does the data, the track record of this hypothesis,
justify it as true? While this is certainly an important activity,
no question about it, at least equally important is the
forward-looking perspective of heuristic appraisal: Whether

52
or not this claim is certifiably true or this practice completely
unproblematic, does it open up new avenues of inquiry that it
would be fruitful to pursue?

One way of bringing out my point is to note that many times

in the history of science research has shifted away from an
entrenched theory or problem or practice not because it was
considered mistaken but because it was thought to be
exhausted of interesting results. Many of the best and
brightest scientists shift problem areas in mid-career when
they believe that most of the interesting work has already
been done in their former area. The question scientists ask of
an approach is not “What have you done for me?” or even
“What have you done for me lately?” but “What can you do
for me tomorrow?” Highly successful scientists, I believe, are
pragmatic opportunists. They do not stick faithfully to an
older approach that has served them well when they perceive
promising new opportunities for exciting work.

Let me now sum up my two, contrasting conceptions of

scientific method. The account that I find attractive

a. is a problem-solving versus a directly truth-seeking

enterprise.

b. is pragmatic and opportunistic rather than

foundational-epistemological.

c. makes constant use of heuristic appraisal and does not

leave theory of justification to epistemic appraisal alone.

d. is forward-looking (prospective, future-directed) rather

than backward- looking (retrospective, historical).

53
e. flourishes on case-based thought and practice over
thoroughgoing rule-based inquiry.

In my next section I shall say a little more about case-based

vs. rule-based problem solving and how exemplars can have
heuristic power.

3. Case-based vs. Rule-based Problem

Solving
In my view the most interesting methodological and
epistemological feature of Kuhn’s Structure of Scientific
Revolutions is not his doctrine of revolutions but his account
of concepts-and-categorization, and his corresponding
account of problem recognition and puzzle solving as
exemplar-based or case-based rather than explicitly
rule-based. Although he sometimes allowed that the physical
implementation of his model might involve complete physical
determinism, and possibly rules, at the neural level, Kuhn
insisted that, at the levels of description to which we humans
have cognitive and methodological access (levels of human
communication and control), scientific inquiry amounts to
learning by example, to case-based pattern-matching of new
puzzles to exemplary problems and solutions already
available. It does so by means of “acquired similarity
relations” rather than rule-based derivations of solutions from
first principles. Hence the priority of paradigms to rules
(Structure §V).

In effect, Kuhn argued that physical scientific inquiry is

driven by case-based rather than rule-based reasoning and
practice. While the terms ‘case-based reasoning’ and

54
‘rule-based reasoning’ (CBR and RBR) are today most
commonly used in artificial intelligence (AI), the basic ideas
have a long history dating back to debates between medieval
logic, with its concern with universality, and rhetoric, with its
concern for exemplum — and even back to Socrates’s attempt
to defend rule-based definition over examples. Although I
think that Kuhn exaggerated the incompatibility of the two
approaches and underestimated the role that rules of various
kinds can play, I shall, for present purposes, continue to draw
the distinction sharply.

To be sure, most extant case-based problem-solving systems

in AI are implemented in terms of rules, but the mode of
problem solving is interestingly different from that of the
older generation of rule-based problem solvers. As Kuhn
already emphasized, there is a genuine difference between
case-based and rule-based practice at consciously accessible
levels of methodological description and advice. For Kuhn,
learning and computation remain largely implicit or tacit at
these levels of description and are the more efficient for being
so. In many problem domains, requiring explicit inference
rules and definitions of concepts blocks the road to inquiry.
Over the past twenty years, work in cognitive science, even in
AI, seems to teach the same lesson, although to what degree
remains controversial.

A rule-based system solves problems by following strategies

expressed in terms of rules, e.g., production rules. The basic
idea is to solve a problem by deriving its solution from first
principles plus any heuristic rules. A problem solution is thus
a kind of heuristic proof, and the problem-solving system a
kind of logic supplemented by heuristic rules and empirical
assumptions. For some, that is what the adjective

55
‘computational’ means and requires. In the heyday of expert
systems research, or knowledge-based computation,
“knowledge engineers” attempted to elicit knowledge-laden
heuristic strategies, in the form of rules, from experts in the
particular field of the AI application. A major difficulty of
this approach was the so-called “knowledge-elicitation
bottleneck” or “Feigenbaum’s problem.” Many experts
claimed not to be using rules at all; and when they did offer
rules that fit the current problem, they often violated these
supposed rules when given new problems. The rules were not
uniformly projectable onto new cases.

Now back in the 1960’s, Kuhn had made exactly the same
prediction of philosophers who tried to reduce scientific
practice to rules. In today’s terminology these philosophers
treated methodology of science either as a rule-based “general
problem solver” (truncated by the logical positivists to a logic
of justification only) or as a kind of content-laden, rule-based,
expert system; whereas, for Kuhn, scientific methodology
(insofar as that enterprise can be defended at all) is a
case-based rather than a rule-based system. Thus AI’s
experience with the knowledge-elicitation bottleneck
confirms Kuhn’s claim that sharing a common practice does
not entail sharing a common, rule-based theory of that
practice (Hoyningen-Huhne, 1993, p. 137). Interestingly, the
“expert systems” philosophers were not only the positivists
and Popperians. They prominently included the new,
historical philosophers of science, who attempted to extract
methodological rules from historical cases. However, the
latter were more interested than the positivists were in
heuristic rules that possess real, problem-solving power.

56
Another difficulty with rule-based reasoning (RBR) is that
rule-based systems do not scale well. Indeed, they typically
become slower and clumsier instead of faster as more
knowledge-based rules are added. They do not degrade
gracefully. Conflicts among rules are hard to avoid, since no
one can see all the implications of present rules or of adding
new rules. (The unavoidable failure to recognize all the
deductive consequences is one aspect of the Meno problem.)
Moreover, in order to be reliable, rule-based systems must be
pretty complete; but that makes them relatively static and
unable to learn from experience, including learning from their
mistakes. These and several other difficulties have dampened
the early enthusiasm for simple, rule-based expert systems as
a model of human scientific inquiry.

A case-based approach can often avoid such difficulties.

Case-based reasoning (CBR) is more contextual or situational
than RBR, since rules must abstract over a variety of contexts
and situations. A case description tends to be more
meaningful and easier to remember than an abstract rule, and
it may even lend itself to story. Another advantage of CBR
over many forms of RBR is that one can solve sufficiently
matching problems with confidence even when one has only a
few cases in the case library, as when one imitates or copies
someone else’s way of handling a task. (This sort of
learning-by-being-shown falls somewhere between the
Skinnerian distinction of contingency-shaped and
rule-governed behavior: see §4 below.) Thus CBR is valuable
in ill-structured domains where algorithms and strong
heuristic rules are lacking — and the research frontier is, by
definition, ill-defined in this way. CBR does not perpetually
solve each problem “from scratch” but efficiently matches
new cases to old and adapts their solutions. It is therefore one

57
form of adaptive problem solving. A case-based approach
can, in principle, handle difficult, nonlinear problems in this
fashion, because it does not require a Cartesian
decomposition of a problem or complex system into its
simplest logical components. Hence CBR represents a major
departure from the traditional method of logical analysis and
synthesis. Even when we are dealing with logic problems,
write Rumelhart et al. (1986, 44), “The basic idea is that we
succeed in solving problems not so much through the use of
logic, but by making the problems we wish to solve conform
to problems we are good at solving.”

This sounds very Kuhnian. Kuhn went on to say that a theory

or paradigm (in the large sense) is basically a set of exemplars
sufficient to provide solutions to an entire, significant domain
of problems. That is, (1) relatively few exemplars provide the
basis for solving a potentially infinite number of concrete
problems; and (2) these exemplars constitute a “basis set” that
“spans” the problem space (my terminology, not Kuhn’s). For
according to Kuhn a paradigm guarantees that all puzzles in
its domain are solvable by clever modeling upon its various
exemplars. In saying that the available exemplars are
sufficient to generate the complete space of admissible
problem solutions, this guarantee amounts to a strong
generatability or discoverability claim (cf. Marshall, 1995;
Nickles, 1985).

Exemplars can therefore have great heuristic power. Since an

exemplar contains a problem-plus-solution and includes
components of skilled practice as well as theory, finding a
suitable exemplar or three that seem to fit a new problem area
can provide considerable heuristic guidance, since much of
the necessary know-how is already built in. The relevant

58
exemplary cases show us how to proceed. In fact, drawing
attention to the applicability to new problems of off-the-shelf
exemplars is probably the most important facet of heuristic
appraisal.

4. Various Difficulties
However, not everything here is sweetness and light. A
case-based account of problem solving faces difficulties of its
own. What is to count as a case? How do we produce the very
first exemplars of a new field? At the other end of this
spectrum, which of the zillion possible cases are to be
retained in the case library? How can relevant cases be
indexed and retrieved? By what mechanism is a new case
recognized to be relevantly similar to one case but not to
another? In scientific research what and where is this case
library, anyway? Is it internal to the individual investigator or
does it reside externally in the informational resources of the
community, e.g., in real libraries or distributed over
communities of investigators? Clearly, retaining in accessible,
episodic memory all instructive, problem-solving experiences
as distinct cases is an impossibility. It is too piecemeal.

So, while it may ameliorate some aspects of the Meno

problem and cognitive economy problems, CBR would seem
to exacerbate others. After all, one main attraction of general
rules is economy: one rule can subsume an infinity of possible
cases, and rules can often be organized into a system.
(However, an example can also stand for a potential infinity
of cases.)

59
We must ask similar, skeptical questions about Kuhnian
exemplars and the processes by which they are constructed,
indexed, retrieved, and activated. His own system of
case-based reasoning turns out to be too simple. In fact, a
CBR perspective already reveals some shortcomings of
Kuhn’s account of exemplars. For example, Kuhn’s
exemplars are all positive achievements, whereas, typically,
some of the most exemplary lessons are negative. Janet
Kolodner (1995) stresses that we need to include the notable
failures also — the “war stories.” Kuhn 1974 does note the
need to learn dissimilarities among things when acquiring the
similarity relations, but his scientific exemplars all seem to be
achievements rather than failures (Hoyningen, 1993). In
neglecting what we learn from negative exemplars, Kuhn’s
position is too far from Popper’s!

Second, surprisingly, Kuhn’s account of exemplars is not

sufficiently historical. He often presented them as fixed
historical achievements, as static anchors for future research,
when, in fact, exemplars themselves have a history. Insofar as
normal research adapts old solutions to new problem
environments in a case-based manner, by replicating them
with variations and then selectively retaining those that work
better, we should expect the set of exemplars to evolve. Think
of the history of Planck’s black-body radiation law and of
Bohr’s atom model, from the original papers through the
various editions of Sommerfeld’s Atombau und Spektrallinien
(cf. Kuhn, 1978)! The whiggish misremembering and
rereading of the past in terms of the present that worried
Kuhn the historian also occurs within the incremental
development of normal science itself. As Kuhn the historian
did recognize, we need to ask not only how the fund of past
experience conditions present perception and practice but also

60
how present problems and commitments shape our recall of
past work. The scientist qua scientist is necessarily a bad
historian (Nickles, 1992). We should expect a mutual
adjusting, a mutual fitting, of old exemplars to new problems,
not a one-way influence. Hence, Kuhn’s account is
dynamically inadequate. This point is directly relevant to the
concerns of this conference on the history and future of
behavioral therapy.

One difficulty faced by a case-study approach is that

scientists’ treatment of cases is quite different from
historians’. Historians aim to understand the past in its own
terms, e.g., to understand what Wilhelm Wundt and J. B.
Watson were trying to do in their own contexts, and not with
an eye to what (the historian knows) came later. Attempting
to understand and evaluate past work in terms of present
conceptions is anathema to historians and is called the whig
fallacy (Butterfield, 1931). It is actually a collection of
several related ways of being unhistorical, but the basic idea
is that whig history judges previous work as good or bad, as
insightful or wrongheaded, insofar as it anticipates our most
recent work. Kuhn and other historians sometimes say that it
would be better if historians did not even know how the story
eventually turned out.

By contrast, the perspective of scientists working at the

frontier of research is necessarily whiggish. They are
interested in exemplary cases, positive and negative, insofar
as those cases look similar to present problems and suggest
solutions. Much of the creative work of research consists in
mutually deforming past exemplars and present problems in
order to achieve an adequate match. We might say that
scientists are forward-looking, whereas historians are

61
backward-looking (or, rather, forward-looking from the point
of view of some time now past). Moreover, with each major
success, indeed with each grant proposal, scientists rewrite
the previous history of their subfield in order to make their
present work look like a logically plausible, if not inevitable,
continuation of previous work. Scientists, as such, use history
to clarify and advance their current projects. In this respect,
good science is bad history, and vice versa (see Nickles,
1992).

We can now see why it is dangerous for historians to take

scientists’ accounts of their earlier work at face value, for
scientists, as such, cannot help but perceive that work through
the filter of everything that has happened since then in their
specialty area, not to mention the filter of their hopes and
expectations for the future. Especially when they do not have
the opportunity to study carefully their own early drafts, lab
notes, and the like, scientists’ recollections are notoriously
inaccurate. And even when they are accurate, one can
question whether the scientists are absolute authorities on
what they were and are doing. For, contrary to the standard
conception, scientific texts are rather like literary texts in not
possessing a single, definitive, “absolute” meaning. Rather,
they are susceptible of variant (yet entirely competent)
readings. For example, Einstein and Ehrenfest found much in
(or read much into) Planck’s work that Planck had not
realized that he put there and even repudiated. And just before
he died, Niels Bohr vehemently denied as crazy some traces
of his earlier ideas that Kuhn and his student, John Heilbron,
claimed to find in Bohr’s famous 1913 papers on the quantum
atomic structure of the elements. Historians and scientists are
frequently at odds over what the scientists themselves
previously thought and did! As in political history, there can

62
be a struggle for authority over the historical record. Are
professional historians or the interested participants
themselves the authorities?

To take another famous example, E. G. Boring’s History of

Psychology (1929 and later) is whiggish in just the ways that I
have described. He wrote history more as a scientist than as a
historian. On his conception, as stated in his prefaces to the
various editions, what we should include in the history of
psychology changes as the field changes. A main criterion of
his selection of older material was how well it related to
current work.

The issue of authority over the history and meaning of a piece

of work provides the transition to a brief discussion of the
intentional fallacy. This is the mistake of taking the author of
a paper or book as the absolute and exhaustive authority on its
content and meaning. On this view, the author’s intentions
determine the meaning completely. People working in literary
theory exposed this mistake long ago. That it is a mistake is
almost an axiom of literary and art criticism, for such
criticism has little point, at least in the case of living artists
and authors, if all one has to do is to ask them what the novel
or the painting or the musical score means. Once again the
question of authority is central. How privileged, if at all, are
the author’s own claims about what a work means?

Both fallacies caution us that anyone wanting to understand

how science is really done has to be very careful of scientists’
own accounts of their past work. When Kuhn said that
genuine historical work could transform our image of science,
he was extolling the virtues of professional history not only
over journalistic, anecdotal history, and skeletal chronologies

63
(i.e., history as a list of established facts and dates), but also
over history as the late-career reflections of practicing
scientists themselves.

However, the important question I want to raise is this: Is it

equally dangerous for you scientists to write case histories
not as genuine history but with methodological intent, for
your own practice and that of your students?

My short response to warnings about the tendency toward

whiggism in this context is: So what?! After all, qua
scientists, you are, and must be, whigs at heart. That seems to
me no sin as long as you recognize that you are teaching your
students to do what you take to be good science rather than to
do good history. This use of “history” is an important part of
the enculturation into a scientific community that Kuhn
himself emphasizes in other passages.

Thus I would urge that we distinguish two rather distinct

kinds or uses of historical cases: the whiggish “potted
histories” reported by you scientists yourselves, on the one
hand, and the cases (sometimes the same cases) as
reconstructed by historians and historical sociologists and
anthropologists, on the other hand. Still a third use is that
made by philosophers in testing their own models of scientific
development. Unfortunately, Kuhn tended to conflate all of
these things in his notion of exemplar.

Another difficulty of the case-study approach is that we

almost cannot resist generalizing from anecdotal cases, an
evidentially sloppy practice that attempts to turn cases into
rules. A related temptation is to fall into the “just one more
case” fallacy, the mistake of forcing new cases into the

64
framework of the few cases already available. This is the
mistake, if your only tool is a hammer, of seeing every
problem as a nail.

A further problem, one relevant to the business of this

conference, is that much scientific knowledge consists of
practical skills or know-how rather than knowledge that can
be readily articulated. As I noted before, Kuhnian exemplars
have a practice dimension. Insofar as this is true, we should
all keep in mind that the stories you scientists tell can capture
only so much of your scientific work. Are these stories fully
effective only to an audience of specialists?

Let me close by raising one last difficulty for a Kuhnian,

case-based approach to method. Some of you, our
distinguished program participants, have done path-breaking
work on rule-governed learning, understanding, and behavior
(see, e.g., Hayes, 1989). And, like Skinner himself, you surely
believe that your accounts of cognition apply not only to your
subjects but also to you yourselves in your scientific behavior.
So, in boosting case-based learning over rule-based learning,
are Kuhn and I flying in the face of your results?

This issue is a large and difficult one, and my response must

be brief. I personally doubt that there is a direct contradiction
in most cases. It seems to me that a case-based approach to
scientific work does not contradict rule-based learning in the
broad senses of that term that many of you employ. For by
‘methodological rule’ Kuhn meant an abstract generalization
of universal applicability, akin to the laws of logic; whereas
‘rule-guided’ for you normally means behavior guided by
verbal instruction (which may involve setting up quite local
contingencies of reinforcement on that basis) rather than

65
shaped directly by the contingencies of direct experience.
Here I am invoking something like Skinner’s distinction
between contingency-shaped versus rule-governed learning
(Skinner, 1989). In any case, Kuhnian exemplars cannot be
fully conveyed by verbal instruction.

Please note that I am not claiming that there are no

methodological rules at all, or that rule-based and case-based
reasoning are purely incompatible. Clearly, routine
procedures do exist, e.g., statistical analysis methods, many of
which have been automated in the form of computer
programs and have themselves become exemplars. And
exemplary cases, as I have described them, often split the
difference between contingency-shaped and rule-governed
behavior as you psychologists use those terms. Rather, I am
claiming that innovative problem-solving research is better
considered as case-based rather than as purely rule-based, in
the philosophers’ sense, which derives from a two-thousand
year logical tradition. For innovative research typically
matches new problems to old problems-plus-solutions, as
nearly as possible, and does not pretend that solutions can be
produced by means of general rules.

One consequence of my Kuhnian view is that

problem-solving is a more local, domain-relative activity than
rule-based methodologists would have us believe. Now if an
entire community of investigators has been brought up on the
same sets of exemplars, as in Kuhnian normal science, then
there will be nearly universal agreement about which
problems are genuine and which solutions are successful. But
in a field such as psychology, in which one finds distinct
schools even within the behavioral approach, we cannot
expect such “unanimity of agreement” and “fullness of

66
communication” (Kuhn, 1962). For recognition of
problem-solving progress is now more a matter of one’s
rhetorical tradition than of logic-plus-empirical data. Science
turns out to be a highly cultural activity!

References
Boring, E. (1929). A history of experimental psychology. New
York: Appleton-Century-Crofts.

Butterfield, H. (1931). The whig interpretation of history.

London: Bell.

Donovan, A., Laudan, L., & Laudan, R. (Eds.). (1988).

Scrutinizing science: Empirical studies of scientific change.
Dordrecht: Kluwer.

Feyerabend, P. (1975). Against method. London: NLB.

Fisch, M. (1982). The writings of Charles S. Peirce: A

chronological edition. Bloomington: Indiana University
Press.

Graham, L., Lepenies, W., & Weingart, P. (1983). Functions

and uses of disciplinary histories. Dordrecht: Reidel.

Hayes, S. C. (Ed.). (1989). Rule-governed behavior:

Cognition, contingencies, and instructional control. New
York: Plenum Press.

67
Hoyningen-Huene, P. (1993). Reconstructing science:
Thomas Kuhn’s philosophy of science. Chicago: University of
Chicago Press.

James, W. (1897). The will to believe and other essays. New

York: Dover, 1956.

Judson, H. (1999, Spring). Why history of recent science?

Recent Science Newsletter. Center for History of Recent
Science, George Washington University.

Kolodner, J. (1993). Case-based reasoning. San Mateo, CA:

Morgan Kaufmann.

Kuhn, T. (1962). The structure of scientific revolutions (1st

ed.). Chicago: University of Chicago Press.

Kuhn, T. (1970). The structure of scientific revolutions (2nd

ed.). Chicago: University of Chicago Press.

Kuhn, T. (1974). Second thoughts on paradigms. In F. Suppe

(Ed.), The structure of scientific theories (pp. 459-82).
Urbana: University of Illinois Press.

Kuhn, T. (1978). Black-body theory and the quantum

discontinuity, 1894-1912. Oxford: Oxford University Press.

Lakatos, I. (1970). Falsification and the methodology of

scientific research programmes. In I. Lakatos & A. Musgrave
(Eds.), Criticism and the growth of knowledge (pp. 91-196).
Cambridge: Cambridge University Press.

68
Latour, B., & Woolgar, S. (1986). Laboratory life (2nd ed.).
Princeton: Princeton University Press.

Laudan, L. (1977). Progress and its problems. Berkeley:

University of California Press.

Laudan, L. (1981). Science and hypothesis. Dordrecht:

Reidel.

Marshall, S. (1995). Schemas in problem solving. Cambridge:

Cambridge University Press.

Nickles, T. (1985). Beyond divorce: Current status of the

discovery debate. Philosophy of Science, 52, 177-206.

Nickles, T. (1987a). Lakatosian heuristics and epistemic

support. British Journal for the Philosophy of Science, 38,
181-205.

Nickles, T. (1987b). From natural philosophy to

metaphilosophy of science. In R. Kargon & P. Achinstein
(Eds.), Kelvin’s Baltimore lectures and modern theoretical
physics: Historical and philosophical perspectives (pp.
507-541). Cambridge: MIT Press.

Nickles, T. (1988). Questioning and problems in philosophy

of science: Problem-solving versus directly truth-seeking
epistemologies. In M. Meyer (Ed.), Questions and
Questioning (pp. 38-52). Berlin: Walter De Gruyter.

Nickles, T. (1992). Good science as bad history: From order

of knowing to order of being. In E. McMullin (Ed.), The

69
social dimensions of science (pp. 85-129). Notre Dame:
University of Notre Dame Press.

Nickles, T. (1995). History of science and philosophy of

science. Osiris 10, 139-163.

Nickles, T. (1998a). Kuhn, historical philosophy of science,

and case-based reasoning. Configurations, 6, 51-85.

Nickles, T. (2000). Kuhnian puzzle solving and schema

theory. Philosophy of Science, 67, S242-S255.

Polanyi, M. (1958). Personal knowledge. Chicago: University

of Chicago Press.

Polanyi, M. (1966). The tacit dimension. Garden City, NJ:

Doubleday.

Price, D. (1963). Little science, big science (2nd Ed.). New

York: Columbia University Press.

Reese, H. (1989). Rules and rule-governance: Cognitive and

behavioristic views. In S. C. Hayes (Ed.), Rule governed
behavior: Cognition, contingencies, and instructional control
(pp. 3-84). New York: Plenum Press.

Rosch, E. (1973). Natural categories. Cognitive Psychology 4,

328-350.

Rouse, J. (1987). Knowledge and power. Ithaca, NY: Cornell

University Press.

70
Rumelhart, D., Smolensky, P., McClelland, J., & Hinton, G.
(1986). Schemata and sequential thought processes in PDP
models. In J. McClelland & D. Rumelhart (Eds.), Parallel
distributed processing (Vol. 2). Cambridge, MA: MIT Press.

Newell, A., & Simon, H. (1972). Human problem solving.

Englewood Cliffs, NJ: Prentice Hall.

Skinner, B. F. (1981). Selection by consequences. Science,

213, 501-4.

Skinner, B. F. (1989). The behavior of the listener. In S. C.

Hayes (Ed.), Rule governed behavior: Cognition,
contingencies, and instructional control (pp. 85-96). New
York: Plenum Press.

Smith, L. (1986). Behaviorism and logical positivism: A

reassessment of the alliance. Stanford: Stanford University
Press.

Sommerfeld, A. (1919). Atombau und Spektrallinien.

Braunschweig: Vieweg.

Whewell, W. (1840). Philosophy of the inductive sciences

founded upon their history. London.

Young, R. (1966). Scholarship and the history of the

behavioral sciences. History of Science, 5, 1-51.

71
Notes
1
I am indebted to the U. S. National Science Foundation for
research support on heuristic appraisal and problem solving,
and to the conference participants for helpful discussion.
Section 3 and parts of section 4 are borrowed from Nickles
(2000).

72
Chapter 2

Joseph Wolpe: Challenger

and Champion for Behavior
Therapy
Roger Poppen

Southern Illinois University

I have been asked to represent Joseph Wolpe, who was also a

competitor, though of a very different and much more
effective sort, as I hope to show in describing his
contributions to behavior therapy. We were provided with a
list of four areas to cover in our presentations and this will
serve as an outline: first, a brief intellectual biography;
second, important developments in the rise of behavioral
therapies; third, a case study of important publications; and
fourth, object lessons for the future. A more extensive
treatment of these and related topics may be found in Poppen
(1995).

Joseph Wolpe’s Intellectual

Development
Roots

73
An old proverb states, “As the twig is bent, so grows the
tree.” The course of Joseph Wolpe’s intellectual development
can be traced not only to his childhood, but even before that
to his family history. His grandparents emigrated from
Lithuania at the end of the 19th century, a time of reprisals
against its Jewish population, and settled in Johannesburg,
South Africa. Lithuanian Jews had a long tradition of
scholarship and respect for learning, and this was true in
Joseph’s own family. His parents, while not religious,
maintained traditional values of self-discipline, hard-work,
and learning. As the oldest of four children, young Joseph did
not disappoint them. He was a precocious and avid reader,
fond of sports and sports stories, but was not a gifted athlete
himself. Rather, he entered and won numerous scholastic
competitions, winning prizes in a wide variety of subjects.
Thus, early on a competitive repertoire was shaped, a
repertoire that persisted throughout his professional career.

In high school Wolpe developed an interest in chemistry,

essentially completing college-level courses while studying
on his own. Again, this may be a significant precursor of his
later achievements. It reinforced his skills in independent
study. And the subject matter itself developed his interest in a
systematic, orderly, quantitative, approach to a topic.

Another factor in his early shaping was the politically liberal

beliefs of his family. South Africa was a country of sharp
class distinctions and institutionalized racism. The ideas of
helping the less fortunate, changing an oppressive system,
seeking a place of freedom, and standing up for one’s beliefs,
can be seen throughout Wolpe’s life.

Forming the Question

74
As a compromise between his wishes to be a chemical
researcher and his parents’ wishes for a respectable career,
Joseph chose to study medicine, following the British model
of a six-year university curriculum after high school.
University provided new intellectual avenues for him to
explore. He discovered the joys of discussion and debate of
philosophical issues with other young medical students. He
was interested in epistemology, finding the empiricist
philosophers to be more satisfactory in explaining the origins
of knowledge and the operations of the mind.

His interest in epistemology moved him toward psychiatry

which, in the 1930’s, was dominated by Freudian theory and
practice. Wolpe wavered in his acceptance of the notions of
unconscious forces that determine behavior, first regarding
them as foolish and then giving them credence. Following his
usual pattern, once a topic captured his interest he plunged in
and read all he could find on it. He even began keeping a
dream diary and analyzed his own dreams.

Wolpe completed his studies in the early days of World War

II and volunteered as a medical officer in the Cape Corps. He
was assigned to a military hospital that saw a considerable
number of men with what was then called “war neurosis.” He
participated eagerly in their treatment, which, according to the
psychodynamic model, involved the release of repressed
memories through infusion of sodium pentothal, or
“narcoanalysis” — the traditional method of psychoanalysis
being much too lengthy to carry out in these circumstances.
Another characteristic of Wolpe’s professional life became
apparent at this time — an interest in outcome. However
dramatic the processes revealed by narcoanalysis, however
strong the emotional “release,” the overall results were

75
disappointing. Few young soldiers were returned to adaptive
functioning as a result of their treatment. As he did earlier,
Wolpe participated in vigorous discussion with his peers. But
rather than broad questions of epistemology, these discussions
focused on the causes and cures of neurosis.

Wolpe continued to read widely, including anthropology,

sociology, and political works. He was interested to find that
the Russian allies totally rejected Freudian theory. Karl Marx
and Ivan Pavlov suggested that environmental rather than
intrapsychic events were important causes of behavior.
Pavlov described “experimental neurosis” and proposed
neurological mechanisms to explain it. The idea that
environmental events resulted in physiologic changes which
mediated an individual’s behavior appealed to Wolpe.

In 1945, Wolpe met a psychology professor at the University

of Cape Town, James G. Taylor, who, upon hearing of his
interest in Pavlov, recommended Clark Hull’s Principles of
Behavior (1943). This book had a tremendous impact on
Wolpe. Hull’s rigorous system described behavior in terms of
elemental bits that were combined in quantitative relations,
analogous to the way physical matter was comprised of basic
particles described by the periodic table of elements.
Although Hull did not directly relate these behavioral
elements to physiology, Wolpe saw the operation of neural
connections. Like Pavlov, Hull’s work was based on animal
experimentation but described basic principles that could be
applied to all species, including humans. Unlike Pavlov,
Hull’s theory did not address neurosis, but it set Wolpe
thinking as to how it could be extended to include such
behavior. Wolpe had rejected Pavlov’s theory of cortical
pathology as responsible for experimental neurosis, as well as

76
Freud’s theories of conflicting unconscious forces, and saw in
Hull’s system an alternative to both. Just how this might take
form was unclear, but it provided him a direction for further
study.

Upon his discharge from the service in 1946, Wolpe returned

to the University of Witwatersrand to pursue the M.D. degree
which required a research dissertation. Of course no one on
the faculty knew anything about conditioning and learning, so
Wolpe again embarked on a course of independent study.
Fortuitously, Leo Reyna was appointed to the Psychology
Department later that same year. Reyna (1946) had just
completed his Ph.D. under Kenneth Spence, Hull’s
collaborator, and was well-versed in the intricacies of
Hull-Spence theory. Wolpe attended Reyna’s seminar on
learning and joined an informal group of students attracted to
this field of study. Reyna’s dissertation had dealt with
extinction of learned behavior; this seemed particularly
relevant to neurosis, which was notoriously persistent.
Discovering the principles that enabled the extinction of
neurotic behavior, they decided, would be the solution that
Freud and Pavlov had sought but failed to find. Reyna
continued to provide guidance and feedback as Wolpe carried
out his dissertation research.

Finding the Answer

Pavlov’s (1927) work on experimental neurosis opened the

door for the empirical investigation of its causes and cures.
Characteristic of his previous endeavors, Wolpe conducted an
exhaustive review of the experimental neurosis literature.
There seemed to be a bewildering variety of species and
procedural differences, making it difficult to come up with a

77
general paradigm. One feature was fairly consistent, however;
once it occurred, experimental neurosis was very difficult to
get rid of. This seeming permanence was responsible for
Pavlov’s suggestion of neurologic damage, a point Wolpe
was not willing to concede.

Wolpe found the research conducted by Jules Masserman

(1943) very relevant. Masserman had applied noxious stimuli
while cats operated a food mechanism, resulting in numerous
“neurotic” behaviors, including food refusal, hyperarousal,
and excessive timidity or aggressiveness. Masserman, in line
with Freudian theory, attributed this result to conflict between
motivational states. Wolpe proposed that “conflict” was
irrelevant; that the behavior was simply the product of
aversive conditioning, as Watson and Rayner (1920) had
described many years previously. Accordingly, he replicated
Masserman’s procedure, shocking cats as they opened a food
box, but he included another group of cats that were shocked
with no food present, thus eliminating the possibility of
“conflicting motivational states.” In both groups the effects
were the same, supporting his contention that aversive
conditioning was responsible for neurotic behavior. The
adverse arousal to shock had been conditioned to stimuli
associated with it. Wolpe called this conditioned arousal
“anxiety.” Part of the anxiety response included food refusal;
all animals refused to eat even though food had not been
paired with shock for some of them. This suggested a means
of treatment.

It is instructive to look at Wolpe’s reasoning as he adapted

Hullian theory to fit his observations of anxiety acquisition,
maintenance, and extinction. The main principal he employed
was Hull’s drive-reduction theory of reinforcement. In brief,

78
anxiety functioned as both a response and a drive. It was
evoked during acquisition by the unconditioned stimulus, and
during maintenance by the conditioned stimulus. When the
stimulus terminated, the anxiety drive was reduced,
reinforcing the anxiety response. Thus even though the
primary aversive stimulus no longer occurred, anxiety
continued to be reinforced through its own reduction, making
it remarkably persistent.

Wolpe proposed another feature of anxiety leading to its

persistence, namely that as an autonomic response, anxiety
generated little “reactive inhibition.” According to Hullian
theory, each occurrence of a motor response generated a
fatigue-like event, reactive inhibition, which could
accumulate into a drive state with repeated responding. If a
motor response occurred without reinforcement, then the
reactive inhibition would build up to the point where its
dissipation would reinforce “not-responding,” and the
response would cease. Thus extinction was seen as resulting
from reinforcement of “not-responding” through the reduction
of reactive inhibition. Since anxiety generated little reactive
inhibition, extinction was greatly delayed.

Faced with such resistance to extinction, it became necessary

to find another means to inhibit anxiety. For this, Wolpe
invoked a principal used by Sherrington in the description of
spinal reflexes, “reciprocal inhibition,” in which activation of
one muscle group, for example flexors, inhibited the
activation of an antagonistic group, for example extensors,
and vice versa. Wolpe noted that the autonomic nervous
system is comprised of two reciprocally inhibitory branches.
Anxiety was largely the response of the sympathetic branch;
digestion was largely mediated by the parasympathetic

79
branch. This explained the food refusal in anxious animals
and it also suggested a means of overcoming anxiety — by
feeding. If feeding could be made to occur in a situation that
evoked anxiety, then anxiety would be inhibited in that
situation and its bond to that stimulus weakened. This set the
stage for the second part of Wolpe’s cat experiment,
eliminating the experimental neurosis by feeding in the
presence of attenuated anxiety-evoking stimuli.

Wolpe’s success with this procedure confirmed the utility of

the reciprocal inhibition principle and provided the
foundation for seeking applications to human neuroses. The
first couple of years were difficult ones, as he set about
finding responses that could inhibit anxiety and developing
ways of presenting anxiety cues in a controllable manner. His
solutions are well known — his use of relaxation (adapted
from Edmond Jacobson, 1938), assertion (adapted from
Andrew Salter, 1949), and sexual arousal (pre-dating Masters
and Johnson, 1970) as anxiety inhibitors, and the anxiety
hierarchy, both imaginal and in vivo, as a means of controlled
exposure.

Wolpe’s clinical success further validated the concept of

reciprocal inhibition, crystallizing his intellectual
development at this point in time. All his medical training and
experience, his philosophical discussions and psychological
readings, his experimental work and clinical practice, came
together — formed a Gestalt (to borrow a term from the
cognitive tradition) that organized his past experience and
provided the context for all his subsequent work. Wolpe
synthesized Hume and Russell, Freud and Hull, Pavlov and
Sherrington. From his reading and philosophical discussions
of epistemology, Wolpe assumed an empirical, logical,

80
materialistic stance, opposed to mysticism and mentalism. His
study of medicine reinforced this position, adding a
physiologic reductionist point of view. From Freud he took
the idea of anxiety as the core of maladaptive behavior, and
also the idea that anxiety was acquired and could be removed
— though not in the ways that Freud proposed. From Pavlov
came the idea that neurosis could be studied experimentally in
the laboratory, its factors teased apart and examined in a
quantitative fashion. Pavlov also advanced the idea of a
neurologic substrate of behavior, though Wolpe did not
accept the mechanisms he suggested. Hull provided the
specific stimulus-response-reinforcement framework to
organize the study of behavior; Wolpe extrapolated these
principles from rats in a runway to people trapped by
crippling fears. Finally, Sherrington’s principle of reciprocal
inhibition suggested how to counteract destructive arousal
and replace maladaptive with adaptive behavior.

Developmental Milestones in Behavior

Therapy
Before Wolpe, there was no behavior therapy. His work was a
major factor in the birth and development of this field. In this
section Wolpe’s contributions to several major milestones in
the growth of behavior therapy are described.

Challenge to Traditional Psychotherapy

Hans Eysenck’s 1952 review of treatment outcomes, in which

he concluded that psychotherapy was no better than general
supportive care, and in some instances was inferior to it, was
the first salvo of a revolution in psychotherapy. Wolpe

81
(1952a) published a similar critique, with the added feature of
describing the success of his therapy procedures. Eysenck’s
article seemed to kick open an anthill and traditional therapy
proponents swarmed to the counterattack. But the challenge
was not simply to prove effectiveness compared to base rates
of recovery; the comparison now was with this newfangled
therapy based on learning theory.

A few words about Wolpe’s relation with Eysenck are in

order. In the early 1950’s, Eysenck established a group of
clinical researchers at Maudsley Hospital, London, who set
about developing and testing therapy interventions based on
Hullian learning theory. Eysenck quickly recognized Wolpe
as a valuable ally and recruited him into the battle against
traditional therapy. Wolpe visited Maudsley in the mid-50’s,
giving training seminars on his procedures and adopting parts
of Eysenck’s personality theory. Stanley Rachman, who had
studied with Wolpe in South Africa, joined the Maudsley
group after completing his Ph.D. and carried out pioneering
research on desensitization and related procedures. Wolpe
continued to communicate with the Maudsley group during
the 60’s and participated in the campaign against
psychotherapy. The critique of Freud’s case of Little Hans, by
Wolpe and Rachman (1960), is a classic example of this
assault.

Eysenck was probably just the irresistible force needed to go

up against the immovable object of traditional psychotherapy.
Though not the first to use the term “behavior therapy,” he
was instrumental in establishing it as the name of the new
therapy, for example as the title of a book reviewing the state
of the art in 1960, and in the title of the first professional
journal (Behaviour Research and Therapy) in 1963. Wolpe

82
provided much of the ammunition in these early battles. In
addition, his competitive juices were stirred by Eysenck’s
combativeness, and he carried on as a leader on the American
front.

Learning-Theory Foundation of Behavior Therapy

Wolpe was a strong proponent of the need for interconnection

between theory, research, and practice. At the most
fundamental level, he maintained that basic principles of
learning, discovered in animal laboratory research, were
applicable to complex human behavior. Theoretical accounts
of research findings informed his clinical practice, as for
example his use of the reciprocal inhibition principle in
developing his therapy procedures. And he sought to include
procedures that were developed by others under his
theoretical umbrella. He was critical of pure eclecticism, or
using techniques that seemed to work with no concern for
why they worked or how they were related to other
procedures. To the extent that behavior therapy is rooted in a
basic science of behavior, it is following the lead of Wolpe.

Of course there had always been widespread disagreement

among learning theorists and researchers as to just what the
basic principles are. Hullian theory faded from the scene just
as Wolpe was extending it to account for neurotic behavior.
No one found his drive-reduction theory of anxiety
compelling, and he was equivocal about it himself in later
years. In the 1960’s, Skinnerian theory and research methods
were popular, and “behavior modification” (aka “applied
behavior analysis”) made its appearance in the treatment of
people with schizophrenia and autism. Wolpe welcomed these
developments but maintained that operant conditioning

83
principles were not relevant for the emotional problems that
characterized neurotic behavior. For their part, most behavior
analysts did not address the types of clinical problems that
Wolpe dealt with and did not consider his physiologic
reductionist approach to be consistent with functional
analysis.

In the midst of this breech, the “cognitive revolution”

gathered momentum in the 1970’s. According to this view,
operant and respondent learning principles derived from the
animal laboratory were at best insufficient, and at worst
misleading, in accounting for complex human behavior.
Extensive systems of cognitive learning processes were put
forth to account for therapeutic change. In Wolpe’s view, this
was a counter-revolution rather than an advancement, a return
to mental mechanisms reminiscent of Freud’s psychodynamic
forces. He challenged cognitive formulations on
philosophical, theoretical, and practical grounds (Wolpe,
1978). Philosophically, he endorsed the approach of Gilbert
Ryle (1949), who regarded cognitive theory as a kind of
spiritualism, appealing to ghostly mental processes outside
the realm of a deterministic physical universe. Theoretically,
Wolpe held that so-called cognitive events could be reduced
to neurophysiologic processes or could be described as covert
behavior, as Skinner (1953) had suggested; no new
hypothetical mentalistic processes were necessary.
Practically, Wolpe held that cognitive therapy procedures
were either ineffective, redundant with behavioral ones, or
could be accounted for by learning principles.

In the past decade, behavior analysts have begun to address

anxiety disorders and other complex clinical problems — for
example, my own work on relaxation training (Poppen, 1998)

84
and, most notably, the work of Steve Hayes and his
colleagues on Relational Frame Theory and Acceptance and
Commitment Therapy (Hayes & Wilson, 1995). It will be
interesting to see if such insurgencies are the beginning of a
counter-counterrevolution.

Therapy as Education

Wolpe’s view of therapy as a learning process cast the

therapist in the role of teacher and the client as student. This
stood in marked contrast to other therapy approaches in
which, for example, the therapist helped to “uncover
conflicts” or “facilitate development,” and is the model which
today characterizes the broad field of behavior and cognitive
therapy.

An educational approach requires a curriculum. Wolpe

developed treatment regimens, to be employed for particular
types of disorders, in which the behavior of the therapist and
the client were clearly specified. Some of the regimens
developed by Wolpe are in use today, notably systematic
desensitization and assertion training. Relaxation training,
which Wolpe introduced as a component of his procedures,
has been successfully extended to the treatment of many
medical disorders (Poppen, 1998). The many variations,
extensions, and new procedures that have been developed by
others continue to follow the education model. The behavior
therapist today has a variety of curricula at his/her disposal, to
be deployed according to client needs, in contrast to
one-size-fits-all procedures such as “empathic listening” or
“interpreting resistance.”

85
An educational approach also requires assessment. Wolpe
proposed to measure the problematic behavior of the client in
response to the environment in which it occurred. As with
treatment procedures, some of the assessment methods he
helped develop are currently useful, such as the Life History
Questionnaire and the Fear Survey Schedule. He advocated
what he initially called a “stimulus-response analysis”
(Wolpe, 1969) and later termed “behavior analysis of case
dynamics” (Wolpe, 1990), that is consistent with what many
now term a “functional analysis.” The goal is to determine the
specific needs of the client which, in turn, determines the
therapy procedures to be employed.

The choice of treatment method and the selection of treatment

goals hinges on assessment. In addition, assessment
throughout the course of intervention allows the teacher/
therapist to fine-tune the curriculum and change procedures or
goals if necessary. Assessment also allows the effectiveness
of therapy to be determined, a necessary requirement for
accountability. These all are standard operating procedures
for behavior therapists that come to us by way of the
educational model that Wolpe was the first to employ and
publicize.

Desensitization Research

Wolpe’s cat experiment came at the end of an era in which

clinical questions were tackled with animal research (Poppen,
1970). However, his clearly specified technique of systematic
desensitization triggered an avalanche of research on human
subjects with common fears that seemed more relevant to
clinical issues. Peter Lang and A. D. Lazovik (1960) were the
first to recognize that snake-avoidant college students

86
comprised a huge pool of potential subjects with whom
methods of anxiety reduction could be investigated. Jerry
Davison’s (1968) dissertation, employing this population, was
a paragon of experimental rigor that demonstrated the
effectiveness of desensitization while controlling for
extraneous factors. Gordon Paul’s (1966) dissertation,
comparing desensitization with insight therapy in college
students with public speaking anxiety, was an immediate
classic. Hundreds if not thousands of studies followed in the
next decade, investigating the variables and parameters that
comprised desensitization and, to a lesser extent,
assertiveness training, along with alternative methods and
explanations. Other procedures, such as flooding and
modeling, were investigated in a similar fashion. The net
result was an immense bulwark of empirically validated
procedures that formed the identifying characteristic of
behavior therapy and distinguished it from traditional
psychotherapies.

A question about analogue research, either with cats or

college students, was its relevance for actual clinical
problems. Wolpe himself questioned the utility of studies on
“weak fears.” The most important conclusion, however, was
that therapy could be, indeed must be, empirically studied. If
analogue studies omitted too many important factors, than
other studies including those factors must be done.

Clinical Trials

The most valid, though most difficult approach, was to study

real therapy with real patients. Wolpe first gained widespread
attention for his reports of unprecedented therapeutic
outcomes. Wolpe’s research method, hardly sophisticated but

87
revolutionary in the 1950’s, was to rate the outcome of all
patients he had seen who had met certain diagnostic and
treatment criteria. Wolpe (1952a, 1954, 1958) reported 90%
cured or much improved, in contrast with the two-thirds
base-rate recovery reported in Eysenck’s (1952) review.
Traditional psychotherapists were challenged to prove
Eysenck wrong, while behavior therapists took up the
challenge to prove Wolpe right. Analogue research promoted
more sophisticated designs than Wolpe’s retrospective case
summarization, and controlled prospective studies were soon
undertaken with clinical populations as well.

One of the first was Arnold Lazarus’ (1961) dissertation,

chaired by Wolpe. Lazarus compared matched patients
receiving psychodynamic interpretation or systematic
desensitization, both administered in a group format, on a
variety of self-report and direct observation measures. The
outcomes were overwhelmingly in favor of desensitization,
but possible experimenter bias was a confounding factor.

A few years later, Wolpe and Lazarus participated in a

benchmark study comparing behavior therapy,
psychodynamic therapy, and waiting-list controls (Sloane et
al., 1975). R. Bruce Sloane recruited Wolpe with this study in
mind, and planning began soon after his arrival at Temple in
1967. Not published until 1975, the results presaged a
continuing controversy in clinical outcome research. One
result, surprising to some but confirming Eysenck (1952), was
the degree of improvement in the wait-list controls. A more
controversial conclusion was that both behavior therapy and
psychodynamic therapy were equally effective. On the one
hand, this forced the traditional psychotherapy community to
admit behavior therapy into the clubhouse, but Wolpe and

88
others were not content to settle for “equality,” they wanted
clear superiority. Closer reading of the data indicated a
definite edge in favor of behavior therapy (e.g. Giles, 1983;
Poppen, 1976), but the “equal effectiveness” conclusion
prevailed.

The same controversy was played out a few years later when
nearly 500 therapy outcome studies were evaluated with the
statistical technique of meta-analysis (Smith, Glass, & Miller,
1980). These authors included the full spectrum of analogue
and clinical studies, identifying three major classes of therapy
(Behavioral, Verbal, and Developmental), made up of six
subclasses, in turn comprised of 18 individual types. The
largest contributor to their data pool was desensitization
research. Behavioral therapies, at all levels of groupings,
showed consistently greater effects than verbal therapies, and
both were much superior to developmental counseling. But
Smith et al. downplayed the effectiveness of behavior therapy
by attributing the differences to measurement bias,
recommending that the “benefit of the doubt should be
granted to theories that lack technologies readily applicable to
outcome evaluation” (p. 31). Traditional psychotherapists felt
that meta-analysis research upheld the equal effectiveness
notion, while Wolpe and others criticized the meta-analysis
procedure, or the studies included in them, or reanalyzed the
data to show the superiority of behavior therapy.

Much therapy research, and meta-analyses of that research,

has been conducted in the past 20 years, and slowly the
superiority of behavioral (and cognitive-behavioral)
procedures is being recognized (e.g., Lambert & Bergin,
1992). Although Wolpe did not conduct the research himself,
not having mastered the grantsmanship so necessary to carry

89
out such projects, his contributions have been immense. His
claims of behavior therapy superiority spurred the showdown
studies that led to grudging acceptance of its equality and
final recognition as the treatment of choice for many
disorders.

Professional Organizations and Journals

A necessary step in any professional movement is for those

who form a loose-knit community of persons with similar
interests to come together in a more formal organization.
Wolpe, through his writing and travels, was certainly
responsible for inspiring interest among a large number of
people dispersed around the world. He was also instrumental
in bringing them together into a professional group.

One of the first steps in this direction was the 1962

conference at the University of Virginia in Charlottesville,
that Wolpe chaired and organized with the help of Leo Reyna
and Ian Stevenson (Wolpe, Salter, & Reyna, 1964). The
conference was called “The Conditioning Therapies: The
Challenge to Psychotherapy,” reflecting Wolpe’s competitive
stance toward traditional therapy. The conference reflected
the past, present, and future of this new field. The past was
represented by Howard Liddell and W. Horsely Gantt on
experimental neurosis, and Andrew Salter on “conditioned
reflex therapy.” Cutting edge research was represented by
Peter Lang’s presentation of his systematic investigation of
systematic desensitization with snake-phobic college students.
The future was represented by Stevenson and Reyna, who
discussed the need for research on psychotherapy and the
issues that would arise in such research. This historic
conference provides a valuable reference point for the current

90
one, and the student would be well advised to compare the
issues then and now; what has changed and what issues
remain?

Another early step was Wolpe’s organization of the “June

Institute,” a one-month summer training program and seminar
which drew people from around the world to participate in
intensive training and discussion. These began in Virginia in
1965 and continued for about 15 years after Wolpe moved to
Temple. The first group included Dorothy Susskind, Alan
Goldstein, and Joseph Cautela, all of whom were to make
notable contributions in the coming years.

Dorothy Susskind, a clinician in New York, hosted continuing

meetings among professionals in the New York area
interested in behavior therapy to discuss professional and
practice issues. Participants in these meetings included
Andrew Salter, Cyril Franks, Joe Cautela, Jerry Davison,
Leonard Krasner, Arnold Lazarus and Wolpe. Franks edited
the group’s newsletter and spearheaded the formalization of
the group, holding a convention in 1967 concurrently with the
American Psychological Association as the “Association for
Advancement of Behavioral Therapies” (later unified as
“Behavior Therapy, “ or AABT). Franks served as the first
president and Wolpe became the second, continuing to be a
featured speaker at AABT conventions throughout his life.
Wolpe tried to do something similar within the American
Psychiatric Association but was rather less successful. In
1970, the AABT Newsletter became the journal, Behavior
Therapy, and in the same year Wolpe and Reyna cofounded
and edited the journal Behavior Therapy and Experimental
Psychiatry.

91
Professional Training

An essential feature of a profession is the establishment of

teaching and training facilities, typically within university
settings. Wolpe’s influence as an educator was, first, in
providing content through his books and articles, and second,
in providing hands-on training through apprenticeship
arrangements in workshops and internships at the June
Institutes and the Behavior Therapy Unit at Temple. Wolpe
was not a daily classroom lecturer but had a marked influence
on those of us who are. Wolpe’s teaching methods included
demonstrations of therapy, live, simulated, or filmed, and
participation with students as a co-therapist. One of his chief
educational contributions was to open the door of the therapy
room. While Carl Rogers had provided audio recordings of
therapy sessions, Wolpe showed that much more was
involved than “talk” therapy. Indeed, Wolpe put the
“behavior” in behavior therapy, and put the behavior of both
therapist and patient on display for educational purposes.

To summarize, behavior therapy was born in competition as

Wolpe sought to play David to Freud’s Goliath. His theories
and methods were major protagonists in the psychotherapy
wars that have raged in the literature for almost half a century.
Although Wolpe was not able to drive psychoanalysis from
the field, he was able to wrest grudging recognition that
behavior therapy was “equally effective.” But Wolpe rejected
the offer of equivalency and continued to press for victory. A
major point for his side was the increased acceptance of
outcome research, as behavior therapists forged ahead on the
empirical front. The revolutionary notion that psychotherapy
should have measurable outcomes carried the day.
Competition continued within the behavior therapy camp, as

92
Wolpe’s theories and methods were subjected to rigorous
investigation. Variations, additions, and alternatives arose
from cognitive and operant learning theories. These
challenged the benchmarks established by Wolpe and greatly
expanded the scope of behavior therapy. Progress is achieved
through competition, and we have all benefited from Wolpe’s
willingness to take controversial positions.

Significant Publications
Wolpe’s publishing career did not begin auspiciously. He at
first was determined to be a Hullian learning theorist.
However his efforts in this direction were not well received.
Only when he published the results of his clinical work did he
begin to be noticed. Thus the reaction Wolpe received to his
initial publications very much influenced the development of
behavior therapy.

Wolpe had dreams of taking Hullian learning theory into the

realm of abnormal psychology, and of earning royalties —
and perhaps an academic position — based on the hard work
he had done on his dissertation. With encouragement from
Reyna, Wolpe compiled his research into a book-length
manuscript entitled Conditioning and Neurosis. This
encompassed a complete review and critique of research and
theories on experimental neurosis, culminating in his own
experiment and his extension of Hullian theory to account for
the origins and treatment of neurotic behavior. Ambitiously,
he sent this to Clark Hull himself for review and, hopefully,
recommendation for publication. Hull, seriously ill and trying
to complete his own books, passed the manuscript along to
Kenneth Spence, who concluded that there was little market

93
for such a work. Interestingly, the following year saw the
publication of Personality and Psychotherapy by John
Dollard and Neil Miller (1950), and Learning Theory and
Personality Dynamics by O. H. Mowrer (1950). As two of the
most prominent students of Hull, Miller and Mowrer had
conducted basic research on anxiety conditioning in the
1940’s and in these works extended Hullian theory to account
for human neuroses. Both also were influenced greatly by
Freudian theory and offered nothing other than
psychoanalysis, couched in learning theory terminology, as
treatment. It should be remembered that, at this time, Wolpe
himself had not developed any new clinical procedures based
on his learning principles. Any therapeutic applications were
purely speculative and it is perhaps fortunate that his
manuscript was not published before he had done the hard
work of translating principles into practice.

Somewhat daunted, Wolpe persisted with a series of seven

theoretical articles, published over a four-year period (1949 to
1952) in the Psychological Review, in which he presented a
“neurophysiological view” of various learning phenomena,
such as reinforcement via drive-reduction, stimulus
generalization, and latent learning. Hull had avoided direct
reference to neural structures, preferring to describe
hypothetical processes that mediated between environmental
stimuli and overt behavior. Nevertheless, it was clear that the
operation of the nervous system was to be inferred from these
processes, as Skinner (1950) had noted in his critique of the
“conceptual nervous system.” Wolpe did exactly what
Skinner had criticized, filling the gap between stimulus and
response with neural sequences having various arrangements
of excitatory and inhibitory synaptic connections. These were
not based on actual physiological evidence but rather were

94
exercises in logic. The recent surge of interest in “neural
network learning theory” (Tryon, 1993) might find it
profitable to revisit these neglected papers of Wolpe.

With no prospects as an academician, Wolpe was forced to

earn a living as a clinician, a difficult task for a brand new
psychiatrist who was shut out of the local referral network
and who, besides, was struggling to turn theoretical concepts
into procedures that the few patients who came his way would
pay for. By 1952 he had developed the main features of
systematic desensitization, assertiveness training, and sexual
therapy, and began to experience success as a clinician. In
that year he published an account of his cat experiment in a
British journal (Wolpe, 1952b) and a criticism of
psychoanalysis and early report of his alternative methods in
a South African journal (Wolpe, 1952a). The latter served
only to further alienate him from the local psychiatric
professional community. In 1954 he published another
outcome survey of a series of patients, reporting 96% cured or
much improved (Wolpe, 1954). This publication attracted
widespread attention. He received over a hundred reprint
requests for this paper, most from the United States. This was
in immense contrast to his theoretical papers, telling Wolpe
that there was a large potential market for this line of
endeavor.

In addition to the attention they generated, these reports were

a major factor supporting his application to the Center for
Advanced Study in the Behavioral Sciences, in Stanford,
California. During his year at the Center, he finished the
manuscript for his first book, Psychotherapy by Reciprocal
Inhibition, published by Stanford University Press the
following year (Wolpe, 1958). This book pulled together his

95
work to date, including his neural theory of conditioning (a
reprise of his “neuropsychological view” papers), a critical
review of the experimental neurosis literature, and a
recounting of his cat experiment. He added a theoretical
account of the acquisition of various types of human
neuroses, including pervasive anxiety, hysteria, and
obsessional behavior, providing illustrations from his patients.
The last half of the book gave specific details of his therapy
procedures. He described the assessment interview and
advocated the use of the Willoughby Personality Inventory,
both as a means to discover information and as an assessment
device to measure therapeutic change. He described his
therapeutic procedures in detail, presenting individual cases to
illustrate their use. Finally he supplemented his earlier
outcome reports with an additional 88 cases, reporting an
overall success rate of 90%.

Wolpe’s detailed specification of procedures in 1958 had a

tremendous impact on clinical researchers and educators. In
1966 he published a book with Arnold Lazarus targeted for
clinicians, Behavior Therapy Techniques, which, as the name
indicates, focused almost entirely on therapeutic procedures.
In 1969, the first of four editions of his textbook, The
Practice of Behavior Therapy appeared (1969, 1973, 1982,
1990). One can look at the proportion of pages in each of
these volumes that are devoted to particular topics as
reflecting the selective effect of the behavior therapy audience
over time. For example, almost half of his 1958 book is
devoted to theoretical exposition. This is reduced to about 6%
by 1973, but takes a jump to about 13% in 1982 and 1990 as
he took on the theoretical issues in cognitive therapy. Wolpe
did not mention cognitive issues until the second edition of
his textbook in 1973. However, his earlier works included

96
brief descriptions of procedures that he later identified as
instances of “cognitive” procedures — if one chose to use that
terminology — namely “clarifying misconceptions” and
“thought-stopping.” These took up about 1% of his first three
books. His two final editions included chapters on “cognitive
therapeutic techniques” and presented critiques of the
“cognitive revolution” in behavior therapy, taking up about
17% of his last book. Finally, Wolpe first recognized operant
conditioning in 1969, where he included a brief chapter in an
effort to be inclusive of the entire field of behavior therapy.
However the topic comprised only 2-3% of this and
subsequent editions. In the final analysis, these books reflect
the increasing battles Wolpe had with cognitive approaches,
his rather cursory recognition of behavior analysis, and his
continued faith in the efficacy of his own approach to
behavior therapy.

Object Lessons for the Future

Wolpe was a pioneer in proclaiming, both in words and by
example, first, that therapy should be measurably effective,
second, that procedures should be clearly specified, and third,
that learning theory should be a useful guide for research and
practice. These are matters of continuing concern that will
shape future developments. I will briefly consider these three
directives from the perspective of Wolpe’s contributions.

Effectiveness of Therapy

Great strides have occurred in assessment of therapy

outcomes. What Wolpe began as therapist ratings of
improvement based on retrospective review of his own or

97
others’ case reports, has evolved into multimethod-multitrait
assessment batteries involving direct observation,
physiological measurement, self-report, significant-other
report, and medical records, in longitudinal, long-term
follow-up, blinded clinical trials. The sophistication and
expense of outcome research has grown and will continue to
do so. However, in his own work and his editorship of
Journal of Behavior Therapy and Experimental Psychiatry,
Wolpe supported the innovative case report and small-n
study. This approach continues to provide the seeds for new
and more effective procedures, and both large and small-scale
studies are necessary for continued development.

A major competitor, or possible ally, in demonstrating

effective therapy outcomes is the drug industry (see Hayes &
Heiby, 1996). Clinical trials of drugs are driven by companies
that earn huge profits from selling pills that soothe dysphoria,
relieve migraine, curb appetite, calm hyperactivity, or
produce sexual arousal. There is a huge demand for drugs that
relieve distress and improve behavior, as the public has come
to believe that “better living through chemistry” is an
inalienable right. There are no comparable behavioral health
companies that profit from promoting behavioral regimens
targeting the same ends, though there are some
approximations to this approach. As a physician, Wolpe felt
that drugs could be useful adjuncts to behavior therapy in
individual cases. That is, he did not prescribe drugs as the
sole means of treating a condition and he lamented the
wholesale shift of psychiatry from the psychoanalyst’s couch
to the prescription pad. Some clinical trials have examined the
combination of drugs and psychotherapy, typically including
a drug alone, therapy alone, and combined drug plus therapy
groups. Occasionally a drug company recognizes the

98
necessity of concurrent behavioral intervention, such as the
combination of behavioral procedures and the nicotine patch
for treating smoking. However, these approaches do not
address the factors that may make drugs useful in one case
and not in another. Nor do they solve the problem of how to
make a profit delivering a behavioral program whether or not
drugs are involved. “Behavioral health” programs are a
definite trend, and their control over treatment is an important
issue for the future.

The rise of managed health care, including mental health

services, and the focus on cutting costs, has raised many
issues of accountability. Not only should treatment be
effective, it should be acceptable to the consumer, safe, and
efficient (Giles, 1993; Pekarik, 1993). Wolpe observed that
many of his patients had undergone years of psychoanalysis
with little improvement; and others have reported harmful
effects of this approach (e.g., Masson, 1988). Wolpe focused
directly on patient complaints, provided a concrete
explanation of their problems, and presented a definite plan of
action, all of which build patient confidence. He reported
number of sessions as an indicator of the efficiency of his
approach (although much greater efficiency is required by
today’s standards). Ironically, traditional approaches have
generated “brief psychotherapy” to meet the demands of the
payers, the goals and procedures of which appear congruent
with behavior therapy: a focus on symptoms and current
functioning, directive therapist activity, homework
assignments, and use of various procedures to meet client
needs (Koss & Butcher, 1986). Convergence of behavior
therapy and brief psychotherapy is likely in the future.

Specification of Therapy Procedures

99
Wolpe opened the door of the consulting room and clearly
described desensitization, assertiveness training, sexual
therapy, and other procedures. Another important step
involved the search for “common factors” across various
therapy approaches. Jerome Frank (1961) is known for this
line of investigation, but early in his career Wolpe (1958, p.
193) recognized that “the various special points of procedure
that the different therapists regarded as so vital to success
were not vital at all, and that the effective factor must have
been something that all the therapeutic situations generated in
common.” Wolpe, and many since, have speculated on what
those common features are and the mechanism of their action
(Arkowitz, 1992). This issue is also likely to continue into the
future.

A major tool in analogue desensitization studies was the

treatment manual that specified the exact procedures to be
followed, the sequence of steps, and even the number and
duration of sessions. Manuals were developed to make sure
that all persons receiving a particular intervention were
treated the same, just as patients in a drug study receive the
same course of medication. Manuals allowed patients to be
treated with placebos or with particular amounts or
combinations of the treatment variables. In short, the
treatment manual insured the integrity of the independent
variable. Although manuals were developed to test theoretical
propositions, the therapist behaviors existed independent of
the theory and, as Gordon Paul (1966) showed, could even be
carried out by therapists who did not believe in the theory.
Treatment manuals were extended to clinical trials with real
patients, allowing equivalent treatments to be carried out at
multiple sites by many different therapists. Finally, treatments
that are empirically shown to be the most effective can be

100
designated as “best practice” to be employed for particular
disorders. Managed care companies are interested in best
practices and treatment guidelines, and it is likely this trend
will markedly increase in the future.

Wolpe (1986) was critical of the use of treatment manuals in

the practice of therapy, in which the choice of method is
based on the patient’s assignment to a DSM diagnostic
category. Wolpe felt that an individualized case analysis was
necessary to determine the factors important for each patient’s
behavior, and that these went far beyond DSM classification.
He felt that treatment should be selected to meet the unique
needs of each person, and that treating a person as a member
of a diagnostic group may well result in ineffective therapy. A
counter-argument states that it is much better to employ an
intervention regimen that is backed by data than by the
clinician’s seat-of-the-pants intuition as to what is best for a
given client. Therapists vary widely in what they might think
is appropriate and proceed with their favorite procedure
regardless of supporting evidence. Perhaps the treatment
manual would not be quite as good as what a highly skilled
therapist could come up with, but it would be considerably
better than the wild guesses of the not-so-good therapists.
This may be seen as another version of the old clinical vs.
statistical judgement issue (Meehl, 1954), this time applied to
treatment decisions rather than assessment. When it comes to
assessment, the evidence favors the statistical “cookbook”
approach, and the same may be true here. But it is an
empirical question. Can a therapist do better — achieve
greater improvement or do it more quickly — by adjusting
therapy to what is perceived as the individual needs of the
patient as opposed to doing it “by the book?” This would
appear to be an important topic for future research. While

101
outcome data on current methods are good, there is still much
room for improvement and it would be premature to close the
book — or the manual — at this point.

The Role of Theory

Shorn of their theoretical trappings, all effective behavioral

and cognitive-behavioral therapies follow the general
procedural framework that Wolpe figured out 50 years ago.
This framework consists of four steps. First, identify elements
of the critical situations — those in which the undesirable or
maladaptive behavior is likely to occur — and present them to
the patient in a controlled fashion. Second, weaken the
undesirable responses. Third, strengthen incompatible
adaptive behavior. Fourth, continue steps one through three in
a progressive manner until the desirable behavior occurs
readily in everyday, uncontrolled environments. This strategy
is followed in desensitization for phobic problems and
assertiveness training for social anxieties, in exposure and
response prevention procedures for obsessive-compulsive
disorders, in cognitive therapy for depression and controlled
drinking for alcoholics, in toilet training for the incontinent
and language training for the autistic, in stress-management
for headache sufferers and pain-management for chronic pain.
Theories have been useful in suggesting methods to employ at
each of these steps. For example, reciprocal inhibition theory
has given us relaxation and graded exposure procedures;
social learning theory has given us modeling and cognitive
theory has given us self-instructions; neurochemical theory
has given us anxiolytics and anti-depressives and behavior
analysis has given us prompting and reinforcement
procedures. Theories will continue to be useful to the extent
that they suggest even more effective methods.

102
No theory has yet provided an overarching account of all
methods at all steps that is acceptable to all of us. This
conference may be helpful in that regard. It shows us that
behavior therapy has progressed along the scientific path that
Wolpe was one of the first to point out. Science is a
cumulative and corrective process, an evolution born of
competitive behavior and professional disputation. Perhaps
there is a young Wolpe in the audience who is saying, “I can
do better than that,” a young person who sees how things fit
together in ways that we currently do not, and who can take
behavior therapy to yet a higher level of effectiveness or
theoretical integration. That would be a fitting legacy to this
great champion of behavior therapy.

References
Arkowitz, H. (1992). Integrative theories of therapy. In D. K.
Freedman (Ed.), History of psychotherapy: A century of
change (pp. 261-303). Washington, DC, American
Psychological Association.

Davison, G. C. (1968). Systematic desensitization as a

counter-conditioning process. Journal of Abnormal
Psychology, 73, 91-99.

Dollard, J., & Miller, N. E. (1950). Personality and

psychotherapy: An analysis in terms of learning, thinking,
and culture. New York: McGraw-Hill.

Eysenck, H. J. (1952). The effects of psychotherapy: An

evaluation. Journal of Consulting Psychology, 16, 319-324.

103
Eysenck, H. J. (Ed.). (1960). Behaviour therapy and the
neuroses. London: Pergamon Press.

Frank, J. D. (1961). Persuasion and healing. New York:

Shocken Books.

Giles, T. R. (1983). Probable superiority of behavioral

interventions I: Traditional comparative outcome. Journal of
Behavior Therapy and Experimental Psychiatry, 26, 241-248.

Giles, T. R. (1993). Consumer advocacy and effective

psychotherapy: The managed care alternative. In T. R. Giles
(Ed.), Handbook of effective psychotherapy (pp. 481-488).
New York: Plenum.

Hayes, S. C., & Heiby, E. (1996). Psychology’s drug

problem: Do we need a fix or should we just say no?
American Psychologist, 51, 198-206.

Hayes, S. C., & Wilson, K. G. (1995). The role of cognition

in complex human behavior: A contextualistic perspective.
Journal of Behavior Therapy and Experimental Psychiatry,
14, 29-32.

Hull, C. (1943). Principles of behavior. New York:

Appleton-Century-Crofts.

Jacobson, E. (1938). Progressive relaxation. Chicago:

University of Chicago Press.

Koss, M. P., & Butcher, J. N. (1986). Research on brief

psychotherapy. In S. L. Garfield & A. E. Bergin (Eds.),

104
Handbook of psychotherapy and behavior change (3rd ed.,
pp. 627-670). New York: Wiley.

Lambert, M. J., & Bergin, A. E. (1992). Achievements and

limitations of psychotherapy research. In D. K. Freedman
(Ed.), History of psychotherapy: A century of change (pp.
360-390). Washington, DC: American Psychological
Association.

Lazarus, A. A. (1961). Group psychotherapy of phobic

disorder by systematic desensitization. Journal of Abnormal
and Social Psychology, 63, 505-510.

Lazovik, A. D., & Lang, P. J. (1960). A laboratory

demonstration of systematic desensitization psychotherapy.
Journal of Psychological Studies, 11, 238-247.

Masserman, J. (1943). Behavior and neurosis. Chicago:

University of Chicago Press.

Masson, J. (1988). Against therapy: Emotional tyranny and

the myth of psychological healing. New York: Athenum.

Masters, W. H., & Johnson, V. E. (1970). Human sexual

inadequacy. Boston: Little, Brown & Co.

Meehl, P. (1954). Clinical versus statistical prediction.

Minneapolis, MN: University of Minnesota Press.

Mowrer, O. H. (1950). Learning theory and personality

dynamics. New York: Ronald Press.

105
Paul, G. (1966). Insight versus desensitization in
psychotherapy. Stanford, CA: Stanford University Press.

Pavlov, I. P. (1927). Conditioned reflexes (G. V. Anrep,

Trans.). London: Oxford University Press.

Pekarik, G. (1993). Beyond effectiveness: Uses of

consumer-oriented criteria in defining treatment success. In T.
R. Giles (Ed.), Handbook of effective psychotherapy (pp.
409-436). New York: Plenum.

Poppen, R. (1970). Counterconditioning of conditioned

suppression in rats. Psychological Reports, 27, 659-671.

Poppen, R. (1976). Review of R. B. Sloane, F. R. Staples, A.

H. Cristol, N. J. Yorkston, & K. Whipple, Psychotherapy
versus behavior therapy. Journal of Behavior Therapy and
Experimental Psychiatry, 7, 101.

Poppen, R. (1995). Joseph Wolpe. Thousand Oaks, CA: Sage

Publications, Inc.

Poppen, R. (1998). Behavioral relaxation training and

assessment (2nd ed.). Thousand Oaks, CA: Sage Publications,
Inc.

Reyna, L. (1946). Experimental extinction as a function of the

interval between extinction trials. Unpublished doctoral
dissertation. University of Iowa.

Ryle, G. (1949). The concept of mind. London: Hutchinson.

106
Salter, A. (1949). Conditioned reflex therapy. New York:
Creative Age Press.

Skinner, B. F. (1950). Are theories of learning necessary?

Psychological Review, 57, 193-216.

Skinner, B. F. (1953). Science and human behavior. New

York: Macmillan.

Sloane, R. B., Staples, F. R., Cristol, A. H., Yorkston, N. J., &

Whipple, K. (1975). Psychotherapy versus behavior therapy.
Cambridge, MA: Harvard University Press.

Smith, M. L., Glass, G. V., & Miller, T. I. (1980) The benefits

of psychotherapy. Baltimore, MD: Johns Hopkins University
Press.

Tryon, W. W. (1993). Neural networks I: Theoretical

unification through connectionism. Clinical Psychology
Review, 13, 341-352.

Watson, J. B., & Rayner, R. (1920). Conditioned emotional

reactions. Journal of Experimental Psychology, 3, 1-4.

Wolpe, J. (1952a). Objective psychotherapy of the neuroses.

South African Medical Journal, 26, 825-829.

Wolpe, J. (1952b). Experimental neuroses as learned

behavior. British Journal of Psychology, 43, 243-268.

Wolpe, J. (1954). Reciprocal inhibition as the main basis for

psychotherapeutic effects. Archives of Neurologic Psychiatry,
72, 205-226.

107
Wolpe, J. (1958). Psychotherapy by reciprocal inhibition.
Stanford, CA: Stanford University Press.

Wolpe, J. (1969). The practice of behavior therapy. New

York: Pergamon Press.

Wolpe, J. (1973). The practice of behavior therapy (2nd ed.).

New York: Pergamon Press.

Wolpe, J. (1978). Cognition and causation in human behavior

and its therapy. American Psychologist, 33, 231-236.

Wolpe, J. (1982). The practice of behavior therapy (3rd ed.).

New York: Pergamon Press.

Wolpe, J. (1986). Individualization: The categorical

imperative of behavior therapy practice. Journal of Behavior
Therapy and Experimental Psychiatry, 17, 145-153.

Wolpe, J. (1990). The practice of behavior therapy (4th ed.).

New York: Pergamon Press.

Wolpe, J., & Lazarus, A. A. (1966). Behavior therapy

techniques: A guide to the treatment of the neuroses. New
York: Pergamon Press.

Wolpe, J., & Rachman, S. (1960). Psychoanalytic evidence: A

critique based on Freud’s case of little Hans. Journal of
Nervous and Mental Disease, 131, 135-148.

Wolpe, J., Salter, A., & Reyna, L. J. (Eds.). (1964). The

conditioning therapies: The challange in psychotherapy. New
York: Holt, Rinehart, & Winston.

108
109
Chapter 3

B. F. Skinner’s Contribution
to Therapeutic Change: An
Agency-less, Contingency
Analysis
Julie S. Vargas

West Virginia University

No one in the 20th century has had more impact on methods

of changing behavior than B. F. Skinner. Whether attacked or
revered, his work continues to affect practices in education,
business, cultural design, and behavior therapy. The science
he developed is the most comprehensive, the most researched,
the most validated analysis of human behavior to date. Yet, in
spite of widespread discussion and publicity, Skinner’s work
is frequently misunderstood.

B. F. Skinner is often erroneously classified as a mechanistic

stimulus-response psychologist. True, Skinner began in the
behaviorist S-R tradition. But like Darwin, who through
detailed observation and recording, found himself unable to
support the biblical versions of creation he initially held,
Skinner, in working with the behavior of rats, found he could
not fit his observations into the classical S-R conception of
behavior. Control over the behavior of his rats lay in the

110
consequences of individual actions. Changes in the rate of bar
pressing occurred through a process of selection by
consequences, not through the pairing of antecedent stimuli.
His functional analysis showed how behavior is related to,
and thus can be controlled by, factors in an individual’s
environment, without appealing to psychological processes in
the “mind” or physiological processes in the brain. Practices
incorporating Skinner’s discoveries rippled throughout
society, impacting particularly education and behavior
therapy. How did a small-town Pennsylvania lad come to
make such an impact on the twentieth century?

Intellectual Background
Skinner grew up in a typical family of four in a small railroad
town in Pennsylvania. Life was good. His father’s law
profession was growing and the family could afford the newly
appearing fruits of science and industrialization — electricity,
the radio, the telephone, silent movies, and when Skinner was
six years old, the automobile. Skinner attended the local
public school. Early on he began challenging conventional
thinking. In his autobiography, Skinner tells of an incident
that happened in eighth grade. His English class was studying
Shakespeare’s As You Like It. At home, the young Fred had
heard someone question the authorship of the plays and he
announced to his class that Shakespeare didn’t write the play
they were reading. His teacher, Miss Graves, told him he
didn’t know what he was talking about. As Skinner puts it,
“that afternoon I went down to the library and found a copy of
Sir Edwin Durning-Lawrence’s Bacon is Shakespeare which I
read in great excitement. The next day, to Miss Graves’s
dismay I knew only too well what I was talking about.” His

111
teacher permitted discussion and Skinner continued to gather
evidence for his case, reading “biographies of Bacon,
summaries of his philosophical position, and a good deal of
the Advancement of Learning, the Essays, and Novum
Organum” (Skinner, 1976, p. 129).

Though Skinner says he doubted that he got much out of

these readings at the time, it is difficult to imagine a boy
continuing to read so much philosophy just to bolster a
position in an English class. More likely, Bacon’s conception
of truth as emanating from the manipulation of objects rather
than from the word of authority must have appealed to a
youth who loved to tinker. Moreover, Bacon’s idealism and
belief in bettering the world through scientific inquiry fitted
well with Skinner’s family’s emphasis on leaving the world a
better place than one found it. Then, too, Skinner’s science
courses supported Baconian principles. High school physics
revolved around experiments: Showing that a feather falls as
fast as a penny in a vacuum, that air has weight, and rolling
balls down inclined planes to verify Gallileo’s equations. In
high school, too, Skinner took botany and chemistry. His
intellectual training had begun.

In college, courses in chemistry and biology extended

Skinner’s scientific training. In biology, Skinner encountered
Mach’s Occam’s razor approach to science. Explanation,
Mach asserted, consisted in empirical observation and
description. Mach was later to play an important role in
Skinner’s thinking. But in college, Skinner majored in
English and his main activities revolved around literature and
the college magazine, of which he was an editor. After
college, and an unhappy year and a half living with his
parents while trying unsuccessfully to write fiction, Skinner

112
took a job at a bookstore in Greenwich Village in New York,
living the bohemian life and subscribing to literary magazines
like the Dial. It was in the Dial that Skinner read a review by
Bertrand Russell in which Russell called Watson’s
Behaviorism “massively impressive.” Skinner bought
Watson’s book and also Russell’s Philosophy. The latter
book, Skinner says, “begins with a careful statement of
several epistemological issues raised by behaviorism
considerably more sophisticated than anything of Watson’s”
(Skinner, 1979, p. 10). Years later, Skinner told Russell that
his book Philosophy had converted him to behaviorism.
Russell exclaimed, “My God, I thought I demolished that
view.” But Skinner had stopped reading at chapter eight
where Russell talks of that nature of the physical world, and
thus missed the last third in which Russell criticizes
behaviorism. Towards the end of his Greenwich Village stay,
Skinner decided to go back to school to study psychology.
Following advice from an old Hamilton professor about
where to go, he applied to Harvard. His training was about to
begin in earnest.

Graduate Training

Skinner arrived at Harvard in the Fall of 1929 at the age of

24. It was ironic that he chose Harvard for his graduate work.
The chair of the Department of Philosophy and Psychology
was E. G. Boring, a disciple of Titchener. Following
Titchener, Boring defined psychology as the study of the
mind or consciousness, exactly the position that Watson had
attacked. Boring’s students studied topics like how people
judged the weights of little pill boxes or how well a person
could tell the difference between loudnesses made by
swinging a ball on a string from various heights and letting it

113
bang against a block of wood. Boring’s presence extended to
every psychology classroom in the form of piped compressed
air for producing low whistles for students to compare. This
was not what Skinner planned to study. But fortunately, when
Skinner arrived, Boring was away on Sabbatical finishing his
book, the History of Psychology. Skinner looked for courses
to take and found one in the biology department that looked
promising. To his delight, the text discussed Pavlov’s work!
The course was taught by Hudson Hoagland, a young
instructor brought in by William Crozier, chair of a newly
created Department of Physiology. Crozier had studied with
Jacques Loeb. (It was a small world. Loeb had corresponded
with Mach and had taught a course that Watson took at the
University of Chicago.) Like Loeb, who was said to “resent
the nervous system,” Crozier had no use for explanations that
appealed to inner processes. Explanation consisted of finding
functional relationships between the manipulations of the
experimenter and the resulting behavior of the organism —
the same platform Watson had espoused. As E. Vargas put it,
the match between Crozier and Skinner was “ a professional
marriage made in heaven: Crozier — caustic, hard-driving
and hard-drinking, impatient, contemptuous of what he called
organ physiology, an advocate of Loeb in biology and Mach
in philosophy; and Skinner — sarcastic, radical and
rebellious, impatient, contemptuous of compromisers, and
eager to put the investigation of behavior on an independent
scientific footing” (Vargas, 1995, p. 108). In his second
course in physiology Skinner was assigned research with a
more senior student on Crozier’s favorite subject, tropisms.
This resulted in Skinner first published research article
(jointly with T. Cunliffe Barnes). Crozier was an editor of the
Journal of General Psychology, and the Barnes and Skinner
article “The Progressive Increase in the Geotropic Response

114
of the Ant Aphaenogaster,” appeared in that journal (Barnes
& Skinner, 1930). Predictably, it references Crozier and Loeb.

After his first year of graduate study, Skinner continued to

sign up for research credits over in physiology. But Crozier’s
fascination with tropisms did not appeal to Skinner. It is a
tribute to Crozier that he supported Skinner’s endeavors and
did not insist that the young student conduct tropism research.
He encouraged Skinner to pursue his own inclinations. As
Skinner describes it,

In my research courses ... I worked entirely without

supervision. No one knew what I was doing until I handed in
some kind of flimsy report. Possibly the psychologists
thought I was being counseled by Crozier and Hoagland, and
they may have thought that someone in psychology was
keeping an eye on me, but the fact was that I was doing
exactly as I pleased. (Skinner, 1979, p. 35)

Freed from a dictated program of research, Skinner was

shaped by what he experienced. His day to day behavior was
controlled by how his rats acted rather than by theories or by
hypotheses to prove or disprove. Experimenting with
behaving organisms provided perfect contingencies for
discovery. Still, it took some time before he broke away from
the stimulus-response tradition that dominated not only
psychology in general, but also the behaviorism of the time.

Skinner was, of course, familiar with Watson’s Behaviorism.

In that book, Watson proclaimed behavior to be a subject
matter in its own right, but he attributed causes to antecedent
stimuli even when manipulating consequences. For example,
Watson described shaping “da” in an infant as saying “da”,

115
and then giving a bottle contingently when the baby uttered
“da.” He analyzed the process, however, as Pavlovian
conditioning to the “da” stimulus preceding responding
(Watson, 1924 p. 182). Watson’s procedure of presenting
stimuli in “trials” also made it difficult to see relationships
between actions and consequences independently of the role
of preceding stimuli.

Skinner started working in this reflex tradition, looking at the

responses of rats to prior stimuli. With his tinkering
proclivities, however, he kept designing new pieces of
equipment for experiment after experiment, and in so doing
stumbled onto procedures that did away with trials. The
apparatuses that Skinner finally perfected — the operant
chamber and the cumulative recorder — enabled him to
observe control by consequences when no particular stimulus
was present. There were no trials in an operant chamber, and
moment to moment changes in rate of bar pressing could thus
be related directly to the way in which consequences were
arranged. Skinner described the evolution of the apparatus,
and the experiments that he conducted in “A Case History in
Scientific Method” (Skinner, 1999).

Discovery of the Operant: Selection by

Consequences
The “Case History” article Skinner wrote affirms his
inductive approach to science. In the early 1950’s Skinner
was asked to contribute a chapter to a book as part of a
“Project A” directed by Sigmund Koch. Contributors were
asked to provide “any set of sentences formulated as a tool for

116
ordering empirical knowledge with respect to some
specifiable domain of events” (Skinner, 1999, p. 108). To
Skinner, this was the reverse of Bacon’s dictum that books
should follow science, science should not follow books. The
“method” of science is what scientists do, not a set of rules or
procedures. Skinner had already published an article in which
he argued against the methodology of statistics and
hypothesis testing, offering his own behavior to illustrate how
discoveries come about. He submitted a revised version of
that article for the book. The only rules provided, such as
“When you run onto something interesting, drop everything
else and study it,” or “Some people are lucky,” are serious
suggestions, though phrased whimsically as if in contempt for
rules of science.

In writing his case history, Skinner spent some time going

through old research notes to reconstruct the sequence of his
activities as accurately as possible. The sheer number of
pieces of apparatus and experiments he built is impressive in
itself. Following his tropism experiments with ants in the
spring semester of his first year as a graduate student, he was
given some rats by a student (Gregory Pincus — the inventor
of the birth control pill). Skinner built what he grandly called
the “Parthenon” because it had steps descending from a
tunnel in a tenuous resemblance to the Greek temple. Skinner
was familiar with maze research, but had been bothered by
the fact that the position of the rat and the noise of the
opening of the door to the maze were not controlled. He
developed a “silent release box”, watched the rat come down
the Parthenon steps, then sounded a click and recorded the
rat’s behavior. Successive clicks produced less and less of an
effect. When some of the rats had babies, Skinner started
studying the behavior of the baby rats. He tore up the

117
Parthenon and built a platform on piano wires to measure the
motion of baby rats when pulled increasingly by the tail.
Extending the work to adult rats and abandoning attachments
to the animals, Skinner build an eight-foot runway in a tunnel,
again recording the movements of the rats as clicks were
sounded. Note that he was still investigating reflexes in a trial
format — the responses of the rats to an antecedent click.

To insure that a rat would travel through his tunnel, Skinner

gave it a bit of wet mash to eat at the end of the runway.
Candidly explaining that he got tired of carrying the rats back
and forth, Skinner describes adding a back alley forming a
rectangular runway. But the rats would pause after eating the
mash. Skinner got interested in those pauses and “dropped
everything else”. But then the runway did not need to be eight
feet long. By January of his second year as a graduate student,
Skinner had constructed a shorter rectangular runway with a
tilting mechanism permitting the rat to make its own records
by moving a needle up and down a moving strip of blackened
paper as the rat traveled from one end of the runway to the
other. Each trial began when the rat ate the single pellet of
food that the tipping of the runway produced. Here, luck
entered. Skinner had used scraps to build equipment and had
not bothered to cut off a spindle when fashioning the food
magazine. One day, his behavior came under control of that
spindle, and he saw that by attaching a weighted string that
would unwind as the experiment proceeded, he could
transform his linear polygraph records into curves. Thus, in
early Spring of 1930, the cumulative recorder was born. At
about this time, too, Skinner shifted from recording running
to recording the rate at which the rat ate pellets of food. One
element critical for operant conditioning had fallen into place:
Rate as a dependent variable. In a trial format the

118
experimenter produces each stimulus to which the organism is
to respond, making rate meaningless. In Skinner’s apparatus,
the rat could produce a pellet at any time without waiting for
any antecedent stimulus. In a letter to his parents he explained
that he had demonstrated that “the rate in which a rat eats
food, over a period of two hours, is a square function of the
time. In other words, what heretofore was supposed to be
”free” behavior on the part of the rat is now shown to be just
as much subject to natural laws as, for example, the rate of his
pulse” (Skinner, 1979. p. 59). But though freed from the
restrictions of “trials”, in the publication of his experiments
Skinner still talked of “eating reflexes”. Time since the
beginning of the eating period served as the antecedent
stimulus. He had not yet broken free of the stimulus-response
tradition. Skinner reported on his work correlating rate of
eating as a function of time in two more articles. The first,
submitted in July 7, 1931, describes apparatus with a door
that the rat pushes open to get at pellets. In the article Skinner
grapples with the problem of lack of consistency between
stimuli and responses that you should get in a reflex: “If it is
in fact true that a rat’s approach to a bit of food is reflex, why
is the response not always evoked by the appropriate
stimulus?” (Skinner, 1932a, p. 32). He appeals to a “third
variable”, conditions such as deprivation of food imposed by
the experimenter to take the place of what other psychologists
would call “drive”, thus avoiding any need to appeal to
internal physiology. In the second article, submitted three
months later, he describes a “problem box” with a lever —
the first mention of an operant chamber in Skinner’s
published works. In this article, submitted three years after
arriving at Harvard, Skinner talks of “eating behavior” as a
chain of reflexes, still in an S-R format: “The stimuli for the

119
initial members of this sequence of reflexes emanate from the
food or the food tray.” (Skinner, 1932b).

The discovery of the control by consequences that shifted

Skinner’s dependent variable from ingestion to bar pressing
began with an accident. One day, the food magazine in the
operant box jammed. As Skinner describes it,

At first I treated this as a defect and hastened to remedy the

difficulty. But eventually, of course, I deliberately
disconnected the magazine. I can easily recall the excitement
of that first complete extinction curve. I had made contact
with Pavlov at last: Here was a curve uncorrupted by the
physiological process of ingestion. It was an orderly change
due to nothing more than a special contingency of
reinforcement. It was pure behavior. (Skinner, 1999, p. 117)

The “contact with Pavlov,” was, in fact, a complete break

with Pavlov’s type of conditioning. When Skinner shifted his
dependent variable to the rate of bar pressing he began
looking at postcedent controls. Here was a “reflex” that did
not seem to be explained by conditioning of the antecedent
“stimulus substitution” type. He must have written his friend
and colleague, Fred Keller about his “discovery” because on
October 2, 1931, Keller wrote him back, saying, “The only
thing that bothered me about your very welcome and newsy
letter was that talk about a brand new theory of learning.”
(Keller, 1931). By February of the following year Skinner
submitted an article making clear the distinction between
Type I conditioning (respondent) and Type II conditioning
(operant), (Skinner, 1932c). At the age of twenty seven, B. F.
Skinner had started onto a line of research that began a whole
new science of behavior.

120
The central relationship Skinner discovered was that of the
operant, a class of behaviors defined by a common effect on
the environment. The bar press in the operant chamber was
defined by the closing of a switch, not by the form of the
motions of the rat. Skinner realized that even if the rat falls
upon the lever, for example when slipping from climbing up
the corner of the box, the depression of the lever is recorded
as a bar press. Such data are just as legitimate as a press by a
paw, but until bar presses are brought under experimental
control of consequences, the actions that close the switch are
not considered part of the operant of bar pressing. Unlike
respondent conditioning, operant conditioning is controlled
not by the stimuli that precede actions, but by those that
follow. Operant behavior is a two-term relationship in which
no identifiable stimulus need precede a response.

Skinner’s dependent variable, probability of behaving,

measured as rate of actions, was recorded, in a sense, by the
rat itself. Every bar press was captured permitting a
fine-grained analysis of functional relationships. Skinner was
fortunate to obtain five years of grants that gave him complete
freedom to do research. In those five years, he investigated,
“deprivation and satiation, reinforcement and non
reinforcement, schedules of reinforcement, differential
reinforcement with respect to properties or stimuli and
responses, aversive consequences, and a few behavioral
drugs” (Skinner, 1938). The results were published in 1938,
along with a sophisticated analysis of his philosophy of
science, in The Behavior of Organisms (Skinner, 1938/1991).

In investigating how the probability of responding could be

altered by manipulating the way in which food was made
contingent upon responding, Skinner opened up a whole new

121
approach to how behavior originates. Reinforcing
consequences select behavior, making it more likely to occur
again under similar circumstances. The process of selection
not only increases a particular response, it shifts a whole
gradient of properties of responding, gradually producing
novel responses, much as new species are created by the
selection of individuals. Just as Darwin’s analysis provides
the mechanism through which the extraordinary variety of
species could arise, Skinner’s analysis gave a plausible
account of the origins of novel or unusual behavior in people
including many of the kinds of behaviors with which
therapists deal. Interestingly, Skinner didn’t draw the
selection analogy until very late in his career. It appears in the
title of a 1981 article in which Skinner talks of three kinds of
selection affecting what people do, natural selection, operant
conditioning, and cultural evolution (Skinner, 1981). Of
course antecedent stimuli come to gain control over operant
behavior, too, but only by being present when
action-consequence relationships occur.

Implications for Therapy — Direct

Manipulation of Consequences
Had Skinner stopped there, he would have contributed a
technology for psychotherapy. By arranging various
contingencies of reinforcement, Skinner was able to shape
behavior of extreme forms, to produce “negative utility,” like
pathological gambling, and to demonstrate the harmful effects
of punishment. Behavior changes according to contingencies
of reinforcement or punishment, providing plausible
explanations of how psychoses might have originated, as well

122
as how behaviors could be changed. Direct manipulation of
consequences of client behavior was all that was needed. To
directly shape behavior, however, the therapist had to interact
with patients as behavior occurred. The therapist who sees a
client only in office visits is not present to set up
consequences following specific problem behaviors of a
client. But in institutions, more continuous interactions were
possible. Thus it was in institutional settings that operant
conditioning as a therapeutic technique first occurred.

As early as 1932, soon after getting his degree, Skinner had

talked about setting up some lever-pressing experiments for
human subjects at Worcester State Hospital, but nothing came
of the talks. His concern with the human condition showed in
Walden Two, but while in Minnesota and Indiana
(1936-1949) he did not address therapy. Even though
extensive work with non-human animals demonstrated the
variety of behaviors that different contingencies of
reinforcement could produce, it wasn’t until a student of
Skinner’s, Ogden Lindsley, took over a project with
institutionalized adults, that the first direct applications of
operant techniques with human beings was conducted
(Lindsley, 1960). Lindsley’s human operant chambers
consisted of separate rooms, a bit larger than a standard office
cubicle. Patients had access to a manipulandum such as a
plunger to pull or a panel to press. Behind the row of
chambers, equipment controlled the schedules for the
dispensing of reinforcers, the latter consisting of everything
from candy to cigarettes to visual stimuli. Cumulative records
were obtained, showing the same kinds of sensitivities to
contingencies found in rats and pigeons. Lindsley had a
steady stream of visitors, many of whom set up their own
laboratories. Therapeutic applications included work on

123
problems as diverse as stuttering (Flanagan, Goldiamond, &
Azrin, 1958, p. 177), multiple tics (Barrett, 1962/66), and
wearing or glasses by a child with autism (Wolf, Mees, &
Risley, 1964/66). By 1964, Lindsley could report that “more
than 100 applications of free-operant methods to human
behavioral pathology have been published (Lindsley, 1966, p.
167).

Moving into environments in which patients lived, albeit still

in institutions, Ayllon and Michael (1959) trained psychiatric
nurses to systematically reinforce or ignore specific behaviors
in the patients with whom they interacted every day. (Ayllon
& Michael, 1959). For all of the clients served, standard
treatments involving counseling or prescription drugs had
failed. Only direct manipulation of immediate consequences
of client behavior worked as a therapeutic technique. The
procedures involved consequences of ongoing behaviors, that
is, the continuous flow of a client’s actions, rather than
behavior as a response to tasks presented in trials. The
procedures thus followed directly from Skinner’s discovery of
the power of consequences over the frequency of behavior.
By 1979, operant conditioning was the technique cited most
by the members of the Association for Behavior Therapy who
responded to a questionnaire about procedures they
“frequently used” (Cautela, 1986, p. 5).

When patients are institutionalized, therapists have access to

behavior as it occurs, and can alter the consequences of that
behavior daily in experimental sessions or on the ward. In the
more common therapeutic format of office visits once a week,
however, the therapist may not experience the behaviors of
concern firsthand, and he or she has little control over the
contingencies that occur outside office interactions. The

124
therapist must rely, therefore, not on direct observation and
shaping of behavior, but on patient report. Thus verbal
behavior enters strongly into the standard therapeutic
relationship.

Skinner’s Analysis of Verbal Behavior,

Covert Behavior, and Feelings
Skinner started working on verbal behavior very early. By the
time his first book, Behavior of Organisms, came out he was
already writing a book on “language.” It took him over 20
years to complete what he often called “his most important
book.” Working on this book, Skinner grappled with topics
that had not yet been satisfactorily analyzed, including
internal feelings and awareness, as well as the reasons people
say, gesture, and write what they do.

The Basic Framework of Verbal Behavior

Verbal behavior is particularly complicated since it has its

effect on the world through the mediation other individuals.
Where reaching for the salt contacts salt directly, asking for
the salt produces salt only through the actions of a second
person. In analyzing verbal behavior Skinner asked “Of what
variables is a particular response a function?” Extending
principles from the laboratory, he sought relationships
between what is said and controls in the environment.
Consistent with his overall approach, he excluded internal
structures or agencies as causes for verbal behavior. That left
three basic sources of control: deprivation or aversive
stimulation (mand); objects or events in the verbalizer’s
presence (tact); and verbal stimuli (intraverbal)1. Interacting

125
with all three categories, a mediator serves both to reinforce
specific verbal responses and to provide a discriminative
stimulus for general aspects of verbal behavior, such as
starting to speak or the language spoken. Complex
relationships such as those involved in “grammatical
construction” were also addressed in the book.

Categorizing verbal behavior by its function, rather than by its

form, Skinner addressed “meaning”. Meaning does not reside
in the particular words uttered, but rather in contingencies —
the relationship between the particular words emitted, the
consequences that have followed similar behavior in the past,
and the context in which a statement occurs. “Fire!” means
one thing when said by a cold camper in the presence of
someone who can make a fire (mand), another when reacting
to the presence of flames (tact), and a third when translating
the French “feu” (intraverbal). “Milk” said when under the
control of “wanting” milk is a different operant (mand) than
saying “milk” as a result of “seeing” milk (tact). Tacting is
widely misinterpreted as reference, but an example by
Skinner illustrates the difference (Skinner, 1986). Saying the
word “fishing” instead of, say, “hunting” in the statement,
“I’m fishing for a letter I want to show you,” is partly a tact if
it occurs because of a large swordfish mounted on an office
wall. But clearly the speaker is not “referring to the fish.”
Conversely, one can refer to Roosevelt without Roosevelt
being present. The latter is reference but not a tact, since
Roosevelt is not currently present.

126
Influence of Skinner’s Analysis of
Verbal Behavior on Therapy
The distinction between mand and tact has led to
technological improvements in the field of special education.
Traditionally, children with autism were taught first to name
objects. When they learned to say “milk” when shown milk, it
was assumed that they “knew the word,” and could ask for
milk if they wanted it, even if milk was not currently visible
to them. Mark Sundberg and his colleagues showed that, just
as Skinner had predicted in Verbal Behavior, the two “milks”
were separate operants (Skinner, 1957). Teaching a child to
tact did not necessarily enable that child to mand. The same
form of response (saying “milk”) under one set of controls did
not always occur under a second set of controls. When
different operants are typically learned far apart in time, as
with speaking and reading, everyone recognizes the
difference in controls: No one would expect a child who can
say “milk” to automatically then be able to read that word.
Reading has to be taught separately. Similarly mands and
tacts may need to be taught separately to children with
autism. Since the mand is maintained by getting the specified
object or event, while tacts are maintained by generalized
reinforcement such as social approval, Sundberg and his
associates found that it was easier to teach mands than tacts.
Children learn much more quickly to “ask for what they
want,” than to “name objects”. Where it could take literally
months to teach a rudimentary tact vocabulary of four to five
terms, by reversing the order and teaching mands first,
children more rapidly learned a basic vocabulary. (See the

127
journal, The Analysis of Verbal Behavior, for additional
applications of Skinner’s analysis.)

The analysis of verbal behavior also provided insight into

some profound dysfunctional social behaviors. Not being able
to tell others what you want can be very frustrating. In
looking at the function of many of the bizarre behaviors of
children, that is, at what usually followed those behaviors, it
became clear that many of these behaviors were mands.
Throwing oneself on the floor and screaming may not look
like verbal behavior, but if it produces a characteristic
consequence such as “teacher attention,” it is likely to be a
mand. By specifically shaping more appropriate ways to bring
about the “characteristic” consequence, therapists have been
able to eliminate many extreme behaviors without resorting to
punishment, which produces its own problems (Bowman,
Fisher, Thompson and Piazza (1997). Similar analyses have
shown many maladaptive adult behaviors to be mands. Using
the “characteristic consequences” as reinforcers to shape more
acceptable mands, therapists have been able to eliminate
long-standing troublesome behaviors.

Skinner analyzed verbal behavior much as he analyzed the bar

press. He looked for factors in the environment responsible
for a particular action. Having researched basic principles of
reinforcement, shaping, discrimination and generalization, he
was sensitive to the nuances of contingencies over behavior.
To be sure, the mediated aspect of verbal behavior required a
special analysis (470 pages in fact), but the basic principles
did not change from one kind of behavior to another. Though
the analysis was not different, it became more complex when
addressing behaviors going on inside an individual.

128
Internal Feelings

In Verbal Behavior Skinner specifically addresses internal

thoughts and feelings. The skin, he maintained, is simply a
physical boundary and, as he put it elsewhere, “The skin is
not that important as a boundary” (Skinner, 1963, p. 953).
Stimuli occurring inside the body can be treated just like
those outside, except that they can only be reported by the
person observing them. The tooth ache you feel inside is just
as much a stimulus as a puncture applied externally.
Considering internal events as valid variables in a science (the
radical behavioral position) set Skinner apart from
methodological behaviorists who follow the logical positivist
tradition of intersubjective invariance, or truth by agreement.
Where methodological behaviorists insist, for example, on
interobserver reliability for reported data, radical behaviorists
look for truth in effectiveness at prediction and control. As
Skinner pointed out, Robinson Crusoe could develop a
science even if his man Friday never turned up.

But the lack of a common contact with internal events poses a

problem for teaching tacting. How does a member of a verbal
community shape up a reasonably accurate response to an
event that only the learner experiences? Skinner described
various ways. In “public accompaniment,” visible stimuli
usually correlated with an internal event help the teacher. A
mother, seeing a skinned knee, may say “that must hurt,” thus
pairing the term “hurt” with whatever the child is feeling.
Secondly, through “collateral responses” (a child may wince,
for example), the term “hurt” is again mentioned. Metaphor
may be used. In asking about a pain, doctors often suggest
terms taught in situations where stimuli are external: “Is it a
sharp pain, or dull? Does it throb?” People with lower back

129
pain often describe it as feeling “like a red hot poker”. Lastly
much behavior is learned at the overt level, and then reduces
to the covert. We learn to read at first aloud, but gradually the
pronunciation of words recedes to the covert level, with, in
some readers, a slight movement of the lips revealing the
origin of the behavior. (It is also interesting that when the
covert level is inadequate, as when reading very difficult
directions or in a noisy environment, the behavior is likely to
reappear at the overt level.) In a similar way, a person may
tact a physical reach for a cigarette in the pocket where she
used to keep cigarettes, then only start to reach, and finally
reach at such a small magnitude that only the ex-smoker can
identify each incident as an “impulse.” In researching the
origins of dozens of terms describing internal feelings, (such
as “anxiety”) Skinner found that they originated as terms
describing overt behaviors. “Anxiety,” for example originated
in the word for the overt action “choke.” Gradually the
word’s meaning shifted to describing events inside the skin.
Still, a verbal community taught its new members to describe
“feeling anxious.”

Awareness or Consciousness

Like verbal behavior, awareness or consciousness also

requires the mediation of others. We become conscious when
those around us arrange contingencies for paying attention to
something, whether that something is internal or external. All
of us have walked up a set of stairs. But you may not be
aware of how many steps you took. If an acquaintance often
asked about numbers of steps (particularly if reporting the
number accurately resulted in some strong consequence such
as winning a $100 bet), it wouldn’t be long before you could
accurately report the number of steps, and even the numbers

130
of other steps in stairs you encountered. The phrase
“conscious of the number of steps” would than be applied to
your behavior. The same process occurs when the
reinforcement is less conspicuous, such as social approval.
Thus the community of people with whom clients interact
determines awareness. One part of that community is the
therapist. In therapy sessions, the therapist can increase a
client’s awareness of aspects of his or her life by the kinds of
statements to which the therapist attends. The only drawback
to working primarily with verbal behavior is that what clients
say may not accurately reflect critical components of a their
life.

Suppose, for example, instead of a fixed $100 for reporting

numbers of steps, you would receive $10 for each step you
reported, and the person handing out the money was unable to
check on your accuracy. The contingencies in such a case
encourage exaggeration, as exemplified by stories of the size
of fish caught. The fact that verbal behavior can be shaped
independently of actual events confronts anyone who must
rely on the description of events he or she has not directly
observed. When a client relates feeling better after several
therapy sessions, is the client’s life better, or has only the
client’s verbal behavior been shaped? Shaping verbal
behavior alone is particularly dangerous with children who
are under strong contingencies of pleasing adults, as shown
by the graphic descriptions of fictitious sexual abuse. What a
therapist reinforces in the interaction during a session thus
affects what the client reports during a session as well as what
clients notice in their daily lives.

Therapeutic Extensions of Skinner’s Analysis of Verbal

Behavior

131
By extending the principles discovered in the laboratory to
verbal behavior and to an analysis of how covert behavior
comes about, Skinner opened the way for therapists to base
therapy on covert behaviors and covert reinforcement. This
made it possible to bring a client’s problems into the
therapist’s office, since scenarios, visualized actions, and
imagined consequences are not restricted to a particular place.
In the 1950’s Wolpe used visualization to desensitize patient
fears in a respondent conditioning format. If respondent
extinction through imaging of antecedent stimuli could help a
patient, why couldn’t imagery be used in an operant
paradigm? In 1966, Joe Cautela began using imagery with
operant behaviors. He described his “covert conditioning”
procedures as “the modification of a behavior by imagining
particular consequences to influence the behavior in the
desired directions” (Cautela, 1986, xii). The “particular
consequences” were reinforcement, extinction, and
punishment, taken directly from Skinner’s work. Covert
conditioning has been used successfully in treating a wide
variety of patient problems, including anxiety, athletic
performance, pain, depression, aggression, and sexual
disorders (Cautela & Kearny, 1993). Skinner’s assertion that
verbal and covert behaviors obey the same laws, and that they
interact with each other and with overt behaviors has been
born out in therapeutic applications.

Lessons for the Future

At the same time that Skinner provided therapists with tools
based upon a contingency analysis, Skinner addressed the
controls over the therapist. What determines what a therapist
does when approached by a client? Certainly training comes

132
into play. Training may guide the initial interview — whether
inventories are given or not, and which ones, or what
questions are asked. But as interactions continue, other factors
enter in. Current cultural practices affect therapeutic practice.
Sometime back, therapists would try to change the sexual
orientation of a homosexual client. Today that is not
necessarily the accepted approach. Standards of ethical
conduct also change according to the times. But ultimately,
like the experimenter shaping behavior in a rat, chimpanzee,
pigeon, or any other convenient platform, and like a teacher
shaping behavior of a baby, child, or adult, the therapist, in
order to be maximally effective, must come under control of
change in behavior of the client. While the precision obtained
by cumulative records may not be feasible for tracking daily
behavior, direct continuous measurement of the rate of
relevant behaviors is the goal to approximate.

Skinner consistently argued against internal agencies.

Certainly physiological processes occur as behavioral
selection is taking place, but the analysis of brain functioning
belongs to another discipline. Furthermore, understanding
what is happening in the brain will never provide a
parsimonious explanation of why a person sings a particular
song at a particular time, or even why a pigeon pecks a
particular disk at a particular rate. Behavior must be explained
at the behavioral level, and that requires looking at factors in
the environment. To change behavior, the therapist must look
at the relationships between clinically relevant behaviors and
the contingencies over those actions. Appealing to agencies,
such as personality traits, short-term and long-term memory
processes, or even measurable physiological activities, not
only fails to explain behavior of interest but draws attention
away from the environment where control actually lies. In his

133
last talk, Skinner warned against “cognitive” approaches,
drawing an analogy between them and “creationism.” Only
by abandoning internal causal agents will therapists turn full
attention to the contingency relationships responsible for the
behaviors they treat.

Summary
Therapy was not Skinner’s field. But, in one sense he was a
therapist. For he contributed an analysis of behavior which,
combined with the conditioning procedures worked out by
Pavlov, encompass all behavior. Skinner discovered the
critical role of postcedent events both in selecting actions that
become part of an individual’s repertoire and in determining
how antecedent events, through pairing with selective
processes, gain control over behavior. He spent thousands of
hours in research, changing every conceivable aspect of
contingencies and noting the effects on the probability of
behavior, measured as rate of actions. By working out the
complexities of how the rate of operant behavior is a function
of environmental variables, Skinner showed how maladaptive
behaviors could have originated, and how alternative
repertoires could be established to take their place.

Psychologists working with institutionalized populations first

grasped the significance of the role of postcedent events. By
directly altering the kinds of consequences following patient
behaviors they were able to solve long-standing behavior
problems. Not only were they able to eliminate behavioral
excesses, such as tantruming, they also improved the life of
patients who had done little but lie about for months or years.
They taught these patients to shower, to dress themselves, to

134
make their own beds, get themselves to meals on time, and
even to enjoy social events — all by changing contingencies.

Skinner extended his laboratory findings to an analysis of

verbal behavior. He categorized verbal operants according to
the functional controls over behaviors mediated by the actions
of other people. Following Skinner’s analysis, therapists have
looked at the function rather than the form of behaviors of
individuals who do not talk. Many extreme behaviors, such as
tantrums and even vomiting, have been found to be verbal. By
teaching alternative ways of “asking” for the consequences
that followed these extreme behaviors, therapists have,
without using punishment, eliminated the maladaptive
behaviors.

The analysis of verbal behavior includes covert actions.

Behavior inside the skin follows the same laws of
reinforcement, extinction, punishment, shaping, and
differentiation as behavior occurring overtly. Like overt
behaviors, covert behaviors are a function of contingencies.
They can be addressed like any other behaviors. Including
internal actions in the analysis leads to effective therapies for
changing dysfunctional thoughts and feelings.

While Skinner addressed all behavior, internal as well as

external, his analysis excluded any kind of internal “agency”
as responsible for action. Unlike thoughts and feelings which
can be recorded by the behaving person, agencies such as “the
mind,” or “self” must be inferred from other behaviors,
usually the ones to be explained and changed. Including them
in an analysis leads to a dead end. At best, attributing
behavior to an agency such as “mental illness” or “poor
self-concept,” raises questions about what caused the mental

135
illness or self concept. At worst, statements that sound like
explanations hinder further inquiry. By excluding agency
entirely, behavior can be functionally related to
environmental events — the same events that an agency
analysis must eventually consider. Until Skinner’s discovery
of the role of consequences in selecting behavior, and the
hundreds of experiments investigating the relationship of
contingencies to the probability of particular actions, the
functional relationship between behavior and environmental
events was not well understood. With the science of
contingent relationships between behavior and environmental
events, the therapist has a comprehensive, well-researched,
and effective approach upon which to base practice.

References
Ayllon, T., & Michael, J. (1959). The psychiatric nurse as a
behavioral engineer. Journal of the Experimental Analysis of
Behavior, 2, 323-334.

Barnes, T. C., & Skinner, B. F. (1930). The progressive

increase in the geotropic response of the ant Aphaenogaster.
Journal of General Psychology, 4, 102-112.

Barrett, B. H. (1962/1966). Reduction in rate of multiple tics

by free operant conditioning methods. In R. Ulrich, T.
Stachnik, & J. Mabry (Eds.), Control of human behavior (pp.
143-150). Glenview, IL: Scott, Foresman and Company.

Cautela, J. R., & Kearney, A. J. (1986). The covert

conditioning handbook. New York: Springer.

136
Flanagan, B., Goldiamond, I., & Azrin, N. (1958). Operant
stuttering: that control of stuttering behavior through
response-contingent consequences. Journal of the
Experimental Analysis of Behavior, 1, 173-177.

Keller, F. S. (1931). Letter to B. F. Skinner: Skinner-Keller

Letters (Vol 1., pp. 1). Cambridge, MA: B. F. Skinner
Foundation Archives.

Lindsley, O. R. (1966). Geriatric behavioral prosthetics. In R.

Ulrich, T. Stachnik, & J. Mabry (Eds.), Control of human
behavior (pp. 156-168). Glenview, IL: Scott, Foresman and
Company.

Lindlsey, O. R. (1960). Characterization of the behavior of

chronic psychotics as revealed by free operant conditioning
methods [Monograph]. Diseases of the Nervous Systems,
66-78.

Skinner, B. F. (1932a). Drive and reflex strength. Journal of

General Psychology, 6, 22-37.

Skinner, B. F. (1932b). Drive and reflex strength II. Journal

of General Psychology, 6, 38-48.

Skinner, B. F. (1932c). On the rate of formation of a

conditioned reflex. Journal of General Psychology, 6,
274-286.

Skinner, B. F. (1938/1991). The behavior of organisms. New

York: Appleton-Century-Crofts. (Reprinted by the B. F.
Skinner Foundation, 1991).

137
Skinner, B. F. (1956/1999). A case history in scientific
method. In Cumulative record: Definitive edition (pp.
108-131). Acton, MA: Copley Publishing Group.

Skinner, B. F. (1957). Verbal behavior. New York:

Appleton-Century-Crofts.

Skinner, B. F. (1963). Behaviorism at fifty. Science, 140,

951-958

Skinner, B. F. (1976). Particulars of my life. New York:

Alfred A. Knopf.

Skinner, B. F. (1986). The evolution of verbal behavior.

Journal of the Experimental Analysis of Behavior, 45,
115-122.

Skinner, B. F. (1979). The shaping of a behaviorist. New

York: Alfred A. Knopf.

Vargas, E. A. (1995). Prologue, perspectives, and prospects of

behaviorology. Behaviorology, 3, 107-120.

Watson, J. B. (1924). Behaviorism. New York: W. W. Norton

& Company.

Wolf, M. M., Ridlsy, T., & Mees, H. (1964/1966).

Application of operant conditioning procedures to the
behavior problems of an autistic child. In R. Ulrich, T.
Stachnik, & J. Mabry (Eds.), Control of human behavior (pp.
187-193). Glenview, IL: Scott, Foresman and Company.

138
Footnotes
1
Skinner called this category “verbal behavior under control
of verbal stimuli” and used “intraverbal” for one special case.
E. A. Vargas suggested using “intraverbal” for the whole
category and in a private conversation Skinner agreed that
such designation would have been better. “But,” he said, “it
would be too hard to change now.” Some of us are still trying.

139
Chapter 4

Jacob Robert Kantor

(1888-1984): Pioneer in the
Development of Naturalistic
Foundations for Behavior
Therapy
Paul T. Mountjoy

Western Michigan University

The name of Kantor has high recognition among

psychologists (see O’Donnell, 1985; Popplestone &
McPherson, 1994/1999), especially among behaviorally
oriented ones, and yet it is commonly stated that he has been
relatively noninfluential. Perhaps Schoenfeld (1969) indicated
an essential point: “It was our wit that was wanting, not his.
We demanded too little of ourselves, and expected too much
from him.” (p. 329)

I add that Kantor suggested a manner of doing scientific work

which represented a different approach from the dominant
one of the early twentieth-century, and, indeed, which
remains in contrast to the most popular approaches of the late
twentieth-century as well. This has made mastery of his
works difficult. Consequently, the appreciation of his

140
contributions has been limited, and I hope to redress that
balance to some extent here.

I accept teaching of (and other efforts at dissemination of)

scientific attitudes, as well as innovation in naturalistic and
behavioral approaches to psychological events to be of equal
importance to that of being an active and/or innovative
behavior therapist for purposes of this presentation. We
behavioral psychologists are, above all, interested in and
devoted to the process of bringing the benefits of a natural
science and technology of psychological events to all
members of the human species.

Precis of Kantor’s Life

Julius, the father of Jacob Robert, emigrated from Vilna,
Lithuania to Harrisburg, PA in about 1880, and there became
Rabbi of the Orthodox Congregation Chisuk Emuna Bene
Russia. Shortly, his wife Mary and their first daughter
emigrated to this country. The first child born in this country
was Jacob Robert (in 1888), and the family soon included
three daughters and four sons. On August 2, 1899 Julius died,
which required the eldest son, Jacob Robert, to leave school at
the age of 11 years to become an economic asset to the
family. About 1900 Mary moved the children to Chicago,
Illinois. Jacob Robert never did graduate from high school,
but he was able to remediate his academic deficiencies in a
year at Valparaiso, so that he was enrolled as a student at the
University of Chicago during the summer quarter of 1911.
Here, he obtained the Ph.D., magna cum laude in 1917, and in
1920 went to Indiana University, where he remained until his
retirement in 1959. He then returned to Chicago to live with

141
his daughter (Helene Juliet Kantor — ”Bobby”), and in this
period created some of his most important works (e.g.,
Kantor, 1981). (See Kantor, 1976; Mountjoy & Cone, in
press; Mountjoy & Hansor, 1986; and Wolf, 1984, for
additional details). During his retirement he traveled and
lectured frequently, including several visits to Mexico, on the
first of which he received a standing ovation and a gold medal
(Ribes, 1984). Death came in February of 1984, seemingly in
the midst of his creative and active life, since he had
published one book and two journal articles in that year, as
well as having prepared a book chapter which was published
posthumously. A Festschrift was presented (Smith, Mountjoy,
& Ruben, 1983) only a few months prior to Kantor’s death.

Kantor’s Approach to Psychological

Science
To sum up Kantor in a single sentence I would say something
like this: (1) Utmost respect for the complexities of things and
events; (2) complete admission of our present ignorance of
many of the interrelationships between various things and
events; and (3), overwhelming optimism concerning our
ability to understand these interrelationships by means of
scientific investigation.

Maybe I should be briefer, like this: Even the best living

scientists must be judged as fundamentally ignorant of things
and events, but given time all will be understood.

Examples of our ignorance are often made obvious by new

discoveries. A recent example from physics concerns the
speed of light through a medium. As an undergraduate I was

142
taught that although light could be slowed slightly by the
medium through which it passed, it was for all practical
purposes constant at about 186,300 miles per second in a
vacuum. Hau, Harris, Dutton, & Behrooz (1999) announced
that with the use of a Bose-Einstein condensate (BEC), they
had slowed light to a velocity of 17 meters per second
(approximately 38 miles per hour). This was accomplished by
shining a “coupling” laser into the opaque BEC, then also
firing a second laser beam into the BEC so that the two beams
interacted in a process called electromagnetically induced
transparency. Only 25% of the luminous energy passed
through the formerly opaque BEC, and it was slowed to only
17 meters per second. They expect the technique to further
allow them to slow light down to 37 meters per hour. Quite an
unusual and unexpected finding.

Another announcement which overthrows a physiological

dogma I also learned as an undergraduate actually is
confirmed by several replications, so that we even know some
of the psychological variables which are involved in this
biological event. The report is of actual neurogenesis in the
adult human brain (Eriksson, et al, 1998), which I was taught
could never occur. The basic research has reported
neurogenesis to be a function of running (Van Praag,
Kempermann, & Gage, 1999), living in enriched
environments (Kempermann, Kuhn & Gage, 1997) and
learning (Gould, Beylin, Tanapat, Reeves, & Shors, 1999).
Obviously, these reports also indicate the inevitable
interactions between biological and psychological events.

However, we must remember that all scientific constructs

(whether utterances, theories, generalizations, or claims) are
actually behavior (or if written, a behavior product) by a

143
person who is behaving within a certain space-time
framework (setting, or context), and who has a unique
interbehavioral (learning, or behavioral) history. This history
consists not only of the specialized professional training
received in both undergraduate and graduate school, but is
also comprised of both the post-training self-learning features
of academic-scientific-research activities, and the cultural
shaping procedures of pre-professional training. After all, we
are usually reasonably law-abiding (follow the rules of our
larger cultural group, and the many cultural sub-groups to
which we belong), but we all are shaped to be politically
conservative or liberal, and some of us are trained to be
religious in various degrees, etc. This baggage of our early
life does follow us about as we attempt to do the scientific
job, and it does to some extent influence our scientific
behavior. However when these social preconceptions intrude
too much we can have a scientific problem. In other words,
science is composed of two complementary and necessary
factors: One is empirical investigation, and the second is the
logical analysis, by means of which we examine among other
things, the influence of cultural factors upon scientific work
(Kantor, 1958/1959, 1963-1969). Throughout the history of
science we see the constant interplay of these two essential
components of our scientific activities (see, e.g., Burtt, 1924/
1932; Dampier, 1929/1966).

Here I shall follow Kantor in his use of the term “assumption”

in the logical/mathematical sense of a parameter within a
scientific system. For example, one of the most important
assumptions of interbehavioral psychology is that: Scientific
behaviors are continuous with everyday behaviors. That is to
say, psychological events may be observed and interpreted by
anyone (and they are in everyday life by every lay person!),

144
and it is possible for any person to be trained to apply
scientific methodology to the observation and interpretation
of those events.

An extremely important role for the scientist to play is that of

a critic. This role may be played out in two ways: One is to
demythologize our culture, or more precisely, to attempt to
abolish the intrusions of cultural myths into our scientific
work. This role was embraced heartily by the early
behaviorists, who concentrated upon religico-politico myths.
Primary interest among these persons was the
religico-politico myth of an internal determiner of behavior
which transcended space and time. Natural scientists had
already partially demolished this myth by demoting Soul to
Mind. That is, the Unified Soul, a permanent (immortal)
entity with powers of action, such as thinking, remembering,
and willing, became an Atomic Mind, in which the atoms
were transient mental states. With Soul thus out of science, it
became possible actually to turn away from the Mental
(Psychic) in an attempt to develop a natural science of
behavior. Unfortunately this frequently involved merely
substituting a conceptual nervous system (CNS) which was
variously endowed with either Soul or Mind properties so that
the net gain was very small (e.g., M. Meyer, 1911). Thus,
Kantor (1947, 1963-1969) called for the abandonment of the
quasi-scientific mythology of the nervous system as a
determiner of behavior in all of its guises. Of course, atomic
ways of thinking now have merely gone deeper into the
unknown and substituted new areas of ignorance for the old
ones, and consequently we see the development and
triumphant trumpeting of the “new” science of genetic
determinism. (See, as an example of a biologist who rejects
this doctrine, Rose, 1997. In contrast, Weiner, 1999, lauds the

145
work of Benzer, for demonstrating the genetics of complex
human behavior with drosophila.)

Actually, the internal inconsistencies of the half-psychic and

half-physical universe were evident soon after the
development of the construct. For example, shortly after the
beginning of the Common Era (CE), Saint Augustine
(354-430 CE) lamented that his Soul commanded his body to
obey and yet it did not — and asked his god, rhetorically,
from whence came this awful thing? (Augustine, ca 399/1912
CE). Augustine, of course, did not allow this empirical
observation to interfere with his faith, and thus demonstrated
the validity of another of Kantor’s generalizations:
Philosophical preconceptions determine in large part the
manner in which data are interpreted.

It seems to me we could all agree that abnormal behavior,

representing as it does a deviation from the norm, might be
characterized as “unique” in the sense that it is so unusual as
to present problems of classification into categories in spite of
the DSM’s valiant attempts to do so throughout its various
editions (American Psychiatric Association, 1952/1994).
Indeed, classification is conventionally regarded as one of the
hallmarks of science. However, Kantor swam against the tide
because he insisted upon the uniqueness of all (that is, each
and every) psychological event. For this reason, among
others, his system seems to me to be uniquely suited for an
approach to psychopathology. And, remember that Kantor
(1963-1969) insisted upon a critico-historical analysis as the
foundation for a natural science of behavior, and subsequently
(1958/1959, 1987) a system of behavior therapy.

146
There is a cautionary tale among historians of science which
runs something like this. A biologist was investigating diets
for cattle at a research station. His superior was a physicalistic
reductionist. One day the biologist showed his superior two
lists of chemical elements. This sort of thing: Al = x%, C =
y%, H = z%, and so forth. The biologist said that these
represented the analyses of two diets for cattle, and asked
which one was better. The superior stated he could see no
difference. The biologist replied that this was strange because
while one represented food intake, the other represented the
outflow of residual metabolic matter.

This story is one I never discussed with Kantor, but I know he

would have approved. I shall follow Kantor’s intentions as
exemplified by the story, and paraphrase the moral as: In the
analysis of psychological events, do not reduce those events
to sterile mathematical abstractions (Kantor, 1917).
Remember that, in treatment, we always deal with an
organism in trouble, and it is embedded in a psychological
setting, within which it interacts maladaptively with a
stimulus object. Kantor agreed with other individuals such as
Goldstein (1939) and A. Meyer (1934) on this point.

Citation of Goldstein and A. Meyer here is intended to remind

you that the topic of abnormal behavior is inextricably
entangled with medicine in a historical sense, and that within
both psychology and medicine, as well as in science in
general, we may discern two major traditions. One is analytic
and reductionistic, while the other also analyzes, but does not
attempt to reduce unitary psychological events to sterile
mathematical abstractions, and thus attempts to do full justice
to the event in its entirety. Another way to contrast these two
positions is to say that the reductionistic approach treats a

147
temporal mechanistic chain of smaller events as explanatory,
while the non-reductionistic attitude treats total events with
multiple factors which comprise the complete causal
description.

Incidently, one rule that both of these scientific traditions in

psychology follow is the emphasis upon the necessity for an
outside observer of the psychological event. This was
suggested early on by M. Meyer (1922, 1927) with the use of
the term “the other one” in the titles of two of his books.

Obviously the reductionistic tradition is dominant today,

especially in the popular reports which communicate current
science to the lay public. However, it is possible to
discriminate the differences between the discrete unit,
reductionistic, approach, and the more continuum oriented,
non-reductionistic, attitude, as far back as the first stirrings of
the philosophy and practice of science among the ancient
Greeks, some five centuries Before the Common Era (BCE).
Illustrative are the contrasting positions of Democritus
(460-370 BCE) (to whom there were irreducible atoms which
combined to form everything else), and Heraclitus (540-475
BCE) (who regarded reality as being composed of a dynamic
flux of constant change and becoming). Well into the early
twentieth century the Heraclitian attitude was still
competitive, but the successful application of atomic theory in
physics (in spite of strenuous opposition by many physicists,
for example, Einstein and Mach) had severe repercussions in
all the other sciences.

And, make no mistake about it: All scientific and technologic

enterprises are interrelated. For example, biology has been
impacted by the successes of atomism in physics so that most

148
biologists are members of the reductionistic school, as
described in Wilson’s (1998) Consilience, which proclaims
the desirability, possibility and feasibility of reducing all of
sociology (read psychology) as well as all of biology to
genetic (i.e., internal) determinants. However, competing
traditions do persist, and actually remain healthy and
influential. The non-reductionistic approach, which is to be
found in Darwin’s (1859/1964) Origin of Species (along with
the opportunity for reductionistic development), also gave rise
to the ecological movement (Brewer, 1960). The most easily
available scientific and philosophical accounts of the status of
non-reductionistic biology are Meyr (1982, 1997) and Rose’s
(1999) Lifelines. Rose, incidently is an active researcher in
the molecular mechanisms of memory, while Wilson is also
famous for his ecological work. We scientists always strive
for consistency, but all too often fail to achieve it.

Examples of the non-reductionistic approach in psychology

are not difficult to discover. An excellent example is
Lewontin, Rose & Kamin (1984), with their contribution to
what has been called “the IQ wars.” (See Hunt, 1999, for a
discussion of the current impact of political allegiances within
contemporary social science.) Marr (1990) has discussed the
issue of reductionism in completely behavior analytic terms,
and comparison of my statements here with his should be
rewarding.

It has been my privilege to not only read the works of three of

the authentic geniuses of the twentieth century, but also to
have actually been able to interact with all three on a personal
level. Alphabetically, Arthur Fisher Bentley (1870-1957),
Jacob Robert Kantor (1888-1984), and B. F. Skinner
(1904-1990). These men have not only shaped my intellectual

149
life, but enriched my social life as well. In this presentation I
shall use some of the technical vocabulary of each, as well as
the common language. Please bear with me as I do not do this
to confuse, but: (1) Because there is so little agreement
regarding technical psychological terminology; and (2),
Because all three of these men were constrained by the nature
of psychological events to address the same issues, even
though they did so with somewhat different terminology and
constructs.

Kantor proposed that events be kept separate from their

scientific descriptions (that is, constructs regarding events are
not the same as the actual events themselves). This rule
simply extends and formalizes the ordinary scientific
convention of first reporting data, and then secondly,
interpreting those data (with the realization that alternative
forms of interpretation of those same data are always
possible). And the derivation from that rule is that we
psychologists should not mistake the constructs of the other
sciences for the data of psychology, as Wilson does.
Psychology is a relatively independent science among the
other relatively independent sciences. We should not use the
constructs of the other sciences as our justification for the
scientific status of psychology. We do have our own data:
The psychological event. This is reflected in Skinner’s (1938)
construct of the “reflex” and in Kantor’s (1924-1926)
construct of the “behavior segment.”

Here, then, are two complementary, nondualistic or natural

science, definitions for the psychological event, and both are
derived from empirical observations. Skinner’s probably
descended more from Sherrington’s (1906) biological
tradition, and Kantor’s probably derived more from the brass

150
instrument aspect of the functionalist school of James (1890).
Bentley’s (1895) definition may be regarded as one early step
toward the elimination of the purported mental explanatory
fiction.

All three are in agreement that we cannot satisfactorily

describe a psychological event in terms of only a portion of
that event, i.e., neither solely in terms of the purported mental
and/or physiological functioning of the organism, nor in terms
of the environment alone. For example, Skinner (1938, p. 35)
clearly stated:

The impossibility of defining a functional stimulus without

reference to a functional response, and vice versa, has been
especially emphasized by Kantor [Italics in original, and
citation is to Kantor, 1933b].

These definitions (rules) do not mean that we cannot isolate a

portion of the psychological event for study (e.g., the
organism and/or the stimulus), but it does mean that we can
neither confuse an isolated physiological aspect of the event
with the entire event, nor argue that this physiological aspect
of the event is the complete explanation and cause of the
entire event. Conversely, we cannot argue that the
environment is the sole determiner of the behavior of the
organism. These rules proscribe not only the reduction of
events to one of their actual components, but also the
imposition of non-scientific cultural abstractions upon events
as the ultimate and final explanation. In psychology the social
conventions of Soul and Mind have been especially
troublesome, even though severely attacked by most
advocates of a natural science of psychology. As we have
attempted to approach the status of a natural science some of

151
us have fallen into the error of simply substituting the
conceptual nervous system (CNS) for the
non-spatio-temporal, and purported, rather than actually
isolatable, explanatory device of the Soul or Mind.

At the present time the role of physiological mechanisms in

behavior in general, and in abnormal behavior in particular, is
controversial, not least among psychologists. I have no doubt
that the controversy may be traced, at least in part, to the
medical model (Ulmann & Krasner, 1969/1975) and the
technology of medicine, as well as to a failure to critically
examine basic preconceptions. It has been a problem in
American psychology at least since James’ Principles (1890).
Kantor’s Problems of Physiological Psychology (1947)
addressed this problem in his usual iconoclastic manner. Is
the science of psychology simply a handmaiden of
physiology? Kantor’s answer to this question was a
resounding “NO!” He regarded psychology as a relatively
independent science among the other relatively independent
physical and biological sciences, and psychology was of
equal validity (Kantor, 1953, 1958/1959). In fact, since
scientific research consisted of the interbehavior of an
organism (a scientist) with things and events (which
essentially was the definition of psychology given by Kantor),
psychologists had a special expertise in evaluation of the
validity of the actions of all scientists!

Probably this is as good a place as any to insist that Kantor

(1947) was correct in all essentials when he argued that
physiology was not the ultimate explanation for behavioral
events. Although he based his analysis upon the literature
prior to that publication year (e.g., Sherrington, 1906, and
especially see Leyton & Sherrington, 1917, among others),

152
many recent reports have upheld his statements of some 52
years ago. For example: Vining, et al. (1997) reported on 58
children who had been hemispherectomized in order to
control life threatening grand mal episodes. Their behavior
was quite normal for such severely injured organisms, and
included remarkable voluntary control of the arm and leg
contralateral to the missing hemisphere.

Emphasis upon observables was one of Kantor’s hallmarks.

He regarded the use of physiological factors which had not
been observed to be as great a deviation from proper scientific
procedures and principals as was the utilization of the, by
definition unobservable, Soul or Mind.

Kantor wrote frequently of the necessity for the denial and

abandonment of what he called the psychophysical
protopostulate, and the substitution of a naturalistic
protopostulate in its place. In different words these attitudes
are shared by all who attempt to develop a natural science
system of psychology. From that point on there may well be
competing (or complementary) objective systems of
psychology, i.e., the behaviorisms of Watson, M. Meyer,
Weiss, Hull, Skinner, and others, as well as the
interbehaviorism of Kantor. These somewhat different
systems which have been proposed for a natural science of
psychology will continue to fight for supremacy in the
scientific free market place of ideas in terms of the essential
criteria of both empirical support, and logical consistency,
and may the best system, or amalgamation of systems, win.
[Note well that in the preceding statement I have allowed
some covert assumptions to sneak in! Logical analysis
indicates that I assume that the optimum form of government
for scientific progress is a democracy, and that the optimum

153
form of economic system for scientific progress is capitalism.
So, in a definite sense I am, as was Kantor, a true descendant
of the eighteenth century enlightenment.]

Remember well that any set of data may be interpreted in

myriad manners, as is exemplified in Mazurs (1957/1974)
presentation of approximately 700 versions of the periodic
table of chemical elements within the last 100 years! These
variations were not solely the result of the discovery of new
elements, but many definitely depended upon different
manners of approaching these data. My son, who was then a
teaching assistant in the graduate program in chemistry at the
University of Michigan, reacted to Mazurs with the statement
that some of the more recent tables had great potential for the
teaching of certain aspects of chemistry as over and above the
standard table posted on all lecture room walls in chemistry
departments.

Perhaps the most productive (heuristic) position to take is that

Kantor provided the best philosophical system up to this point
in time for an objective and naturalistic science of behavior,
while Skinner provided the best investigative system so far
for this type of endeavor within science. Note well that, for
Kantor, science was applied philosophy (Kantor, 1917), and
that philosophy was essentially logical and empirical in nature
(Kantor, 1969), and did not deal with specious metaphysical
problems such as the nature of god, the existence of evil, etc.
(Kantor, 1945-1950, 1981). This attitude or assumption, of
course, obviates the usual distinction between the so called
“talking cure” which aims to deal with the psychic, and the
behavioral therapies which regard language as a type of
behavior. Psycholinguistic behaviors comprise both extremely
powerful stimuli and very powerful responses (see Kantor,

154
1936, 1953, 1977, as well as Skinner (1957), and many others
too numerous to cite here).

Bentley may be unknown to many, so I simply state that he is

the only non-psychologist I know of who as a philosopher of
science has consistently and repeatedly denied the
psycho-physical protopostulate (1908/1949, 1926/1936, 1932,
1935, 1954). One of Bentley’s last works was coauthored
with John Dewey (Dewey & Bentley, 1949).

A basic issue is that if the psychophysical protopostulate is

denied, abandoned, or rejected, what is to be substituted for
it? Some individuals are able to reject only portions of it, and
in the false and misguided belief that they have actually
completely rejected it, proceed on a reductionistic course of
reducing the Soul or Mind to the central nervous system, or as
Skinner called it, the conceptual nervous system (CNS).
Perhaps concept is a more desirable term than assumption.
Kantor’s solution (and he did better than most of us have been
able to in completely rejecting that mystical concept of a
universe which is half physical and half psychic which so
permeates our culture and language) was the construct of the
coordinate stimulus-response function (1924-1926). Perhaps a
quotation from a competent psychologist (long time professor
at the University of Michigan) who confesses that he could
not comprehend this aspect of Kantor’s teaching will make
this clear (Walker, nd ca 1993).

To me his position sounded indistinguishable from Bishop

Berkeley’s subjective idealism. However, Kantor was
vehement in denying that identification. He seemed to be
saying there was no real world, therefore no mind-body
problem of any stripe. While I ‘believed’ intensely in what

155
Kantor was teaching us, I was never able to escape the forms
of the English language to think, speak and write on the
Kantorian theoretical plane (p. 184).

Walker did come close to the point here. To Kantor, there was
no problem of a “real world.” The so-called problem of
whether there is “an external reality” is a pseudo-problem
created by that weird, but venerable, politico-religious
assumption of a psycho-physical universe. There is no need to
assume a dual universe, and to become entangled in problems
of “internal” and “external” realities. The simplest solution
seems to be to ignore the dual universe assumption, since it
leads to scientifically insoluble problems and is itself
inherently self-contradictory. In addition, it is not derived
directly from scientific investigation, but only indirectly
through considerations of social control and cultural
cohesiveness. What is needed are other assumptions which
will allow the scientist to proceed in a clear, coherent, and
productive manner. To the development of these other
assumptions, and the explication of their consequences,
Kantor was to devote the largest portion of his professional
career.

One solution of psychologists to an assumed psychophysical

universe is to simply devote themselves to the accumulation
of empirical factoids, serene in the misguided and misleading
belief that the facts somehow speak for themselves
independently of theory. The history of science is filled with
such individuals (Dampier, 1929/1966, and others). However,
it has long been known that unacknowledged philosophical
biases are the fatal flaw in most such persons (Burtt, 1924/
1932, Kantor, 1958/1959). Another solution, the one chosen
by Kantor, was the mirror image of this, and it was to develop

156
a philosophy suitable for the pursuit of a natural science of
psychology, and to this endeavor he remained true throughout
his life. For him there were human beings (organisms) who
interacted with their surroundings. One adjustment technique
was called science, and this was the most satisfactory of all
interactions for producing rules which resulted in more and
more effective types of interactions with things and events.
He elected to: (1) observe what scientists actually did; and
(2), develop a set of rules; which (3), if followed, would result
in better scientific work. That is, he argued that if one
observed scientists at work, and then accurately described the
behaviors they performed, and then reacted both critically and
constructively, the inevitable result would be better, and more
naturalistic science. Up to the point in time at which Kantor
began to study there was excess pseudo-scientific verbiage
which was not really based upon observation, and which
therefore just got in the way of the scientific worker since the
job of the scientist, properly conceived, was to observe and
describe things and events. The point was discovery. And to
discover most effectively the worker must not approach with
notions which were not derived from previous observations.
Political and religious concepts of a dualistic (half Mental and
half Physical) world had been developed to organize and
control cultural groups. That may have been a legitimate
behavior for people in the distant past, whose aim was to
produce a coherent and cohesive society with a high
probability of survival in a hostile environment, but it was not
scientific work. The scientist does things differently than the
priest or politician. And the behavior of priests, politicians,
and even scientists, was a proper domain of psychological
events which are important objects for observation and study
by the psychological natural scientist (see Kantor, 1963-1969

157
for explication of his natural science approach to the
interbehaviors of scientists with their data).

Kantor chose to present his system in the formal manner

which is associated with Euclid’s Elements (ca 300 BCE).
Those who have forgotten their high school geometry will
find that Artmann (1999) gives a useful review, which should
make readers more sympathetic to Kantor’s (1958/1959)
formal statements of the assumptions and theorems of
interbehavioral psychology.

Note well that Kantor was always interested in developing

solutions to problems of maladjustment. The first two years
after his Ph.D. (1918-1920) were spent as an instructor at the
University of Chicago. During this time he taught abnormal
psychology two times, experimental psychology once, as well
as introductory several times, and published his first
independent works. Two of his first four published papers
(1918, 1919) were devoted to the topic of abnormal behavior.
To be sure, upon reading these it is clear that he was really
presenting his first somewhat fumbling attempts to escape
psychophysical dualism, and was not yet as successful as he
would become in later years. However, a mature presentation
was not to be long in appearing. The last chapter of the
Principles (1924-1926) was a well developed description in
terms of both: (1) a lack of necessary reactions; and (2) the
acquisition of unsuitable ones (p. 459). Obviously, there are
many details which I cannot explicate here.

His last (posthumously published) paper (1987) was an

exposition of how interbehavioral psychology qualified as a
therapeutic approach. During his long tenure at Indiana
University he regularly taught “principles of

158
psychopathology.” In addition, he treated abnormal behaviors
in both of his introductory textbooks (Kantor, 1933a; Kantor
& Smith, 1975). Thus, we can see that his early interest in
psychological maladjustment was maintained throughout his
scholarly life and career. Other publications which support
this generalization include his 1923 paper on personality
(which certainly influenced Lundin’s treatment of personality
(1961, 1969), and which was reiterated, for example, in his
introductory texts (1933a, 1975).

Actually the first public presentation of Kantor’s naturalism

was a presentation to “the Psychological Seminar in the
University of Minnesota, 1916-17” (Kantor, 1920, p. 260).
Here he attacked what was widely thought to be the cutting
edge of psychological science at that time: The testing
movement (see Mountjoy and Cone, in press) Talk about
swimming against the tide! Why, he even defended the
concept that women have equal intelligence as compared to
men, in this paper.

A Summary of Kantor’s System in

Brief
There are naturally occurring things and events, and these
differ in their duration and inumberable other details. Events
are short lived and examples would be my uttering a word, or
a sentence. Things are maybe best thought of as slow events,
such as a human being. That is, thinking of a normal human
life span of say 70 years, as compared to the duration of mere
seconds of a verbal utterance. Our time scale in scientific
terms has expanded to billions of years for some things. The

159
earth, for example, has existed as an entity for approximately
four and one-half billions of years. Now, since we human
beings interact with other things (or events), and produce
behavior products which may, as in the case of writing,
endure for many years, it seems useful to speak of the most
effective of these interactions as science.

Scientific interbehaviors simply attempt to discover what is

really going on in this concatenation of things and events of
which we are an integral part. Yet, everyone has a different
history of contacts with things and events, and these
inevitably result in different reactions to them. The most
deviant reactions serve as the best examples. Some people
seek out classes of stimuli which most of us avoid: For
example, those we term masochists who obtain sexual orgasm
only when tissue injury accompanies other stimulation; or
those we term sadists who obtain sexual gratification only by
inflicting tissue injury upon others. We attempt to avoid
having our scientific work influenced by personal histories
which produce behaviors such as these, and among those
personal histories we must include exposure to the
politico-religico mythologies of our specific cultural group
and/or sub-group. For example, bad science is one result of
our western European cultural bias toward a single god,
which deity some opine, may be influenced by prayer.
Anthropologists of western European cultural background
(i.e., Christians) have seized upon this notion of
intercessionary prayer to separate their “advanced religion”
from the “primitive magic” of “uncivilized savages,” on the
ground that the “primitive” belief is that with magic one can
force the gods to bend to one’s will. The “advanced “ belief is
that the deity must be supplicated, and cannot be forced to
obey. Why “advanced?” Only because of their membership in

160
a cultural group (Christians) which already so believes. Now,
if I could discern the religious origin of this bit of
so-called-science as an undergraduate, why is it that Ph.D.’s
in anthropology still have problems? Well, in brief, they have
not totally rejected the psychophysical protopostulate (I love
how that phrase rolls off the tongue).

Serious debate concerning the relationships between scientific

work and religious institutions emerged with the advent of the
heliocentric hypothesis of our solar system developed by
Copernicus (1473-1543). It is well known that Galileo
(1564-1642) spent the last years of his life in confinement for
advocating the Copernican hypothesis. These battles between
science and religion have been described by Draper (1874)
and White (1898/1995). An historical analysis of the
abandonment of religion during recent times is provided by
Turner (1985). A more contemporary viewpoint may be
found in a recent issue of a popular magazine (Frazier, 1999).

Kantor’s Influences Upon the

Behavior Therapies
Name recognition, and Smith’s (1990/1993) citation study,
both lead to the conclusion that Kantor’s influence is a
reasonable topic for investigation. Here I present some
preliminary data which should be examined in order to
determine the extent to which Kantor has been influential, or
alternatively, to what extent he should have been influential,
but was not. My own examination continues, but replication
by independent observers is one of the hallmarks of a natural
science. Be reminded that the study of the history of a natural

161
science is itself an example of working as a natural scientist
(Kantor, 1963-1969, 1976).

Event Driven (No Direct Contact with Kantor)

Many scientists work in a manner which is compatible with

the interbehavioral system, but apparently have had no direct
contact with Kantor or his writings. I prefer to call these
individuals “event driven” because frequently in our
conversations Kantor argued that since he himself was both
event oriented, and relatively free of handicapping cultural
mythology, all scientists would eventually conform to his
position. Examples may be found in Smith (1990/1993) and
Smith & Smith (1996), while Rose’s (1997) discussion of
biological processes serves as a most compelling example.

Formal Students

Kantor rejected the “great man” approach to history, and with

equal vigor, the notion that we are all passive victims of our
culture. Instead he opted for an interactional approach which
included the cultural variables, and each person’s individual
developmental history, which in combination with biological
conditions produced a unique individual. He argued that an
interaction between these two variables might combine to
produce a scientist who would produce a unique solution to
scientific problems (Kantor, 1976; Mountjoy & Cone, in
press). In his formal classes he attempted to facilitate a
problem solving approach to scientific work rather than a
slavish acceptance of his own system. Often he stated that he
would be a failure if his students did not progress further than
he himself had been able to in the development of a natural
science of psychology. However, he did insist upon students’

162
appreciating his own critical approach to the commonly
accepted cultural mythology of a psychophysical universe. In
my opinion, there was a common core emphasis upon a
problem solving approach which was shared by all of the
faculty in the psychology department at Indiana University
when I was in residence between 1949 and 1954. I assume
that Kantor played a role in producing that approach among
his colleagues. At any rate, the department did produce a
number of individuals who were achievers within psychology.

The students whose dissertations and theses could be

ascertained to have involved Kantor as Chair were listed by
Hearst and Capshew (1988) (records of faculty directors were
not kept carefully in early years), and I present them here in
chronological order: 1934: Jerry W. Carter (M.A.); 1936:
Paul M. Schroeder (M.A.); 1937: Ignacio T. Briones (M.A.);
1938: Ignacio T. Briones (Ph.D.); 1938: Jerry W. Carter
(Ph.D.); 1942: J. W. Bowles (M.A.); 1942: John Bucklew
(Ph.D.); 1943: Robert W. Lundin (M.A.); 1944: Annemarie
Lehndorff (M.A.); 1944: Nicholas H. Pronko (Ph.D.); 1947:
David T. Herman (Ph.D.); 1947: Robert W. Lundin (Ph.D.);
1947: D. Morgan Neu (M.A.); 1948: Harris E. Hill (Ph.D.);
1948: Irvin S. Wolf (Ph.D.); 1951: Marjorie P. Mountjoy
(M.A.); 1952: J. W. Bowles (Ph.D.); 1953: Paul T. Mountjoy
(M.A.); 1954: Solomon Weinstock (Ph.D.); 1957: Paul T.
Mountjoy (Ph.D.). That is a total of 14 individuals, and
perhaps more, since the early records (roughly the decade of
the 1920’s) are so incomplete.

Here, I cite some publications of selected individuals from the

above list who have published items which appear to me to be
worth consulting: Carter (1937a, 1937b, 1937c, 1938, 1939,
1968), Louttit & Carter (1939); Briones (1937), Bowles &

163
Pronko, (1949), Bucklew (1941, 1943, 1958), Bucklew &
Hafner (1951; Lundin (1961, 1965, 1969); Mountjoy (1957,
1976, in press a, in press b), Mountjoy & Cone (1995, 1997,
in press); Mountjoy & Ruben (1983); Pronko (1946, 1980, ),
Pronko & Hill (1949), Pronko & Bowles (1951), Pronko &
Herman (1982); Herman et al. (1957); Hill (1944a, 1944b,
1945); and Wolf (1958a, 1958b). The major contributions of
Mountjoy are not well represented by the references cited
above for him. In fact, his editorial work on The Record, and
especially his participation in the development of the
behavioral graduate program at Western Michigan University,
constitute contributions which will impact the scientific
culture some years into the future.

Other graduate students at Indiana who were influenced by

Kantor (but did not have him as the Chair of their committees,
and there are no available public records that I know of as to
whether he served on their committees) include (in
alphabetical order): Charles Boltuck, John F. Brackman, Sam
L. Campbell, Robert S. Daniel, Donald Doehring, Paul Fuller,
Frederick P. Gault, Joel Greenspoon, John Grossberg, Adolph
Jack Hafner, Lloyd E. Homme, Billy L. Hopkins, Lawson
Hughes, Gilbert R. Johns, Arthur Kahn, Fred Kanfer, Neil D.
Kent, Parker E. Lichtenstein, David O. Lyon, Marion W.
McPherson, James D. Miller, Edith Neimark, Stanley C.
Ratner, Samuel H. Revusky, Julian B. Rotter, Max S.
Schoeffler, William A. Shaw, Dewey A. Slough, Edward L.
Walker, and Don Zimmerman. In addition, Cromwell
(Cromwell & Snyder, 1993), has stated that his undergraduate
senior seminar course with Kantor influenced his entire career
(see also Wynne, Cromwell & Matthysse, 1978, and
Anonymous, 1995). Noel W. Smith also completed the
undergraduate senior seminar under Kantor, but was

164
prevented from further academic work with Kantor by his
retirement in 1959. My criteria for inclusion of the above list
are both personal communications and publications that I
have been able to discover.

Here, I cite some publications of selected individuals: Fuller

((1949, 1973, 1987); Greenspoon (1955); Grossberg (1972,
1981), Hafner (1958); Kanfer & Phillips (1970); Lichtenstein
(1983, 1984); Mountjoy (1957, 1976, 1980, 1987, in press a,
in press b), Ratner (1957), Ratner & Rice (1963), Ratner,
Gawronski & Rice (1964); and Rotter (1942, 1954).

Examination of Hearst & Capshew’s (1988) listing of MA

and Ph.D.’s granted at Indiana reveals that studies of verbal
conditioning constitute a noticeable category. While Skinner
(1957) certainly played a role here because of his tenure as
chair of the department for two years, it is important to note
that Kantor treated language interbehaviors naturalistically in
a series of papers beginning in 1921 and 1922, and
culminating in his book of 1936 (and later in 1977).

Colleagues of Kantor

Many professional psychologists, and other scientists who

had already completed their formal training at institutions
other than Indiana University, became interested in Kantor as
a result of personal contacts/reading, and so on. Herewith, I
list some of these in alphabetical order: Don Baer, S. Howard
Bartley, Sid Bijou, Joseph V. Brady, Al Cone, Donna Cone, J.
J. Gibson, Israel Goldiamond, Emilio Ribes, Nate Schoenfeld,
B. F. Skinner, William Stephenson, Paul Swartz, William S.
Verplanck, and A. P. Weiss. This list contains 14 names, but

165
many other examples may be found in Smith, Mountjoy and
Ruben (1983) and Smith (1990/1993).

Here I shall be extremely selective in citing publications:

Bijou (1981), Bijou & Ghezzi (1994), Bijou & Ribes (1996);
Brady (1970, 1975); Schoenfeld (1969, 1974, 1993); Swartz
(1963); Verplanck (1983); and Weiss (1925/1929).

In some cases, both undergraduates and graduate students at

Indiana University were influenced by Kantor, but completed
their graduate work at other institutions. These include David
Bakan (1952, 1966), Rue Cromwell, Edward Walker, and
Donald W. Zimmermann (1979, 1982). Bakan told me in
1997 that he went to Indiana University because he felt that
Davis and Kantor were the outstanding psychologists in this
country at that time.

Students of Kantor’s Students

Also, there are the students of Kantor’s students. For

convenience and brevity, I list only some of whom I know,
however, these have been remarkably productive in terms of
publications: Dennis J. Delprato (via Stan Ratner); Linda (nee
Parrot) J. Hayes (via Paul Mountjoy); Ed Morris (via Parker
E. Lichtenstein, Irvin S. Wolf, and Sid Bijou); Gerald L.
Shook (via Mountjoy); and Douglas Ruben (via Paul
Mountjoy).

Selected references include: Delprato, (1995), Delprato and

McGlynn (1988), Delprato and Midgley (1992), Ruben and
Delprato (1987); Hayes, (Parrot, 1984), Hayes and Ghezzi,
(1997); Morris, (1982); Ruben, (1983, 1984a, 1984b, 1985a,
1985b, Barrer & Ruben, 1984, Ruben and Delprato, 1987,

166
Koziol, Stout, & Ruben 1993). Although references could be
cited for Shook, it appears most appropriate to indicate that
the Society for the Advancement of Behavior Analysis
presented a twenty-five year career award for Public Service
in Behavior Analysis to him at the 1999 convention of The
Association for Behavior Analysis in Chicago, IL. This award
was bestowed primarily for his work on developing testing
and legislative procedures for National Certification of
Behavior Therapists.

The Psychological Record

It may be that the best known of Kantor’s actions to further

the natural science of psychology was his founding of The
Psychological Record in 1937 (Mountjoy and Cone, 1997, in
press). This was the first psychological journal to actively
seek manuscripts which had an authentic natural science
orientation. After its rebirth in 1956 by Swartz (Bartlett,
1997), Wolf and Mountjoy accepted the responsibility of
editing it, and later passed that duty on to Charles Rice, who
is the current Editor. At the time of this writing, The Record is
in its 49th year of publication, and thus has contributed to the
dissemination of psychological research and theory for nearly
50 years. Many papers which may be labeled Experimental
Analysis of Behavior and/or Applied Analysis of Behavior
types have been published in The Record, both prior to and
following the establishment of the journals which bear the
acronyms of JEAB and JABA. Additionally, many cognitive
behavior therapy papers have appeared in the pages of The
Record.

Kantor remained deeply involved in the editorial affairs off

The Record until his death.

167
Interbehavioral Research
Kantor not only encouraged laboratory work by his students;
he also regarded experimentation as essential to scientific
work and progress (Kantor, 1978, 1959/1960). In so far as he
was concerned, each and every thing and event was subject to
experimental manipulation and scrutiny in principle. He
himself had clinical skills, and actually was the hypnotist in
the Davis and Kantor (1935) report on changes in
physiological measurement associated with hypnosis. Thus it
seems appropriate to cite reports which are either directly
related to the interbehavioral framework and/or at least
compatible with it. One example which rewards attention is
Kantor’s analysis of the experimental analysis of behavior
(1960).

Other selected examples include these: Wahler was

introduced to interbehaviorism by Verplanck. Wahler and his
associates have vigorously evaluated the role of setting
factors across a spectrum of situations (e.g., Wahler &
Dumas, 1989). Hart & Risley (1995, 1999) have diligently
recorded the actual (literally millions of them!) linguistic
interactions of young children learning to speak. They
deserve high praise for these heroic activities. Their two
reports clearly indicate the misleading nature of purely
physicalistic descriptions of “environments” as influences
upon behavior. The data necessary to understand the
interbehavioral evolution of language are the sequential
interactions of children with their stimulating surroundings.
That is, the important sets of data include the smallest details
of interactions. If Kantor was alive to peruse these reports I

168
am confident that he would have found them to be
praiseworthy.

Plomin (1994) is a geneticist/psychologist who has attempted

to transcend the sterility of the conceptual dichotomy between
“nature” and “nurture” by actual description of the “interplay”
between genetic and experiential factors. This is a stimulating
work in spite of what I regard as certain conceptual
deficiencies. Diamond (1997) has presented an interesting
analysis of the role of geographic factors in cultural evolution.
Kantor has spoken of such aspects of events as either
hindering or facilitating conditions.

Smith has conducted both empirical (1976a, 1976b) studies

and theoretical analyses (1982, 1984), as well as having
attempted a generalized description of interbehavioral therapy
procedures (1978).

Stephenson (1984) reported a long lasting positive regard for

Kantor, and has long associated the Q-Methodology which he
has developed with the interbehavioral system. Q-Sorts are
one manner of studying those difficult to observe
interbehaviors in which the stimulus object is not readily
apparent. These events are frequently referred to as
“subjective” or “private.”

An interesting review of developments within the

interbehavioral system of psychology (as reflected in The
Psychological Record from 1937 to 1983) has been presented
by Ruben (1984b). This would serve as useful background
and a solid foundation for a review which would include
additional published sources.

169
Summary and Conclusions
Robert Kantor gave us a way of looking at (perceiving) the
world which allowed us to bypass the hoary traditions of
nearly two millennia of conceiving reality as having two
levels of existence: The tangible world of science and its
exact linguistic opposite, the purportedly intangible world of
spirit, which is the ultimate and absolute cause not only of
behavior, but indeed of everything.

We may conceive of his writings as in large part consisting of

rules of scientific behavior. Somewhat loosely we may regard
them as functioning in the Skinnerian sense of rule governed
behavior. Of course, Kantor himself preferred the term
“assumption” with a very specific meaning of having been
derived from interaction with things and events rather than
just some proposition(s) which someone dreamed up
arbitrarily out of the blue. (See Artmann, 1999, as well as
Kantor, 1945-1950, 1958/1959, for discussions of the roles of
assumptions, or axioms, and their origins, in both the
deductive sciences and the inductive sciences.)

Assumptions may be ranged on a continuum from the most

abstract to the most concrete. For example, from science is
the best way to learn about things and events (Kantor, 1958/
1959) through many intermediate stages to the discussion of
“Applied Subsystems” to which he devoted chapter 17
(Kantor, 1958/1959, p. 170-178).

The psychological event is not an expression of the inner

essence of the person, but a complex concatenation, or
interaction, among the actual factors involved in, or present

170
in, the event. We can categorize these in a general way as (1)
an organism or biological entity, (2) a stimulus object, (3) the
surroundings and limits, and (4) the history of interactions of
that Organism with that Stimulus Object.

Note that we may substitute other terms as equivalences

which are used by other individuals. Surroundings and limits
are environment, reinforcement, etc. That is, all scholars have
always been interacting with the same things and events that
Kantor interacted with. Almost all scholars are limited by
concepts that do not truly reflect things and events but instead
reflect concepts derived from the general religio-politico
conditions of Western European culture to which I have so
frequently alluded already, and which Kantor (1963-1969)
described so eloquently.

Kantor always sought the most general statement possible.

For example, “things and events” are a continuum from
relatively stable things such as the earth (four-and one-half
billions of years old — though constantly changing) to events
such as the relatively fleeting psychological events (such as a
linguistic utterance) which may occupy only some portion of
a second, or up to a few seconds of time.

Thus, interbehavioral therapy does not deal with the Soul or

Mind but with behaviors, with behavioral events, the totality
of complete events. No therapist knows prior to interaction
with the patient what aspect(s) of the behavioral events may
need to be changed or modified to produce the desired
(socially acceptable) outcome. And, what assumptions may
(must) be made in arriving at the concept of the desired
outcome? After all, that concept of a desired outcome is
actually a behavior performed by the therapist. Hopefully the

171
construct of a desired outcome is based upon the expectations
of the social/cultural group(s) to which the client belongs, not
necessarily only those to which the therapist belongs! (See M.
Meyer, 1922, 1927 re: “the other one,” which is his
mnemonic for the necessity of an observer external to the
psychological event.)

Choice of words is quintessentially important if we are to

communicate with our audience, whether that audience be
lay, or patient-client, or professional peers. My
interbehavioral history with respect to many words is
extremely different from that of other individuals. For
example, psychology is the study of the Mind or Soul to the
majority of both lay and professional Americans, rather than
the study of the interactions of organisms with objects as it
was to Kantor, and I devoutly hope it is to me.

Postscript
The evidence discussed above leads me to conclude that
Kantor played a reasonably important role in the development
of the theoretical foundations for behavior therapy through his
efforts at construction of a system for a natural science of
psychology. In addition, he devoted a remarkable amount of
energy to the dissemination of supportive research activities
with his founding and support of The Psychological Record,
and his teaching activities. Even after retirement, until his
hearing impairment made it impossible to react to questions
from the audience, he continued to travel and lecture. Indeed,
he devoted his life to the pursuit of a natural science of
psychology. What better epitaph could there be?

172
Immediately prior to his fatal seizure, Robert was preparing a
never to be completed paper for The American Psychologist,
and his daughter later found a handwritten note on his desk
which may have been intended as a title within the text, or
perhaps a general concluding statement for that paper. I quote
it in full as it does summarize his life work in one pithy
statement (Mountjoy and Hansor, 1986, p. 1297):

No spirits, wraiths, hobgoblins, spooks, noumena,

superstitions, transcendentals, mystics, invisible hands,
supreme creator, angels, demons...[ca 1984].

References
With some annotations. Certain of these references may be of
antiquarian interest only, but others are quite seminal. This
list is, of course, historically incomplete as it centers upon my
discovery of items which are by and/or relatable to J. R.
Kantor. For a more complete, though, of course now dated by
the inevitable passage of relentless time, list of references
concerning behavior therapy, see Kazdin (1978). I must admit
that my selection criteria might be argued to have been not
always consistent. At times a selection was made based upon
its early date, and at other times I considered content as more
important. For some individuals, I listed more works in order
to characterize their career. In my own defence, I state that I
used my judgement as to what criteria would serve this
audience best, and still produce a reference list of a
manageable size. The annotations serve as a substitute for a
fuller discussion of the work of many of the individuals cited
due to constraints upon space available. Some might object to
my inclusion of “popular” works, written for the lay audience,

173
however, I regard the education of mankind into the nature of
science to be essential for all citizens. After all, they vote for/
against the people who provide the monies to allow us to do
our work.

American Psychiatric Association. (1952/1994). Diagnostic

and statistical manual of mental disorders (4th ed.), 1994.
Washington, DC: American Psychiatric Association. (1st ed.,
1952.)

Anonymous. (1995). An interbehavioral approach to the study

of psychopathology: An interview with Rue Cromwell. The
Interbehaviorist, 23, 9-11.

Artmann, B. (1999). Euclid: The creation of mathematics.

New York: Springer-Verlag. (The first seven chapters
(especially) indicate the importance of axioms and definitions
in mathematics; a position which Kantor and others have
extended to the empirical sciences.)

Augustine, Saint. (ca. 399/1912). Confessions. (W. Watts, tr.;

Loeb Classical Library). London: Heinemann.

Bakan, D. (1952). The exponential growth function in Herbart

and Hull. American Journal of Psychology, 65, 307-308.

Bakan, D. (1966). The influence of phrenology on American

psychology. Journal of the History of the Behavioral
Sciences, 2, 200-220.

Barrer, A. E., & Ruben, D. H. (1984). Readings in brain

injury. Guilford, CT: Special Learning Corp. (Useful for
education of lay public, especially family members, because

174
of emphasis upon rehabilitation, social adaptive living, and
community reintegration.)

Bartlett, N. R. (1997). The Psychological Record: Rebirth in

1956. The Psychological Record, 47. 21-24.

Bentley, A. F. (1895). The units of investigation in the social

sciences. The Annals of the American Academy of Political
and Social Sciences, Philadelphia, 5(6), 915-941. (Publication
version of his Ph.D. dissertation.)

Bentley, A. F. (1908/1949). The process of government.

Chicago: The University of Chicago Press. (Reissued 1935,
Bloomington, IN: The Principia Press; and, 1949, in a new
edition, with an introduction by H. T. Davis, Evanston, IL:
Principia Press of Illinois.)

Bentley, A. F. (1926/1936). Relativity in Man and Society.

New York: G. P. Putnam’s Sons. (Reissued, 1936,
Bloomington, IN: The Principia Press).

Bentley, A. F. (1932). Linguistic analysis of mathematics.

Bloomington, IN: The Principia Press. (This contains his
answer to the Whitehead & Russell Principia Mathematica
[1910-1912/1926-1927], which argues that all knowledge
may be reduced to a few mathematical and logical statements.
The converse, that all knowledge may be deduced from a
limited number of propositions, is also proposed. Bentley
disagrees vehemently.)

Bentley, A. F. (1935). Behavior, knowledge, fact.

Bloomington, IN: The Principia Press.

175
Bentley, A. F. (1954). Inquiry into inquiries. Boston: Beacon
Press. (Contains complete list of his publications.)

Bijou, S. W. (1981, May). Child development and

interbehavioral psychology. Presented at the meeting of the
Association for Behavior Analysis, Milwaukee, WI.

Bijou, S. W., & Ghezzi, P. M. (1994). Outline of J. R.

Kantor’s psychological linguistics. Reno, NV: Context Press.

Bijou, S. W., & Ribes, E. (1996). New directions in behavior

development. Reno, NV: Context Press.

Bowles, J. W., & Pronko, N. H. (1949). Reversibility of

stimulus function under hypnosis. The Journal of Psychology,
27 41-47.

Brady, J. V. (1970). Some conceptual problems and

psychophysiological experiments. In M. B. Arnold (Ed.),
Feelings and emotions, the Loyola symposium. New York:
Academic Press.

Brady, J. V. (1975). Toward a behavioral theory of emotion.

In L. Levi (Ed.), Emotions: Their parameters and
measurements. New York: Raven Press.

Brewer, R. (1960). A brief history of ecology: Part I -

pre-nineteenth century to 1919. Occasional Papers of the C.
C. Adams Center for Ecological Studies, 1, 1-18.

Briones, I. T. (1937). An experimental comparison of two

forms of linguistic learning, The Psychological Record, 1,
205-214. (I must confess that I am deplorably ignorant

176
concerning his later career, although I know that he occupied
a special place in Kantor’s life. Tears came to the eyes of
Bobby when she spoke to me about Briones after her father’s
death. Briones may have returned to his native land to teach.)

Bucklew, J., Jr. (1941). An experimental set-up for the

investigation of language problems. Journal of Experimental
Psychology, 28, 534-536.

Bucklew, J., Jr. (1943). An exploratory study in the

psychology of speech reception. Journal of Experimental
Psychology, 32, 473-494.

Bucklew, J, Jr. (1958). Evidence from retrograde amnesia for

a unit of behavior higher than the stimulus-response. The
Psychological Record, 8, 13-16.

Bucklew, J., Jr., & Hafner, A. J. (1951). Organismic versus

cerebral localization of biological defects in
feeblemindedness. The Journal of Psychology, 32, 69-78.

Burtt, E. A. (1924/1932). The metaphysical foundations of

modern physical science: A historical and critical essay.
London: Routledge & Kegan Paul Ltd. (Second, revised
edition, 1932.) (Argues that it is impossible to understand
modern physics without understanding the philosophy of
Isaac Newton.)

Carter, J. W., Jr. (1937a). An experimental study of the

stimulus function. The Psychological Record, 1, 33-48.

Carter, J. W., Jr. (1937b). A case of reactional dissociation

(hysterical paralysis). American Journal of Orthopsychiatry,

177
7, 219-224. (Is this the first approximation toward behavior
therapy? Perhaps. The date is right, and the content awaits
analysis by a competent behavior therapist who is interested
in the history of psychology.)

Carter, J. W., Jr. (1937c). A new serial presentation apparatus.

The Journal of General Psychology, 17, 409-414.

Carter, J. W., Jr. (1938). An experimental study of

psychological stimulus-response. The Psychological Record,
2, 33-92.

Carter, J. W., Jr. (1939). Manual for the psycho-diagnostic

blank: A guide for diagnostic interviewing in psychological
clinic work. The Psychological Record, 3, 249-290.

Carter, J. W., Jr. (1968). Research contributions from

psychology to community mental health. New York:
Behavioral Publications, Inc.

Cromwell, R. L., & Snyder, C. R. (Eds.). (1993).

Schizophrenia: Origins, processes, treatment, and outcome.
New York: Oxford University Press.

Dampier, W. C. (1929/1966). A history of science and its

relations with philosophy & religion. London: Cambridge
University Press. (Before 1956 there were 15 editions and
printings.)

Darwin, C. R. (1859/1964). On the origin of species. London:

W. Clowes and Sons. (A facsimile of the first edition with an
introduction by Ernst Mayr, Cambridge, MA: Harvard
University Press, 1964.)

178
Davis, R. C., & Kantor, J. R. (1935). Skin resistance during
hypnotic states. Journal of General Psychology, 13, 62-81.
(Kantor was the hypnotist in this study.)

Delprato, D., & McGlynn, F. D. (1988). Interactions of

response patterns and their implications for behavior therapy.
Journal of Behavior Therapy and Experimental Psychiatry,
19, 199-205.

Delprato, D. J., & Midgley, B. D. (1992). Some fundamentals

of B. F. Skinner’s behaviorism. American Psychologist, 47,
1507-1520.

Delprato, D. J. (1995). Interbehavioral psychology: Critical,

systematic, and integrative approach to clinical services. In
W. O’Donohue & L. Krasner (Eds.), Theories of behavior
therapy: Exploring behavior change (pp. 609-636).
Washington, D.C., American Psychological Association.

Dewey, J., & Bentley, A. F. (1949). Knowing and the known.

Boston: Beacon Press.

Diamond, J. (1997). Guns, germs, and steel: The fates of

human societies. New York: W. W. Norton & Company.

Draper, J. W. (1874). History of the conflict between religion

and science. New York: D. Appleton and Company. (About
42 editions and printings up to 1937.)

Eriksson, P. S., Perfilieva, E., Bjork-Eriksson, T., Alborn, A.,

Nordborg, C., Peterson, D. A., & Gage, F. H. (1998).
Neurogenesis in the adult human hippocampus. Nature
Medicine, 4, No. 11, 1313-1317.

179
Frazier, K. (Ed.). (1999, July-August). Skeptical Inquirer,
23(4), 1-84. This special issue on science and religion is cited
here by the name of the editor, and contains papers by many
scientists, philosophers, and popular writers.

Fuller, P. R. (1949). Operant conditioning of a vegetative

human organism. American Journal of Psychology, 69,
587-590.

Fuller, P. R. (1973). Professors Kantor and Skinner: The

“Grand Alliance” of the 40’s. The Psychological Record, 23,
318-324.

Fuller, P. R. (1987). From the classroom to the field and back.

In D. H. Ruben & D. J. Delprato (Eds.), New ideas in therapy.
Westport, CT: Greenwood Press.

Goldstein, K. (1939). The organism. New York: American

Book Co.

Gould, E., Beylin, A., Tanapat, P., Reeves, A., & Shors, T. J.
(1999). Learning enhances adult neurogenesis in the
hippocampal formation. Nature Neuroscience, 2, No. 3,
260-265.

Greenspoon, J. (1955). The reinforcing effect of two spoken

sounds on the frequency of two responses. The American
Journal of Psychology, 68, 409-416.

Grossberg, J. M. (1972). Brain wave feedback experiments

and the concept of mental mechanisms. Journal of Behavior
Therapy and Experimental Psychiatry, 3, 245-251.

180
Grossberg, J. M. (1981). Comments about cognitive therapy
and behavior therapy. Journal of Behavior Therapy &
Experimental Psychiatry, 7, 25-33.

Hafner, A. J. (1958). Rorschach test behavior and related

variables. The Psychological Record, 8, 7-12.

Hart, B., & Risley, T. R. (1995). Meaningful differences in

the everyday experiences of young American children.
Baltimore, MD: Paul H. Brookes Publishing Co.

Hart, B., & Risley, T. R. (1999). The social world of children

learning to talk. Baltimore, MD: Paul H. Brookes Publishing
Co.

Hau, L. V., Harris, S. E., Dutton, A., & Behrooz, C. H.

(1999). Letter to Nature, 397, 594-598.

Hayes, L. J., & Ghezzi, P. M. (Eds.). (1997). Investigations in

behavioral epistemology. Reno, NV: Context Press.

Hearst, E., & Capshew, J. H. (1988). Psychology at Indiana

University: A centennial review and compendium.
Bloomington, IN: Indiana University Department of
Psychology.

Herman, D. T., Lawless, R. H., & Marshall, R. W. (1957).

Variables in the effect of language on the reproduction of
visually perceived forms [Monograph]. Perceptual and Motor
Skills, 7, 171-186.

181
Hill, H. (1944a). Stuttering: I. A critical review and
evaluation of biochemical investigations. The Journal of
Speech Disorders, 9, 245-261.

Hill, H. (1944b). Stuttering: II. A review and integration of

physiological data. The Journal of Speech Disorders, 9,
289-324.

Hill, H. (1945). An interbehavioral analysis of several aspects

of stuttering. The Journal of General Psychology, 32,
289-316.

Honig, W. K. (1959). Perspectives in psychology XII.

Behavior as an independent variable. The Psychological
Record, 9, 121-130.

Hunt, M. (1999). The new know-nothings: The political foes

of the scientific study of human nature. New Brunswick, NJ:
Transaction Publishers.

James, W. (1890). Principles of psychology (2 vols.) New

York: Holt.

Kanfer, F. H., & Phillips, F. S. (1970). Learning foundations

of behavior therapy. New York: Wiley. (After Kantor’s death
a signed presentation copy of this well known book was
found in Kantor’s personal library: from Frederick Kanfer to
J. R. Kantor.)

Kantor, J. R. (1917). The functional nature of the

philosophical categories. Unpublished Doctoral Dissertation.
The University of Chicago.

182
Kantor, J. R. (1918). Conscious behavior and the abnormal.
Journal of Abnormal Psychology, 13, 158-167.

Kantor, J. R. (1919). Human personality and its pathology.

Journal of Philosophy, Psychology, Scientific Method, 16,
236-246.

Kantor, J. R. (1920) Intelligence and mental tests. Journal of

Philosophy, Psychology, Scientific Method. 17, 260-268.

Kantor, J. R. (1921). An objective interpretation of meanings.

American Journal of Psychology, 32, 231-248.

Kantor, J. R. (1922). An analysis of psychological language

data. Psychological Review, 29, 267-309.

Kantor, J. R. (1923). Does psychology need a new conception

of personality? [Abstract]. Psychological Bulletin, 20, 80-81.

Kantor, J. R. (1924; 1926). Principles of psychology (Vols.

1-2). New York: Alfred A. Knopf. (Reissued by Principia
Press, 1949. See especially the chapter on Abnormal
Reactions and Psychopathic Personalities.)

Kantor, J. R. (1933a). A survey of the science of psychology.

Bloomington, IN: Principia Press. (His first introductory text.)

Kantor, J. R. (1933b). In defense of stimulus-response

psychology. Psychological Review, 40, 324-336.

Kantor, J. R. (1936). An objective psychology of grammar.

Bloomington, IN: Indiana University. Reissued by Principia
Press, 1952.

183
Kantor, J. R. (1945; 1950). Psychology and logic (Vols. 1-2).
Bloomington, IN: Principia Press.

Kantor, J. R. (1947). Problems of physiological psychology.

Bloomington, IN: Principia Press. (See especially for his
pithy reactions to, and demolitions of, the specious and
conventional arguments for the superiority of the human CNS
to that of non-human animals, and against the construct of
centers in the brain [read powers of the Soul.])

Kantor, J. R. (1953). The logic of modern science.

Bloomington, IN: Principia Press.

Kantor, J. R. (1959). Interbehavioral psychology (2nd ed.).

Bloomington, IN: Principia Press. (See especially Chapter 17,
Applied Subsystems, the Subsystems of Psychotechnology,
Educational Psychology, and Clinical Psychology, pp.
170-178.)

Kantor, J. R. (1963; 1969). The scientific evolution of

psychology (Vols. 1-2). Chicago and Granville, OH: The
Principia Press.

Kantor, J. R. (1969). Scientific psychology and specious

philosophy. The Psychological Record, 19, 15-27;395-312.

Kantor, J. R. (1970). An analysis of the experimental analysis

of behavior (TEAB). Journal of the Experimental Analysis of
Behavior, 13, 101-108.

Kantor, J. R. (1976). The origin and evolution of

interbehavioral psychology. Mexican Journal of Behavior
Analysis, 2, 120-136.

184
Kantor, J. R. (1977). Psychological linguistics. Chicago, IL:
Principia Press.

Kantor, J. R. (1978). Experimentation: The acme of science.

Mexican Journal of Behavior Analysis, 4, 5-15.

Kantor, J. R. (1981). Interbehavioral philosophy. Chicago,

IL: Principia Press.

Kantor, J. R. (1987). What qualifies interbehavioral

psychology as an approach to treatment? In D. H. Ruben & D.
J. Delprato (Eds.), New Ideas in Therapy: Introduction to an
interdisciplinary approach. Westport, CT: Greenwood Press.

Kantor, J. R., & Smith, N. W. (1975). The science of

psychology: An interbehavioral survey. Chicago, IL: Principia
Press.

Kazden, A. E. (1978). History of behavior modification.

Baltimore, MD: University Park Press.

Kempermann, G., Kuhn, H. G., & Gage, F. H. (1997). More

hippocampal neurons in adult mice living in an enriched
environment. Nature, 386, 493-495.

Koziol, L. F., Stout, C. E., & Ruben, D. H. (Eds.). (1993)

Handbook of childhood impulse disorders and ADH: Theory
and practice. Springfield, IL: C. C. Thomas.

Leyton, A. S. F., & Sherrington, C. S. (1917). Observations

on the excitable cortex of the chimpanzee, orangutan, and
gorilla. Quarterly Journal of Experimental Physiology, 11,
135-222. (Reports on 22 chimpanzees, three orangutans, and

185
three gorillas, whose motor cortices were mapped by
electrical stimulation, portions ablated, and then allowed to
recover from the surgery. Upon re-examination, recovery of
function was recorded. All experimental subjects were quite
young, which is congruent with the report of Vining, et al
(1997) concerning recovery of function in young human
beings following hemispherectomy.)

Lewontin, R. C., Rose, S., & Kamin, L. (1984). Not in our

genes: Biology, ideology, and human nature. New York:
Pantheon.

Lichtenstein, P. L. (1983). The interbehavioral approach to

psychological theory. In N. W. Smith, P. T. Mountjoy, &
Rubin, D. H. Reassessment in psychology. Washington, DC:
University Press of America.

Lichtenstein, P. L. (1984). Interbehaviorism in psychology

and in the philosophy of science. The Psychological Record,
34, 455-475.

Louttit, C. M., & Carter, J. W., Jr. (1939). The

psychodiagnostic blank. Indiana University Psychological
Clinics, 2(7), 12.

Lundin, R. W. (1961). Personality, an experimental

approach. Toronto, Ontario: Macmillan.

Lundin, R. W. (1965). Principles of psychopathology.

Columbus, OH: C. E. Merrill Books.

Lundin, R. W. (1969). Personality: A behavioral analysis.

Toronto, Ontario: Macmillan.

186
Marr, M. J. (1990). Behavioral pharmacology: Issues of
reductionism and causality. In J. E. Barrett, T. Thompson, &
P. B. Dews (Eds.), Advances in behavioral pharmacology.
Hillsdale, NJ: Lawrence Erlbaum Associates, Publishers.

Mayr, E. (1982). The growth of biological thought: Diversity,

evolution, and inheritance. Cambridge, MA: Harvard
University Press.

Mayr, E. (1997). This is biology. Cambridge, MA: Harvard

University Press.

Mazurs, E. G. (1957). Graphic representations of the periodic

system during one hundred years. University, AL: The
University of Alabama Press. (Second, revised edition, 1974.)

Meyer, A. (1934). The psychobiological point of view. In M.

Bentley & E. V. Cowdry (Eds.), The problem of mental
disorder. New York, McGraw-Hill.

Meyer, M. (1911). The fundamental laws of human behavior:

lectures on the foundation of any mental or social science.
Boston: Richard G. Badger, The Gorham Press.

Meyer, M. (1922). Psychology of the other-one: An

introductory text-book of psychology. Columbus, MO: The
Missouri Book Company.

Meyer, M. (1927). Abnormal psychology: When the other-one

astonishes us. Columbia, MO: Lucas Brothers.

187
Morris, E. K. (1982). Some relationships between
interbehavioral psychology and radical behaviorism.
Behaviorism, 10, 187-216.

Mountjoy, P. T. (1957). Differential behavior in monozygotic

twins. The Psychological Record, 7, 65-69.

Mountjoy, P. T. (1976). Science in psychology: J. R. Kantor’s

field theory. Revista Mexicana de Analisis de la Conducta, 2,
3-21.

Mountjoy, P. T. (1980). An historical approach to

comparative psychology. In. M. R. Denny (Ed.), Comparative
psychology (pp. 128-152). New York: John Wiley & Sons.

Mountjoy, P. T. (1983). A history of psychological

technology. In N. W. Smith, P. T. Mountjoy, & D. H. Ruben,
(Eds.), Reassessment in psychology, Washington, DC:
University Press of America, Inc.

Mountjoy, P. T. (1987). The first systematic account of

comparative avian behavior. In E. Toback (Ed.), Historical
perspectives and the international status of comparative
psychology (pp. 5-14). Hillsdale, NJ: Lawrence Erlbaum
Associates, Inc.

Mountjoy, P. T. (in press a.) Biographical Entry; Kantor,

Jacob Robert. American National Biography. New York:
Oxford University Press.

Mountjoy, P. T. (in press b). Biographical Entry; Kantor,

Jacob Robert. The encyclopedia of psychology. Washington,
DC: The American Psychological Association.

188
Mountjoy, P. T., & Hansor, J. D. (1986). Jacob Robert Kantor
(1888-1984). American Psychologist, 41, 1296-1297.

Mountjoy, P. T., & Ruben, D. H. (1983). Behavioral genesis:

Readings in the science of child psychology. Lexington, MA:
Ginn Custom Publishing. (Used in Psychology 160, Child
Psychology, Western Michigan University.

Mountjoy, P. T., & Cone, D. M. (1995). The functional nature

of the philosophical categories: Jacob Robert Kantor’s
doctoral dissertation. The Interbehaviorist, 23, 5-8.

Mountjoy, P. T., & Cone, D. M. (1997). Another new

journal? The Psychological Record: Volumes I-V;
1937-1945. The Psychological Record, 47, 3-20.

Mountjoy, P. T., & Cone, D. M. (in press). A biographical

sketch of Jacob Robert Kantor. In Morris, E. K. & Midgley,
B. D. (Eds.), Modern perspectives on J. R. Kantor and
interbehaviorism. Westport, CT: Greenwood Press.

O’Donnell, J. M. (1985). The Origins of Behaviorism:

American Psychology, 1870-1920. New York: New York
University Press.

Parrot, L. J. (1983). Complex behavior: A systematic

reformation of radical behavioral analysis. Unpublished
Doctoral Dissertation. Western Michigan University. (As was
the case with Kantor’s (1917) dissertation, she presents a
program for making psychology into a natural science by
analyzing the fundamental problem of the linguistic behaviors
of scientists and philosophers.)

189
Parrot, L. J. (1984). J. R. Kantor’s contributions to
psychology and philosophy: A guide to further study. The
Behavior Analyst, 7, 169-81.

Popplestone, J. A., & McPherson, M. W. (1994/1999). An

illustrated history of American psychology. Madison, WI:
Brown & Benchmark, Publishers. (Second Edition, 1999,
Akron, OH: The University of Akron Press.)

Plomin, R. (1994). Genetics and experience: The interplay

between nature and nurture. Thousand Oaks, CA: Sage
Publications.

Pronko, N. H. (1946). Language and psycholinguistics: A

review. Psychological Bulletin, 43, 189-239.

Pronko, N. H. (1980). Psychology from the standpoint of an

interbehaviorist. Monterey, CA: Brooks/Cole Publishing Co.

Pronko, N. H., & Hill, H. (1949). A study of differential

stimulus function in hypnosis. The Journal of Psychology, 27,
49-53.

Pronko, N. H., & Bowles, J. W. (1951). Empirical

foundations of psychology. New York: Rinehart & Company,
Inc. (With the collaboration of D. T. Herman, H. Hill, & J.
Bucklew, Jr.)

Pronko, N. H., & Herman, D. T. (1982). From Dewey’s reflex

arc concept to transactionalism and beyond. Behaviorism, 10,
229-54.

190
Ratner, S. C. (1957). Toward a description of language
behavior: I. The speaking action. The Psychological Record,
7, 61-64.

Ratner, S. C., & Rice, F. E. (1963). The effect of the listener

on the speaking interaction. The Psychological Record, 13,
265-268.

Ratner, S. C., Gawronski, J. J., & Rice, F. E. (1964). The

variable of concurrent action in the language of children:
Effects of delayed speech feedback. The Psychological
Record, 14, 47-56.

Ribes, E. (1984). Obituario: J. R. Kantor (1888-1984). Revista

Mexicana de Analisis de la Conducta, 10, 15-36.

Rose, S. (1997). Lifelines: Biology beyond determinism. New

York: Oxford University Press.

Rotter, J. B. (1942). A working hypothesis as to the nature

and treatment of stuttering. Journal of Speech Disorders, 7,
263-288.

Rotter, J. B. (1954). Social learning and clinical psychology.

Englewood Cliffs, NJ: Prentice-Hall.

Ruben, D. H. (1983) The validation of a behavioral

programmed text for increasing self-control attitudes.
Unpublished MA Thesis, Western Michigan University.

Ruben, D. H. (1984a). Drug abuse and the elderly: An

annotated bibliography. Metuchen, NJ: Scarecrow Press.

191
Ruben, D. H. (1984b) Major trends in interbehavioral
psychology from articles published in The Psychological
Record 1937-1983). The Psychological Record, 34, 589-617.

Ruben, D. H. (1985a). Philosophy journals and serials: An

analytic guide. Westport, CT: Greenwood Press.

Ruben, D. H. (1985b). Progress in assertiveness, 1973-1983:

An annotated bibliography. Metuchen. NJ: Scarecrow Press.

Ruben, D. H., & Delprato, D. J. (Eds.). (1987). New ideas in

therapy: Introduction to an interdisciplinary approach.
Westport, CT: Greenwood Press.

Schoenfeld, W. N. (1969). J. R. Kantor’s Objective

psychology of grammar and psychology and logic: A
retrospective appreciation. Journal of the Experimental
Analysis of Behavior, 12, 329-47.

Schoenfeld, W. N. (1974). Notes on a bit of psychological

nonsense: “Race differences in intelligence.” The
Psychological Record, 24, 17-32. (His Presidential Address to
the Eastern Psychological Association, 1937. Compatibility
with Kantor is obvious, even though he is not cited in the
references.)

Schoenfeld, W. N. (1993). Religion and human behavior.

Boston: Authors Cooperative, Inc. (Written in about four
months in the Spring of 1971, it was distributed in manuscript
form as a “test edition” in 1982, see prefaces, pp vi-xxvi. The
volume may be interpreted on many levels as it is largely
written in the vernacular.)

192
Sherrington, C. S. (1906). The integrative action of the
nervous system. New Haven: Yale University Press.

Skinner, B. F. (1938). The behavior of organisms. New York:

Appleton-Century-Crofts.

Skinner, B. F. (1957). Verbal behavior. New York:

Appleton-Century-Crofts.

Smith, N. W. (1976a). Twin studies and heritability. Human

Development, 19, 65-68.

Smith, N. W. (1976b). Longitudinal personality comparison

in one pair of identical twins. JSAS Catalog of Selected
Documents in Psychology, 6(4), 106.

Smith, N. W. (1982). Brain, behavior, and evolution. The

Psychological Record, 32, 483-490.

Smith, N. W. (1984). Fundamentals of interbehavioral

psychology. The Psychological Record, 34, 479-494.

Smith, N. W. (1990/1993). Greek and interbehavioral

psychology: selected and revised papers of Noel W. Smith
(2nd rev. ed. 1993). Lanham, MD: University Press of
America. (Contains a complete list of Kantor’s publications,
as well as a citation study of references to his works.)

Smith, N. W., & Shaw, N. E. (1979). An analysis of

commonplace behaviors: Volitional acts. The Psychological
Record, 29, 179-186. (Describes an interbehavioral approach
to behavior therapy.)

193
Smith, N. W., Mountjoy, P. T., & Ruben, D. H. (Eds.).
(1983). Reassessment in psychology: The interbehavioral
alternative. Washington, DC: University Press of America.
(This is a Festschrift for Kantor, and includes contributions
both by his students and also by individuals who were not
students in the formal sense, but were influenced by his
works, and sometimes engaged in correspondence and other
personal interactions with him. Although marred by numerous
typographical errors due to hasty production, for which
Mountjoy takes full responsibility, it was delivered into
Kantor’s hands by Mountjoy and Ruben late in 1983, shortly
before Kantor’s death in early 1984.)

Smith, N. W., & Smith, L. L. (1996). Field theory in science:

Its role as a necessary and sufficient condition in psychology.
The Psychological Record, 46, 3-19.

Stephenson, W. (1984). Methodology for statements of

problems: Kantor and Spearman conjoined. The
Psychological Record, 34, 575-588.

Swartz, P. (1963). Psychology: The study of behavior.

Princeton, NJ: D. Van Nostrand Company, Inc. (An
interbehavioral introductory text which has many interesting
features.)

Turner, J. (1985). Without God, without creed: The origins of

unbelief in America. Baltimore, MD: The Johns Hopkins
University Press.

Ullmann, L. P., & Krasner, L. (1969/1975). A psychological

approach to abnormal behavior, Englewood Cliffs, NJ:
Prentice-Hall.

194
Van Praag, H., Kempermann, G., & Gage, F. H. (1999).
Running increases cell proliferation and neurogenesis in the
adult mouse dentate gyrus. Nature Neuroscience, 2, 266-270.

Verplanck, W. S. (1983). Preface. In N. W. Smith, P. T.

Mountjoy, & D. H. Ruben (Eds.), Reassessment in
psychology. Washington, DC: University Press of America.

Vining, E. P. G., Freeman, J. M., Pillas, D. J., Uematsu, S.,

Carson, B. S., Brandt, J., Boatman, D., Pulsifer, M. B., &
Zuckerberg, A. (1997). Why would you remove half a brain?
The outcome of 58 children after hemispherectomy—the
Johns Hopkins experience: 1968 to 1996. Pediatrics, 100,
163-171.

Wahler, R. G., & Dumas, J. E. (1989). Attentional problems

in dysfunctional mother-child interactions: An interbehavioral
model. Psychological Bulletin, 105, 116-130. (Reviews about
100 empirical and theoretical papers. Concludes that
environmental stressors, classifiable as interbehavioral setting
factors, are important variables.)

Walker, E. L. (nd., ca. 1993). CHAFF. Mill Creek, WA:

Hedgehog Press. (Walker’s birth & death dates are
1914-1997. He was long time Professor at the University of
Michigan, and a member of the board of editors of The
Psychological Record. These reminiscences were privately
printed and distributed).

Weiner, J. (1999). Time, love, memory: A great biologist and

his quest for the origins of behavior. New York: Alfred A.
Knopf. (A Pulitzer Prize winning author of popular scientific

195
works describes Seymour Benzer’s reduction of complex
behavior to genetic components.)

Weiss, A. P. (1925/1929). A theoretical basis of human

behavior. Columbus, OH: Adams. (Ranks Hunter, Kantor,
and Lashly as the three eminent behaviorists of that time.)

White, A. D. (1896/1955). A history of the warfare of science

with theology in Christendom. London: Macmillan and
Company, and New York: D. Appleton & Company. (About
25 editions and printings up to 1955. The most convenient
edition is the reissue of 1955, New York: Braziller. Because
of its exhaustive treatment of the historical record it remains
the definitive source for this topic. White was soon to be the
first president of Cornell University, the first American
university founded upon sectarian rather than secular
principles.)

Whitehead, A. N., & Russell, B. (1910-1912/1925-1927).

Principia mathematica (Vols. 1-3). Cambridge: Cambridge
University Press. (Second Ed., 1925-1927.)

Wilson, E. O. (1998). Consilience: The unity of knowledge.

New York: Alfred A. Knopff.

Wolf, I. S. (1958a). Stimulus variables in aphasia: I. Setting

conditions. Journal of the Scientific Laboratories, Denison
University, 44, 203-217.

Wolf, I. S. (1958b). Stimulus variables in aphasia: II.

Stimulus objects. Journal of the Scientific Laboratories,
Denison University, 44, 218-228. (These two papers of Wolf

196
indicate the important roles of environmental conditions upon
the behaviors of brain injured, aphasic, patients.)

Wolf, I. S. (1984). J. R. Kantor, 1888-1984. The

Psychological Record, 34, 451-453. (This obituary is in the
fall issue, pp 448-634, which is a special issue devoted to the
commemoration of Kantor’s scholarly contributions to the
natural science of psychology; includes his founding of The
Psychological Record in 1937, and his continuing association
with its editorial policies until his death.)

Wynne, L. C., Cromwell, R. L., & Matthysse, S. (Eds.).

(1978). The nature of schizophrenia; New approaches to
research and treatment. New York: John Wiley & Sons.

Zimmerman, D. W. (1979). Quantum theory and

interbehavioral psychology. The Psychological Record, 29,
473-485.

Zimmerman, D. W. (1982). The universe — an unscientific

concept. The Psychological Record, 32, 337-347.

197
Acknowledgments
My long term friends and colleagues Dr. Donna M. Cone, Dr.
Dennis Delprato, and Dr. Noel W. Smith have read an early
draft of this paper and their comments have contributed
greatly to its current status. All errors remain, of course, my
own responsibility. Dr. Howard E. Farris, Chair, Department
of Psychology, Western Michigan University, and Dr. R. R.
Hutchinson, President of The Foundation for Behavioral
Resources, have provided logistical support in terms of office
space, computer access, photocopying and mailing privileges.

198
Chapter 5

Child Behavior Therapy:

Early History
Sidney W. Bijou1

University of Nevada, Reno

Introduction
In keeping with the purpose of this conference, I will focus
my remarks on the early history of child behavior therapy
considered in its broadest sense. To this end I will present
them in terms of the four topics suggested by William
O’Donohue: (1) my intellectual biography, (2) my perception
of the important developments in the rise of child behavior
therapy, (3) detailed case studies of selected publications, and
finally, (4) possible object lessons learned for the future.

Intellectual Biography
In reviewing my rather long history, I find that my intellectual
biography with respect to the early history of child behavior
therapy easily falls into two phases. The first pertains to my
training and experience as a clinical psychologist, the second,
to my training and experiences as a child psychologist
focusing on the psychopathologies.

199
Phase One

The first phase began when I enrolled as a graduate student in

psychology at Columbia University in 1935. At that time the
department was proudly theoretically eclectic with a faculty
consisting, among others, of Henry E . Garrett, Elizabeth
Hurlock, Otto Klineberg, A. T. Poffenberger, A. J. Warden,
and Robert S. Woodworth. Two events outside of my courses
moved me in the direction of child clinical psychology: One,
a self-selected internship in mental retardation, the other my
thesis research.

The internship consisted of a summer at Letchworth Village,

a large residential institution for severely retarded persons,
which served families and social agencies mostly in the New
York City area. Sponsored and supported by the New York
State Department of Mental Health, the purpose was to attract
more professionals to the field of mental retardation.
Consequently, students in medicine, dentistry, nursing,
education, social work, and psychology were recruited.

That summer provided a heady experience for me in that I

had many contacts with the members of the staff and
particularly the head psychologist, Elaine F. Kinder. It also
afforded me an opportunity to learn more about the problems
encountered in treating low-functioning persons in a large
institution as well as a chance to interact with students in
allied professions. But it was disappointing to see how limited
the contributions of the clinical psychologists were, their only
function being to administer psychological tests (mainly
intelligence tests) to help the medical staff diagnose and
classify the residents. Their training in areas other than test
administration was completely wasted.

200
Incidentally, Uri Bronfenbrenner, known particularly for his
ecological child development (e.g., Bronfenbrenner, 1992),
grew up in the environment of Letchworth Village where his
father was a well-respected research pathologist.

The second event that nudged me further in the direction of

clinical psychology — my thesis research — was in fact a
compromise. After reading two books by John B. Watson
(1919, 1930), I became intrigued with his views and decided
that I would like to do a behavioral experimental study with
children. In my enthusiasm and naivete, I wrote to Watson,
then vice president of the J. Walter Thompson Advertising
Company, telling him of my interest in the behavioral
approach to child development and asking him if he would be
willing to suggest a topic for my master’s degree research. He
replied promptly with the suggestion that I study how young
children learn “muscle sense.” How, for example, does little
Jimmy know that his arms are stretched out at shoulder height
when his eyes are closed? He mentioned also that for subjects
I might contact Patty Hill-Smith, director of the model
kindergarten at the nearby Horace Mann School at
Columbia’s Teachers College. Although this kind of problem
appealed to me, I could not find a member of the psychology
department who was sufficiently interested to serve as my
advisor. I was sorely disappointed but to progress in my
graduate studies I undertook to complete a project begun by a
faculty member, Louise E. Poole, with whom I had been
taking a course. This project involved the measurement of
nonverbal intelligence in young retarded children (Bijou,
1938), thus giving me an opportunity to test many retarded
children and to learn more about the measurement literature.

201
After receiving my degree in 1937, I took a position as
clinical psychologist at the Delaware State Mental Hygiene
Clinic on the campus of the Delaware State Hospital, near
Wilmington. The Clinic provided diagnostic services to adult
patients in the hospital and to children and adults with mental
health problems throughout the small state of Delaware. My
job was to administer psychological tests and prepare reports
for the psychiatrists to help them make diagnoses and
recommendations.

The upside of the appointment was that for the first time I was
exposed to patients of all ages and covering the entire range
of mental health problems, since the Clinic served all the
social and educational agencies in the state. I administered 33
different types of psychological tests to over 2000 patients.
The downside was similar to that of the clinical psychologists
at Letchworth Village: my duties were largely limited to
testing and report writing.

It was at the Delaware Mental Hygiene Clinic that I worked

with Joseph Jastak, the chief psychologist, to develop a
clinical test of school achievement, the Wide Range
Achievement Test (Jastak & Bijou, 1938). The need for such
a test arose from Jastak’s conviction that all “mental”
diagnoses should be based, not on a single score, but on a
profile of scores including basic school achievement (Jastak,
1934). Because a brief achievement test of reading, writing,
and arithmetic was nonexistent, we constructed one that could
be administered and scored in about fifteen minutes.

In addition to its use by many clinical psychologists, and later

by psychometricians, the Wide Range Achievement Test was

202
found to be a valuable screening measure for young children
nationwide when the Head Start program was launched.

I left Delaware in 1939 to complete my graduate training.

Because of Columbia’s lack of theoretical focus, I decided to
go to a different university. Before leaving Columbia, I
confided my disappointment to one of my professors, George
W. Hartmann, who suggested I consider studying with Kurt
Lewin (1935, 1936), a brilliant neo-Gestaltist who had
recently immigrated from Germany and was now at the
University of Iowa. Acting on Hartmann’s advice, I applied to
and was accepted by the department of psychology and to my
surprise was offered an assistantship. On my arrival in Iowa
City, I learned to my chagrin that Lewin could not serve as
my advisor because he was not a member of the psychology
department; his appointment was in the Child Welfare
Research Station (later the Institute of Child Development).
The chairperson, John McGeoch, informed me that if I
wanted to study psychology and keep the assistantship, I
would have to select an area of concentration in psychology. I
decided on learning since it was closest to my interest in child
clinical psychology. Kenneth Spence was assigned as my
adviser.

I did, though, take a course from Lewin titled, “Theory of

Psychology and Personality Development.” While Lewin was
indeed a stimulating although somewhat disorganized
lecturer, the content of his course was appealing inasmuch as
it dealt with the behavior of children. But what bothered me
was that Lewin’s analysis of a child’s behavior was in terms
of correlated relationships. All the conditions of an interaction
— situation, response, and motivation — were entered into a
“field” on the basis of an individual’s perceptions, hence the

203
relationships among the variables were correlational rather
than functional. My interest was on functional relationships.
Besides, I was losing interest in Lewin’s theory. I was
beginning to believe, through the intensive learning course
with Spence, that the learning theories of Hull, Tolman, and
Guthrie held more promise than Lewin’s for the future of
psychology as a natural science.

The Iowa department at that time (1939-1941) was an

exciting place for graduate students. Spence, armed with
Hullian learning theory, was warring against Lewin and his
field theory: Spence and his students (among them Isador
Farber, Robert Grice, Arthur Irion, Howard and Tracy
Kendler, and Benjamin Underwood) were doing research on
learning and motivation using animals in mazes while Lewin
and his students (among them Leon Festinger, Ronald Lippit,
and Ralph White) were researching the similar topics with
children in socially structured situations.

Among the highlights of my training at Iowa were Spence’s

two-semester course in animal learning and conditioning and
his informal seminar , The Monday Night Group, which was
devoted entirely to an intensive chapter-by-chapter review of
Hull’s manuscript for his Principles of Behavior (Hull, 1943);
Gustav Bergmann’s course in philosophy of science (mostly
logical positivism); E. F. Lindquist’s course in advanced
statistics (mostly small sample research designs); and John
McGeoch’s courses, one in systematic psychology and the
other in human learning (mostly rote memory studies).

For my stipend, I worked in the Department’s clinic as an

assistant to Charles Strother, administering psychological
tests. The only feature of this experience that was new to me

204
was working with many children and adults with speech
problems.

Because of my interest in abnormal behavior, Spence

permitted me to do my thesis on experimental neurosis in rats,
modeled after Pavlov’s studies with dogs . I constructed an
apparatus that held the animal in a stationary position but
allowed it to raise its head to lift a bar in response to a visual
stimulus (Bijou, 1942, 1943). The experimental procedure
followed the instrumental rather than the classical
conditioning model. In this situation, the animals showed
strong emotional behavior when required to make fine visual
discriminations (Bijou, 1951).

By the time I received my degree, I was convinced with Hull

that the objective of the science of psychology was to develop
a general theory of behavior that would account for events
between the stimulus input and the response output of an
organism and that research should be on testing hypotheses to
tease out the properties of the internal variables and their
relationships. Although the theory dealt with the behavior of
individuals, the research typically involved small-sample,
group designs. The fact that, thus far, Hull’s approach was
based on the behavior of nonverbal organisms was not a
matter of concern for him for he believed that the complex
human behavior of verbal organisms would eventually be
incorporated into the system. In this respect, the task ahead
was to redefine Freud’s concepts and principles in
scientifically acceptable ways. Some of Hull’s students had
already begun to do that. Sears and his colleagues, for
example, had been using Freudian concepts in their studies of
children (Sears, Whiting, Nowlis, & Sears, 1953).

205
There was an interlude of about five years — from 1941 to
1946 — between receiving my degree and my first academic
appointment. During that time, I spent two years as a research
psychologist at the Wayne County Training School in
Northville, Michigan, a residential school for high-grade
retarded children, and three years in the military service, first
administering intelligence tests at Induction Centers, then Air
Force aptitude tests for selecting pilots, bombardiers, and
navigators and finally, supervising psychological services at
Convalescent Centers in resorts throughout the country.

Phase Two

The second phase of my experience and training began in

1946 with an academic appointment at Indiana University. I
was recruited by B. F. Skinner, then chairman of the
psychology department, to serve as assistant professor and
director of the newly formed clinical program. The
Department and Skinner wanted a clinical program with an
experimental-learning orientation. My clinical experience in
Delaware and in the military service plus my training in
learning with Spence qualified me for this position.

I took on the first group of graduate students in clinical

psychology with a course in experimental psychopathology
which was saturated with references by Hull’s students,
among them Carl Hovland, Hobart Mowrer, Neal Miller, and
Robert Sears. I also set up an animal laboratory to continue
my thesis research on conflict. Once when I complained to
Skinner that I could not get the differences in approach and
avoidance gradients postulated by Miller (1944), he smiled
and simply said, “The animal is always right.”

206
Indiana was as intellectually exciting as Iowa, although in a
different way. Here, frequent heated discussions took place
among faculty and students over the merits and limitations of
the Skinnerian, Hullian, and Kantorian views. There was a
Skinner-Kantor, standing-room-only, joint seminar plus
informal competing seminars by the students of Skinner and
Hull. At the same time, ground-breaking research was being
done by students, such as Bill Estes, Norman Gutman, Paul
Fuller, and Joel Greenspoon.

During this time, I audited several lectures of J. R. Kantor’s

undergraduate class on psychopathology and had discussions
with him about theory construction. I quickly learned that he
took issue with Hull’s view that psychology should study,
almost exclusively, the presumed variables and events
between stimulus input and response output of an organism.
He argued, furthermore, that Hull’s system was physically
reductionistic and questioned his emphasis on the
hypothetico-deductive method for theory construction and
research, averring that both deductive and inductive methods
are proper tools for psychology as a natural science.

In 1948, when Skinner left to take a permanent appointment

at Harvard, I left to go to the University of Washington where
I was appointed associate professor and Director of the
Institute of Child Development. I was pleased to join the
faculty at Washington for among other attractions (e. g., the
ocean and mountains), I would get to know Edwin Guthrie,
learn more about his learning theory and his research with
George Horton on cats in a puzzle box. Unfortunately,
Guthrie was no longer teaching; he was now vice-president of
the University. But occasionally “just to keep his hands in” he
would give a seminar. I audited one of these rare occasions

207
which consisted of his telling stories, some related to
learning, some not. When students raised questions critical of
his theory, he would laugh and counter with one even more
devastating. The students obviously enjoyed his stories.
Incidentally, the undergraduates good-naturedly referred to
the psychology department as offering only two courses: An
introductory course on the behavior of cats in a puzzle box
and a developmental course on the behavior of kittens in a
puzzle box.

The Institute, which was established in 1910 to provide

state-wide services to children, actually consisted of a
two-room clinic, part of the Psychology Department in Denny
Hall. The responsibility of the small staff was to provide
psychological test services under the supervision of
Stevenson Smith, a student of Guthrie. My conception of a
child institute was rather different. I visualized it as a research
organization consisting of a clinic, a preschool, and a child
study unit. With the support of the Institute’s Board of
Directors, I arranged to move the Institute into the building
which then housed only the University Nursery School. By
1950 the “new” Institute consisted of a Child Development
Clinic, a two-unit Nursery School and a Research Laboratory.

With this expansion I launched what I hoped would be a

research program on children based on Hullian principles, one
similar to what Sears and his colleagues were doing at Iowa
and Stanford. This group was engaged in two types of
research: Correlational group studies of the relationship
between parents and their young children (e.g., Sears,
Whiting, Nowlis, & Sears, 1953) and laboratory studies of
individual children using doll play techniques (e.g., Sears,
1947). I was primarily interested in their studies with

208
individual children. On the basis of a review of the literature,
I decided to carry out a study which had implications for the
frustration-aggression hypothesis (Bach, 1945). The study
would be concerned with the relationship between aggression
in a social situation and fantasy aggression as manifested in a
doll-play situation (fantasy aggression). In planning the
experiment I realized that I would have to obtain teachers’
ratings of children’s aggressiveness, prepare a doll house
simulation of a preschool, find or make a teacher doll and
three children dolls, one being of the opposite sex of the
subject, train two observers to record data from behind a
one-way glass, then analyze their protocols in terms of
frequency, direction and latency of aggression, and train a
student to perform the task of an “experimenter,” the person
who interacts with the subject in the course of an experiment.
All these requirements made me realize that laboratory
research with children in the Hullian tradition was quite
different from animal research which was relatively simple
and included good control over the antecedent and consequent
conditions of an objectively recordable response. The study
was never done. Instead I turned to the only other procedure I
knew — Skinner’s operant conditioning model.

I built a toy-like apparatus, shown in Figure 1, consisting of

three parts: a response mechanism, a trinket dispenser, and a
recording pad. To make a response the child would simply
take a ball from the receptacle and drop it into the hole above.
The ball, a Spaulding handball, would activate the dispenser
to deliver a trinket and then return to the receptacle for the
next response, thus creating a free operant situation with
objective, countable responses, a clearly discernible and
manipulable consequence, and a potentially objective
antecedent condition (Bijou, 1955).

209
Figure 1. The ball dropping apparatus.

With this crude device, I carried out two experiments (Bijou,

1957a) each showing that intermittent reinforcement was
more resistive to extinction than continuous reinforcement in
four -year -boys). However, I was disappointed with the
results because the research design, which involved
comparison of small groups, produced a relatively small
number of responses (the means ranged in the order of 15.3 to
26.2 responses) compared to the large numbers of responses
generally seen in operant infra-human studies. It was apparent
that the design was more appropriate for theory testing than
for demonstrating functional relationships.

For the next study (Bijou, 1957b), I built another apparatus

(Fig. 2) which would allow me to follow operant conditioning

210
principles scrupulously. Because the ball dropping response
on occasion created recording problems (e. g., a child would
miss the hole and go scrambling to retrieve it ), I substituted a
lever (an O-Cedar sponge mop handle) for a response. In
addition, two sophisticated commercial instruments
(Gerbrands) replaced my home-made devices: One for
recording responses and stimulus events, the other for
delivering trinkets as well as other small objects.

Figure 2. The lever-pressing aparatus.

With this set-up, I studied the effect of several fixed-internal

and fixed ratio-schedules of reinforcement on four-year-old
boys and obtained hundred of responses in orderly
relationships to the schedules. Samples of the data obtained
are shown in Figure 3.

211
Figure 3. Cumulated responses of the first, third, and fifth
sessions of two four-year old boys. Sequence A is on a
fixed-interval schedule of reinforcement. Sequence B is on a
fixed-ratio schedule of reinforcement.

On my next trip to the East Coast, I stopped to see Skinner in

Cambridge and showed him some of these and other data. He
was both surprised and pleased. Incidentally, in one of his last
papers, Sears recognized that the externalization of the
reinforcement mechanism (drive) in Skinner’s system helped
the behavioral approach move ahead.

I was now convinced that laboratory studies with children

should be conducted in controlled settings like infra-human
animal studies. To augment the Institute facilities, I built a
mobile laboratory, a converted house trailer, shown in Figure

212
4, that could be easily towed to any nursery school in the
Seattle area where additional studies could be carried out
(Bijou, 1958a).

Figure 4. Mobile child study laboratory.

I also set up a research laboratory, similar to the one at the

Institute, at the nearby Rainier State School for operant
studies with retarded children (Orlando, Bijou, Tyler, &
Marshall, 1960).

During the 1950’s and 60’s I recruited several new faculty

members for the Institute staff, all of whom had training and
experience in experimental studies with animals and/or
children. Included were Donald Baer, who did his thesis
research with kittens as subjects at the University of Chicago

213
under Howard Hunt; Montrose Wolf who studied remedial
reading with Arthur Staats at Arizona State University; Jay
Birnbrauer who carried out learning studies at Indiana
University, and Robert Orlando who did discrimination
learning studies on children with retardation under David
Zeaman at the University of Connecticut.

During this period the Institute attracted a group of highly

talented graduate students, among them, Douglas Kenny, Ivar
Lovaas, Robert Peterson, Donald Pumroy, Shirely Ann
Spence Pumroy, Stephanie Stolz, Todd Risley, James
Sherman, Robert Wahler, and Ralph Wetzel. A note regarding
the training of graduate students at Washington: I had the
unique experience of sponsoring the Ph. D. thesis of the
daughter of the professor who sponsored my Ph. D. thesis.
Shirley Ann Spence Pumroy, the daughter of Kenneth
Spence, carried out an operant study on the effects of
reinforcement on resistance to extinction and emotional
behavior with preschool children (Pumroy, 1954). As far as I
know her father did not object.

Innovative research now flourished in all divisions of the

Institute. From the laboratories came studies by Baer on
escape and avoidance on two schedules of reinforcement
(1960), on the effect of positive reinforcement on
extinguishing responses (1961), on laboratory control of
thumb sucking by withdrawal and re-presentation of
reinforcement (1962a), and behavior avoiding reinforcement
withdrawal (1962b); by Baer, Peterson, and Sherman on the
development of imitation by reinforcing behavior to a model
(1967; by Baer and Sherman on the control of generalized
imitation (1964); by Bijou on patterns of reinforcement and
resistance to extinction in young children (1957a), on operant

214
extinction after fixed-interval schedules (1958b), on
discrimination as a baseline for individual analysis of young
children (1961); by Bijou, Lovaas, and Baer on experimental
procedures for analyzing the interaction of symbolic social
stimuli and children’s behavior (1965); by Bijou and Oblinger
on responses of normal and retarded children as a function of
the experimental situation (1960); by Bijou and Orlando on
single and multiple schedules of reinforcement with retarded
children (1960, 1961); and by Bijou and Sturges on positive
reinforcers for experimental studies with children —
consumables and manipulables (1959).

From the nursery school came research after the teachers

witnessed the results of a study which showed the power of
the application of behavior principles. The director of the
nursery school had confided in Montrose Wolf that her
teachers had been unsuccessful in eliminating regressive
crawling of a three-and-a-half -year-old girl even though they
were giving her attention and support when she was crawling
or sitting on the floor or ground on the assumption that the
behavior was a symptom of stress. Wolf suggested that they
do a little study of the problem in which they would give the
child attention and support when she was standing and
walking and would planfully ignore her when she was
crawling or sitting on the floor or ground. Reluctantly, she
and the teachers followed his suggestion and in a short time
the crawling and floor sitting were eliminated. A brief
reversal of conditions convinced them that the intervention
was indeed effective (Harris, Johnston, Kelley, & Wolf,
1964). Working with members of the research staff, the
nursery school staff then used the same method and procedure
to eliminate operant crying and whining, isolate play,
excessive passivity and excessive aggression and the

215
development of motor skills (Hart, Allen, Buell, Harris &
Wolf, 1964; Harris, Wolf, & Baer, 1964: Hart, Reynolds,
Baer, Browley, & Harris, 1968); Johnston, Kelley, Harris, &
Wolf, 1966; Sloane, Johnston, & Bijou, 1967).

From the Clinic came studies by Wahler on child-child

interactions in free field settings (1967), infant-mother
interactions (1969a), and oppositional children (1969b), and
by Wahler, Winkel, Peterson, and Morrison on procedures for
training mothers to serve as therapists for their own children
(1965). A spin-off of this series was one in which a mother
was trained in her home to be a therapist for her four-year-old
boy with serious aggressive behavior (Hawkins, Peterson,
Schweid, & Bijou, 1966). And finally, came the now classic
clinical study by Wolf, Risley and Mees (1964) on a
three-and- a -half -year-old, hospitalized, “schizophrenic” boy
who needed training to wear his glasses to compensate for the
surgical removal of cataracts.

These studies and their applications provoked discussions

about the role of norm-reference and personality tests in child
behavior therapy. It was noted that the investigators and
practitioners did not use intelligence or projective tests, did
not take into account psychiatric diagnoses, and did not make
predictions about the outcome of treatment . All true. These
previously almost mandatory practices were unnecessary
because treatment was not linked with psychiatric diagnostic
categories. Needed were only reliable assessment techniques
for evaluating initial behavior repertoires, treatment progress,
and follow-up. (Bijou & Peterson, 1971; Marholin & Bijou,
1978) It should be noted, nevertheless, that criterion-reference
tests, such as checklists, were useful in many instances for
planning treatment and academic programs.

216
In 1961 I spent a sabbatical year at Harvard with Skinner on a
Senior National Institute of Mental Health Fellowship. I set
up a laboratory at the Fernald State School near Cambridge to
study complex discriminative behavior in retarded children.
Using a match-to-sample apparatus developed by James
Holland, I explored the programming of antecedent stimuli to
facilitate the learning of right-left visual discrimination
(Bijou, 1968). I also audited Skinner’s large undergraduate
lecture course which was the basis of his Science and Human
Behavior (Skinner, 1953) and his weekly meetings of the
“Pigeon Staff”, an informal seminar for psychologists in and
around Harvard for displaying and discussing data from their
research with pigeons. Each session ended on a high note
with a round of beer. And I spent considerable time in
Skinner’s Teaching Machine Project (Skinner, 1968), learning
principles and techniques of programmed instruction.

While at Harvard I visited several special classes and

programs for young retarded children in the Northeast. All
seemed uninspired and relatively ineffectual. That impression
together with my exposure to programmed instruction led me,
on my return to Seattle, to construct an experimental
classroom to individualize instruction for retarded children in
residence at the Rainier State School. I recruited two
behaviorally trained teachers — Cecilia Tague and John
Kidder — and arranged for Montrose Wolf and Jay
Birnbrauer to serve as consultants. Together they developed
both a motivational system based on a token economy and a
programmed curriculum for teaching reading, writing,
arithmetic, and related subjects. The children attending the
classes were considered serious behavior problems with IQ’s
in the low 60’s. All not only made good progress, but their
classroom problem behaviors were eliminated, and they

217
actually enjoyed coming to class (Bijou, Birnbrauer, Kidder,
& Tague, (1966). After the project was terminated, the
reading program, now known as the Edmark Reading
Program, was adapted for use with a computer touch- screen
and is distributed by IBM ( International Business Machine
Company).

As the research of the Institute became nationally known

(e.g., Kazdin, 1978), faculty members began to receive
attractive offers from other universities. Since the majority of
the members of the Psychology Department did not
appreciate their accomplishment (many even disparaged their
work as “unscientific”), they were unwilling to match their
offers. As a result Baer, Wolf, Birnbrauer, and Orlando soon
departed. Without the wholehearted support of the
Department, it was difficult to recruit replacements. However,
I was able, fortunately, to entice Howard Sloane to leave the
Johns Hopkins School of Hygiene and Public Health and join
the staff. The continued negative attitude of the Department
led Howard and me to accept offers from the University of
Illinois in 1965.

At Illinois I established the new Child Behavior Laboratory

which consisted of six research rooms and two experimental
classrooms for teaching and treating young handicapped
children. Robert Peterson and Thomas Sajway were added to
the staff in 1966 and research continued in the Laboratory
(Peterson, 1968; Peterson, Cox, & Bijou, 1971; Sajway,
Twardosz, & Burke, 1972; Sloane & MacAulay, 1968) and in
the home (Peterson & Peterson, 1968; Zeilberger, Sampen, &
Sloane, 1968). Among the talented graduate and post-doctoral
students were Richard Amato, Tadashi Azuma, Leroy Ford,
Arthur Miller, Edward Morris, Susan O’Leary, Joseph

218
Parsons, Ely Rayek, Emilio Ribes, Howard Rosen, Grover
Whitehurst, Barbara Wilcox, and Koaru Yamaguchi.

In the late 1960’s I began to disseminate our knowledge in

other countries. Most of my effort in this endeavor was
devoted to psychologists and educators in Mexico and Japan.
My venture into Mexico began with an invitation to give talks
and workshops at the University of Veracruz at Xalapa and at
the National Autonomous University of Mexico in Mexico
City in 1967. In order to spread “the word” to larger
audiences, Emilio Ribes, Professor of Psychology at the
University of Veracruz at Xalapa, persuaded his department
to appoint him and me as co-presidents of a new organization,
The International Symposium on Behavior Modification, to
sponsor conferences throughout Mexico and South America.
At each meeting we arranged to have six to eight
psychologists present their research and describe service
programs on a topic selected to be of central interest by the
local psychologists. In the ten years, between 1971 and 1981,
13 well-attended conferences were held in Xalapa, Mexico
City, Caracas, Panama City, Bogota, Lima, and Brasilia,
covering such topics as standards for paraprofessional
training, early education, delinquency and aggression,
self-control, and various clinical issues.

My dissemination activities in Japan began after I had

participated in the 1972 meeting of the International Congress
of Psychology held in Tokyo. Kaoru Yamaguichi, Professor
of Special Education, Tokyo Gakugei University, invited me
to give an address in Tokyo and Tadashi Azuma, Professor of
Psychology at the University of Soporo to give one in
Hokaido. During the following years, I gave lectures in
Nagasaki, Kumonto, Papu, Osaka, and other cities. Then in

219
1980, Yamaguchi arranged for me to attend a meeting of the
Personnel Committee of the Asian Conference on Mental
Retardation which was concerned with plans for the future
treatment of mentally retarded children. With Yamaguchi’s
support, I advised the Committee to invest their countries’
limited resources to training teachers in the Portage Project, a
behaviorally based teaching method, so that they would be
able to train parents to treat their young retarded children in
their homes or in community centers rather than dissipate
their funds on building large institutions. Yamaguichi himself
took action by forming the Japan Portage Association,
translating and the Portage Program into Japanese, and
adapting it for his culture. The Association now has 34
chapters throughout the country. Among them, 28 chapters
provide individualized teaching of mothers for about 300
retarded infants and young children; the other six chapters
carry out parent training in small groups called Day Care
Centers. In Tokyo, about 70 handicapped infants and young
children and their parents come to the headquarters facility
once or twice a week for individualized instruction and
Portage teachers home-visit about 40 families and teach
parents of about 60 special children in Day Care Centers.

Our recommendation to the Personnel Committee with

respect to Asian countries was largely implemented by the
establishment of the Portage International Portage
Association in 1980 by Yamaguichi and Portage leaders from
England and the United States. This organization, which
meets biannually, has been instrumental in having the Portage
Model translated into 34 languages and used in about 90
countries, many of which have a national Portage association.

220
Important Developments in the Rise of
Child Behavior Therapy
Looking back over my experiences, I believe there were
several indications of the rise of child behavior therapy. The
first was a study by a graduate student at Indiana University
(Fuller, 1949) showing that the behavior of a profoundly
low-functioning institutionalized patient, who, according to
the hospital staff, was unable to learn, could in fact learn to
increase his arm movements when they were followed by a
mixture of warm milk and sugar and to decrease them when
the contingency was withheld. This simple demonstration
impressed not only the hospital staff but also many
psychologists because never before had operant principles
been shown to be effective with a human being.

Another indicator was a study by Greenspoon (1955) in

which he created an experimental situation parallel to the
operant conditioning animal model and demonstrated that the
verbal behavior (saying words) of college students could be
manipulated by the experimenter’s delivery of contingent
verbal behavior — “uh-hum” — and visual and auditory
stimuli. This finding had a general impact similar to the
Rheingold, Gewirtz, and Ross study (1959) which showed
that human vocalization and mild physical touching can
function as reinforcers for normally developing infants. It also
had a specific impact on the then ongoing controversy
between Rogers and Skinner (1956) as to whether the
therapist is influential in guiding a client’s conversation.

Still another harbinger were the studies by Ferster and

DeMyer (1961,1962) showing that young autistic children

221
could learn simple responses to vending-type machines in a
laboratory setting. The fourth indicator was the previously
mentioned clinical study by Wolf, Risley, and Mees (1964) in
which operant principles were applied to the treatment of a
hospitalized three-and-a-half-year-old “schizophrenic” boy
with a serious visual handicap. This was indeed a remarkable
study in that the team had not only to create the treatment
programs but also had to train the attendants, nurses, and
parents to carry out the programs and to record data in a form
amenable to a quantitative presentation. This study pointed
out the route to the future that child behavior therapy would
take.

Case Studies of Publications

Of all the publications that flowed from the work at

Washington and Illinois, two appear to have had the greatest
impact on psychologists and students: One was the
methodology for field studies (Bijou, Peterson, & Ault, 1968;
Bijou, Peterson, Harris, Allen, & Johnston, 1969); the other a
behavior analysis of child development (Bijou & Baer, 1961).

Methodology for Field Studies

The methodology for studying children’s behavior in natural

settings was an expansion and adaptation of the laboratory
operant conditioning method. It developed with the
previously mentioned study in which the regressive crawling
of a preschool girl was quickly eliminated (Harris, Wolf, &
Baer, 1964). The method used was refined and expanded with
the many subsequent studies in the nursery school and clinic.

222
In addition to the development of a field methodology, these
studies had two powerful side effects. One was that the staff’s
over-arching view of child development and child
management was changed from the Freudian-Rogerian
position to the behavioral approach (Allen & Goetz, 1982;
Goetz & Allen, 1983). The other was that these studies served
to redefine the activities of the staff: They were not only
teachers; they were now researchers as well. One of the
nursery school teachers, Betty Hart, was so inspired by the
research that after taking a Ph. D. degree at the University of
Kansas, she carried out a series of studies on the role of
verbal behavior in the development of children and finally
embarked on a multi-year, pioneering study with Todd Risley
on how children learn to talk in their home settings (Hart &
Risley, 1995, 1999).

Books on Child Development

The other publication that had the most telling effect on

psychologists — the small paper-back book by Baer and me
— was intended for our large undergraduate courses in child
development which both of us were teaching at the time.
Dissatisfied with the textbook (Mussen, Conger, & Kagan,
1956) we had been using, we wanted one that would present
child development from a behavioral point of view and would
also serve as a background for the research being carried out
at the Institute.

Our plan was to write a series, the first being a summary of

the basic principles, the others, descriptions of their
application to the successive stages of development. Hence,
the second volume (Bijou & Baer, 1965) presents and
describes the application of principles to the prenatal,

223
neonatal, and infancy stages, and the third to the early
childhood stage (Bijou, 1976). To our regret, we learned
through a review of the literature that we could not extend the
series to middle childhood and beyond because of the paucity
of studies amenable to a functional analysis at those levels.
We therefore supplemented the series with a book of readings
to show the application of principles in others areas related to
child development (Bijou & Baer, 1967).

Volume One was revised twice. The first revision (Bijou &
Baer, 1978) emphasized that the relationship between
behavior and stimulating conditions was mutual and
reciprocal, and extended the analysis to complex behavior,
such as self-management and problem-solving. The second
revision (Bijou, 1993) included an analysis of verbal behavior
and elaborated on the meaning of setting factors.

I learned from personal contacts that these books were used

not only in courses on child development, but also in clinical
and introductory courses. The series enjoyed a much wider
circulation than we had anticipated.

Lessons Learned for the Future of

Child Behavior Therapy
One lesson learned was that insufficient attention was being
devoted to the joint treatment of parents and child. This lack
of attention refers not to parent training per se or to training
parents to deal with their child’s problem behavior (Dangle &
Polster, 1984) but to the treatment of parents and child in
relation to each other. Considering that the earliest child
therapies evolved from the psychoanalytic-medical model, the

224
oversight is understandable. Believing that the parents were
either the cause of the problem or part of it, early-day
therapists tended to exclude parents from treatment on the
assumption that they would either delay or retard the child’s
treatment program.

Assuming that parents are the most important social influence

in a child’s early life, child behavioral therapists should insist
on parents’ involvement, the extent to which would depend
on the severity of the problem. With severe disturbances,
autism for example, treatment should be planned and carried
out with the child and parents as a social unit.

Another lesson learned is that although it was recognized

early on that there is a need for training and experience
standards for those who practice child behavior therapy,
nothing was done about it. As early as 1971 there was an
expression of such a need. It was at the first meeting of the
International Symposium on Behavior Modification held in
Xalapa, Mexico. The entire panel of Mexican and American
psychologists was concerned with standards for the training
of behavioral paraprofessionals. It is encouraging that six
states in the United States now have such requirements and a
strong movement exists in the Association for Behavior
Analysis to support this movement at the national level.

Summary
The early advances in child behavior therapy have been
nothing short of spectacular. A review of therapeutic
techniques for children (Bijou, 1954) described five
approaches, three of which were offshoots of psychoanalysis:

225
Child analysis (A. Freud, 1946; Kline, 1949), the briefer
analytic child therapies (Newell, 1941), and the expressive
therapies (Levy, 1939). Of the other two, one was based on
the Rankian psychology (Taft, 1933; Allen, 1942), the other
on Rogers’ client-centered approach (Axline, 1947). There
was no mention of learning or behavior therapies. An update
of that review, 12 years later (Bijou & Sloane, 1966),
included both classical and operant conditioning child
therapies. The classical conditioning therapies emphasized the
systematic desensitization of children with phobias (e.g.,
Bentler, 1962; Lazarus, 1959), whereas the operant
conditioning therapies stressed a range of problems from
autism (Wolf, Risley, & Mees, 1964) to everyday “normal”
problems (Harris, et al., 1964) to the rehabilitation of simple
motor skills (Johnston, Kelley, Harris, & Wolf, 1966). So
rapidly had the field expanded during the next ten years that
the editor of the 1975 edition of American handbook of
psychiatry saw fit to include a separate chapter on the child
behavior therapies (Bijou & Redd, 1975).

All the described research and applications served to set the

foundation for intensive studies on specific techniques for
specific problems which we are now witnessing.

References
Allen, F. H. (1942). Psychotherapy with children. New York:
W. W. Norton.

Allen, K. E., & Goetz, E. M. (Eds.). (1982). Early childhood

education: Special problems, special solutions. Rockville,
MD: Aspen Systems.

226
Axline, V. M. (1947). Play therapy. New York: Houghton
Mifflin.

Bach, G. R. (1945). Young children’s play fantasies.

Psychological Monographs, 59(2), 1-69.

Baer, D. M. (1960). Escape and avoidance response of

preschool children to two schedules of reinforcement
withdrawal. Journal of the Experimental Analysis of
Behavior, 3, 155-160.

Baer, D. M. (1961). Effect of withdrawal of positive

reinforcement on an extinguishing response in young
children. Child Development, 32, 67-74.

Baer, D. M. (1962a). Laboratory control of thumb sucking by

withdrawal and re-representation of reinforcement. Journal of
the Experimental Analysis of Behavior, 5, 525-528.

Baer, D. M. (1962b). A technique of social reinforcement for

the study of child behavior: Behavior avoiding reinforcement
withdrawal. Child Development 33, 847-858.

Baer, D. M., Peterson, R. F., & Sherman, J. A. (1967). The

development of imitation by reinforcing behavioral similarity
to a model. Journal of the Experimental Analysis of Behavior,
10, 405-416.

Baer, D. M. & Sherman, J. A. (1964). Reinforcement control

of generalized imitation in young children. Journal of the
Experimental Analysis of Behavior, 1, 37-49.

227
Bentler, P. M. (1962). An infant’s phobia treated with
reciprocal inhibition therapy. Journal of Child Psychology
and Psychiatry, 3, 185-189.

Bijou, S. W. (1938). The performance of normal children on

the Randall’s Island Performance Series. Applied Psychology,
22, 186-191.

Bijou, S. W. (1942). The development of a conditioning

methodology for studying experimental neurosis in a rat.
Journal of Comparative Psychology, 44, 91-106.

Bijou, S. W. (1943). A study of experimental neurosis in the

rat by the conditioned response technique. Journal of
Comparative and Physiological Psychology, 36, 1-20.

Bijou, S. W. (1951). A conditioned response technique to

investigate experimental neurosis in the rat. Journal of
Comparative and Physiological Psychology, 44, 84-87.

Bijou, S. W. (1954). Therapeutic techniques with children. In

L. A. Pennington & I. A. Berg (Eds.), An introduction to
clinical psychology (2nd ed.). New York: Ronald Press.

Bijou, S. W. (1955). A systematic approach to an

experimental analysis of young children. Child Development,
26, 161-168.

Bijou, S. W. (1957a). Patterns of reinforcement and resistance

to extinction in young children. Child Development, 28,
47-54.

228
Bijou, S. W. (1957b). Methodology for an experimental
analysis of child behavior. Psychological Reports, 3, 243-250.

Bijou, S.W. (1958a). A child study laboratory on wheels.

Child Development, 29, 425-427.

Bijou, S. W. (1958b). Operant extinction after fixed-interval

schedules with young children. Journal of the Experimental
Analysis of Behavior, 1, 25-29.

Bijou, S. W. (1961). Discrimination performance as a

baseline for individual analysis of young children. Child
Development, 32, 163-170.

Bijou, S. W. (1968). Experimental analysis of left-right

concepts in young children. International Review of Research
in Mental Retardation (Vol. 3, pp. 65-96). New York:
Academic Press.

Bijou, S. W. (1976). Child development: The basic stage of

early childhood. Englewood Cliffs, NJ: Prentice-Hall.

Bijou, S. W. (1993). Behavior analysis of child development

(2nd rev.). Reno, NV: Context Press.

Bijou, S. W., & Baer, D. M. (1961). Child development: A

systematic and empirical theory (Vol. 1). Englewood Cliffs,
NJ: Prentice-Hall.

Bijou, S. W., & Baer, D. M. (1965). Child development: II.

Universal stage of infancy. New York:
Appleton-Century-Crofts.

229
Bijou, S. W., & Baer, D. M. (Eds.). (1967). Child
development: Readings in experimental analysis. Englewood
Cliffs, NJ: Prentice-Hall.

Bijou, S. W., & Baer, D. M. (1978). Behavior analysis of

child development. Englewood Cliffs, NJ: Prentice-Hall.

Bijou, S. W., Birnbrauer, J. S., Kidder, J. D., & Tague, C. E.

(1966). Programmed instruction as an approach to the
teaching of reading, writing, and arithmetic to retarded
children. Psychological Record, 16, 505-522.

Bijou, S. W., Lovaas, O. I., & Baer, D. M. (1965).

Experimental procedures for analyzing the interaction of
symbolic social stimuli and children’s behavior. Child
Development, 36, 237-248.

Bijou, S. W., & Oblinger, B. (1960). Responses of normal

and retarded children as a function of the experimental
situation. Psychological Reports, 6, 447-454.

Bijou, S. W., & Orlando, R. (1960). Single and multiple

schedules of reinforcement in developmentally retarded
children. Journal of the Experimental Analysis of Behavior, 4,
339-348.

Bijou, S.W., & Orlando, R. (1961). Rapid development of

multiple schedule performances of retarded children. Journal
of the Experimental Analysis of Behavior, 4, 7-16.

Bijou, S. W., & Peterson, R. F. (1971). The psychological

assessment of children: A functional analysis. In P.
McReynolds (Ed.), Advances in psychological assessment.

230
(Vol. 2, pp. 63-78). Palo Alto, CA: Science & Behavior
Books.

Bijou, S. W., Peterson, R. F., & Ault, M. H. (1968). A method

to integrate descriptive and experimental field studies at the
level of data and empirical concepts. Journal of Applied
Behavior Analysis, 1, 175-191.

Bijou, S. W., Peterson, R. F., Harris, F. R., Allen, A. K., &

Johnston, M. S. (1969). Methodology for experimental
studies of young children in natural settings. Psychological
Record, 19, 177-210.

Bijou, S. W., & Redd, W. H. (1975). Child behavior therapy.

In S. Arieti (Ed.), American Handbook of Psychiatry, (Vol. 5 ,
2nd Ed., pp. 579-585). New York: Basic Books.

Bijou, S. W., & Sloane, H. N., Jr. (1966). Therapeutic

techniques with children. In L. A. Pennington & I. A. Berg
(Eds.), An introduction to clinical psychology (3rd Rev.).
New York: Ronald Press.

Bijou, S. W., & Sturges, P. T. (1959). Positive reinforcers for

child experimental studies with children — consumables and
manipulables. Child Development, 30, 151-170.

Bronfenbrenner, U. (1992). Ecological systems theory. In R.

Vasta (Ed.), Six theories of child development (pp. 187-249).
London: Jessica Kingsley Publishers.

Dangle, R. F., & Polster, R. A. (Eds.). (1984). Parent

training: Foundations of research and practice. New York:
The Guilford Press.

231
Ferster, C. B., & DeMyer, M. K. (1961). The development of
performances in autistic children in an automatically
controlled environment. Journal of Chronic Diseases, 13,
312-345.

Ferster, C. B., & DeMyer, M. K. (1962). A method for the

experimental analysis of the autistic child. American Journal
of Orthopsychiatry, 32, 89-98.

Freud, A. (1946). Psychoanalytic treatment of children.

London: Imago Press.

Fuller, P. R. (1949). Operant conditioning of a vegetative

human organism. American Journal of Psychology, 62,
587-590.

Goetz, E. M., & Allen, K. E. (Eds.). (1983). Early childhood

education: Special environmental, policy, and legal
considerations. Rockville, MD: An Aspen Publication.

Greenspoon, J. (1955). The reinforcing effect of two spoken

sounds on the frequency of two responses. American Journal
of Psychology, 68, 409-416.

Harris, F. R., Johnston, M. K., Kelley, C. S., & Wolf, M. M.

(1964). Effects of positive social reinforcement on regressed
crawling of a nursery school child. Journal of Educational
Psychology, 55, 35-41.

Harris, F. R., Wolf, M. M., & Baer, D. M. (1964). Effects of

adult social reinforcement on child behavior. Young Children,
20, 8-17.

232
Hart, B. M., Allen, K. E., Buell, J. S., Harris, F. R., & Wolf,
M. M. (1964). Effects of social reinforcement on operant
crying. Journal of Experimental Child Psychology, 1,
145-153.

Hart, B. M., Reynolds, N. J., Baer, D. M., Brawley, E. R., &

Harris, F. R. (1968). Effect on contingent and non-contingent
social reinforcement of the cooperative play of a preschool
child. Journal of Applied Behavior Analysis, 1, 73-76.

Hart, B. M., & Risley, T. R. (1995). Meaningful differences in

the everyday experiences of young children. Baltimore: Paul
H. Brooks Publishing Co.

Hart, B. M., & Risley, T. R. (1999). The social world of

children: Learning to talk. Baltimore: Paul H. Brooks
Publishing Co.

Hawkins, R. P., Peterson, R. F., Schweid, E., & Bijou, S. W.

(1966). Behavior therapy in the home: Amelioration of
problem parent-child relations with the parent in a therapeutic
role. Journal of Experimental Child Psychology, 4, 99-107.

Hull, C. H. (1943). Principles of behavior. New York:

Appleton-Century Co.

Jastak, J. F. (1934). Variability of psychometric performances

in mental diagnosis. New York City.

Jastak, J. F., & Bijou, S. W. (1938). Wide Range

Achievement Test. Wilmington, DE: Guidance Associates.

233
Johnston, M. S., Kelley, C., Harris, F. R., & Wolf, M. M.
(1966). An application of reinforcement principles to the
development of motor skills of a young child. Child
Development, 37, 379-387.

Kazdin, A. E. (1978). History of behavior modification:

Experimental foundations of contemporary research.
Baltimore: University Press.

Klein, M. (1949). The psychoanalysis of children. London:

Hogarth Press.

Lazarus, A. A. (1959). The elimination of children’s phobias

by deconditioning. Medical Proceedings of South Africa,
261-265.

Levy, D. M. (1939). Release therapy. American Journal of

Orthopsychiatry, 9, 913-936.

Lewin, K. (1935). A dynamic theory of personality. New

York: McGraw-Hill.

Lewin, K. (1936). Principles of topological psychology. New

York: McGraw-Hill.

Marholin II, D., & Bijou, S. W. (1978). Behavioral

Assessment: Listen when the data speak. In D. Marholin II
(Ed.), Child behavior therapy (pp. 13-36). New York:
Gardner Press.

Miller, N. E. (1944). Experimental studies of conflict. In J.

McV. Hunt (Ed.), Personality and the behavior disorders
(Vol. 1 pp. 431-465). New York: Ronald Press.

234
Mussen, P. H., Conger, J. J., & Kagan, J. (1956). Child
development and personality (1st ed.). New York: Harper &
Row.

Newell, H. W. (1941). Play therapy in child psychiatry.

American Journal of Orthopsychiatry, 11, 245-251.

Orlando, R., Bijou, S. W., Tyler, R. M., & Marshall, D. A.

(1960). A laboratory for the experimental analysis of
developmentally retarded children. Psychological Reports, 7,
261-267.

Peterson, R. F. (1968). Imitation: A basic behavioral

mechanism. In H. N. Sloane, Jr. & B. D. MacAulay (Eds.),
Operant procedures in remedial speech and language
training (pp. 61-76). Boston: Houghton Mifflin.

Peterson, R. F., Cox, M. A., & Bijou, S. W. (1971). Training

children to work productively in classroom groups.
Exceptional Children, 37, 419-500.

Peterson, R. F., & Peterson, L. W. (1968). The use of positive

reinforcement in the control of self-destructive behavior in a
retarded boy. Journal of Experimental Child Psychology. 6,
351-360.

Pumroy, S. A. S. (1954). The effects of amount of

reinforcement on resistance to extinction and emotional
behavior with preschool children. Unpublished doctoral
dissertation, University of Washington.

235
Rheingold, H. L., Gewirtz, J. L., & Ross, H. W. (1959).
Social conditioning of vocalizations in the infant. Journal of
Comparative and Physiological Psychology, 52, 68-73.

Rogers, C. R., & Skinner, B. F. (1956). Some issues

concerning the control of human behavior. Science, 124, No.
3231, 1057-1066.

Sajway, T. E., Twardosz, S., & Burke, M. (1972). Side effects

of extinction procedures in a remedial preschool. Journal of
Applied Behavior Analysis, 5, 163-175.

Sears, R. R. (1947). Influence of methodological factors on

doll play performance. Child Development, 18, 190-197.

Sears, R. R. (1975). Your ancients revisited: A history of

child development. In E. M. Hetherington (Ed.), Review of
Child Development Research (Vol. 5, pp. 1-73). Chicago:
University of Chicago Press.

Sears, R. R., Whiting, J. W. M., Nowlis, V., & Sears, P. S.

(1953). Some child rearing antecedents of aggression and
dependency in young children. Genetic Psychology
Monographs, 47, 135-234.

Skinner, B. F. (1953). Science and human behavior. New

York: The Macmillan Co.

Skinner, B. F. (1968). The technology of teaching. New York:

Appleton-Century-Crofts.

Sloane, H. N., Jr., Johnston, M. K., & Bijou, S. W. (1967).

Successive modification of aggressive behavior and

236
aggressive fantasy play by management of contingencies.
Journal of Child Psychology and Psychiatry, 8, 217-226.

Sloane, H. N. ,Jr. & MacAulay, B. D. (Eds.). (1968). Operant

procedures in remedial speech and language training.
Boston: Houghton Mifflin.

Taft, J. (1933). The dynamics of therapy in a controlled

relationship. New York: The Macmillan Co.

Wahler, R. G. (1967). Child-child interaction in free field

settings: Some experimental analysis. Journal of
Experimental Child Psychology, 5, 278-293.

Wahler, R. G. (1969a). Infant social development: Some

experimental analyses of an infant-mother interaction during
the first year of life. Journal of Experimental Child
Psychology, 7, 101-113.

Wahler, R. G. (1969b). Oppositional children: A quest for

parental reinforcement control. Journal of Applied Behavior
Analysis, 2, 159-170.

Wahler, R. G., Winkel, G. H., Peterson, R. F., & Morrison, D.

C. (1965). Mothers as behavior therapists for their own
children. Behaviour Research and Therapy, 3, 113-124.

Watson, J. B. (1919). Psychology from the standpoint of a

behaviorist. Philadelphia: Lippincott.

Watson, J. B. (1930). Behaviorism. (Rev. ed.) Chicago: The

University Press.

237
Wolf, M. M., Risley, T. R., & Mees, H. (1964). Application
of operant conditioning procedures to the behavior problems
of an autistic child. Behaviour Research and Therapy, 1,
305-312.

Zeilberger, J., Sampen, S. E., & Sloane, H. N., Jr. (1968).

Modification of a child’s problem behaviors in the home with
the mother as therapist. Journal of Applied Behavior Analysis,
1. 47-53.

Footnote
1
Many thanks to my wife, Janet, for her careful and thorough
editing of the manuscript and my son, Bob, for preparing the
figures.

238
Chapter 6

Studies in Behavior Therapy

and Behavior Research
Laboratory: June 1953-1965
Ogden R. Lindsley

University of Kansas and Behavior Research Company

The Beginning
I was born on August 11th, 1922 in Jane Brown Hospital,
Providence, Rhode Island. My father was a young Harvard
Law School graduate practicing corporate law in Edwards and
Angel, Providence’s most prestigious law firm. My mother
had dropped out of Wheaton College at the end of her junior
year to marry my father upon his graduation from law school.
We wintered at 282 Wayland Avenue on the upper income
East Side of Providence. We summered on the west shore of
Narragansett Bay in Quidnessett, Quonset Point, Rhode
Island, on our 365-acre farmstead with a house built in 1804
and its own Allen’s harbor.

One of my earliest ambitions was to be like my

great-grandfather, Isaac Lindsley, who invented hair cloth
looms to weave the manes and tails of horses into a
stain-proof fabric to cover Victorian furniture. Stories about

239
his genius were family legends. He sailed to London,
England, to design looms to weave hair cloth royal coat of
arms fabric to cover chair backs and seats in Buckingham
Palace. He spent endless hours inventing in a shop behind his
house. He invented a very fast early bicycle, and a machine to
roll cigars. There was a wonderful oil painting of him in long
mustache and beard over the large mahogany Stella music
box in grandmother Lindsey’s home. The music box stood 37
inches from the floor and 24 inches from the wall. When I
was so small that I had to reach up to the edge of the music
box, I would pull with both hands standing tippy-toe to peer
over the top at Isaac Lindsley’s face. I had been given his
brass microscope with glass slides in a handsome wooden box
with an engraved brass nameplate which read, “Isaac
Lindsley, Inventor.” I wanted to grow up to be like him!

We had two rooms and a bath on our third floor attic. One
was Helen, our maid’s bedroom, and my mother, Mildred
Flagg Monroe Lindsley, made the other into a little school
room with three desks, a chalkboard, a table, and book cases
containing first and second grade public school books. I still
have some of those books: the McCall Speller -Intermediate,
Mother Westwind’s Children, the Hiawatha Primer, The
Elson Readers Book Three, Stories of American Discovers
for Little Americans. Every afternoon my mother taught me
school up there. Often one or two of my playmates came over
to be taught by my mother. It was wonderful, she was so kind
and had such a great sense of humor. Most of the time we
were learning and laughing. I learned so much that when it
came time for school, the principal at Moses Brown Lower
School had me skip first grade so I went into second grade at
six years of age. We wore little navy blue short pant suits
with white Eaton collars, and little navy blue caps with

240
MBLS, for Moses Brown Lower School, embroidered in
white. We had to walk home through some streets where
tough poor kids from Fox Point walked home from Nathan
Bishop Junior High School. The tough teens chased and
terrorized us, throwing rocks and horse chestnuts at us and
yelling “Momma’s Baby Lemon Sucker,” referring to the
MBLS on our caps. For a while we had police escorts until
the Fox Pointers gave up.

My father, Ogden Richardson Lindsley, taught me to

discover. He never directly answered my questions. He
always answered by saying, “How could we find out?” When
I asked him to teach me how to tie my shoelaces, he said “See
if you can figure it out by yourself.” When I tiptoed into his
room in the morning as he was dressing to see how he tied his
shoes, he noticed me and said, “No spying, figure shoe tying
out by yourself!” I did, and I now tie my shoes by a very
unique method, different from most people’s way.

On May 20th, 1935 my father was killed in a head-on

automobile accident in a car driven by my mother. He had
taken a job as a vice president and sales manager for Everett
and Baron, a shoe polish company for which he had done
superb legal work. His salary was about three times his old
law firm salary. He was late for a keynote speech against
Roosevelt’s National Recovery Administration at a shoe
retailers convention in Boston, about 60 miles on a two-lane,
blacktop Route One from Providence. My mother was driving
about 70 miles per hour in a brand new Chrysler Airflow
Imperial Sedan. She went to pass an eighteen wheeler truck
on a curve and ran head-on into a large LaSalle limousine
coming the other way. My father writing his speech in the
passenger seat took the full force of the collision dying

241
instantly, his body totally crushed. My mother had multiple
fractures; her hips were fractured in five places and she was in
the hospital almost a year-long enough for the guppy fish pair
that my brother and I gave her to have multiplied, filling a
bowl in almost every room in the hospital!

Nannie, my grandmother Lindsley, blamed and never forgave

my mother for killing her only child. I blamed neither my
mother nor my father. I blamed business for driving him so
fanatically that he was urging my mother to speed ever faster
while he wrote his speech notes in the passenger seat. Had he
accepted the Harvard Law Review and stayed in academia, I
doubted that he would have been killed strolling across
Harvard Yard. This early interpretation of the cause of my
father’s death laid the groundwork for my avoidance of
business as a career.

A few weeks before my father’s death I asked him if he

believed in life after death. He answered in his typical
fashion, “How could we find out?” I had just become a Boy
Scout in Troop 8, Providence, and learned Morse code and
American Indian smoke signaling. I had had trouble
separating L (.-..) from F (..-.), and my father suggested the
memory aid, “L is our Lindsley family, mother a dot, father a
dash, Ogden a dot, and Brad a dot. And you know that
gentlemen always have ladies go first.” I remember this to
this day. I told my father that if one of us dies the survivor
will go by the body on a nice quiet evening with no wind and
make a fire with wet leaves so the smoke will go straight up.
The dead one may be able to interrupt the column of smoke
with short and long interruptions — the dots and dashes of
Morse code — and send a message to the survivor that way.

242
My grandmother and mother could not agree where to bury
my father, so for a while he was in a tomb in Swan Point
Cemetery, the ritziest one in Providence. I sneaked over to his
tomb on a still fall night and built a nice wet fire with lots of
smoke beside the tomb and waited for signals from my dead
father. Suddenly I felt a hand on my shoulder! A male voice
boomed out, “What are you doing here, Lad!” I shook with
fright, turned around and saw a large, scowling, uniformed
Providence cop. I blurted out, “I’m keeping warm. My dead
father is in there!” The cop beamed and said in an Irish
brogue, “Sure and the fine young lad is holding a wake for his
dear departed father!” He took me to McDonald’s drugstore,
bought me ice cream, and then to the East Side police station
on Sessions Street. At the station he told all the cops what a
brave little man I was sitting awake by his father’s tomb. I did
not have the heart to break his myth and tell him that I was
trying to communicate with the dead!

We lost all of our money. The shoe polish company did not
pay my father’s salary from January to May 1935, when it
was customary for executives to get the full year’s salary for
the year in which they died. A partner in rental housing did
not honor the partnership and we received no money from
that source. It was the depth of the depression. My father did
not believe in insurance, saying he was smarter than insurance
people and could better invest his money. My mother had a
giant policy drawn up on my father that was to have sent me
to Switzerland for prep school if he died. But he had been too
busy cornering the shoe polish market to take the physical
examination, so this life insurance policy was not in force.
We moved to our summer home, the farm in Quidnesset,
full-time and became country boys. I put my arm around my
little brother’s shoulders and said “Don’t cry, Brad. I will take

243
care of you!” I never teased Brad or hit him again. I never
cried after my father’s death, and I assumed his role. I
dropped the “junior” from my name and stopped people from
calling me “Sonny.” I became “Ogden,” or “Oggie.”

From East Side little rich kids, my brother and I became

South County poor country boys. We moved to our former
summer home. We grew our own vegetables. We tended our
own ponies. We sold city kids pony rides on the beach. I
trapped skunks and sold their pelts for money to buy long
pants for Brad and me to wear to North Kingstown High and
Wickford Grammar schools. We walked half a mile up our
road to the mailbox to catch the school bus each day.
Whenever we were cut we went to our beach and soaked our
wounds in the clear salt water until the bleeding stopped. I
raised turkeys from day old chicks and sold them in town for
Thanksgiving and Christmas dinners. We raised vegetables
and sold them in town. We became country kids.

Mr. Brown, track coach and cross-country coach, taught me

to run through pain. I won Class C medals in the mile and
cross-country. I could not go out for team sports because
there was no late school bus that went north to our farm. I
lettered in track and expected to get a nice, leather armed,
brown and white, NKHS letter jacket on Class Day. I almost
cried and then got super angry when I was handed a little
plastic envelope with small felt NKHS letters inside. Track
was a minor sport and did not earn the big chenille NKHS
that the team sport lettermen got! I became high school paper
cartoonist, lead singer in the minstrel show, and held several
class offices.

244
In 1937 my horse, Prince, got a thorn in his head and
developed a severe throat infection. It was before penicillin,
so he was rapidly dying. I had to walk him about a quarter
mile down to our sandpit, where it was easy digging, and dig
his grave. Then I injected Prince with the strychnine that the
vet had given me to kill him. It was one of the hardest things I
had to do up to that point in my life. Having my father killed
was hard, but having to kill Prince was even harder!

I continued Boys Scouts in a small town, Troop 3 East

Greenwich, and became a patrol leader by recruiting and
starting a new patrol for the troop. I earned Eagle Scout and
many merit badges beyond. I became a leader at summer
scout camp, Yawgoog, and went on 500 mile canoe trips. I
won State Jamboree competition with fire by flint and steel by
using flint from the family heirloom flintlock rifle and
charred linen from my grandmother’s linen wedding slip. I
learned how to raise and train wasps and hornets to attack
other kids who entered my shop without permission. I loved
the behavior of farm and wild animals. I spent hundreds of
hours stalking them, imitating their calls, and learning their
habits. I trapped eels, crabs, skunks, muskrats, mink,
woodchuck, rabbits, guinea hens, quail, and pheasants. I dug
clams, quohaugs, mussels, and oysters. I caught saltwater fish
and lobsters. We were taught not to eat freshwater fish, even
though my grandfather had stocked the ice house pond with
pickerel. Mother would not let us swim in fresh water. It
could be polluted and dangerous. Salt water healed and was
always safe. We were north Atlantic kids.

In the great hurricane of 1938, I was out running

cross-country practice and the wind blew us down. It was
raining salt water! I thought it was the end of the world. We

245
had to walk about 12 miles home through blown over trees
across the highway. Refuse from broken Oceanside homes
and boats littered our front lawn and pasture. My brother and
I found three dead bodies washed up on our side of the
harbor. We called the National Guard who had declared an
emergency and sealed off our area to prevent looting.

In June 1939 I graduated from North Kingstown High School

with a class of about 35 students. I was 16 years of age and
went to Providence to work as a mechanic in a gasoline
station for a year, to make money and mature a bit before
attending Brown University the following year. Atlantic
Refining Company checkers came around to the stations in
unmarked cars, and if you approached them with the correct
Tom’s River Lubrication Service spiel they gave you a silver
dollar. “Good morning, Sir! Shall I fill it with Atlantic White
Flash Plus? Or do you prefer the Ethyl?” “Thank you, Sir! I’ll
also make sure your oil and water are safe!”

In September 1940 I entered Brown University as a freshman

to attain a degree in electrical engineering. The family plan
was to train as an engineer but at a liberal arts college. Join a
fraternity. Meet all the right people. Build social skills and
become a highly paid sales engineer. Richard Fink, my
stepfather, was a top salesman for Cook, Dunbar, and Smith
manufacturers of rolled gold plate. He often took me on his
sales visits when I was in high school.

At Brown I was in the class of ‘44 and undisciplined. I drank

too much, pledged Zeta Psi fraternity and with my piano
playing buddy, Phil Simpson, sang duets all night at fraternity
parties. I fell asleep in 8:00 A.M. classes, having slept not at
all the night before. I got either the highest grade in the class

246
(Analytical Geometry) or the lowest (Advanced Calculus). If
I loved it, I exceeded. If I hated it, I flunked. There was no
middle ground! The war escalated in Europe. High school
friends joined the merchant marine on the Murmansk run.
Others joined our U.S. Marines. One was a Navy flier. I
thought about joining the war effort every day. I envied
uniforms and combat ribbons and wings! My heart was in
flying a spitfire as an American volunteer for the Royal Air
Force over England and chasing Messerschmitts back to
Germany. I would rather die in air combat, a hero, than sit
here in Providence in a lecture seat.

In January 1942 the U.S. Army Air Force dropped its age
requirement for cadets from 21 to 18 years. I immediately
joined and was in Class 42J Aviation Cadets at Maxwell
Field, Montgomery, Alabama. I went through preflight at the
top of my class, but washed out in June on a flight physical
eye exam with prism divergence to exceed six diapers. The
medics giving the physicals urged us to erase the check marks
on the examination forms we carried from exam post to exam
post. Some of the cadets did erase physical problem checks. I
chose not to erase my prism divergence check. I was so
young, naive, and moral that I thought the physicians knew
more than I did. I did not want to take nine men on my crew
to their death because I wanted silver wings. So I left the
check mark on my form, and was washed out. The
ophthalmologist said my eyes were not good enough to be a
navigator or bombardier. My eyes were not even good enough
for glider pilot school! I was discharged from Cadets. I was
ashamed to go home as a civilian, so I enlisted in the Army
Air Corps as a private at Maxwell field. In July the Air Corps
found there were not enough perfect eyes in the country to
fight the war, so they dropped mission requirements for

247
flying. Prism divergence dropped from six diopters to 12!
Mine was eight diopters! I was so young and inexperienced
that I did nothing about it. I let them assign me to the military
police and be a guard at the main Maxwell field gate. Because
I had been top in my cadet preflight class I really could polish
brass, buttons and buckles, stand rather ramrod straight, and
salute smartly.

I got promoted to corporal, and drove a jeep on night patrol

around the air base perimeter. I still love the throb of that
little flat head four engine. I got caught by the officer of the
day while lying on the hood of that MP jeep with a beautiful
young nurse and counting the stars while on duty! They sent
Juanita to combat in Africa and busted me to private and put
me on the back of a GI truck cab with a sawed-off shotgun
guarding prisoners while they picked up paper along the
Birmingham highway fence. I was now a prison chaser! I
started drinking too much beer, spending every night in town
with an MP pass. One day I fell asleep on the truck cab under
my umbrella. When I looked up I saw only three of my four
prisoners. The fourth was running toward the airfield fence. I
yelled halt three times, pumped a shell into the shotgun,
aimed in the direction of the prisoner now running faster, and
pulled the trigger. A buck shot hit him in the leg. We put him
in the back of the truck and drove him to the base hospital
while the other three prisoners swore at me and pledged they
would get me some day for this.

On the advice of older, peace time GIs, I had been going

about once a week to both Provost Marshal and Base
Chaplain complaining that I had army general classification
test scores above 150 and should be better used than as a
prison chaser. The combination of these appeals, my test

248
scores, my year-and-a-half college engineering experience,
and the shooting got me sent to airplane mechanics school at
Keesler Field, Biloxi, Mississippi. I studied hard and became
a student instructor who went to class an hour before the other
students, and then taught a squad of about 15 students the
topic of the day. I graduated first in a class of about 800 and
was awarded my certificate and engineers badge by the major
general in front of the entire school student body massed at
attention. As a reward for being top in the class, I was being
sent to Helicopter Engineering Officer School at Twenty Nine
Palms while all the other graduates were off to gunnery
school at Tyndall Field, Florida, to become combat flight
engineers. The first sergeant came running out of the orderly
room yelling, “Lindsley, we’ve got to cut you a new set of
orders! We have to re-stencil your barracks bags! You can’t
go to Twenty Nine Palms! You’re on detached service from
the 831st Military Police Company and have to go back to
Maxwell!”

Back at Maxwell I wasn’t even in the Air Force anymore. The

831st Guard Squadron had been transferred out of the Army
Air Corps into the Army Military Police Branch and made the
831st Military Police Company. We no longer wore a
propeller and wings on our uniforms, but wore crossed
pistols! And I was an airplane mechanic! I had blue triangles
with a gold rotary engine insignia sewn on my right sleeve! I
got teased by the former cops and the company. They would
bend over, point to their butt and say, “Hey! Engineer! Come
over here and check my oil!”

I went back to my provost marshal and chaplain monthly,

bitching, and soon got transferred back into the Air Force
82nd training squadron on Maxwell Field as a flight engineer

249
on B24 bombers. I worked up to be crew chief for Major
Buttman on Army 00, The Flying Goose, the squadron
commander’s ship. Soon our squadron was transferred to pilot
transition training at Smyrna Field, Tennessee. An instructor
pilot and an engineer took up two rookie pilots just out of
twin-engine flight school, shot a few landings by each student
pilot, then the instructor got out. The flight engineer stayed
while the student pilots learned to fly the heavy four engined
B24 bomber. We would shoot 30 to 60 landings a day!
Landing and takeoff accidents were common. I was flying
every day, but not in a spitfire over England chasing
Messerschmitts back to Germany.

Tech Sergeant Perez, a former peace time ground machine

gunner at Schofield Barracks, Hawaii, had volunteered to
switch over to the Air Corps and lay on his belly in the back
of an early B-17 and fire a flexible 50 caliber machine gun
out of its sawed off tail, becoming one of the B-17’s first tail
gunners. Perez had flown a tour of duty in the Pacific, told
great war stories, and wore an air medal with oak leaf clusters
and a yellow and red Asia-Pacific combat ribbon with several
battle stars. I admired Perez. He really knew the Air Corps
enlisted flight crew life. We were good drinking and woman
chasing buddies. One day just as I landed with two rookie
pilots, the B-24 in front of us went off the end of the runway
and nosed down, twin tail sticking up in the air at a 45-degree
angle. A little smoke rose out of one of the starboard engines.
We taxied over to operations while civilian fire trucks raced
to the nosed up B-24. The waist windows were closed and
jammed shut by the fuselage twisting from the crash. You
could hear the men inside yelling and pounding the sides of
the aircraft. Flames were now coming from the starboard
engine. The civilian crash crews had forgotten their asbestos

250
suits and a truck raced back to the hangars to get them. Just
then there was a terrible swoosh and a wave of hot air as the
plane went up in flames from the spilled 120 octane gas.

Back at operations, my student pilots asked, “Who were the

student pilots in that plane?” I asked, “Who was the
engineer?” The operations officer replied, “Tech Sergeant
Perez.” I shouted, “S—t! Not Perez! Combat in the Pacific,
and now ashes at the end of a runway in Smyrna, Tennessee!”
I went into Nashville to our favorite bar, got drunk, and
refused to fly. The flight surgeon made me come to his office
every day and sign a yellow sheet under the words, “I am
yellow. I refuse to fly for my country.” I did, but each day
added “with rookie pilots in Smyrna. I want to fly in combat!”
above my signature. After about two weeks of this, I got sent
to gunnery school at Tyndall Field, Florida.

I went through gunnery school at Tyndall, air crew make up

and assignment at Savannah, Georgia, and staging for high
altitude bombardment at Langley Field, Virginia. We flew our
factory new silver B24-J from Langley to Newark, New
Jersey; to Bedford, Massachusetts; to Bangor, Maine; to
Gander, Newfoundland; the Azores; Marrakech, Morocco; El
Aouina, Tunisia; to the U.S. air base in Lecce, Italy. We were
a replacement crew to the 415th squadron, 98th Bomb Group,
15th Air Force when Major Habegar, our new squadron
commander, took our shiny new plane for his own crew,
leaving us to fly as replacements in battle scarred, patched,
veteran ships from the African campaign. I was shot down
with a crew that I had never seen before on my first real
combat mission to the Asta Romani oilfields in Ploesti,
Rumania on July 22, 1944.

251
After parachuting out over the North Albanian Alps we were
traded by partisans to the Croatians for guns and then handed
over to the Germans in Dubrovnick, Yugoslavia. We were
interrogated by the Hungarian Gestapo in Pestvideki Prison in
Budapest and imprisoned in Gross Tychow, Pomerania,
between Danzig and Berlin. In January of 1945 our guards
marched us from Luft IV to Hamburg across northern
Germany to keep us away from the advancing Russian
armies. We had only two bowls of hot soup in 63 days. I went
from 175 to 114 pounds. I sneaked through the wires into the
French prisoners’ compound as my fellow Air Force POW’s
were marched back into Germany. A few days later, dressed
as a French POW, I escaped from a wood picking-up detail in
the forest with two French POWs. We went through the
German front lines to the Queen’s Regiment, British 2nd
Army. I escaped rather than be marched back into Germany
with the rest from Stalag Luft IV. I was afraid that when
Germany was only a few hundred kilometers wide, there
would be no room for hated prisoners of war, and we might
all be machine gunned as we marched.

Mary Elizabeth Moore and I were married in early November

1945 while I was still in the Air Force. I had dated her a few
times before the war, and we dated while I was recovering
from malnutrition and Pleurisy at Cushing General Hospital
in Framingham, Massachusetts. I got discharged in November
1945 and went back to Brown University on the GI Bill. Mary
got a job as a secretary in a lawyer’s office. We had a small
basement apartment on George Street in Providence. Having
difficulty back in Advanced Calculus, I chose taking courses
that I enjoyed with content that I liked. Most of these courses
were in experimental psychology and biology, so that meant
dropping out of engineering. I had a double undergraduate

252
major in experimental psychology and histochemistry. I had
been influenced by Flanders Dunbar’s book, “Psychosomatic
Medicine.” I planned to personally solve the mind-body
problem by becoming expert in both. I became president of
Zeta Psi fraternity and help fill the house with returned
combat veterans. Notable among them was Dick Check,
former chief quartermaster on the aircraft carrier Bunker Hill.
I was also proud of getting the first Jewish man into our
chapter by blackballing the whole delegation until his box
passed with all white balls. With Ben Latt, whose dad was a
union organizer, I started the Lincoln Society with its motto
of “Fellowship without Fences.” I became interested in liberal
causes and folk music, and learned to sing and play a
six-string guitar. I graduated with Highest Honors in
Psychology in June 1948. I did not make Phi Beta Kappa or
Cum Laude because Brown averaged in my grades and
incompletes (which had become E’s) from before the war!

My main mentor at Brown during graduate training was Carl

Pfaffman, an electro physiologist who had studied with Lord
Adrian while on a Rhodes Scholarship and had earned a Ph.D.
from Cambridge University. Carl taught me the details and
personal discipline of laboratory science. His fame came from
isolating and recording the nerve fibers for taste in the chorda
tympani nerve that ran from the anterior two thirds of the
tongue. My undergraduate Honors Thesis on handedness in
rats showed that you could determine which paw they
preferred to use by biasing their first one handed reaches in an
angled chute. I received highest honors, because I had the
honesty to write in my thesis that I was forced to discard one
of the 20 odd rats who was slowest to learn, because in a fit of
rage at his slowness I pulled him by his tail from the

253
apparatus and threw him across the room. That was the last
time that I angered at slow learners.

I admired Lord Adrian for his creativity from a story that Carl
had told me about Adrian’s first recording of the electrical
discharge from an eel’s eye. It was in the days before reliable
electronic amplifier tubes. Adrian physically amplified the
electrical signal by using a light and a movable mirror. He
went into the basement of the longest cathedral he could find
and set up his equipment. In the dark a small mirror moved a
miniscule distance when the eel eye was stimulated with light.
At the other end of the hundred foot long dark cathedral
photographic paper was pasted on the wall which recorded
the amplified signal. This classy, simple, physical solution
fascinated me. I wanted to create similar simple solutions to
my own psychological research problems.

I designed and built the psychology department’s

histochemistry lab for staining nerve fibers. My master’s
thesis reported the diameters and the conduction velocities of
the C fibers used for taste in the chorda tympani nerve of the
rat. Carl had taught me the tease method of single nerve fiber
recording. You anesthetized a rat, put him under a microscope
in a moist chamber and shielded room, and using iridectomy
tools surgically exposed the nerve running across its eardrum.
While stimulating its tongue from dripping salt or sugar
solutions, you watched a cathode ray oscilloscope recording
from silver—silver chloride electrodes touching the
nerve. You very gradually tweezed off pieces of the nerve
until you got down to seeing only a few fiber’s spikes
showing on the oscilloscope tube from each stimulation. Then
you watched with fingers crossed hoping the nerve fibers
would die one at a time leaving a single fiber firing. At that

254
moment you stimulated and took as many pictures as you
could before it also died. This took 3 to 4 hours, most of a
Saturday, and a lot of patience. About 1 in 5 or 1 in 10
operations got a single nerve fiber dose response curve! Here
I learned scientific discipline and patience.

About this time Floyd Ratliff, a former artillery officer, and

fellow graduate student studying with Lorrin Riggs,
conducted a simple creative solution similar to Adrian’s eel
eye signal amplification in the cathedral basement. Floyd
proved that the vibration of the eye is necessary for sight by
gluing a small mirror to a contact lens and aiming the visual
signal at the mirror on the eye so the eye’s motion moved the
signal also. The mirror sent the signal to a screen and the
signal disappeared. An elegantly simple solution to an age old
vision question that even Helmoltz could have solved but did
not! It looked like all that you had to do was design one
brilliant experiment and your place in textbooks was assured
and your professorship was granted and tenured. Such was
my plan.

In my courses I learned Behaviorism from Walter Hunter.

From him I also learned that it is impossible to improve the
name of a discipline once it is accepted by society. Hunter’s
attempt to change the name of Psychology to Anthroponomy
was a dismal failure. His claims to fame were building the top
experimental behavioral program at Brown from scratch, and
his research design of the double alternation maze. I fought
young professor Greg Kimble, fresh from an Iowa Ph.D., who
taught us Hull-Spence learning theory when he said you did
not need to know what went on in the nervous system to
understand and predict behavior. Just before the end of his
course, I finally agreed that he was right. He gave me my only

255
B grade at Brown. To this day Greg does not believe he
taught me anything, but I know how much he changed me.

I set to work on my doctoral dissertation apparatus. I planned

to do with smell what Carl had done with taste. I had learned
to surgically expose a rat’s olfactory epithelium and nerves
under anesthesia. I was designing glass tubing and fans to
blow odors over the epithelium at constant flow, temperature,
and moisture. I had taken all the courses for the doctorate in
experimental at Brown — over 60 credit hours. Everything
went smoothly. Then disaster struck! The dean of the
Graduate School dropped dead with a coronary. The acting
dean was Barnaby C. Keeney, a history professor with all
three degrees from different universities. He was convinced
that made him a superior scholar, and he put into action “no
more three degrees at Brown.” Hunter, a strong man, was on
sabbatical. Harold Schlosberg, famous for coauthoring a
textbook and a weaker, more compromising person than
Hunter, was acting department chair. Schlosberg let me and
my office mate and friend, William Kessen, be ordered to find
another university for our doctorates. Kessen, whose advisor
was Kimble, went off to Yale along with Kimble. I had to
find a place to conduct my electro-physiological research.
There were three options. Johns Hopkins to study with Eliot
Stellar who had just written a textbook on Physiological
Psychology; McGill to study with Donald Hebb, a wonderful
student-supporting professor who let his students conduct any
creative research they chose; Harvard with Robert Galambos
who had just perfected his micro-electrode method of single
nerve fiber recording. I applied to all three. My wife Mary
preferred Harvard because she grew up in Weston, a Boston
Suburb. I preferred McGill because I loved Canada and knew
that all Hebb’s students loved him. I was offered a fellowship

256
with preference to Mayflower descendants at Harvard, and
Galambos accepted me as a student. Our family — by now
Mary Elizabeth and Deborah Melinda had been born —
moved to a housing project in Watertown, Massachusetts in
June 1950. I wanted an early start at Harvard to learn the
ropes and to learn micro-electrode recording from Galambos
before classes started in September.

Galambos put me to work sitting at a table measuring nerve

discharges on film along with Kathy Safford, another
graduate student who knew nothing about electro physiology.
The crew in the pit doing the operations on auditory nerve
preparations in cats was a closed shop. They were all
experienced researchers — Galambos, Walter Rosenblith, and
Jerzy Rose. They did not even have lunch with us. They did
not invite me to watch their operations and recording. I was
their recording slave, and they kept pushing us to read and
measure more film records faster and faster. They seemed
desperate. So went my summer.

In September I was put in Psychology 101 with Law School

drop outs who had never had a psychology course, and started
graduate school all over again with absolutely no credit for
any course (over 60 graduate credit hours) taken at Brown. I
soon saw why. There was almost no overlap in content with
courses with the same title from Brown. Harvard was more
past oriented, covering a rich history of European and
especially German experimental psychology from E. G.
Boring’s influence. Harvard was also psychophysics oriented
from S. S. Stevens’ influence. I essentially embarked on a
second Ph.D. program. In June I went to Galambos and asked,
“Now will you teach me your micro electrode recording
method this summer?” Galambos smiled and said, “Lindsley,

257
you’ve blown it! I’m going to Walter Reed in 2 weeks.” I
said, “Wow! And you’re taking me with you?” Galambos
said, “Nope! You are a graduate student and your fellowship
is for here!” I said, “But my fellowship is to study micro
electrode recording with you!” He said, “Yes, but I have not
asked Walter Reed to support you.” Such was the life of a
graduate student at Harvard. So different from Brown.

Meanwhile, B. F. Skinner had asked me to assist him teaching

sections of his undergraduate course, “Natural Science 114.”
The dittoed text for this course later became Skinner’s book
“Science and Human Behavior.” The class had a weekly
lecture by Skinner with a demonstration of pigeons or rats
illustrating the topic of the week. The other two meetings of
the class each week were in smaller sections run by graduate
students that covered the weekly topic in more detail. Skinner
asked me to lead one section. I said, “I haven’t even read the
text!” He said, “I know, but at least you’re a behaviorist
having been taught by Hunter at Brown. I am being
undermined by graduate student section leaders who tell the
Harvard undergraduates “I don’t believe it either, but that is
what Skinner wants you to say on the final exam.” So that
spring semester I taught a section of Natural Science 114.
Fred assigned me my first job to shape up a high jumping rat
for a class demonstration of the astonishing results of shaping.
Armed with the Halloween clicker Skinner gave me, I
promptly got four naive male rats from the animal colony,
two ring stands and a meter stick from the biology labs, and a
pellet dispenser from Skinner’s pigeon-rat apparatus room. I
soon had all four rats magazine trained; when they heard the
pellet drop they ran to the pan and ate it.

258
After several sessions of pairing the clicks with the pellet
dispenser the rats were clicker trained, running back to the
place where they were last clicked. The four rats were soon
stepping over the meter stick lying on edge. When I raised the
meter stick up on the ring stands, one rat kept pulling down
on one end. I immediately put him back in his home cage and
built him a different apparatus, pivoting and putting an
adjustable sliding weight on the meter stick. I was going to
teach this rat to lift weight. I put a ring handle on the end of
the meter stick, because his paws kept slipping off. Within a
week of daily training, Samson Rat was pulling down a
weight equal to his own body weight. He jumped up, hung on
the handle with both fore paws, and tried to bounce down the
weight. One time his back feet swung forward and his toe
nails stuck into the plywood wall and with this leverage he
was able to pull down a weight heavier than his own weight. I
put him in his home cage, ran to the Harvard Coop across
Harvard Square, getting there just before closing at 5:00 pm. I
bought a rubber stair tread to use as an exercise mat on the
wall of Samson’s weight lifting apparatus, so his back feet
would not slip. I was in a hurry to see if Samson’s exercise
mat would help him lift more weight. Now, what to use for
fast drying glue? I remembered the very, very expensive
Ambroid, non conductive electrode cement, in the psycho
acoustic electro physiological operating room. It cost about
$40 a quart but dried in an instant. I smeared the back of the
stair tread with Ambroid, ran to the pigeon and rat labs in the
other end of Memorial Hall basement, and stuck the tread on
Samson’s wall. I rushed Samson from his home cage back to
his lifting apparatus with its new non-slip mat. Samson leaped
up, swung on his ring, planted both hind feet on his rubber
mat, arched his back, and pulled the weight all the way down.
HOORAY! Samson had taught me what he needed! Within

259
another week he was pulling down weight over two and a half
times his own body weight!

Skinner has often said that he did not make free operant
conditioners, the rats did! Samson had just made me a free
operant conditioner. It was the speed of learning, and the
precise control of Samson’s behavior that bowled me over. I
was a laboratory scientist. I had to admit that I had more
control over a whole, free roaming, rat than I ever did over a
single rat nerve in a temperature controlled moist chamber. I
never again did a physiological experiment. I still kept all my
expensive custom designed iridectomy forceps and knives.

For several years I rationalized that I would combine

Skinner’s methods to improve my physiological research. I
dreamt up many such experiments and wrote them down in
my notebooks. One such experiment was to teach rats to hold
their breath to a sound signal to clean up their motor cortex
for study. The giant Betz motor cells in the brain make more
electrical discharge than the other cells. As the rat breathes,
these motor cells discharge and swamp out the delicate small
sensory neural discharges that you are trying to record.

I often discussed these possibilities with Jim Olds, a fellow

graduate student. Jim went on to discover the reinforcement
center in the brain as a result of combining free operant with
neurophysiological research.

When it came time for my doctoral dissertation research, Fred

said he had two things I might work on. One, put together a
pigeon box and take it over to the medical school where Otto
Krayer in the pharmacology department is interested in
measuring the effects of drugs on free operant behavior. This

260
was what Peter Dews later did. The other was with Walter
Jetter who was a state pathologist and professor of legal
medicine at Boston University medical school and who had a
grant from the atomic energy commission to study the effects
of irradiation on beagle dog’s physiology, exercise and
behavior. Fred said, “I have been receiving a few hundred
dollars a month from him in consultation, which has helped
keep Julie in Putney School. All I have done is design a
hamburger magazine that Ralph Gerbrands has about half
built.” I chose to work with the dogs rather than the pigeons
because they seemed more social and closer to my goal of
researching human behavior. I also knew that dog blood is
closer to human blood than that of other animals, and
preferred by many pharmacologists for research.

At Boston University I designed and built the first dog

operant conditioning apparatus. I studied the dogs in a
one-hour behavior sample that included a baseline of pressing
a panel on a one minute variable interval schedule for raw
hamburger reinforcement. They learned to stop pressing when
a light flashed (visual discrimination). They also learned to
pause when a buzzer signaled a loud, aversive horn blast
(conditioned suppression), going back to work immediately
after the horn stopped. This was the first use of loud noise as
an aversive stimulus. I ran 63 two year old male beagle dogs
for 50 or more one hour long daily sessions to stabilize their
behavior samples.

I injected the dogs with alcohol, Nembutal,

methamphetamine, and saline placebos to calibrate the effects
of well known drugs on this behavior sample. I did this so
that we could say enough alcohol to make them stagger had
this effect, and enough irradiation to kill them had that effect.

261
The well known drug effects provided a comparison. Then the
dogs were taken over to Massachusetts Institute of
Technology (MIT) in small body tight cages where they were
given 300 Roentgen units of total body irradiation (150 units
on each side) from a Van der Graph Generator. That amount
of radiation will kill half of the dogs (Lethal Dose of 50% —
LD50) from Leukopenia (loss of white cells) about 15 days
after irradiation. The survivors gradually recover, taking
another 15 days to regain strength and health. Hunter, named
for Walter Hunter and the quickest to learn of the 63 dogs,
survived and became Fred Skinner’s family dog. Hunter lived
to a ripe old age, although Eve Skinner was convinced that
Hunter’s embarrassing flatulence had been caused by his
irradiation.

Our free operant behavior sample picked up an immediate

effect of the irradiation an hour afterwards by the dogs
anticipating and extending their conditioned suppression
(experimental anxiety) period when the buzzer signaling the
noise blast sounded. Of course, when the dogs became
physically ill, their responding dropped off, but their visual
discrimination and conditioned anxiety suppression were not
changed by their physical deterioration. The same dogs were
run every day on an overhead maze by Albert Dimascio, and
Nathan Azrin, two BU psychology graduate students. There
were no discernible immediate effects of irradiation on their
maze times or errors. Azrin was so impressed with the
sensitivity of the free operant that after my urging he applied
to Harvard’s Psychology Department and was accepted to
study for his doctorate with Skinner.

This research became my doctoral dissertation, accepted in

June 1957. From Walter Jetter I learned the advantage of

262
keeping experimental animals in top physical shape. Our dogs
were examined and treated by a veterinary physician every
day. They had their temperatures and blood sampled every
day. They were washed and combed every day, Their home
cages were hosed down and sterilized every day. They
exercised on a treadmill at 20 miles per hour at 20 percent
grade for 20 minutes every day!

From our free operant research I was strongly rewarded for

apparatus design and procedure innovation. I had introduced a
new species to the free operant! From the sensitive effects of
extended anxiety one hour after irradiation, but not three
hours after irradiation, I learned the incredible power and
sensitivity of the free operant method. This strengthened my
resolve to use the free operant with people.

Whenever Fred and I met to discuss my doctoral dissertation

on the effects of total body irradiation and drugs on beagle
dog discrimination and fear (Lindsley, 1957a), we always
strayed from the topic.

263
Figure 1. Hunter, the fastest learner of 67 beagle dogs,
pressing a panel to get a bit of hamburger delivered up
through the magazine hole to his right. Hunter survived his
dose of 50% lethal total body radiation and became the
household pet of Julie and Debby, B. F. Skinner’s daughters.

We wondered whether the catatonic schizophrenic standing in

a corner all day was the result of total extinction. We
wondered whether the hebephrenic was reinforced for
giggling on a variable ratio schedule. If so, all we had to do
was find a reinforcer and shape them back to their normal
performance.

Fascinated, I promised Fred that if he could get funds, I

would give human free operant research with psychotics five
years of my life. If it didn’t pan out, my parachute plan was to
go to Ringling Brothers Circus and shape Gargantua the

264
gorilla to play a piano and simple card games. I ended up
spending eleven and a half years studying psychotics from the
back wards of Metropolitan State Hospital.

Skinner got support from the Milton Fund of Harvard and the
U. S. Office of Naval Research. Harry Solomon, chair of
Psychiatry at Harvard and Commissioner of Mental Health
for Massachusetts agreed to serve as a co-investigator. I
started in June 1953 with $7,500 for the year in “A”
Basement (an abandoned hydro therapy unit) of Metropolitan
State Hospital, Waltham, Massachusetts. We studied both
acute and chronic male and female psychotic patients, autistic
children, and even the most violent patients from the locked
wards (it was before routine drug therapy). We studied
attendants, adult volunteers, and school children as normal
controls.

265
Figure 2. Chronic psychotic patient, diagnosed catatonic
schizophrenic, standing in his usual position in a corridor of
Harvard’s Behavior Research Laboratory waiting for his
daily session in one of the experimental rooms behind the
doors to the left.

Figure 3 shows ten minutes of the one hour cumulative record

of session 91 for patient number 46 at the right pf the photo.
The record shows he pulled the plunger 220 times for 13
reinforcements (diagonal hatch marks) on a fixed ratio 20
schedule. He paused 11 times (flat places on record), but not
immediately after reinforcement as do normal animals and
normal human adults and children. The patient showed two
aspects of normal fixed ratio behavior, high speed and pauses.

266
However the third aspect of fixed ratio behavior, pausing
immediately after reinforcement was deficient.

Project Name Selection

Skinner named our project “Experimental Analysis of the
Behavior of Psychotic Patients” (Skinner, Solomon, &
Lindsley, 1954; Lindsley & Skinner, 1954). Hospital staff,
patients, and parents reacted negatively to the words
“Experimental Analysis.” I made a list of over 12 possible
names and chose “Studies in Behavior Therapy.” Market tests
of the name with staff, patients, and family were positive. I
liked the words because they meant we treated behavior
problems with behavior. Skinner and Solomon approved this
name for our project (Lindsley, Skinner, & Solomon 1953,
1954a, 1954b). This was the first use of the name “Behavior
Therapy.”

267
Figure 3. Head banging psychotic on his way from male
violent ward to Behavior Research Laboratory for his 91st
experimental session. The chart at the right shows ten minutes
of his experimental session, which is described in the text.

After two years we were accepted as part of the hospital staff

by patients and families. Our stationary and business cards
were used up. We felt secure enough to change our name to

268
“Harvard Medical School Behavior Research Laboratory”
(Lindsley, Skinner, & Solomon 1955). Strangers would read
it in the telephone book and call up and ask, “Do you do
laboratory research on behavior?” Mary Hall, our secretary,
would laugh and answer, “Of course!” This was the first use
of the name “Behavior Research.”

Operandum Design
Many chronic psychotic patients occasionally become highly
destructive with no advance warning. They throw objects,
smash chairs, and break windows. This is the behavior that
keeps them in the hospital, and this is the behavior we must
study and understand. In order to record the behavior of such
patients while in their destructive episodes, we needed
indestructible rooms, signals (stimuli), operating switches
(operanda), and reward delivery magazines. The operanda had
to be able to be moved easily at frequencies above 300 per
minute so there would be no ceiling on response frequency.
No commercially available switches met these demands, so
we designed and Ralph Gerbrands built, our “Lindsley
Operandum.” Other laboratories later purchased this
operandum from Gerbrands, Inc.

Figure 4 shows the condition of the experimental room after a

violent patient had a destructive episode during the
experimental session. Note the chair is destroyed, but the
plexiglass windows to left and right used to yoke rooms or
display picture stimuli and rewards are intact. The work panel
with a signal light above each of two plunger operanda and
the magazine delivery chute on the right is intact. The patient

269
broke his chair against the work panel, but did not harm our
indestructible panel.

Symptom Recorders
To record psychotic symptoms we used electrical mats on the
floor to record pacing, and voice operated relays hidden in the
ceiling to record vocal hallucinating (talking and yelling to no
one). These frequencies were recorded minute to minute on
cumulative recorders with electrical counters for the hourly
and daily totals. Three recorders ran through each session, a
manual work recorder, a pacing recorder, and a vocal
hallucinating recorder (Lindsley, 1959, 1963a).

Figure 5 diagrams the room and apparatus for directly and

continuously recording the vocal and pacing symptoms and
the manual plunger pulling for reinforcements. You can see
the cumulative recorders at the bottom of the relay racks at
the right apparatus alley in figure 6.

Reinforcer Search
I designed and Gerbrands built universal magazines that
would carry anything from a penny, an M&M candy, a
cigarette, or a slice of apple, to a package of cigarettes, and
deliver them rapidly down a chute into the experimental
room. You can see four universal feeders mounted up the
outside back of the experimental rooms at the left of figure 6.
We never found adequate rewards for several of the patients
(Lindsley 1956a). In vain we tried projected 2 x 2 slides of
various subjects, including nude women for the men. We tried

270
various music selections and movies, but the silent periods in
the image and sound when the frequency of responding
dropped disturbed the viewers.

Figure 4. Interior of experimental room at end of one hour

experimental session during which a violent patient had a
destructive episode.

271
Figure 5. Diagram of experimental room and apparatus for
recording vocal symptoms, manual plunger pulling for
reinforcements, and pacing symptoms.

272
Figure 6. Apparatus Alley, a long corridor behind the six
experimental rooms contained the controlling and recording
apparatus. Details described in the text.

Figure 6 shows apparatus alley. You can see the universal

magazines mounted vertically along the left back wall of the
experimental rooms. The near magazine contains assorted
candies on the trays about to be delivered. At the second room
down William Nichols stands with his head under a periscope
observation hood secretly watching the patient working inside
the room. At the right you see vertical wheeled relay racks
with the controlling apparatus, counters, timers and recorders.
See four reset able counters at the top of the near rack. At the
top of the next rack hang two timing tapes. Note the
cumulative recorders on shelves with baskets to o catch their
records as they roll out. I am the young bearded man wearing

273
hospital keys on a chain and standing while recording some
counts on a standard data card resting on top of the fourth
recorder down.

Conjugate Reinforcers
We designed and built the first conjugate reinforcer to
continuously present narrative movies and music without the
brief pauses that destroy narration and mood. The reward is to
bring the image or sound louder, closer, or more in focus. The
rate of response is directly linked to the intensity of the video
or audio channel. The faster the patient presses one switch the
louder the audio; the faster they press the other switch the
brighter the projected image.

Figure 7 diagrams the two rooms linked by conjugately

controlled closed circuit televised psychotherapy sessions.
The patient pressed one switch to listen to her therapist and
the other switch to look at her therapist (Lindsley 1963b,
1969). We found looking and listening were independent,
sometimes occurring together and at other times singly. The
looking and listening were related to therapeutic content.
Conjugately reinforced loud noises went deeply into sleep
(Lindsley 1957b), anesthesia (Lindsley, Hobika, & Etsten,
1961), coma (Lindsley & Conran, 1962), and infancy (Lipsitt,
Pederson, & Delucia, 1966), but still did not generate
behavior from our most withdrawn chronic psychotics.
Conjugately reinforced television commercials were easily
calibrated (Lindsley 1962c), and the conjugate schedule is
sensitive enough to record preference for stereo over mono
phonic music (Morgan & Lindsley, 1966). Over a hundred

274
studies using the conjugate schedule have since been
reviewed (Rovee-Collieer & Gekoski, 1979).

Reinforcer Behavior Therapy

We eliminated and reduced symptom frequencies in some
patients with differential reward methods (Barrrett, 1962;
Lindsley, 1959). However, there were some patients for
whom we never found a useful reinforcer.

Figure 7. Diagram of experimental rooms and apparatus for

recording a patient’s listening and looking at her
psychotherapist throughout their daily session.

275
Social Reinforcers
In attempts to see if patients were too guilty to work to reward
themselves, we tried giving them the opportunity to feed a
hungry kitten as a reward (Lindsley, 1956b). We yoked two
experiment al rooms to see whether patients would work to
reward a friend, attractive member of the opposite sex, or a
stranger (Cohen, 1962; Cohen & Lindsley, 1964). We
generated cooperation between children without giving
instructions by using reward contingencies alone (Azrin &
Lindsley, 1956). None of these attempts were effective with
our most depressed inactive patients.

Simultaneous Discrimination and

Differentiation SIDAD
One of our most powerful diagnostic methods used a panel
with two signal lights. Each was lit for one minute as they
switched back and forth. A plunger under each light could be
pulled singly or both at once. This work panel is shown in
figure 4. Pulling the left plunger with the left light on was
reinforced on a fixed ratio 10 schedule (every tenth response
rewarded) with a coin or candy. Pulling the left plunger with
the right light on, or the right plunger with the left light on, or
the right plunger with the right light on was never reinforced.
We recorded each of these four reflexes separately on
counters and cumulative recorders. A fifth recorder
continuously recorded simultaneous plunger pulls (within 125
milliseconds of each other) which were never reinforced
(Barret & Lindsley, 1962).

276
Simultaneous learning to discriminate (tell the lights apart)
and differentiate (tell the plungers apart) could be seen
developing on the five recorders. Learning deficits in
nonverbal and violent patients were easily diagnosed and
compared with the learning of normal children and adults
(Lindsley, 1962a).

Free Operant Equation, COLAB and

IS — DID
It was difficult to determine what part of free operant
behavior was deficient in the patients who could not perform.
The components of a single operant reflex are, in order: (1)
stimulus; (2) response; (3) reinforcing contingency; and (4)
consequence or reinforcement. In diagramming the analysis
we used capital letters as the symbols for each component and
separated them by dashes:

In analyzing a deficient operant to determine which

component may be causing the deficiency, each component
must be independently tested for operant function on each
individual. We gave another set of names to events whose
operant function is still unproven, to clearly separate them
from components with proven function. These terms are: (1)
programmed event; (2) movement; (3) arrangement; and (4)
arranged event. They are diagrammed as follows:

277
An example shows how this notation system helped us
determine which component of a deficient operant was not
functioning. If a child pulled a plunger on the wall when the
light was on, every tenth pull produced a piece of candy. The
child pulled the plunger and got some candy, but she pulled
when the light was off as much as when it was on. What can
we say, except that the child could not learn to pull only when
the light was on? The operant equation now reads:

Any one of these four components could be deficient. An

operant equation with four unknowns cannot be solved for
one unknown.

A prosthetically oriented educator would not assume the child

has a visual problem. Nor would he try other rewards. Nor
would he change the contingency arrangement. Instead he
alters the movement component, because he had notice the
plunger pulling seemed “rhythmic” or “mechanical.” So, our
educator substitutes jumping-on-a-pedal for
pulling-the-plunger as the movement component:

Our child only jumps for candy when the light is on! Our
educator has now discovered four operant components and

278
can assign operant functions to the components in the child’s
equation:

The child has visual discrimination Ability, jumping as an

operant response, will work on fixed-ratio 10, and is
reinforced by candy. She has a response function deficit using
her hands, but does not have restricted manual movement.

Our educator analyst can now use the three components with
proven function to test to see whether the child will work for
tokens:

This operant equation, containing three components with

known function, can be solved for the unknown consequence
(token) being tested. We must have only one unknown in the
behavior equation we are trying to solve. These analytic
procedures were further detailed in Lindsley, 1964. They later
were named a Common Language for Analyzing Behavior
(COLAB), and still later named IS-DOES (a set of terms for
what it IS, and a set for what it DOES). Still later, I named
them IS-DID in an attempt to make people not use the DOES
terms until they had actually proven them with that child.

279
Ten Year Data Histories
Our core group of 50 male chronic psychotic patients
participated in our rooms each weekday for as long as 10
years. Several patients had 25 to 29 day rhythms in their
performance which we tried to relate to phases of the moon,
sun spots, temperature and humidity fluctuations, but none of
these held up. Patients whose 10 year histories were without
rhythms were good for measuring drug effects because we
could rule out mood swings, hospital events, and family
visits.

Seven Hour Drug Sessions

Since most psychotic episodes of hyper active shouting to no
one or stereotyped pacing in circles lasted from 15 to 45
minutes, we needed sessions as long as two hours to capture
an episode from beginning to end. To record the onset and
duration of an oral or intra muscular injection of a drug our
nurse entered the room after 15 minutes and injected the drug
or a placebo. Usually the drug had its onset effects within 15
to 30 minutes after injection. After about 5 hours the effects
wore off. Therefore we needed 7 hour sessions to record the
full effect of a drug on a patient’s work, pacing, and vocal
symptoms (Lindsley, 1962b).

Figure 8 shows that fifteen minutes into session 746 our nurse
injected 20 mg. of Benactyzine intramuscularly and the
effects appear in the top panel. Fifteen minutes into session
747 our nurse injected a lactose placebo whose effects are
shown in the bottom panel. Notice that the Benactyzine

280
injection produced sustained vocal hallucinating for a about
four hours. During this time the manual work was depressed.
Also notice that the vocal cumulative record is less steep for
the first 2 hours representing a rate of about 12 per minute.
During this time his plunger pulling work fell off from his
normal rate of 160 per minute to 6 per minute. During the
second 2 hours after injection the vocalizing almost doubled
to 21 per minute while his plunger pulling work gradually
returned to normal. Benactyzine produced a psychotic type
episode lasting almost 4 hours. In the placebo session 747 a
short 25 minute and a longer 40 minute psychotic episode
occurred. This patient number 7 displayed 15 to 45 minute
psychotic episodes of this type every few hours on his ward.

The immediate effects of injected drugs on seven hour

sessions produced the same effects as did routine clinical oral
administration which usually took weeks to develop. These
immediate effects permitted us to screen new drugs with
suggested psycho active potential 5 or 6 times faster than the
usual oral administration and ward behavior observation.

Coextensive Reflex Emission

These seven hour sessions permitted us to view and quantify
the interactions between episodes of psychotic vocalizing and
pacing with manual working. With 16 patients these episodes
were independent as with normal adults and children. In 6
patients, all diagnosed schizophrenic, these episodes
coextended of the same time interval. They appeared linked
together (See the records of P20, P59, P56 in figure 9), or
alternated (See P58 in figure 9). All normal controls manually
worked throughout the seven hours with no vocalizing and

281
only a very few brief pacings (See NC 45 in figure 7). The
normals did not stop working during their brief pacing
episodes.

Figure 8. Six hour cumulative response records of never

rewarded hallucinatory vocalizing (VOC EXT) and manual
plunger pulling reinforced with candy on a 1 minute variable
interval schedule (MAN CAN 1’VI) of Benactyzine session
746 and lactose placebo session 747 for Patient number 7.

282
Figure 9. Bar diagrams of 7 hour sessions comparing
emission of never rewarded vocalizing (thin top bar), manual
working for rewards (thick middle bar), and never rewarded
pacing (thin bottom bar) for 8 psychotics, 1 retarded child
(RG), and 1 normal child (NC).

Figure 9 displays bar diagrams for the never rewarded

vocalizing (thin top bar), manual plunger pulling for rewards
(thick middle bar)., and never rewarded pacing (thin bottom
bar) over continuous 7 hour sessions. The performance of
eight adult male psychotic patients, a retarded child (RC), and
a normal child (NCV). The retarded child also had emotional
problems and had recently been thought of as emotionally
disturbed. Note the linked coextension during the last 2 hours
of his session.

283
Except for the vocal hallucinating and pacing symptoms, this
reflex coextension was the first emergent diagnostic item that
we found. All other diagnostic items that we found were
deficiencies, a decrease or absence of some aspect of normal
performance.

Folly of Drug Screening

Our seven hour intra muscular injection drug sessions
predicted the response of patients to long term (3 months) oral
clinical administration of the same drug. This meant we could
screen new drugs at the rate of one a week on ten psychotic
patients. This screening efficiency appealed to drug
companies and to the National Institute of Mental Health. We
had a screening device. We could screen one new drug a
week on the normal work, pacing and hallucinating of chronic
and acute psychotic patients.

We wondered what was the past success of scientists with

screening devices in searching for effective active
compounds? How many drugs must we try before we would
find one that would reduce psychotic symptoms and at the
same time restore normal work performance? Edison said he
screened 3,000 materials before he found the carbon filament
for his electric bulb (Josephson, 1959). Marie Curie tried
hundreds of salts, oxides, and ores before she found the radio
active pitchblende and chacolite (Pflaum, 1993). Salvarsan™
(arsphenamine), an early treatment for syphilis, was the 606th
compound the company tried. The range seemed to be 3,000
to 300 barren attempts before screening success.

284
If we tried a new compound a week and were lucky we might
expect to find a psychotic treatment drug in 300 trials, or 6
years, at the rate of one trial a week. The hooker was that the
drug companies were only producing 5 to 10 per year that
were of human toxicity. That would take 30 years if we were
as lucky as Curie and 300 years if we had Edison’s luck!

The kiss of death to our drug screening came in 1957 to 1962

when Thalidomide taken in even a single dose caused
pregnant women to miscarry or give birth to horribly
deformed babies. Only 17 Thalidomide babies were born in
the United States to women who got the drug illegally, since
it had not been approved for use in the US. However, the
Thalidomide scare caused a shut down in new drugs approved
by the United States Health Service for human trial. New
drugs available for trial with psychotics went from 5 to 10 per
year to 1 in 5 years. With these restrictions, if we had Curie’s
luck, we would find an effective drug in 5 x 300 or 1500
years! So ended our drug screening plans.

Lab Visitors
Our laboratory guest book shows that in the twelve years
from 1953 through 1964 ninety eight university classes from
ten universities with a total of 1857 students spent day long
field trips in our laboratory. Nine Hundred thirty seven
professionals visited our laboratory from the United States
and over seventeen other countries. Visiting psychiatrists and
psychologists included Carl Rogers, Harry Harlow, Roy
Menninger, Piere Pichot, Paolo Nuzzi, Koji Sato, Hudson
Hoagland, Otto Kernberg, Andey Snejnenski, Carl Pfaffman,
Frank Beach, Donald Lindsley, Carl Pribram, Joe Zubin, and

285
Timothy Leary. Visiting behavior analysts included Don
Baer, Harold Weiner, William Morse, Ted Allyon, Dale
Brethower, Matthew Israel, Charlie Catania, Thom Verhave,
Joe Brady, and Charles Ferster. Sidney Bijou visited in April
1957, November 1961, and April 1962. Bijou built a similar
laboratory for children at the University of Washington, as
did Ferster at the University of Indiana Medical School, Azrin
at Anna State Hospital in Southern Illinois, and Barrett at
Fernald State School, Waltham, Massachusetts.

Research Trainees
Larry Fane and Donald Cohen conducted undergraduate
honors theses. Julie Rich and Barbara Morgan conducted their
Master’s theses. Post Doctoral trainees were Nathan Azrin,
Beatrice Barrett, Peter Nathan, Martha Mednick, and Paul
Blachly. Tom Gilbert conducted research in our laboratory
while on a University of Georgia sabbatical.

Awards
In 1962 the American Psychiatric Association awarded one of
our research papers its annual Hofheimer Research Prize
(Lindsley, 1960). In 1964 the American Academy of
Achievement awarded us its Golden Plate Achievement
Award. These twelve pioneering years were cited in two
recent awards, the 1998 Thomas F. Gilbert Distinguished
Professional Achievement Award from the International
Society for Performance Improvement and the 1999 Award
for Distinguished Service to Behavior Analysis from the
Society for the Advancement of Behavior Analysis.

286
Appliers Abandon Frequency
Most of the behavior analysts who visited our laboratory and
then set up studies in clinics, hospitals, and schools did not
record their learner’s rate of response; they recorded percent.
To me this was a crisis because we had proven that frequency
was as much as l0 times more sensitive than percent. I
considered rate of response and the cumulative self-recorder
to be Skinner’s greatest contributions, and both were
discarded by the appliers. Azrin went so far as to say “suit the
metric,” which meant use a different measure for every
behavior you work with. It may have been easier to do and
much easier to sell, but such a loss of measurement standards
rules out real science which requires standard measures

Application research grew like wildfire compared to the

behavioral laboratory research that had originally triggered it.
The laboratories were expensive, hard to fund, and ignored by
both clinicians and small animal laboratory researchers.
Behavior modification, behavior therapy, and applied
behavior analysis were clearly going to dominate the field.
Unfortunately they left behind behavior frequency and self
charting on standard charts, Skinner’s most powerful
discoveries.

The crisis was clear. If something was not done soon,

frequency and standard self charting would die with Skinner!
A few began to apply operant methods to regular and special
education. They too did not use frequency or standard self
charting. They said teachers preferred percent correct, and
percent time on task. They said teachers were dead set against

287
student self charting. Clearly education was a larger industry
with far greater market potential than mental health.

Since I could not convince others to do it, I realized that I

would have to put frequency and standard self charting into
school classrooms. The combination of too few new drugs to
try, increased university overhead charges, increased
competition for smaller and smaller government research
grants, lack of interest in our results, and the crisis of losing
frequency and standard self charting to multiplying
applications, made continuing our laboratory research a poor
choice.

In January 1965, I closed our laboratory and parachuted

further into teacher education at the University of Kansas
Medical Center than the percent correct appliers had yet gone.
My mission was to get teachers using frequency and students
charting their own performance. There we developed
Precision Teaching and the Standard Celeration Chart used by
students to self chart their learning and make their own
improvement decisions (Lindsley 1972, 1996, 1997). But
that’s another story.

References
Azrin, N. H., & Lindsley, O. R. (1956). The reinforcement of
cooperation between children. Journal of Abnormal and
Social Psychology, 52, 100-102.

Barrett, B. H. (1962). Reduction in rate of multiple tics by

free operant conditioning methods. Journal of Nervous and
Mental Disease, 135, 187-195.

288
Barrett, B. H., & Lindsley, O. R. (1962). Deficits in
acquisition of operant discrimination and differentiation
shown by institutionalized retarded children. American
Journal of Mental Deficiency, 67, 424-436.

Cohen, D. J. (1962). Justin and his peers: An experimental

analysis of a child’s social world. Child Development, 33,
697-0717.

Cohen, D. J., & Lindsley, O. R. (1964). Catalysis of

controlled leadership in cooperation by human stimulation.
Journal of Child Psychology and Psychiatry, 5, 119-137.

Jetter, W. W., Lindsley, O. R., & Wohlwill, F. J. (1953). The

effects of irradiation on physical exercise and behavior in the
dog: Related hematological and pathological control studies
(AEC Contract AT No. 30-1, pp. 1201). Boston University
Medical School.

Josephson, M. (1959). Edison: A biography. New York:

McGraw-Hill. pp. 207.

Lindsley, O. R. (1950). Neural components of the chorda

tympani of the rat. Unpublished masters thesis, Brown
University, Providence, RI.

Lindsley, O. R. (1956a). Operant conditioning methods

applied to research in chronic schizophrenia. Psychiatric
Research Reports, 5, 118-139.

Lindsley, O. R. (1956b). Feeding a kitten — a social

reinforcer. In annual technical report #3, November, Contract

289
N5-Ori-07662, Office of Naval Research. Waltham MA:
Harvard Medical School, Behavior Research laboratory.

Lindsley, O. R. (1957a). Conditioned suppression of behavior

in the dog and some sodium pentobarbital effects.
Unpublished doctoral dissertation, Harvard University.

Lindsley, O. R. (1957b). Operant behavior during sleep: A

measure of depth of sleep. Science, 126, 1290-1292.

Lindsley, O. R. (1959). Reduction in rate of vocal psychotic

symptoms by differential positive reinforcement. Journal of
the Experimental Analysis of Behavior, 2, 269.

Lindsley, O. R. (1960). Characteristics of the behavior of

chronic psychotics as revealed by free-operant conditioning
methods [Monograph]. Diseases of the Nervous System, 21,
66-78.

Lindsley, O. R. (1962a). Operant conditioning methods in

diagnosis. In J. H. Nodine & J. H. Moyer (Eds.),
Psychosomatic medicine: The first Hahnemann symposium
(pp. 41-54). Philadelphia: Lea & Febiger.

Lindsley, O. R. (1962b). Operant conditioning techniques in

the measurement of psychopharmacologic response. In J. H.
Nodine & J. H. Moyer (Eds.), Psychosomatic medicine: The
first Hahnemann symposium (pp. 373-383). Philadelphia: Lea
& Febiger.

Lindsley, O. R. (1962c). A behavioral measure of television

viewing. Journal of Advertising Research, 2, 2-12.

290
Lindsley, O. R. (1963a). Direct measurement and functional
definition of vocal hallucinatory symptoms. Journal of
Nervous and Mental Disease, 136(3), 293-297.

Lindsley, O. R. (1963b). Free-operant conditioning and

psychotherapy In J. H. Masserman (Ed.), Current psychiatric
therapies (Vol. 3) (pp. 47-56). New York: Grune and
Stratton.

Lindsley, O. R. (1964). Direct measurement and prosthesis of

retarded behavior. Journal of Education, 147, 62-81.

Lindsley, O. R. (1969). Direct behavioral analysis of

psychotherapy sessions by conjugately programmed
closed-circuit television. Psychotherapy: Theory, Research,
and Practice, 6, 71-81.

Lindsley, O. R. (1972). From Skinner to Precision Teaching:

The child knows best. In J. B. Jordan & L. S. Robbins (Eds.),
Let’s try doing something else kind of thing: Behavioral
principles and the exceptional child (pp. 1-11). Arlington,
VA: Council for Exceptional Children.

Lindsley, O. R. (1996). Is fluency free-operant

response-response chaining? The Behavior Analyst, 19(2),
211-224.

Lindsley, O. R. (1997). Precise instructional design.:

Guidelines from Precision Teaching. In C. R. Dills & A J.
Romiszowski (Eds.), Instructional development paradigms
(pp. 537-554). Englewood Cliffs, NJ: Educational
Technology Publications.

291
Lindsley, O. R., & Conran, P. (1962). Operant behavior
during EST: A measure of depth of coma. Diseases of the
Nervous System, 23, 407-409.

Lindsley, O. R., Hobika, J. H., & Etsten, B. E. (1961).

Operant behavior during anesthesia recovery: A continuous
and objective method. Anesthesiology, 22, 937-946.

Lindsley, O. R., & Jetter, W. W. (1953). The temporary

elimination of discrimination and fear by sodium
pentobarbital injections (dog). American Psychologist , 8,
390.

Lindsley, O. R., & Skinner, B. F. (1954). A method for the

experimental analysis of the behavior of psychotic patients.
American Psychologist, 9, 419-420.

Lindsley, O. R., Skinner, B. F., & Solomon, H. C. (1953).

Study of psychotic behavior, Studies in Behavior Therapy,
Harvard Medical School, Department of Psychiatry,
Metropolitan State Hospital, Waltham, MA, Office of Naval
Research Contract N5-ori-07662, Status Report I, 1 June 1953
- 31 December 1953.

Lindsley, O. R., Skinner, B. F., & Solomon, H. C. (1954a).

Study of psychotic behavior, Studies in Behavior Therapy,
Harvard Medical School, Department of Psychiatry,
Metropolitan State Hospital, Waltham, MA, Office of Naval
Research Contract N5-ori-07662, Status Report II, 1 January
1954 - 31 May 1954.

Lindsley, O. R., Skinner, B. F., & Solomon, H. C. (1954b).

Study of psychotic behavior, Studies in Behavior Therapy,

292
Harvard Medical School, Department of Psychiatry,
Metropolitan State Hospital, Waltham, MA, Office of Naval
Research Contract N5-ori-07662, Status Report III, 1 June
1954 - 31 December 1954.

Lindsley, O. R., Skinner, B. F., & Solomon, H. C. (1955).

Study of psychotic behavior, Behavior Research Laboratory,
Harvard Medical School, Department of Psychiatry,
Metropolitan State Hospital, Waltham, MA, Office of Naval
Research Contract N5-ori-07662, Status Report IV, 1 January
1955 - 31 August 1955.

Lipsitt, L., Pederson, L. J., & Delucia, C. A. (1966).

Conjugate reinforcement of operant responding in infants.
Psychonomic Science, 4, 67-68.

Morgan, B., & Lindsley, O. R. (1966). Operant preference for

stereophonic over monophonic music. Journal of Music
Therapy, 3, 135-143.

Pflaum, R. (1993). Marie Curie and Her Daughter Irene.

Minneapolis: Lerner Publications

Rovee-Collier, C. K., & Gekoski, M. J. (1979). The

economics of infancy: A review of conjugate reinforcement.
Advances in Child Development, 13, 195-255.

Skinner, B. F., Solomon, H. C., & Lindsley, O. R. (1954). A

new method for the experimental analysis of the behavior of
psychotic patients. Journal of Nervoius and Mental Disease,
120(5), 403-406.

293
Chapter 7

A Brief Personal Account of

CT (Conditioning Therapy),
BT (Behavior Therapy) and
CBT (Cognitive-Behavior
Therapy): Spanning Three
Continents
Arnold A. Lazarus

Rutgers University and the Center for Multimodal

Psychological Services

Conditioning Therapy
In South Africa, the forerunners of what came to be known as
“Behavior Therapy,” were Joseph Wolpe (a general medical
practitioner), James Taylor (a university-based psychologist),
Abe Adelstein (an epidemiologist and methodologist),
Cynthia Adelstein (a psychologist), and Leo Reyna, an
American psychologist who served as a senior lecturer at the
University of the Witwatersrand in Johannesburg from 1946
to 1950. The foregoing individuals guided and inspired
Wolpe and were the driving forces behind the well-known

294
experiments that he conducted on cats. Wolpe first submitted
the dissertation that resulted from his animal experiments to
the Department of Psychiatry at the University of the
Witwatersrand. Given that he had never received any formal
training in psychiatry, they turned it down. Similarly, the
Department of Psychology rejected it because Wolpe had
never enrolled in any psychology courses. Finally, given that
he had graduated from the medical school with an M. B., B.
Ch. (Bachelor of Medicine and Bachelor of Surgery), the
Department of Medicine awarded him an M.D. degree in
1948.

In 1956 when I was a graduate student in psychology at the

University of the Witwatersrand one of the senior lecturers
arranged for Wolpe to give talks and demonstrations of his
CT (conditioning therapy). Wolpe was in full-time private
practice, but was not allowed to charge specialist fees because
he was not a licensed psychiatrist. This factor coupled with
the untenable political climate in South Africa led him to seek
for greener pastures. Thus, when he received an appointment
in 1960 at the University of Virginia School of Medicine, he
emigrated from South Africa. This did not occur before he
had gathered together a coterie of clinicians and theoreticians
who were interested in learning what the establishment
thinkers called “ideas from the lunatic fringe.” I became a
member of this august group, although it took some time to
shed the psychoanalytic precepts that had been drilled into
me. S. Rachman, also a graduate student, was an important
member of this cabal. He proclaimed Wolpe “The King,” and
we vied for the position of Crown Prince. Rachman went to
London to study with Eysenck, whereas I continued working
with Wolpe who chaired my doctoral dissertation. In 1960,
when Wolpe left for America, I, a newly minted Ph.D., (but a

295
registered clinical psychologist) inherited his private practice.
I also became the leader of the training group seminar in
Johannesburg that Wolpe had launched. Later, Rachman was
awarded a Ph.D., from the University of London, soon
became Eysenck’s right hand man, and went on to a
distinguished career as a writer, editor and clinical
experimenter.

Whereas the mainstream theorists and clinicians in South

Africa focused essentially on putative dynamic insights and
psychoanalytic concepts, the CT meetings were devoted to
learning theory and the use of exciting new techniques – all
old hat today – imaginal and in vivo desensitization,
relaxation procedures, assertiveness training, thought
stopping, aversion relief conditioning, and the like. Animal
analogues played a major role at this juncture. Methods or
ideas that could not be tested in the animal laboratory were
deemed unacceptable.

In 1957, at one of the Wolpe group meetings I proposed that

we drop the term “Conditioning Therapy” because too many
people tended to attach pejorative connotations to it of bells,
whistles and salivating dogs. Instead, why not call ourselves
“Behavior Therapists” and describe our area of interest as
“Behavior Therapy?” After all, I stated, our main focus is on
behavior and the remediation of maladaptive behaviors. This
suggestion was not well received. The main objections, as I
recall, were that we should not pander to ignorance. Why
permit the prejudices of unenlightened people to sway us?
Moreover, the New York-based practitioner Andrew Salter
(another worthy contender for the title of the father of
behavior therapy) reported no down side from calling his
book Conditioned Reflex Therapy (1949). Wolpe’s

296
commitment to CT is exemplified in the proceedings of an
important conference held at the University of Virginia that
was published under the title The Conditioning Therapies
(Wolpe, Salter & Reyna, 1964). The conditioning label
remained quite popular. Thus, Franks (1964) edited a book he
called “Conditioning Techniques in Clinical Practice and
Research.” Nevertheless, I had published an article (Lazarus,
1958) in which I put forward the terms “behavior therapy”
and “behavior therapist.” Later, Eysenck (1959)
independently used these terms in print, but I lay claim to
having been the first person to use them in a scientific journal.
(Wolpe, 1968, discovered that in 1953, Skinner, Lindsley and
their associates, working at the Metropolitan State Hospital,
Waltham, Massachusetts had some mimeographed status
reports on operant conditioning with psychotic inmates that
they referred to as “Studies in Behavior Therapy.”)

Behavior Therapy
After Eysenck (1960, 1964) edited two books on behavior
therapy, the term became more widely disseminated and won
out over Conditioning Therapy and other designations that
had been proposed — e.g., “Behavioristic Psychotherapy,”
“Objective Psychotherapy,” and “Reciprocal Inhibition
Therapy.” Subsequently, Cyril Franks founded the
Association for Advancement of the Behavioral Therapies in
1966 (soon afterward called the Association for Advancement
of Behavior Therapy) and he later edited a highly significant
book on behavior therapy (Franks, 1969). In the 1960’s and
1970’s, despite denouncements from critics, skeptics and
detractors, there was a high degree of interest in behavioral
methods among researchers, theoreticians and clinicians. In

297
1963 when Albert Bandura invited me to spend a year at
Stanford University training graduate students in behavioral
theories and methods, enthusiasm for this novel and
promising approach was high. (My favorite student was a
very bright, personable and energetic fellow named G. C.
Davison who, as you well know, has played a major role in
the field.)

Interestingly, the emphasis in behavioral methods in South

Africa and England was primarily Pavlovian and Hullian (see
Eysenck, 1956) and focused on respondent conditioning,
whereas in the United States, the work of Skinner on operant
conditioning served as the mainstay. Well-known names in
this arena include Ayllon, Goldiamond, Ferster, Wolf, Risley,
Patterson, Baer, Azrin, Lindsley, Bijou, and many others. The
term “behavior modification” was preferred to “behavior
therapy.” An influential book edited by Ullmann and Krasner
(1965) blended the two concepts. Bandura’s (1969) seminal
studies on social learning theory and modeling served to
broaden the base of behavioral interventions. Perry London’s
(1964) book The Modes and Morals of Psychotherapy was
widely read and served to pave the way for technical
eclecticism and a more elegant behavioral tradition.

From 1960 onward, data driven research on behavioral

theories and methods was conducted at many centers. The
annual conventions of the Association for Advancement of
Behavior Therapy (AABT) traversed such topics as the use of
reinforcement schedules, modification of smoking behavior,
aversion relief treatment of obsessive neurosis, the use of
positive and aversive imagery, the specific effects of
modeling and role playing, community-based operant
learning environments, and other innovative procedures (see

298
Rubin, Fensterheim, Lazarus, & Franks, 1971). Populations to
which these new behavioral methods were applied included
such areas as schizophrenia, alcoholism, mental retardation,
geriatrics, school settings, and juvenile delinquency. Indeed,
the proliferation of books, articles, and popular publications
on behavior therapy led Franks and Wilson (1973) to bemoan
the fact that “quantity is accelerating at a far greater rate than
quality” (p. vii). Thus, they launched their Annual Review of
Behavior Therapy to provide an integrated distillation of the
vast literature. They also offered trenchant commentaries that
placed the many developments in perspective. These erudite
800-page tomes were extraordinarily illuminating, and it
always astonished me that Franks and Wilson managed to
address and fulfill all their other academic and clinical
demands and still produce these volumes. Each year from
1973 to 1979 the monumental volumes appeared on time.
Gradually, it became evident that the commentaries by Franks
and Wilson were becoming equal in length to the reprinted
articles. It also became clear that most readers were much
more interested in the commentaries than in the reprinted
material. Consequently, for Volume 8, Kendall and Brownell
were recruited to add their specific areas of expertise to the
enterprise. With the advent of Volume 12 written (not edited)
by Franks, Wilson, Kendall and Foreyt (1990), the question
was posed as to whether there was still a need for these
publications. The cost-benefit ratio seemed disproportionate
and the series ended with Volume 12. At this stage, formal
behavior therapy had been in existence for over 30 years.

299
Cognitive-Behavior Therapy
Some of the personages associated with Cognitive-behavior
therapy (CBT) include Ellis, Beck, Meichenbaum, Goldfried,
Mischel, Davison and Mahoney. My book Behavior Therapy
and Beyond (1971/1996) is arguably one of the first books on
cognitive-behavior therapy. It soon became a Citation Classic.
In the 1977 Annual Review of Behavior Therapy, the
transition from BT to CBT was discussed, and Cyril Franks
used the term “cognitive-behavior therapy” for the first time
in his overview.

One of my first forays into cognitive zones occurred when I

pointed out to Wolpe that during imaginal desensitization, a
finger signal or a head nod, or whatever other nonverbal sign
of distress had been agreed upon, merely indicated that
discomfort, or displeasure was being experienced. It did not
speak to the content of the uneasiness or anxiety. When a
client signaled distress during a desensitization procedure, the
standard response was to say, “Stop picturing that scene, take
in a few deep breaths and go back to the pure relaxation.” But
I pointed out that by examining the meaning behind the finger
signal, this often shed light on hitherto unknown components
and associations. Wolpe claimed that to conduct discussions
in the middle of the desensitization procedure would dilute
the process and interrupt the relaxation. This procedural
difference was perhaps the first indication that within a few
years, an extensive parting of the ways between us would take
place.

The aforementioned procedural shift was the harbinger of a

significant modification in my thinking. My understanding of

300
the process of change had gone from “insights into putative
unconscious complexes,” to “reciprocal inhibition,
counterconditioning, and extinction.” At this juncture, I
started to view cognitive restructuring as one of the primary
psychotherapeutic change agents. I was influenced by Ellis’s
(1962) Reason and Emotion in Psychotherapy, and London’s
(1964) The Modes and Morals of Psychotherapy, and began
to embrace the notion that the power of a person’s beliefs can
often override his or her operant or respondent conditioning.
When I stated that in addition to focusing on behavior,
elegant therapy called for attention to cognitive processes –
beliefs, attitudes, values, and opinions – my behavioral peers
were unimpressed. They saw it as an atavistic regression to
mentalism, and I was wrongly accused of being a closet
psychoanalyst.

I wrote a series of articles on what I termed “broad-spectrum

behavior therapy,” that culminated in my 1971 book Behavior
Therapy and Beyond. This book places emphasis on the
notion that “effective psychotherapy must teach people to
think, feel, and act differently“ (p. 166). This was the
beginning of an even broader application of treatment
dimensions that I termed “multimodal therapy” (see Lazarus,
1997) which Cyril Franks (1997) dubbed “behavior therapy in
one of its most advanced forms” (p. xii). The multimodal
orientation goes beyond the scope of this paper.

A Pivotal Case

A clinical event that played a crucial role in my outlook

occurred in 1966. I was treating a severely agoraphobic and
fearful 35-year-old woman. She was a cooperative and highly
motivated client who responded well to a standard range of

301
behavioral techniques consisting primarily of deep muscle
relaxation, imaginal and in vivo desensitization, behavior
rehearsal, and assertiveness training. After 5 months of
therapy she was able to enjoy taking long walks alone,
shopping, visiting and traveling without distress. Important
changes had accrued above and beyond the client’s capacity
to venture out of her home — she was no longer socially
submissive and enjoyed a wider range of social outlets; she
found that relaxation and positive imagery were capable of
quelling any residual fears or anxieties; and her marriage
relationship and sexual experiences were more gratifying.
Nevertheless, although the client was delighted by her
newfound ability to remain anxiety-free while traveling and
engaging in the niceties of social interaction, she continued to
view herself as a worthless person. She referred to herself as
being like a 12-year-old who was now able to cross the street
alone, but was contributing nothing to society. At this
juncture, what is now called 'cognitive therapy” was clearly
indicated, and we launched into an assessment of her more
fundamental attitudes and beliefs. This led her to conclude:
“If you want to feel useful, you have to be useful.”
Consequently, she founded an organization that distributed
basic essentials such as food and clothing to impoverished
people. This behavior, based upon her attitudes and
self-concept led her to view herself as “eminently
worthwhile.” In a follow-up interview she stated: “Thanks to
the fact that I exist and care, thousands of people now derive
benefit,” and she proclaimed herself “eminently worthwhile.”
This case is described in greater detail in Behavior Therapy
and Beyond (Lazarus, 1971/1996).

It was this woman who first led me to realize that “behavior

therapy” alone might be insufficient (to use a football

302
analogy) to take people into the end zone. It became quite
apparent to me that it was often necessary to venture beyond
the customary parameters of behavioral interventions into
such territory as values, attitudes and beliefs. Although the
early books on behavior therapy discussed the need to
“correct misconceptions” (Wolpe, 1958; Wolpe & Lazarus,
1966), the focus was solely on erroneous ideas and did not
address the realm of the client’s self-talk, his or her basic
values, or other cognitive processes. When I emphasized the
need to explore and modify such concerns, and when I drew a
distinction between what I termed “narrow band behavior
therapy,” and “broad-spectrum behavior therapy,” the
reactions from Wolpe and many of my fellow behavior
therapists were less than positive. Eysenck (1970), who was
after all a theorist who had never treated a patient in his life,
wrote a strident criticism and said that my ideas would lead to
“nothing but a mishmash of theories, a huggermugger of
procedures, a gallimaufry of therapies, and a charivaria of
activities having no proper rationale, and incapable of being
tested and evaluated” (p. 145). It took about ten years before
the need to add cognitive interventions to standard behavioral
methods became widely recognized. Goldfried and Davison
(1994) in their updated edition of Clinical Behavior Therapy
(first published in 1976) state: “One no longer needs to argue
for the admissibility of cognitive variables into the clinical
practice of behavior therapy. Indeed, more than two-thirds of
the membership of the Association for Advancement of
Behavior Therapy now view themselves as
“cognitive-behavior therapists’” (p. 282). As an aside, I took
Eysenck thoroughly to task in a chapter I called “On Sterile
Paradigms and the Realities of Clinical Practice,” (Lazarus,
1986).

303
As we move into the 21st Century, the emphasis on
empirically supported methods and the use of carefully
crafted treatment manuals have already begun to clarify the
issue of treatments of choice for specific disorders. They
supply much needed information about the active ingredients
of therapeutic techniques (see Wilson 1995, 1998). Add
various breakthroughs that have occurred in the biological
arena to the greater precision that is now taking place in the
field of cognitive-behavior therapy, and it becomes evident
that quantum leaps are in the immediate offing.

So what about the training for future generations? In my own

view, in addition to a thorough grounding in psychology and
biology, students must fully comprehend science,
methodology, and gain a meticulous understanding of the
difference between “data” and “anecdotes” (as well as a
fundamental schooling in treatments of choice for specific
disorders). Moreover, students need to be taught how to
implement relationships of choice (Lazarus, 1993). Much has
been written about “techniques of choice,” whereas
“relationships of choice” have more or less been taken for
granted. Trainees need to acquire a flexible repertoire of
relationship styles to enhance treatment outcomes. Decisions
regarding different relationship stances include when and how
to be directive, supportive, reflective, cold, warm, tepid,
formal, informal and so forth. Talking in generalities about
rapport, good working alliances, empathy, compatibility and
the like will not suffice.

One other important consideration is worth underscoring.

Good therapists, in my estimation, have few (if any)
“buttons” – hypersensitivities that can trigger untoward
affective reactions. For example, recently, two of my students

304
took offense at ethnic slurs uttered by clients during a therapy
session. In one case, the client had made an anti-Semitic
remark. The Jewish trainee took exception to it and handled
the matter emotionally, not clinically. In the other instance, a
Hispanic student refused to continue working with a client
who made an ethnic remark about Puerto Ricans. My
recommendation was that, at the very least, these students
should be urged to undergo a course of systematic
desensitization.

The foregoing considerations are idealistic. The vast majority

of practitioners are drawn from the ranks of counselors, social
workers, and psychotherapists who have scanty scientific
backgrounds, a distant awareness of psychological principles,
and very little understanding of basic behavioral facts and
factors. The marketplace is flooded with extremely poorly
trained personnel. Nevertheless, it is my contention that if we
can move beyond empirically supported methods and develop
explicit and well-established treatments of choice, it will
become mandatory for anyone who practices psychotherapy
to know how, when and where to administer what needs to be
employed. Having lived to see behavior therapy evolve (in the
span of about 40 years) from a laughable, lunatic fringe, naïve
and mechanistic joke, to an orientation that almost occupies
center stage, I can but hope that our findings will be
consolidated and extended throughout the 21st Century to the
benefit of all.

References
Bandura, A. (1969). Principles of behavior modification. New
York: Holt, Rinehart, & Winston.

305
Ellis, A. (1962). Reason and emotion in psychotherapy. New
York: Lyle Stuart.

Eysenck, H. J. (1956). Modern learning theory. Proceedings

of the Royal Society of Social Medicine, 49, 1024-1026.

Eysenck, H. J. (1959). Learning theory and behaviour

therapy. Journal of Mental Science, 105, 61-75.

Eysenck, H. J. (Ed.). (1960). Behaviour therapy and the

neuroses. Oxford: Pergamon Press.

Eysenck, H. J. (Ed.). (1964). Experiments in behaviour

therapy. Oxford: Pergamon Press.

Eysenck, H. J. (1970). A mish-mash of theories. International

Journal of Psychiatry, 9, 140-146.

Franks, C. M. (Ed.). (1964). Conditioning techniques in

clinical practice and research. New York: Springer.

Franks, C. M. (Ed.). (1969). Behavior therapy: Appraisal and

status. New York: McGraw-Hill.

Franks, C. M. (1997). Foreword. In A. A. Lazarus, Brief but

comprehensive psychotherapy: The multimodal way (pp.
ix-xii). New York: Springer

Franks, C. M., & Wilson, G. T. (Eds.). (1973). Annual review

of behavior therapy. New York: Brunner/Mazel.

Franks, C. M., Wilson, G. T., Kendall, P. C., & Foreyt, J. P.

(1990). Review of Behavior Therapy. New York: Guilford.

306
Goldfried, M. R., & Davison, G. C. (1994). Clinical behavior
therapy. New York, NY: Wiley.

Lazarus, A. A. (1958). New methods in psychotherapy: A

case study. South African Medical Journal, 32, 660-664.

Lazarus, A. A. (1971). Behavior therapy and beyond. New

York: McGraw-Hill.

Lazarus, A. A. (1986). On sterile paradigms and the realities

of clinical practice: Critical comments on Eysenck’s
contribution to behaviour therapy. In S. Modgil and C.
Modgil (Eds.), Hans Eysenck: Consensus and controversy.
London: The Falmer Press.

Lazarus, A. A. (1993). Tailoring the therapeutic relationship,

or being an authentic chameleon. Psychotherapy, 30,
404-407.

Lazarus, A. A. (1996). Behavior therapy and beyond.

Northvale, NJ: Jason Aronson.

Lazarus, A. A. (1997). Brief but comprehensive

psychotherapy: The multimodal way. New York: Springer.

London, P. (1964). The modes and morals of psychotherapy.

New York: Holt, Rinehart & Winston.

Rubin, R. D., Fensterheim, H., Lazarus, A. A., & Franks, C.

M. (Eds.). (1971). Advances in behavior therapy. New York:
Academic Press.

307
Salter, A. (1949). Conditioned reflex therapy. New York:
Creative Age Press.

Ullmann, L. P., & Krasner, L. (Eds.). (1965). Case studies in

behavior modification. New York: Holt, Rinehart, &
Winston.

Wilson, G. T. (1995). Empirically validated treatments as a

basis for clinical practice: Problems and prospects. In S. C.
Hayes, V. M. Follette, T. Risley, R. D. Dawes, & K. Grady
(Eds.), Scientific standards of psychological practice: Issues
and recommendations. Reno, Nevada: Context Press.

Wilson, G. T. (1998). Manual-based treatment and clinical

practice. Clinical Psychology: Science and Practice, 5,
363-375.

Wolpe, J. (1958). Psychotherapy by reciprocal inhibition.

Stanford, CA: Stanford University Press.

Wolpe, J. (1968). From the president. Newsletter of the

Association for Advancement of Behavior Therapy, 3, 1-2.

Wolpe, J., Salter, A., & Reyna, L. J. (Eds.). (1964). The

conditioning therapies. New York: Holt, Rinehart, &
Winston.

Wolpe, J., & Lazarus, A. A. (1966). Behavior therapy

techniques. Oxford: Pergamon Press.

308
Chapter 8

Swimming Against the

Mainstream: The Early
Years in Chilly Waters
Albert Bandura

Stanford University

Behaviorally-oriented approaches evolved in an inhospitable

historical climate. In the early 50’s, the field of personal
change was dominated by the psychodynamic drive model of
human behavior. There were several variants to this model,
but they shared three characteristics.

They all emphasized psychic determinism as their guiding

causal model. In this approach, behavior was regulated by the
interplay of inner impulses and complexes. Most of this inner
life operated subterraneously below the level of
consciousness, disguised by defensive mental operations. To
circumvent the defensiveness, surreptitious projective
methods were devised to reveal them. People were asked to
respond to inkblots, ambiguous pictures, sentence stems, or
simply to free associate. Although the theory postulated a
thoroughgoing psychic determinism, the unconscious inner
life was only loosely linked to behavior. The same inner
dynamics could produce any type of behavior, including
opposite styles of responsivity. Thus, a hostile impulse could

309
spawn either irascibility or sweetness. The theory was not
only shrouded in conceptual fog, but the proposed causal
structures were essentially indeterminant and strewn with
loopholes. Such theories were used mainly as post hoc
explanatory devices.

Experimental efforts to verify the basic tenets of

psychodynamic theory were like tilting at windmills.
Proponents of psychodynamic theory dismissed experimental
investigations as entirely unsuitable because the core
determinants were not amenable to experimental variation.
An Oedipal complex was not manipulatable. In their view, the
theory could be tested only through clinical validation. In a
letter to Freud, Rosensweing asked whether the interview
content could be tainted by the therapists’ influences. Freud
argued that the therapist serves as a blank screen on which the
psychic dynamics are projected. Therefore, the interview
content remains uncontaminated. The claim of
immaculateness of method had no foundation in fact.
Interactional analyses showed that therapists were actively
shaping the content through their suggestive interpretations
and selective positive and negative reactions (Bandura,
Lipsher, & Miller, 1960; Murray, 1956).

The second major feature of this approach was the adoption

of a quasi-disease model of deviant behavior. Styles of
behavior that diverged from prevailing norms were viewed as
a symptom of an underlying pathology. However, the disease
was psychic or metaphorical. Problems of living and
unconventional patterns of behavior got labeled as
symptomatic expression of a psychic pathology. We are now
similarly witnessing a widespread medicalization of
psychosocial problems.

310
The third distinguishing characteristic was the heavy reliance
on the interview as the vehicle of personal change. It was
assumed that by analyzing clients’ reports of their recurrent
conflicts and the problems they reenacted with their therapists
they gain insight into their inner dynamics and develop better
ways of behaving. Self-insight would promote change.

In the early 50’s, Hullian theorists were translating the

psychoanalytic doctrine into Hullian learning terms and
processes to render it empirically testable. Dollard and
Miller’s (1950) publication, Personality and Psychotherapy
became the bible for the times on which most of us were
imprinted. Dollard and his Yale colleagues proposed a set of
testable propositions concerning aggression cast in terms of a
frustration-aggression model encompassing instigative drive
forces and displacement processes governed by excitatory and
inhibitory generalization gradients (Dollard, Doob, Miller,
Moirer, & Sears, 1939). Whiting and Child (1953), and Sears
and his colleagues extended the theory to the developmental
sphere (Sears, Whiting, Nowlis, & Sears, 1953).

The psychodynamic theories fared poorly when subjected to

close empirical scrutiny. They lacked predictive power.
Self-appraisals and actuarial systems proved to be better
predictors of future behavior than psychodynamic predictions,
which supposedly had privileged access to the unconscious
determinants (Dawes, Faust, & Meehl, 1989; Shrauger &
Osberg, 1982).

Many outcome studies conducted in the 60’s showed that it is

difficult to change human behavior by talk alone. People
gained all kinds of insights but exhibited little change in
behavior. It is easier to alter people’s beliefs about the causes

311
of their behavior than to change their behavior enduringly.
For example, alcoholics can be more readily persuaded that
they drink because of fixated orality than to get them to give
up booze. Interview modes of treatment showed limited gains
in actual behavioral functioning, but some benefits on
self-ratings of change. However, such measures exaggerate
the level of behavioral change (Williams & Rappoport, 1983).
On self-ratings even most nontreated controls report gains.

Each theoretical approach had its own favored brand of

insight. One could reliably predict the types of insights and
inner dynamics clients would find in the course of therapy
from knowledge of their therapists’ theoretical orientation.
Conversational therapies seemed to be promoting conversion
to belief systems in the guise of self-discovery. One could
also predict whether the clients would find an unconscious
mind and what is in it. A Freudian unconscious is different
from a Jungian one, and Rogerians never unearthed an
unconscious mind. Given the arbitrariness of the self-insights,
it was hardly surprising that they were usually
unaccompanied by behavior changes. Although people who
underwent such treatments often reported some benefits, they
usually did not change any more than comparable cases who
received no formal treatment. Eysenck (1952) did a thorough
demolition job on the efficacy of psychodynamic therapies.

Following the old adage that one should light a candle rather
than curse the conceptual darkness, Dick Walters and I set
forth an alternative view of human behavior in the book,
Social Learning and Personality Development (Bandura &
Walters, 1963).

312
It underscored the influential role of modeling and direct
consequential experiences in the acquisition of behavior
patterns, and their regulation through the complex interplay of
contextual, incentive, and self-regulatory influences rooted in
personal standards.

During this period, I was teaching the psychotherapy courses

at Stanford. I became intrigued by cases in which direct
modification of problem behavior not only produced lasting
improvements in people’s lives, but often had positive
spillover effects on their nontreated areas of functioning.
Once alcoholics were helped to gain sobriety, for example,
the quality of their lives changed dramatically for the better. I
spent several months tracking down such treatments
published in obscure journals housed in the musty catacombs
of the university library. In 1960, I emerged bleary-eyed but
inspired from the catacombs to publish an analysis of
psychotherapy in terms of acquisitional and regulatory
mechanisms in the Psychological Bulletin entitled,
Psychotherapy as a Learning Process. The conceptual
scheme and accompanying psychosocial applications were
organized around six basic principles of behavioral change
(Bandura, 1961). The time was apparently ripe for a new
direction in the conceptualization and treatment of behavior. I
was flooded with reprint requests from home and abroad.

Based on this article, Eysenck invited me to contribute a

chapter to a volume he was editing on behavior therapy. The
chapter kept enlarging until it outgrew the assigned page
allotment. Instead, it turned into a voluminous book in the
making under the title, Principles of Behavior Modification
(1969). The volume addressed the influential role of

313
symbolic, vicarious, and self-regulatory mechanisms in
human adaptation and change.

While working on these projects, I was invited to join the

study section at the National Institute of Mental Health that
reviewed grant proposals in developmental psychology,
personality, and psychotherapy. Proposals for research on
behavior therapy were being routinely rejected. They were
dismissed as simply removing symptoms rather than treating
their underlying causes. Moreover, it was claimed that they
risked spawning more serious symptom substitution.
Apocryphal stories were floated about the potential dangers of
such approaches. In one of the more fanciful scenarios, a
behavior therapist allegedly got a husband to quit grinding his
teeth while sleeping, only to murder his wife the next week!
This turned out to be a conceptual rather than a material
homicide because inquiries never produced a spousal body.

I negotiated an understanding with my colleagues on the

study section that the quality of a research proposal should be
judged against the tenets of the theory on which it is founded
rather than by the causal models favored by theoretical rivals.
My first primary reviews and site visit reports upon joining
the study section were Bijou and Baer’s project at the
University of Washington, and Ivar Lovaas’ autism project at
UCLA. Both were approved and funded.
Behaviorally-oriented studies were now receiving a fair
hearing based on the conceptual, methodological, and social
merits of the proposals.

There were other paths of involvement in the growing

network of proponents of the alternative theories and
psychosocial practices. Wolpe (1958) submitted his

314
manuscript on Psychotherapy By Reciprocal Inhibition to the
Stanford Press for consideration. The Press sent it to me for
advisory evaluation. In my supportive review I predicted that
it will have modest sales at the outset, but continue to garner
significant sales in years to come in the evolving field of
behavior therapy. I invited Arnold Lazarus to Stanford during
my sabbatical leave, which greased his relocation from South
Africa.

For their dissertations, Gerald Davison (1968) verified the

facilitative role of relaxation in the desensitization mode of
treatment, and Bernard Perloff (1970) demonstrated that
positive imagery works better than muscular relaxation in
modifying phobic behavior. Both of these well crafted
experiments, which underscored the influential role of
symbolic and imagery processes, received dissertation
awards. Wolpe (1974) cast the rationale for this treatment in
terms of conditioning relaxation responses to anxiety cues.
People who had been thoroughly desensitized to phobic
threats in imagery, nevertheless varied in their actual coping
behavior. In microanalysis of possible mechanisms governing
therapeutic change, we found that symbolic desensitization
enhances coping behavior to the extent that it raises perceived
self-efficacy (Bandura & Adams, 1977).

The 60’s ushered in radical changes in the explanation and

modification of behavior. Causal analysis shifted from
unconscious psychic dynamics to transactional social
dynamics. Troublesome behavior was viewed as divergent
rather than diseased behavior. Functional analysis of human
behavior replaced diagnostic labeling that categorized people
into psychopathologic types. Laboratory and controlled field

315
studies of the determinants and mechanisms of behavioral
functioning replaced content analyses of interviews.

The modes of treatment were altered in the content, locus, and

agents of change. With regard to content, therapeutic efforts
were directed mainly at modifying the actual problems for
which people sought help through mastery experiences rather
than conversing about their problems and their
psychodynamic origins. In the transactional models that were
adopted, the determinants of human behavior do not reside
solely in the individual. People are both products and
producers of their life conditions. The model of bidirectional
causation had implications for the locus of change. Efforts
were also directed at changing social practices that contribute
to aversive and dysfunctional styles of behavior. Many human
problems are institutional not simply individual. Collective
problems require social solutions. A major issue for science
and practice of psychosocial change was whether efforts
should be centered mainly in treating the casualties of adverse
social practices, or also altering the social practices producing
the casualties.

A comprehensive approach of high social utility also requires

prevention as well as remediation. As another aspect of the
locus of change, to enhance successful results, treatments
were typically carried out in the natural settings in which the
problems arise. Some of the corrective practices were
conducted in homes, schools, workplaces, or in the larger
community, depending upon the sources of the contributing
determinants. With regard to the agents of change,
behaviorally-oriented approaches did not view professionals
as the exclusive dispensers of treatments. By drawing on the
vast resources of people to implement change programs under

316
professional guidance, practitioners greatly expand the scope
of their impact. If professionals had to implement every
aspect of their treatments, their contribution would be but a
tiny ripple in the vast sea of human problems.

Not all the behaviorally-oriented folks worshipped at the

same theoretical alter. Some went the operant route as
providing the best glimpse of the promised land. Others went
the sociocognitive route. Vigorous epistemological battles
were fought over cognitive determinants and the legitimacy of
alternative forms of scientific inquiry (Bandura, 1995).
Scientific advances are promoted by two kinds of theories
(Bandura, 1996). One form seeks to identify relations
between directly observable events, but shies away from the
underlying mechanisms governing the observable events. The
second form seeks to elucidate the mechanisms that explain
the relations between observable events. Operant analysts
took the view that the only legitimate scientific enterprise is
one that links directly observable events.

In commenting on the issue of observability in scientific

inquiry, Nagel (1961) explains that some of the most
powerful theories of the natural sciences are not about factors
that are “observable.” Physicists, for example, have done
remarkably well with atomic theory even though atoms are
not given to public view. The major issues in contention
regarding cognitive determinants were not about the
legitimacy of inner causes, but about the types of inner
determinants that are given favored status. Operant analysts
dismissed internal determinants in the form of cognitions as
explanatory fictions or functionally merely epiphenomena of
conditioned responses. But they increasingly placed the
explanatory burden within their own scheme on determinants

317
inside the organism, namely the implanted history of
reinforcement. Like other internal determinants, ontogenic
history is neither observable nor directly accessible.

Multifaceted Applicability of Social

Cognitive Theory
Social cognitive theory lends itself readily to social
applications. The factors it posits are anchored in indices of
functioning and are amenable to change. The determinants
and mechanisms through which the influences operate are
spelled out, so the theory provides explicit guidelines on how
to promote personal and social change. The models of change
we developed drew heavily on our knowledge of modeling,
self-regulatory, and self-efficacy mechanisms.

Guided Mastery Treatments

The initial applications of modeling to the treatment of phobic
disorders eventually evolved, with the creative contributions
of Ed Blanchard and Bruni Ritter, into a powerful guided
mastery treatment (Bandura, Blanchard, & Ritter, 1969). It
relied on mastery experiences as the principal vehicle of
change. When people avoid what they dread, they lose touch
with the reality they shun. Guided mastery quickly restores
reality testing in two ways. It provides disconfirming tests of
phobic beliefs by convincing demonstrations that what the
phobics dread are safe. Even more important, it provides
confirmatory tests that phobics can exercise control over what
they fear.

318
Intractable phobics, of course, are not about to do what they
dread. Therapists must, therefore, create environmental
conditions that enable phobics to succeed, despite themselves.
This is achieved by enlisting a variety of performance
mastery aids. Threatening activities are repeatedly modeled to
demonstrate coping strategies, and to disconfirm people’s
worst fears. Intimidating tasks are reduced to graduated
subtasks of easily mastered steps. Treatment is conducted in
this stepwise fashion until the most intimidating activities are
mastered. Joint performance with the therapist enables
frightened people to do things they would refuse to do on
their own. Another method for overcoming resistance is to
have phobics perform the feared activity for only a short time.
As they become bolder the length of engagement is extended.
Protective conditions can also be introduced to weaken
resistances that retard change.

Initially, therapists use whatever mastery aids are needed to

restore coping behavior. As treatment progresses, supportive
aids and protective controls are faded until clients manage the
most intimidating activities on their own. After bold
functioning is fully restored, self-directed mastery activities
are arranged in which clients manage different versions of the
threat on their own under varying conditions.

Self-directed mastery was designed to serve three purposes.

By affirming the participants’ personal capabilities,
self-directed accomplishments would eliminate any
misattribution of the successes to the mastery aids. By this
means, misattribution problems can be easily eliminated
without sacrificing the substantial benefits of powerful
mastery procedures. Self-directed mastery experiences further
strengthen and generalize restored coping capabilities.

319
Moreover, multiple diverse successes serve as a vehicle for
building resilience to the negative effects of adverse
experiences. The capacity of an aversive experience to
reinstate dysfunctions depends, in part, on the pattern of
experience in which it is embedded rather than on its
properties alone. A lot of neutral or positive experiences can
neutralize the negative impact of an aversive event and curtail
the spread of negative effects (Hoffman, 1969). For example,
if after treatment, a dog phobic has no contact with dogs, an
aversive encounter will quickly reinstate the phobia. But if the
phobic had many neutral and positive experiences with
different varieties of dogs, an aversive encounter is likely to
produce a circumscribed effect — avoidance of the
threatening animal without phobic generalization to the other
canine types.

The initial tests of the efficacy of guided mastery showed it to

be an unusually powerful treatment (Bandura, et al., 1969;
Bandura, Taylor, Williams, Mefford, & Barchas, 1985). It
eliminated severe snake phobias in everyone within a couple
of hours. It eradicated experienced anxiety, autonomic
reactivity, and secretion of stress-related hormones. It
changed attitudes toward the phobic objects from loathing to
more positive ones. The phobics had been plagued by
recurrent nightmares for 20 or 30 years. The mere sight or
mention of reptiles activated perturbing ruminations that the
phobics felt helpless to turn off. Guided mastery wiped out
their nightmares and aversively intrusive ruminations. The
transformation of dream activity was fascinating to observe.
As clients gained mastery, the phobic object changed from
pursuing and terrorizing them in their dreams to beneficent
creatures. As one woman gained mastery over her snake
phobia, she dreamt that the boa constrictor befriended her and

320
was helping her to wash the dishes. Eventually reptiles faded
from her dreams and ruminative activity ceased.

In comparative outcome studies, for phobics who achieved

only partial improvement with desensitization, modeling
alone, or cognitive behavior therapy, guided mastery
eradicated the phobia in everyone in a short time (Bandura &
Adams, 1977; Bandura, et al., 1969; Biran & Wilson, 1981;
Thase & Moss, 1976). Follow-up assessments five years later
found participants just as bold and unperturbed as they were
at the end of treatment.

I was invited to present our program of research to the

Langley Porter Clinic in San Francisco, a stronghold of
psychodynamic adherents. It began with an inhospitable
introduction to the effect that this young upstart is going to
tell us seasoned pros how to cure phobias! While my host was
proclaiming the virtues of the psychodynamic approach and
flogging a behavioristic caricature, I was trying to figure out
how to begin my sermonette with acknowledgment of the
inimical atmospherics. I explained that my host’s “generous”
introduction reminded me of a football contest between Iowa
and Notre Dame in South Bend. Iowa scored a touchdown,
which tied the score. As the player ran on the field to kick the
extra point, coach Evashevski turned to his assistant coach
and remarked, “Now there goes a brave soul, a Protestant
attempting a conversion before 50,000 Catholics!”

Powerful treatments are typically multifaceted. The next

analytic task was to evaluate the contribution of the various
components to the therapeutic outcomes. In experiments
conducted with Bob Jeffery, we demonstrated that the
mastery-aids component accelerated the rate of therapeutic

321
change, and the self-directed mastery component enhanced
the generality and durability of the changes (Bandura, Jeffery,
& Gajdos, 1975; Bandura, Jeffery, & Wright, 1974).

The next phase in this program of research was to test the

generalized utility of this mode of treatment across different
types of dysfunctions. In an extended series of studies, Lloyd
Williams showed that guided mastery was similarly powerful
with one of the most incapacitating anxiety disorders,
agoraphobia (Williams, 1990; 1992). Guided mastery proved
more powerful than the exposure treatment in vogue, in which
phobics repeatedly confront threatening situations without the
benefit of mastery performance aids (Williams, Dooseman, &
Kleifield, 1984; Williams, Turner, & Peer, 1985; Williams &
Zane, 1989).

Self-Efficacy Theory
Self-efficacy theory was an outgrowth of our research
described earlier designed to build resilience to adversity. In
follow-up assessments, we were discovering that the
participants not only maintained their therapeutic gains, but
made notable improvements in domains of functioning quite
unrelated to the treated dysfunction. Thus, for example, after
mastering an animal phobia, participants had reduced their
social timidity, became bolder in public speaking, expanded
their competencies in various spheres of their lives, and
boosted their venturesomeness in a variety of ways. Success
in overcoming, within a few hours of treatment, a phobic
dread that had constricted and tormented their lives for twenty
or thirty years produced a profound change in participants’
beliefs in their personal efficacy to exercise better control

322
over their lives. They were acting on their belief, putting
themselves to test and enjoying their successes much to their
pleasant surprise.

I redirected my research efforts to gain a deeper

understanding of personal efficacy. To guide this new
mission, I formulated a theory that addressed the key aspects
of human efficacy (Bandura, 1977). These aspects include the
origins of efficacy beliefs, their structure and function, their
diverse effects, the psychosocial processes through which
they produce these effects, and the modes of influence by
which they can be created and strengthened for personal and
social benefit.

This belief system is the foundation of human agency. Unless

people believe they can produce desired effects by their
actions they have little incentive to act or to persevere in the
face of difficulties. Whatever other factors serve as
motivators, they are rooted in the core belief that one has the
power to effect changes by one’s actions. Using diverse
methodologies converging evidence from causal tests
demonstrated that, indeed, efficacy beliefs play a
determinative role in human functioning. In one approach
(Bandura, Reese, & Adams, 1982), perceived self-efficacy
was raised in phobics to differential levels simply by having
them observe modeled coping strategies until the preselected
level of perceived efficacy is attained. Higher levels of
perceived self-efficacy were accompanied by higher
performance accomplishments.

A number of experiments were conducted in which

self-efficacy beliefs are altered by bogus feedback unrelated
to one’s actual performance. Using this type of induction

323
procedure, Weinberg, Gould and Jackson (1979) raised the
self-efficacy beliefs of one group by telling them that they
had triumphed in a competition of muscular strength, and
lowered the self-efficacy beliefs of another group by telling
them that they were outperformed by their competitor. The
higher the instilled illusory beliefs of physical strength, the
more physical endurance the participants displayed during
competition on a new task measuring physical stamina.
Failure in a subsequent competition spurred those with a high
sense of perceived self-efficacy to even greater physical
effort, whereas failure further impaired the performance of
those whose perceived self-efficacy had been undermined.

Bogus normative feedback was another variant on the social

comparative mode of altering beliefs of personal efficacy.
Individuals were led to believe that they performed at high or
low percentile ranks of a reference group, regardless of their
actual performance. Those whose perceived efficacy was
heightened by this means outperformed the individuals whose
perceived efficacy was lowered (Litt, 1988; Jacobs,
Prentice-Dunn, & Rogers, 1984).

Another approach to the test of causality is to control, by

selection, level of ability but to vary perceived self-efficacy
within each ability level. Collins (1982) selected children who
judged themselves to be of high or low mathematical efficacy
at each of three levels of mathematical ability. They were then
given difficult problems to solve. Within each level of
mathematical ability, children who regarded themselves as
efficacious were quicker to discard faulty strategies, solved
more problems, chose to rework more of those they failed,
and did so more accurately. Perceived self-efficacy thus
exerted a substantial independent effect on performance.

324
A fifth approach to causality is to introduce a trivial factor
devoid of information to affect competency, but that can alter
perceived self-efficacy. Cervone and Peake (1986) used
arbitrary anchor values to influence self-appraisals of
efficacy. Self-appraisals made from an arbitrary high starting
point biased students’ perceived self-efficacy in the positive
direction, whereas an arbitrary low starting point lowered
students’ appraisals of their efficacy. The higher the instated
perceived self-efficacy, the longer individuals persevered on
difficult and unsolvable problems before they quit.
Mediational analyses revealed that the anchoring influence on
performance motivation was entirely mediated by perceived
self-efficacy.

Still another approach to the verification of causality

employed a contravening experimental design in which a
procedure that can impair functioning is applied, but in ways
that raise perceived self-efficacy. Holroyd and his colleagues
(Holroyd, et al., 1984), used this mode of verification with
sufferers of tension headaches. In biofeedback sessions, they
trained one group to become good relaxers. Unbeknownst to
another group, they received feedback signals that they were
relaxing whenever they tensed their muscles. They became
good tensors of facial muscles, which, if anything, would
aggravate tension headaches. Regardless of whether people
were tensing or relaxing their musculature, bogus feedback
that they were exercising good control over muscular tension
instilled a strong sense of efficacy that they could prevent the
occurrence of headaches in different stressful situations. The
higher their perceived self-efficacy, the fewer headaches they
experienced. The actual amount of change in muscular
activity achieved in treatment was unrelated to the incidence
of subsequent headaches.

325
The final way of verifying the contribution of efficacy beliefs
to human functioning is to test the multivariate relations
between relevant determinants and subsequent performances
using structural equation modeling. Such analyses indicate
how much of the variation in performance is explained by
perceived self-efficacy when the influence of other
determinants, including past performance, is controlled. The
results of numerous studies revealed that efficacy beliefs
contribute to performance both directly and through their
impact on other determinants (Bandura & Jourden, 1991;
Locke, Frederick, Lee, & Bobko, 1984; Ozer & Bandura,
1990; Wood & Bandura, 1989).

These diverse causal tests conducted with different modes of

efficacy induction, varied populations, using both
interindividual and intraindividual verification designs, and
all sorts of domains of functioning provided supporting
evidence that perceived self-efficacy contributes significantly
to level of motivation and performance accomplishments.

Other lines of research advanced understanding of the

processes through which efficacy beliefs regulate human
functioning. The results of these studies showed that they do
so through their impact on cognitive, motivational, affective,
and choice processes. Specifically, efficacy beliefs influence
whether people think strategically, pessimistically or
optimistically; how well they motivate themselves and their
staying power in the face of obstacles; their emotional
well-being; and the life choices they make.

We conducted a series of studies to test whether different

modes of treatment work in part by instilling and
strengthening beliefs of personal efficacy (Bandura & Adams,

326
1977; Bandura, Adams, & Beyer, 1977; Bandura, Adams,
Hardy, & Howells, 1980). The results of these microanalytic
studies were consistent in showing that the self-efficacy belief
system is a common pathway through which diverse
interventions effect changes.

Self-efficacy theory and its diverse personal and social

applications are extensively reviewed in Self-Efficacy: The
Exercise of Control (Bandura, 1997). It presents the structure
of the theory, documents the centrality of control beliefs in
people’s lives, specifies how to build a resilient sense of
efficacy, and analyzes the processes through which such
beliefs affect human motivation and accomplishments.
Efficacy beliefs promote successful adaptation and change
throughout the life course. The theory lends itself readily to
social applications because it provides explicit guides on how
to effect change. These include applications to education,
health, clinical dysfunctions (i.e., anxiety, phobias,
depression, eating disorders, substance abuse), personal and
team athletic attainments, organizational productivity, and
people’s collective efficacy to improve their lives through
united effort. Meta-analyses amply document the influential
role of perceived self-efficacy in human adaptation and
change (Holden, 1991; Holden, Moncher, Schinke, & Barker,
1990; Multon, Brown, & Lent, 1991; Stajkovic & Luthans,
1998).

Self-Regulatory Mechanism
People are not only knowers and performers. They are also
self-reactors with a capacity to motivate, guide, and regulate
their activities (Bandura, 1986; 1991). Another line of

327
research that I was pursuing was designed to advance
understanding of self-regulatory mechanisms. Self-regulation
operates through a set of psychological subfunctions that must
be developed and mobilized for self-directed change. People
have to learn to monitor their behavior and judge it against an
adopted performance standard. The cognitive comparison sets
the occasion for self-reactive influence. Self-reactions provide
the mechanism by which personal standards regulate
motivation and action.

Much human behavior is regulated anticipatorily by expected

material and social outcomes. Social cognitive theory
broadened this functionalism to include self-evaluative
outcomes. After people adopt personal standards, they
influence their own motivation and behavior by the positive
and negative consequences they produce for themselves. They
do things that give them satisfaction and a sense of
self-worth, and refrain from actions that evoke self-censure.
Studies conducted in collaboration with Mike Mahoney,
Carol Kupers, Karen Simon, and Bernard Perloff shed light
on how personal standards are acquired, documented the
regulative power of self-administered consequences, and
identified conditions under which self-evaluative outcomes
override external ones (Bandura & Kupers, 1964; Bandura &
Mahoney, 1974; Bandura, Mahoney, & Dirks, 1976; Bandura
& Perloff, 1967; Mahoney & Bandura, 1972; Simon, 1979a,
1979b).

328
Self-Management Models with Social
Utility
Our knowledge of self-regulatory and self-efficacy
mechanisms was used to devise efficacious self-management
models with high social utility. Applications to health
promotion and disease prevention is but one such example.

The recent years have witnessed a major change in the

conception of human health and illness, from a disease model
to a health model. It is just as meaningful to speak of levels of
vitality as of degrees of impairment. The quality of health is
heavily influenced by lifestyle habits. By exercising control
over a few health habits, people can live longer, healthier, and
slow the process of biological aging. Exercise, reduce dietary
fat, refrain from smoking, keep blood pressure down, and
develop effective ways of coping with stressors. If the huge
health benefits of these few lifestyle habits were put into a
pill, it would be declared a spectacular breakthrough in the
field of medicine.

Effective self-management of health behavior requires

development of self-regulatory skills to influence one’s own
motivation and behavior. In such programs, people have to
track their behavior and the social and cognitive conditions
under which they engage in it; set proximal goals for guiding
and controlling their behavior; draw from an array of coping
strategies rather than rely on a single technique; create
self-motivating incentives to sustain their efforts; and apply
multifaceted self-influence consistently and persistently.

329
Efficacy beliefs play an influential role in every phase of
personal change —whether people even consider changing
their health habits; whether they enlist the motivation and
perseverance needed to succeed should they choose to do so;
their vulnerability to relapse; their success in recovering
control after a setback; and how well they maintain the habit
changes they have achieved (Bandura, 1997; 1998).

The self-regulatory subfunctions and their self-efficacy

underpinning were built into a self-management model
devised by DeBusk and his colleagues to reduce health risks
and promote health (DeBusk, et al., 1994). It equips
participants with the skills and personal efficacy to exercise
self-directed change. They are provided with guidelines on
how to change detrimental health habits. To motivate and
regulate their actions, they monitor their health habits, set
explicit proximal goals, and apply self incentives to sustain
their efforts. They receive periodic feedback of progress
toward their goals, and instructive guides on how to manage
troublesome situations.

Self-efficacy ratings identify areas in which self-regulatory

skills must be developed and strengthened if desired changes
are to be achieved and maintained.

The productivity of this self-management system is vastly

expanded by combining self-regulatory principles with the
power of computer-assisted implementation. A single
implementer, assisted with a computerized coordinating and
mailing system, provides intensive individualized training in
self-management of large numbers of people simultaneously.

330
In tests of this system, employees in the workplace lowered
high cholesterol levels by altering eating habits high in
saturated fats (Bandura, 1997). They achieved even larger
reductions if their spouses participated. The greater the room
for dietary change, the greater the reduction in plasma
cholesterol. A single nutritionist implemented the entire
program at minimal cost for large numbers of employees.
Sodium intake is linked to hypertension in people who are
sensitive to this mineral, a sensitivity that increases with age
as the body loses some of its efficiency. West and his
colleagues demonstrated with patients suffering from heart
disease that the self-management system enhances their
self-regulatory efficacy and gets them to cut back on their
level of sodium intake to desired levels and to maintain the
low sodium diet stably over time (West, et al., 1999). At each
successive point in the self-change program, the stronger the
perceived self-regulatory efficacy, the greater the reduction in
sodium intake.

Haskell and his colleagues used this system to promote

lifestyle changes in patients suffering from coronary artery
disease, which places them at high risk of heart attacks
(Haskell, et al., 1994). At the end of four years, those
receiving medical care by their physicians showed no change
or a worsening of their condition. In contrast, those aided in
self-management of health habits achieved large reductions in
risk factors. They lowered their intake of fat, lost weight,
lowered their bad cholesterol, raised their good cholesterol,
exercised more, and increased their cardiovascular capacity.
The program also altered the physical progression of the
disease. Those receiving the self-management program had
47% less plaque on artery walls, and a higher reversal of

331
arteriosclerosis. They also had fewer hospitalizations for
coronary heart problems, and fewer deaths.

The success of this self-management system is currently

being compared against the standard medical post-coronary
care to reduce morbidity, and mortality in patients who have
already suffered a heart attack (DeBusk, et al., 1994). The
self-regulatory system is more effective in reducing risk
factors, and increasing cardiovascular functioning than the
standard medical care.

The self-management system is very well received by patients

because it is individually tailored to their needs; provides
continuing personalized guidance and informative feedback
that enables them to exercise considerable control over their
own change; it is a home-based program that does not require
any special facilities, equipment, or attendance at group
meetings that usually have high drop-out rates; can serve
large numbers of people simultaneously under the guidance of
a single implementer; is not constrained by time and place;
combines the high individualization of the clinical approach,
with the large-scale applicability of the public health
approach; and provides valuable health-promotion services at
low cost.

Linking the interactive aspects of the self-management model

to the Internet can vastly expand its availability for
preventive, and promotive guidance. Moreover, this model
lends itself readily to a triage strategy of application that
further enhances its scope and productivity. Many people may
succeed with enabling interactions through the mail alone that
provide them with sufficient structured guidance to
accomplish the changes they seek. Successful self-changers

332
with minimal guidance have a high sense of efficacy that they
can get themselves to adopt healthful habits and to stick to
them. Those who distrust their ability to succeed give up
trying when they run into difficulties. They need additional
support and guidance via telephone contact to see them
through tough times. And finally, those who believe that their
health habits are beyond their personal control need a great
deal of guidance in a stepwise mastery program. Graduated
successes build belief in their ability to exercise control and
bolster their staying power in the face of difficulties and
setbacks.

The self-management of chronic diseases provides another

example of translation of self-regulatory and self-efficacy
theory to highly cost-effective implementation models with
high social utility. Chronic disease is the dominant form of
illness and the major cause of disability. This is a growing
health problem because, with people living longer, there is
more time for detrimental health habits to spawn chronic
diseases. The treatment of chronic disease must focus on
self-management of physical conditions over time. The goal
is to retard the biological progression of impairment to
disability and to improve the quality of life of people with
chronic disease.

Holman and Lorig (1992) devised a prototypic model for the

self-management of chronic diseases. People are taught
cognitive pain control techniques, self-relaxation, and
proximal goal setting combined with self incentives as
motivators to increase level of activity. They are also taught
problem solving and self-diagnostic skills for monitoring and
interpreting changes in their health status, skills in locating
community resources and managing medication programs.

333
The way health care systems deal with clients can alter their
sense of efficacy in ways that support or undermine their
restorative efforts (Bandura, 1998). Clients are, therefore,
taught how to take greater initiative for their health care and
dealings with health personnel. These capabilities are
developed through modeling of self-management skills,
guided mastery practice, and informative feedback.

In a four-year follow-up of arthritic patients, the

self-management program retarded the biological progression
of disease, raised perceived efficacy to exercise some control
over one’s condition, reduced pain and substantially
decreased the use of medical services, and improved the
quality of life (Lorig, 1990). Perceived self-efficacy predicted
the health benefits four years later. The self-management
program produced similar health benefits for people suffering
from other types of chronic diseases (Lorig, et al., 2000).

Macrosocial Applications
Social modeling also plays a paramount role at the
society-wide level in enabling people to improve their lives
and social conditions. Modeling is not only an important
component in most modes of change, but an essential one. It
shortcuts the tortuous process of competency development.
The revolutionary advances in the communications
technologies have vastly expanded the power of symbolic
modeling. In modern day life, the electronic media, feeding
off the communications satellites, are shaping lifestyles
worldwide, transforming institutional practices, and serving
as a major vehicle of sociopolitical change (Bandura, 1997;
Braithwaite, 1994).

334
Symbolic modeling is readily applicable to macrosocial
applications through creative use of the electronic media. The
soaring population growth and the environmental devastation
it produces is the most urgent global problem. The world
population is doubling at an accelerating rate. It will seriously
strain the earth’s carrying capacity and degrade the quality of
life if left unchecked (Ehrlich, Ehrlich, & Daily, 1995).

Sabido (1981) creatively translated several social cognitive

principles into engrossing and influential radio and television
dramatic serials that are being applied internationally with
notable success in raising the status of women and in
stemming the massive population tide. Culturally admired
television models exhibit the beneficial styles of behavior.
Social attraction increases the impact of modeling influences.
Characters representing different segments of the viewing
population are shown adopting the beneficial attitudes and
behavior patterns. Seeing people similar to oneself succeed
enhances the power of modeling. The episodes include
positive models exhibiting beneficial lifestyles, negative
models exhibiting detrimental lifestyles, and transitional
models changing from detrimental to beneficial styles of
behavior. Contrasting modeling highlights the personal and
social effects of different lifestyles. Viewers draw inspiration
from seeing others change their lives for the better. Vicarious
motivators, in the form of the benefits of favorable practices
and the costs of detrimental ones, are vividly depicted.
Depicted benefits provide incentives for change.
Melodramatic and other emotional devices are used to sustain
high attentional involvement in the dramatic presentations.
Epilogues and summarization of the modeled messages are
used as symbolic coding aids to underscore the importance of

335
the social practices that are enacted and to enhance their
recallability.

It is of limited value to motivate people for change if they do

not have the needed resources and environmental supports to
realize those changes. Environmental guides and supports are
therefore provided to expand and sustain the changes
promoted by the media. This format informs people, enables
them with strategies and sustaining self-beliefs, and motivates
them for personal and social change. The story lines model
family planning, women’s equality, beneficial health
practices, and a variety of effective life skills.

Worldwide applications of this creative format in Africa,

Asia, and South America are raising people’s efficacy to
exercise control over their family lives, enhancing the status
of women, and lowering the rates of childbearing (Brown &
Cody, 1991; Singhal & Rogers, 1989; Vaughan, Rogers, &
Swalehe, 1995). A controlled study in Tanzania compared
changes in family planning and contraception use in parts of
the country that received a radio dramatic series with the rest
of the country that did not. The radio series significantly
increased perceived efficacy to exercise control over family
size. Families in the broadcast area adopted family planning
and contraceptive methods at a higher rate.

Some of the story lines centered on safer sexual practices to

prevent the spread of AIDS, where infection rates are high
among long-distance truckers and prostitutes at truck stops.
Responding to increased demand, the National AIDS Control
Program distributed considerably more condoms in the
broadcast region than in the control region. Those in the
broadcast area also reduced the number of sexual partners.

336
The greater the exposure to the modeled patterns, the stronger
the effects on perceived efficacy to control family size and
risky sexual practices.

Kenya provides another example of the impact of mass

communications on reproductive behavior (Westoff &
Rodriguez, 1995). The heavier the exposure to media
messages, the stronger the preference to limit family size, and
the greater the use of contraceptives. The relationship remains
after multiple controls for demographic and socioeconomic
factors. These macrosocial applications illustrate how a small
collective effort can make a huge difference in an urgent
global problem.

Concluding Remarks
The present article traces the evolution of social cognitive
theory from a chilly tributary to part of the mainstream. The
value of a psychological theory is judged by three criteria. It
must have explanatory power; predictive power and, in the
final analysis, it must demonstrate operative power to
improve the human condition. Well-founded theory provides
solutions to human problems. This brief retrospective report
addresses some of the facets of an agentic sociocognitive
approach to human understanding and betterment, and
documents some of the applications of this theory at both
individual and macrosocial levels.

337
References
Bandura, A. (1961). Psychotherapy as a learning process.
Psychological Bulletin, 58, 143-159.

Bandura, A. (1977). Self-efficacy: Toward a unifying theory

of behavioral change. Psychological Review, 84, 191-215.

Bandura, A. (1986). Social foundations of thought and action:

A social cognitive theory. Englewood Cliffs, NJ:
Prentice-Hall.

Bandura, A. (1991). Self-regulation of motivation through

anticipatory and self-reactive mechanisms. In R. A.
Dienstbier (Ed.), Perspectives on motivation: Nebraska
symposium on motivation (Vol. 38, pp. 69-164). Lincoln, NE:
University of Nebraska Press.

Bandura, A. (1995). Comments on the crusade against the

causal efficacy of human thought. Journal of Behavior
Therapy and Experimental Psychiatry, 26, 179-190.

Bandura, A. (1996). Ontological and epistemological terrains

revisited. Journal of Behavior Therapy and Experimental
Psychiatry, 27, 323-345.

Bandura, A. (1997). Self-efficacy: The exercise of control.

New York: Freeman.

Bandura, A. (1998). Health promotion from the perspective of

social cognitive theory. Psychology and Health, 13, 623-649.

338
Bandura, A., & Adams, N. E. (1977). Analysis of
self-efficacy theory of behavioral change. Cognitive Therapy
and Research, 1, 287-308.

Bandura, A., Adams N. E., & Beyer, J. (1977). Cognitive

processes mediating behavioral change. Journal of
Personality and Psychology, 35, 125-139.

Bandura, A., Adams, N. E., Hardy, A. B., & Howells, G. N.

(1980). Tests of the generality of self-efficacy theory.
Cognitive Therapy and Research, 4, 39-66.

Bandura, A., Blanchard, E. B., & Ritter, B. (1969). Relative

efficacy of desensitization and modeling approaches for
inducing behavioral, affective, and attitudinal changes.
Journal of Personality and Social Psychology, 13, 173-199.

Bandura, A., Jeffery, R. W., & Gajdos, E. (1975).

Generalizing change through participant modeling with
self-directed mastery. Behaviour Research and Therapy, 13,
141-152.

Bandura, A., Jeffery, R. W., & Wright, C. L. (1974). Efficacy

of participant modeling as a function of response induction
aids. Journal of Abnormal Psychology, 83, 56-64.

Bandura, A., & Jourden, F. J. (1991). Self-regulatory

mechanisms governing the impact of social comparison on
complex decision making. Journal of Personality and Social
Psychology, 60, 941-951.

339
Bandura, A., & Kupers, C. J. (1964). Transmission of patterns
of self-reinforcement through modeling. Journal of Abnormal
and Social Psychology, 69, 1-9.

Bandura, A., Lipsher, D. H., & Miller, P. E. (1960).

Psychotherapists’ approach-avoidance reactions to patients’
expression of hostility. Journal of Consulting Psychology, 24,
1-8.

Bandura, A., & Mahoney, M. J. (1974). Maintenance and

transfer of self-reinforcement functions. Behaviour Research
and Therapy, 12, 89-97.

Bandura, A., Mahoney, M. J., & Dirks, S. J. (1976).

Discriminative activation and maintenance of contingent
self-reinforcement. Behaviour Research and Therapy, 14, 1-6.

Bandura, A., & Perloff, B. (1967). Relative efficacy of

self-monitored and externally-imposed reinforcement
systems. Journal of Personality and Social Psychology, 7,
111-116.

Bandura, A., Reese, L., & Adams, N. E. (1982).

Microanalysis of action and fear arousal as a function of
differential levels of perceived self-efficacy. Journal of
Personality and Social Psychology, 43, 5-21.

Bandura, A., Taylor, C. B., Williams, S. L., Mefford, I. N., &

Barchas, J. D. (1985). Catecholamine secretion as a function
of perceived coping self-efficacy. Journal of Consulting and
Clinical Psychology, 53, 406-414.

340
Bandura, A., & Walters, R. H. (1963). Social learning and
personality development. New York: Holt, Rinehart &
Winston.

Biran, M. & Wilson, G. T. (1981). Treatment of phobic

disorders using cognitive and exposure methods: A
self-efficacy analysis. Journal of Counseling and Clinical
Psychology, 49, 886-899.

Braithwaite, J. (1994). A sociology of modeling and the

politics of empowerment. British Journal of Sociology, 45,
445- 479.

Brown, W. J., & Cody, M. J. (1991). Effects of a prosocial

television soap opera in promoting women’s status. Human
Communication Research, 18, 114-142.

Cervone, D., & Peake, P. K. (1986). Anchoring, efficacy, and

action: The influence of judgmental heuristics on self-
efficacy judgments and behavior. Journal of Personality and
Social Psychology, 50, 492-501.

Collins, J. L. (1982, March). Self-efficacy and ability in

achievement behavior. Paper presented at the annual meeting
of the American Educational Research Association, N.Y.

Davison, G. C. (1968). Systematic desensitization as a

counterconditioning process. Journal of Abnormal
Psychology, 73, 91-99.

Dawes, R. M., Faust, D., & Meehl, P. E. (1989). Clinical

versus actuarial judgment. Science, 31, 1668-1674.

341
DeBusk, R. F., Miller, N. H., Superko, H. R., Dennis, C. A.,
Thomas, R. J., Lew, H. T., Berger III, W. E., Heller, R. S.,
Rompf, J., Gee, D., Kraemer, H. C., Bandura, A., Ghandour,
G., Clark, M., Shah, R. V., Fisher, L., & Taylor, C. B. (1994).
A case-management system for coronary risk factor
modification after acute myocardial infarction. Annals of
Internal Medicine, 120, 721-729.

Dollard, J., Miller, N. E., Doob, L. W., Mowrer, O. H., &

Sears, R. R. (1939). Frustration and aggression. New Haven,
CT: Yale University Press.

Ehrlich, P. R., Ehrlich, A. H., & Daily, G. C. (1995). The

stork and the plow: The equity answer to the human dilemma.
New York: Putnam.

Eysenck, H. J. (1952). The effects of psychotherapy: An

evaluation. Journal of Consulting Psychology, 16, 319-325.

Haskell, W. L., Alderman, E. L., Fair, J. M., Maron, D. J.,

Mackey, S. F., Superko, H. R., Williams, P. T., Johnstone, I.
M., Champagne, M. A., Krauss, R. M., & Farquhar, J. W.
(1994). Effects of intensive multiple risk factor reduction on
coronary atherosclerosis and clinical cardiac events in men
and women with coronary artery disease. Circulation, 89,
975-990.

Holman, H., & Lorig, K. (1992). Perceived self-efficacy in

self-management of chronic disease. In R. Schwarzer (Ed.),
Self-efficacy: Thought control of action (pp. 305-323).
Washington, D.C.: Hemisphere.

342
Holroyd, K. A., Penzien, D. B., Hursey, K. G., Tobin, D. L.,
Rogers, L., Holm, J. E., Marcille, P. J., Hall, J. R., & Chila,
A. G. (1984). Change mechanisms in EMG biofeedback
training: Cognitive changes underlying improvements in
tension headache. Journal of Consulting and Clinical
Psychology, 52, 1039-1053.

Hoffman, H. S. (1969). Stimulus factors in conditioned

suppression. In B. A. Campbell & R. M. Church (Eds.),
Punishment and aversive behavior (pp. 185-234). New York:
Appleton- Century-Crofts.

Holden, G. (1991). The relationship of self-efficacy appraisals

to subsequent health related outcomes: A meta-analysis.
Social Work in Health Care, 16, 53-93.

Holden, G., Moncher, M. S., Schinke, S. P., & Barker, K. M.

(1990). Self-efficacy of children and adolescents: A
meta-analysis. Psychological Reports, 66, 1044-1046.

Jacobs, B., Prentice-Dunn, S., & Rogers, R. W. (1984).

Understanding persistence: An interface of control theory and
self-efficacy theory. Basic and Applied Social Psychology, 5,
333-347.

Litt, M. D. (1988). Self-efficacy and perceived control:

Cognitive mediators of pain tolerance. Journal of Personality
and Social Psychology, 54, 149-160.

Locke, E. A., Frederick, E., Lee, C., & Bobko, P. (1984).

Effect of self-efficacy, goals, and task strategies on task
performance. Journal of Applied Psychology, 69, 241-251.

343
Lorig, K. (1990, April). Self-efficacy: Its contributions to the
four year beneficial outcome of the arthritis self- management
course. Paper presented at the meeting of the Society for
Behavioral Medicine, Chicago, IL.

Lorig, K. R., Holman, H. R., Sobel, D. S., Laurent, D. D.,

Minor, M. M., & Gonzalez, V. M. (2000). Living a healthy
life with chronic conditions: Self-management of heart
disease, arthritis, diabetes, asthma, bronchitis, emphysema
and others. Palo Alto, CA: Bull Publishing Company.

Lorig, K. R., Sobel, D. S., Stewart, A. L., Brown, Jr., B. W.,

Bandura, A., Ritter, P., Gonzalez, V. M., Laurent, D. D., &
Holman, H. R. (1997). Evidence suggesting that a chronic
disease self-management program can improve health status
while reducing health care utilization and costs: A
randomized trial. Submitted for publication.

Mahoney, M. J. & Bandura, A. (1972). Self-reinforcement in

pigeons. Learning and Motivation, 3, 293-303.

Multon, K. D., Brown, S. D., & Lent, R. W. (1991). Relation

of self-efficacy beliefs to academic outcomes: A
meta-analytic investigation. Journal of Counseling
Psychology, 38, 30-38.

Murray, E. J. (1956). A content-analysis method for studying

psychotherapy. Psychological Monographs, 70, (13, Whole
No. 420).

Nagel, E. (1961). The structure of science. New York:

Harcourt, Brace and World.

344
Ozer, E. M., & Bandura, A. (1990). Mechanisms governing
empowerment effects: A self-efficacy analysis. Journal of
Personality and Social Psychology, 58, 472-486.

Perloff, B. (1970). Influence of muscular relaxation, positive

imagery, and neutral imagery on extinction of avoidance
behavior through systematic desensitization. Unpublished
doctoral dissertation, Stanford University.

Sabido, M. (1981). Towards the social use of soap operas.

Mexico City, Mexico: Institute for Communication Research.

Sears, R. R., Whiting, J. W. M., Nowlis, V., & Sears, P. S.

(1953). Some child-rearing antecedents of aggression and
dependency in young children. Genetic Psychology
Monographs, 47, 135-234.

Shrauger, J. S., & Osberg, T. M. (1982). Self-awareness: The

ability to predict one’s future behavior. In G. Underwood &
R. Stevens (Eds.), Aspects of consciousness (Vol. 3).
Awareness and self-awareness (pp. 267-330). New York:
Academic Press.

Simon, K. M. (1979a). Effects of self comparison, social

comparison, and depression on goal setting and
self-evaluative reactions. Unpublished manuscript, Stanford
University, Stanford, CA.

Simon, K. M. (1979b). Relative influence of personal

standards and external incentives on complex performance.
Unpublished doctoral dissertation, Stanford University,
Stanford, CA.

345
Singhal, A., & Rogers, E. M. (1989). Pro-social television for
development in India. In R. E. Rice & C. K. Atkin (Eds.),
Public communication campaigns (2nd ed., pp. 331-350).
Newbury Park, CA: Sage.

Stajkovic, A. D., & Luthans, F. (1998). Self-efficacy and

work-related performance: A meta-Analysis. Psychological
Bulletin, 124, 240-261.

Thase, M. E., & Moss, M. K. (1976). The relative efficacy of

covert modeling procedures and guided participant modeling
on the reduction of avoidance behavior. Journal of Behavior
Therapy and Experimental Psychiatry, 7, 7-12.

Vaughan, P. W., Rogers, E. M., & Swalehe, R. M. A. (1995).

The effects of “Twende Na Wakati,” an
entertainment-education radio soap opera for family planning
and HIV/AIDS prevention in Tanzania. Unpublished
manuscript, University of New Mexico, Albuquerque.

Weinberg, R. S., Gould, D., & Jackson, A. (1979).

Expectations and performance: An empirical test of
Bandura’s self-efficacy theory. Journal of Sport Psychology,
1, 320-331.

West, J. A., Bandura, A., Clark, E., Miller, N. H., Ahn, D.,
Greenwald, G., & DeBusk, R. F. (1999). Self-efficacy predicts
adherence to dietary sodium limitation in patients with heart
failure. Unpublished manuscript, Stanford University.

Westoff, C. F., & Rodriguez, G. (1995). The mass media and

family planning in Kenya. International Family Planning
Perspectives, 21, 26-31.

346
Whiting, J. W. M., & Child, I. L. (1953). Child training and
personality. New Haven: Yale University Press.

Williams, S. L. (1990). Guided mastery treatment of

agoraphobia: Beyond stimulus exposure. In M. Hersen, R. M.
Eisler, & P. M. Miller (Eds.), Progress in behavior
modification (Vol. 26, pp. 89-121). Newbury Park, CA: Sage.

Williams, S. L. (1992). Perceived self-efficacy and phobic

disability. In R. Schwarzer (Ed.), Self-efficacy: Thought
control of action (pp. 149-176). Washington, D.C.:
Hemisphere.

Williams, S. L., Dooseman, G., & Kleifield, E. (1984).

Comparative power of guided mastery and exposure
treatments for intractable phobias. Journal of Consulting and
Clinical Psychology, 52, 505-518.

Williams, S., L., & Rappoport, A. (1983). Cognitive treatment

in the natural environment for agoraphobics. Behavior
Therapy, 14, 299-313.

Williams, S. L., Turner, S. M., & Peer, D. F. (1985). Guided

mastery and performance desensitization treatments for
severe acrophobia. Journal of Consulting and Clinical
Psychology, 53, 237-247.

Williams, S. L., & Zane, G. (1989). Guided mastery and

stimulus exposure treatments for severe performance anxiety
in agoraphobics. Behaviour Research Therapy, 27, 238-245.

Wolpe, J. (1958). Psychotherapy by reciprocal inhibition.

Stanford: Stanford University Press.

347
Wood, R. E., & Bandura, A. (1989). Social cognitive theory
of organizational management. Academy of Management
Review, 14, 361-384.

348
Chapter 9

The Rise of Cognitive

Behavior Therapy
Albert Ellis

Albert Ellis Institute for Rational Emotive Behavior Therapy

Biographical Introduction
I can’t say that I originated cognitive behavior therapy
because you can always find some ancient sage who practiced
a form of it many hundreds of years ago. Thus, Gautama
Buddha became enlightened and founded Buddhism more
than 2500 years ago; and to his cognitive teachings were soon
added the behavioral exercises of the Zen Buddhists and other
groups. Hindu Yoga practices are also definitely cognitive
behavioral, as are the combinings of religious philosophies
with various behavioral rituals of the ancient Hebrews and the
early Christians.

Indeed, many religious groups intent on creating what might

be called therapeutic change in their members seem to use a
combination of philosophic education, emotional exercises,
and behavioral practices in order to encourage their adherents
to devoutly follow their precepts. We might speculate that
their combination of cognitive, emotive, and behavioral
methods tends to show that in order to make and maintain

349
profound personality changes, humans often have to strongly
and persistently follow these combined methods. Combined
cognitive, emotional, and activity procedures are almost
essential to basic personality change.

When I started to regularly practice psychotherapy in 1943, I

was already prejudiced in favor of cognitive behavior therapy.
Unlike the vast majority of psychologists, I was well on my
way to being a sexologist; and, in addition, I was very
interested, not only in marriage and family relationships, but I
had made a special study of love relationships. Since 1939 I
had trained myself in these areas by reading thousands of
books and articles on sex, love, and marriage and by
voluntarily counseling my friends and relatives on their
problems in these areas.

As a clinical sexologist, I followed the procedures of the early

twentieth century practitioners — especially Iwan Bloch
(1908), August Forel (1922), Havelock Ellis (1936), and W.
F. Robie (1925), who were physicians and who practiced
what could be called cognitive-behavior sex therapy. They
educated their patients sexually, helped minimize their shame
and guilt, and gave them practical in vivo homework
assignments. Following their procedures, I found that I could
help my early clients to overcome their sex problems, as well
as many of their love and marital difficulties, often in just a
few active-directive sessions.

I practiced this kind of cognitive behavior therapy from 1943

to 1947 and then mistakenly thought that I could do a deeper
and more intensive kind of treatment by getting trained in
liberal psychoanalysis, and practicing it for six years. I was
really an existentialist analyst, since my supervisor and

350
trainer, Richard Hulbeck, was a training analyst of the Karen
Horney Institute and was also a leading existentialist
(Hoellen, 1993). Moreover, on my own I was opposed to
Freudian analysis, since I thought that Freud (1965) was
exceptionally unscientific, knew very little about the origin of
people’s sex and love problems, and wrote brilliant fiction
which he presented as scientific non-fiction. So I mostly
followed the neo-Freudian views of Alfred Adler (1929),
Franz Alexander and Thomas French (1946), Erich Fromm
(1955), Karen Horney (1950) and Harry Stack Sullivan
(1953).

I found that even this kind of liberal psychoanalysis was

exceptionally inefficient and superficial, because it was
passive, little educational, and lacking behavioral elements
which I had effectively used as a sex therapist and love and
marriage therapist. So I abandoned it in 1953, did an intensive
study of many techniques of therapy in the next two years
(Ellis, 1955a, 1955b) and started formulating a more efficient
form of psychotherapy.

In January 1955, I started to do what I called rational

psychotherapy (Ellis, 1957a, 1957b, 1958a, 1958b), later
called rational-emotive therapy (Ellis, 1962b; Ellis & Harper,
1961), and finally retitled rational emotive behavior therapy
(REBT) (Ellis, 1993). Although there were a few cognitive
behavior therapies before REBT — such as those of Herzberg
(1945) and Salter (1949) — they were highly unpopular in
1955. Freudian therapy ruled the roost (Freud, 1965)
especially in the United States, and Rogerian Therapy, which
I had been trained in while in graduate school (Rogers, 1943,
1951) was immensely popular.

351
Why was REBT, right from its start, heavily behavioral?
Mainly because I had used behavior therapy on myself before
I even thought of becoming a therapist; and because I had
used it successfully for a dozen years in my specialty of sex,
love, and marriage therapy. I first used it when I was 19 years
of age and had a severe phobia of public speaking. I was the
youth leader of a radical political group, but never dared give
a public speech. But by reading the early experiments of John
B. Watson and his associates (Jones, 1920; Watson, 1919;
Watson & Raynor, 1920), I learned that they deconditioned
young children in a few sessions to overcome their fears of
animals by in vivo desensitization. So I forced myself, very
uncomfortably, to speak and speak in public and within ten
weeks got completely over my phobia — and since that time,
as I frequently tell my workshop audiences, you can’t keep
me away from the public speaking platform!

Thrilled by my success in getting almost a hundred percent

over my panic about public speaking, I soon tackled my
second paralyzing phobia — social anxiety. I could talk to
young women comfortably after being introduced to them by
a friend, but panicked at the thought of approaching them by
myself and starting up a conversation. I never did so, just as I
had avoided all public speeches up to the time I used
exposure to rid myself of this fear. So I gave myself the
homework assignment, in the month of August when I was on
vacation from college, of sitting next to every young woman I
saw sitting alone on a bench in Bronx Botanical Gardens and
giving myself no more than one minute — yes, one lousy
minute — to start a conversation with her. No nonsense,
Albert — one minute!

352
I actually, in one month, sat next to 130 young women —
which I had practically never done in my life before, so afraid
was I of being rejected. Whereupon 30 out of the 130
immediately got up and walked away. But that left me an
even sample of 100 — good for research purposes! Nothing
daunted, I opened a conversation with the remaining 100
women — for the first time, again, in my entire young life. I
spoke about the weather, the birds and the bees, the flowers
and the trees, the book they were reading — truly, about
everything and anything.

As I have often related, I got absolutely nowhere with my

efforts to befriend, to date, and perhaps even to marry a few
of these women with whom I conversed. If Fred Skinner, who
was then teaching at Indiana University, had known of my
futile efforts, he would have predicted that I would have been
extinguished. Of the hundred women I talked to, I only made
one date — and she didn’t show up for it. She kissed me in
the park, promised to meet me again later in the evening, and
didn’t appear. Being a novice at this kind of dating, I forgot to
take her telephone number. So I never did discover what
happened to her. Thereafter, I always took the phone number
of the women I arranged to date!

Anyway, I found out that nothing terrible happened when I

got rejected by 100 women in one month. No one took out a
stiletto and cut my balls off. Women only do that these days!
No one vomited and ran away. No one called a cop. I had
many interesting and pleasant conversations. I found out a
great deal about women. And I got so relaxed about talking to
strange women in strange places that I got good at
conversing, and with my next hundred tries, I actually made
three dates. Better yet, I got completely over my fear of

353
approaching women for the rest of my life and have espoused
the pickup technique of meeting new partners to hundreds of
my clients over the years.

Well, in vivo desensitization really worked for me at the age

of 19. So when I started to practice psychotherapy in 1943,
when I was 30 years old, I used it, with much success, with
many anxious and phobic clients. Some of them achieved
remarkable cures of long-standing panic disorders in just a
few sessions. So my attempts to use this behavioral technique
were nicely reinforced; and REBT has used it more than the
other cognitive behavioral therapies, which followed it about
a decade after I started to use REBT in 1955.

Important Developments in the Rise of

Cognitive Behavioral Therapies
As I have noted, the use of cognitive behavioral methods in
personality change is centuries old. Modern therapy, which
started to become popular in the late 18th century with the
experiments of Franz Anton Mesmer, was almost always
cognitive behavioral. Hypnotists like James Braid, Jean
Martin Charcot, and Hippolyte Bernheim used
educational-persuasive and emotive methods to put their
clients into trances and then often gave them activity
homework assignments to help them work against their
disturbed symptoms. Bernheim (1897/1947) was quite
cognitive, in that he realized that hypnotism worked mainly
because clients took the suggestion of the hypnotist and
decided to follow it. Emile Coué (1923) developed the
cognitive aspects of hypnotism in the early years of the 20th

354
century by realizing that suggestion was not only at the heart
of hypnotic therapy, but also that people’s negative
self-suggestion — or what he called autosuggestion — was a
prime element in creating neurotic disturbance and that they
could consciously choose to replace it with positive
autosuggestion to solve many of their emotive and behavioral
problems. Coué also invented positive visualization to help
disturbed people improve their psychophysical functioning.
So he was definitely a cognitive behavioral therapist.

When did cognitive behavior therapy (CBT) start to be used?

Probably in 1953, when I abandoned psychoanalysis and
started to develop REBT. After several futile attempts to
reform psychoanalytic thinking and to make it more scientific
— in a series of articles I wrote from 1947 to 1953 (Ellis,
1950, 1956) — I abandoned it and looked for an alternative
system of psychotherapy. I went back to philosophy,
especially the philosophy of human happiness, which had
been one of my main hobbies since the age of 16, and
rediscovered the ancient Asian and Greek and Roman
philosophers. I was particularly taken with Epicurus, who
preached the philosophy of disciplined hedonism, and with
Epictetus (1899), who brought the stoic philosophy from
Greece to Rome in the first century A. D. Most of these
ancient writers were constructivists who differed from the
Freudian idea that childhood traumas caused early and later
emotional disturbance. They were also opposed to the
somewhat similar idea of Watson (1919), who stated that if he
trained or conditioned a child during its first five years to
behave in a certain way it would take on that personality
pattern for the rest of its life.

355
On the contrary, the ancient philosophers were
constructivists. They largely maintained, along with Epictetus
(1899), who wrote in The Enchiridion or Manuel in the first
century, “People are disturbed not by things, but by the views
which they take of them.” This philosophy gives humans
some choice in making themselves disturbed and undisturbed;
and it was solidly reiterated by the existentialist philosophers
— such as Kierkegaard (1953), Heidigger (1962), Sartre
(1968), and Tillich (1953) — in modern times.

Taking this constructivist or choice theory to heart, I created

and started practicing REBT in January, 1955. I was not
influenced by George Kelly (1955), whose brilliant
Psychology of Personal Constructs appeared later in 1955 and
which I didn’t read until 1957. But I was thrilled to see, when
I read it, how much his theory of personal choice overlapped
with that which I had already incorporated in rational emotive
behavior therapy.

REBT, as I noted in my first paper on it at the American

Psychological Association Convention in Chicago in August
1956 (Ellis, 1958a), went beyond the previous cognitive
therapies of Janet (1898), Dubois (1907), and Adler (1929),
all of whom worked to change what Janet called the ideés fixe
— the fixed ideas — of disturbed people. REBT specifically
described twelve common irrational or dysfunctional beliefs
— which I derived from my clients’ formulations of their
problems during the first year and a half that I used it. It
hypothesized that these and related irrational beliefs (IBs)
almost invariably accompanied and helped to instigate
people’s neurotic feelings and behavior. As its main cognitive
method, it actively-directively showed clients how to
empirically, logically, and pragmatically dispute these IBs.

356
This was much more specific than the previous cognitive
therapies, and was largely adopted by most of the other
cognitive behavioral systems that originated about a decade
later, such as those of Glasser (1965), Beck (1967),
Meichenbaum (1974, 1977), and Mahoney (1974).

Moreover, my hypothesized irrational or dysfunctional

Beliefs — which I soon raised to about 50 common IBs —
were put into paper and pencil questionnaires. More than
1,000 research studies using these questionnaires have now
been published that tend to confirm my hypothesis that when
people hold more irrational or dysfunctional beliefs, and hold
them strongly, they are more seriously disturbed than those
who hold fewer of them and hold them weakly (Beck, 1991;
Clark, 1997; Ellis, 1979; Glass & Arnkoff, 1997; Hollon &
Beck, 1997).

Moreover, well over a thousand empirical studies have been

published by REBT and CBT researchers that tend to show
that when clients are shown their irrational and dysfunctional
beliefs, and are taught how to use cognitive behavioral
methods of changing them, they tend to become less neurotic
and even less afflicted with severe personality disorders
(Beck, 1991; Hollon & Beck, 1994; Lyons & Woods, 1991;
McGovern & Silverman, 1984; Meichenbaum, 1977;
Silverman, McCarthy, & McGovern, 1992).

So the REBT theories of emotional-behavioral disturbance

and their cognitive behavioral treatment have been backed by
many empirical studies.

What about REBT and its place in behavior therapy? I also

clearly stated in my first presentations on REBT (Ellis, 1957a,

357
1958a, 1958b, 1960, 1962) that it is both highly
emotive-evocative and behavioral; and I changed its name in
1961 from rational therapy (RT) to rational emotive therapy
(RET). In the 1960’s, following the work of Perls (1969) and
Shutz (1967), I also added many experiential exercises to
REBT, including my famous shame-attacking exercises (Ellis,
1973). So, from the start, REBT differed from most other
therapies in its regular use of many cognitive, many emotive,
and many behavioral methods; and in this respect it has
always been, to use Arnold Lazarus’ (1989) term, multimodal
(Kwee & Ellis, 1997).

This is an important aspect of cognitive behavioral therapy: In

theory as well as practice, it is eclectic and integrationist. I
said in my first paper on REBT (Ellis, 1958a) that thinking,
feeling and behaving overlap and are interrelated. The second
and third paragraphs of this article state:

The human being may be said to possess four basic processes

— perception, movement, thinking, and emotion — all of
which are integrally interrelated. Thus, thinking, aside from
consisting of bioelectric charges in the brain cells, and in
addition to comprising remembering, learning,
problem-solving, and similar psychological processes, also is,
and to some extent has to be, sensory, motor, and emotional
behavior. Instead, then, of saying, ‘Jones thinks about this
puzzle,’ we should more accurately say, ‘Jones
perceives-moves-feels-THINKS about this puzzle.’ Because,
however, Jones’ activity in relation to the puzzle may be
largely focused upon solving it and only incidentally on
seeing, manipulating, and emoting about it, we may perhaps
justifiably emphasize only his thinking.

358
Emotion, like thinking and the sensorimotor processes, we
may define as an exceptionally complex state of human
reaction which is integrally related to all the other perception
and response processes. It is not one thing, but a combination
and a holistic integration of several seemingly diverse, yet
actually closely related, phenomena (Ellis, 1958, p. 35).

I think that all behavior therapy is really cognitive behavioral,

since even in its purest form it consists of teaching, educating,
and persuading clients to experiment with new behaviors. In
pointing this out in a discussion I had with Joe Wolpe in
1995, I got Wolpe to state:

There is an important point that Al Ellis made about cognitive

events in therapy. Yes, cognition enters into everything we
do. When we have a conversation, there is cognition on both
sides. When Al Ellis and I are talking to you now, we are
thinking that you are thinking, and all this is cognition. When
I am telling a person to assert himself in certain situations, I
am using his intellect, he is taking on what I say cognitively,
and will later by using his judgment in carrying out assertive
action (Ellis & Wolpe, 1997, p. 116).

Behavior Therapy, then, is just about always cognitive

behavioral; and cognitive-behavior therapy practically always
tends to be integrational, because its theory and practice hold
that there are many roads to treating disturbed individuals,
and these include various psychodynamic, interpersonal,
person-centered, and other therapeutic methods (J. Beck,
1992; Ellis, 1958, 1962, 1998, 1999; Goldfried 1980, 1985).
Even the radical behaviorists, such as Hayes (1994; Hayes,
Strosahl, & Wilson, 1999) have fairly recently included
distinctly cognitive and emotive methods in their form of

359
behavior therapy and seem to be increasingly headed in that
direction. So today, more than ever before, behavior and
cognitive behavior therapy are following REBT and Lazarus’
multimodal therapy in becoming wide-ranging in their
therapeutic procedures.

Case Study of Reactions to Rational

Emotive Behavior Therapy
My most cited publication on REBT was my book Reason
and Emotion in Psychotherapy (Ellis, 1962). The reactions to
it by leading therapists all over the world were almost
uniformly negative. Fritz Perls carried on a feud with me for
many years, contending that the book was boringly
intellectual and completely omitted any emotional element.
Carl Rogers never mentioned it publicly but, according to his
intimates, was solidly against it. Several leading
psychoanalysts called it superficial and made snippy remarks
about it.

Psychologists were much kinder. Francis Ilg and Louise B.

Ames, prominent child psychologists, called it “a most
important, unusually interesting, and at times terribly amusing
book.” The unorthodox psychoanalyst, Harry Bone, a leading
Sullivanian, was quite enthusiastic about it, and wrote:

Aside from his contributions to therapy, Ellis has made many

important thoroughly scholarly researches which do not have
as many readers as they deserve. I unhesitatingly recommend
his unique contribution to psychotherapy and his excellent
exposition of his highly original system. It seems to me that
Ellis’s basic principle of complete absence of blame of not

360
“blaming anyone for anything at any time,” is essentially
identical with Carl Rogers’ principle of unconditional positive
regard. The thoroughness with which they espouse this
principle and its implications together with their respective
ways of effectively implementing it, distinguishes their
systems from other systems. This is the source of their
potency and economy. I consider Ellis’ Reason and Emotion
in Psychotherapy the most important contribution to the field
since Carl Rogers’ contribution (Bone, 1968, p. 174).

Reviews of Reason and Emotion in Psychotherapy were

almost nonexistent; and Contemporary Psychology only
reviewed it when John Gullo, already a practicing
psychologist who used REBT, convinced the editors to
review it and gave it a very favorable review. Otherwise, the
book would have been ignored by this journal — as it was by
all the other professional journals. Despite this fact, Reason
and Emotion sold more copies over the years than almost any
other professional book of its day and a number of therapists
learned how to do REBT mainly by reading it. To my
surprise, although clearly written for the psychological
profession, it became popular in the self-help field and I have
many endorsements of it by readers who found it more
helpful than some of my other books for the public.

Speaking of my popular books, they have been much more

widely reviewed than my professional ones. A Guide to
Rational Living (Ellis & Harper, 1997) first published in
1961, has sold almost two million copies, has been largely
praised highly, and has been one of the books that therapists
have recommended most to their clients. It has received very
favorable reviews by many mental health professionals,
including Cyril Franks, Daniel Wiener, Thomas W. Allen,

361
Harold Greenwald, and Rowena and Heinz Ausbacher. Frank
Richardson said that “It is still perhaps the single best
‘self-help’ book available to lay persons and psychotherapy
clients” (1977, p. 271). Sol Gordon noted, “Still, in my
judgment, the most sensible and usable of the self-help
books” (1980, p. 203).

My popular books which applied REBT methods to sex, love,

and marriage problems have sold very well, but been heavily
criticized for their liberal views. Conservative professionals
and critics have often objected to them strongly and have
sometimes objected to the cognitive behavioral techniques
they presented, often for the first time, to the public (Ellis,
1958b, 1960). They have been influential in the field of sex
therapy, and have encouraged William Masters and Virginia
Johnson, Helen Kaplan, Joseph LoPicolo, Lonnie Barbach,
Bernie Zilbergeld and other authorities to adopt cognitive
behavioral methods. But they also have prejudiced some
professionals against REBT theory and practices in
non-sexual areas, because of their objections to my sexual
liberalism. On the other hand, my forthright sex writings have
influenced many members of the public and mental health
professionals to favor general CBT theory and practice.
Prejudice, apparently, goes both ways!

Possible Object Lessons for the Future

of Cognitive Behavior Therapy
Cognitive behavior therapy has come a long way since I first
started to do REBT in 1955. It is perhaps the most common
form of psychological treatment that therapists actually do

362
today, no matter what system of therapy they say they follow.
Many of its common procedures — such as cognitive
restructuring and in vivo shame-attacking exercises — are
widely used by many different kinds of therapists.

Moreover, in subtle or conscious ways, the use of various

kinds of cognitive instruction restructuring is commonly
employed in various forms of CBT (such as the constructivist
therapies of Mahoney [1991] and Neimeyer [Neimeyer &
Mahoney, 1995], in fairly pure behavior therapy (such as that
of Wolpe 1990), and in radical behavior therapy (such as that
of Hayes (Hayes, Stroshahl & Wilson, 1999)). So, again,
important elements of CBT are almost universally used today
in most psychotherapies (Alford & Beck, 1997; Ellis, 1987,
1994, 1999a, 1999b, 1999c).

This is exactly, I think, what preferably should happen in the

future. Cognitive behavior therapies had better be tested for
their effectiveness in their own right — though, actually, this
is difficult to do, since they include a number of different
cognitive, emotive, and behavioral techniques. But they also
can be at least partially integrated with methods derived from
psychodynamic, interpersonal, person-centered, and other
schools of therapy.

This kind of integration has always been experimentally tried

by many therapists. Even Freud (1965) gave occasional
activity homework assignments; and REBT practitioners have
at times used pollyannaish, unrealistic, and irrational
methods, to which normally they are allergic (Ellis, 1994,
1996, 1999a, 1999b, 1999c). Whatever works works! Though
as Hayes (1994) and I (Ellis, 1999c) have pointed out, some

363
workable methods can also interfere with clients’ using
deeper and more elegant methods of treatment.

Cognitive behavioral therapy is probably here to stay — and

to be constantly revised and improved. That is the way of
scientific endeavor, and therefore the way to continue to go!

References
Adler, A. (1929). The science of living. New York:
Greenberg.

Alford, B. A., & Beck, A. T. (1997). The integrative power of

cognitive therapy. New York: Guilford.

Alexander, F., & French, T. M. (1946). Psychoanalytic

therapy. New York: Ronald.

Beck, A. T. (1967). Depression. New York: Hoeber-Harper.

Beck, A. T. (1991). Cognitive therapy: A 30-year

retrospective. American Psychologist, 46, 382-389.

Bernheim, H. (1987/1947). Suggestive therapeutics. New

York: London Book Company.

Bloch, I. (1908). The sexual life of our time. New York:

Rebman.

Bone, H. (1968). Two proposed alternatives to psychoanalytic

interpreting. In E. Hammer (Ed.), Use of interpretation in
treatment (pp. 169-196). New York: Grune and Stratton.

364
Clark, D. A. (1997). Twenty years of cognitive assessment:
Current status and future directions. Journal of Consulting
and Clinical Psychology, 65, 946-1000.

Cove, E. (1923). My method. New York: Doubleday.

Dubois, P. (1907). The psychic treatment of nervous

disorders. New York: Funk and Wagnalls.

Ellis, A. (1950). An introduction to the scientific principles of

psychoanalysis. Genetic Psychology Monographs, 41,
147-212.

Ellis, A. (1955a). New approaches to psychotherapy

techniques. Brandon, VT.

Ellis, A. (1955b). Psychotherapy techniques for use with

psychotics. American Journal of Psychotherapy, 9, 452-476.

Ellis, A. (1956). An operational reformulation of some of the

basic principles of psychoanalysis. In H. Feigl & M. Scriven
(Eds.), The foundations of science and the concepts of
psychology and psychoanalysis (pp. 131-154). Minneapolis:
University of Minnesota Press. (Also: Psychoanalytic Review,
43, 163-180).

Ellis, A. (1957a). How to live with a neurotic: At home and at

work (Rev. ed.). Hollywood, CA: Wilshire Books.

Ellis, A. (1957b). Outcome of employing three techniques of

psychotherapy. Journal of Clinical Psychology, 13, 344-350.

365
Ellis, A. (1958a). Rational psychotherapy. Journal of General
Psychology, 59, 35-49.

Ellis, A. (1958b). Sex without guilt. North Hollywood, CA:

Wilshire Books.

Ellis, A. (1960). The art and science of love. New York: Lyle
Stuart & Bantam.

Ellis, A. (1962). Reason and emotion in psychotherapy.

Secaucus, NJ: Citadel.

Ellis, A. (Speaker). (1973). How to stubbornly refuse to be

ashamed of anything (Cassette recording). New York: Albert
Ellis Institute.

Ellis, A. (1979). Rational-emotive therapy: Research data that

support the clinical and personality hypotheses of RET and
other modes of cognitive-behavior therapy. In A. Ellis & J.
M. Whiteley (Eds.), Theoretical and empirical foundations of
rational-emotive therapy (pp. 101-173). Monterey, CA:
Brooks/Cole.

Ellis, A. (1987). Integrative developments in rational-emotive

therapy (RET). Journal of Integrative and Eclectic
Psychotherapy, 6, 470-479.

Ellis, A. (1993). Changing rational-emotive therapy (RET) to

rational emotive behavior therapy (REBT). Behavior
Therapist, 16, 257-258.

Ellis, A. (1994). Reason and emotion in psychotherapy (Rev.

ed.). Secaucus, NJ: Carol Publishing Group.

366
Ellis, A. (1996). Better, deeper, and more enduring brief
therapy. New York: Brunner/Mazel.

Ellis, A. (1999a). How to make yourself happy and

remarkably less disturbable. San Luis Obispo, CA: Impact
Publishers.

Ellis, A. (1999b, May 29). The importance of cognitive

processes in facilitating accepting in psychotherapy. Invited
address to the 25th Anniversary Annual Convention of the
Association for Behavior Analysis, Chicago.

Ellis, A. (2000). A continuation of the dialogue on counseling

in the postmodern era. Journal of Mental Health Counseling,
22, 97-106.

Ellis, A., & Harper, R. A. (1997a). A guide to successful

marriage (Rev. ed.). North Hollywood, CA: Wilshire Books.
1997.

Ellis, A., & Harper, R. A. (1997b). A guide to rational living.

North Hollywood, CA: Wilshire Books.

Ellis, A., & Wolpe, J. (1997). Discussion by Albert Ellis and

response by Joseph Wolpe. In J. K. Zeig (Ed.), The evolution
of psychotherapy. The third conference (pp. 115-119). New
York: Brunner/Mazel.

Ellis, H. (1936). Studies in the psychology of sex (2 vols).

New York: Random House.

Epictetus. (1890). The works of Epictetus. Boston: Little

Brown.

367
Forel, A. (1922). The sexual question. New York: Physician’s
and Surgeon’s Book Company.

Freud, S. (1965). Standard edition of the complete

psychological works of Sigmund Freud. New York: Basic
Books.

Fromm, E. (1955). The sane society. New York: Rinehurst.

Glass, C. R., & Arnkoff, D. B. (1997). Questionnaire methods

of cognitive self-statement assessment. Journal of Consulting
and Clinical Psychology, 65, 911-927.

Glasser, W. (1965). Reality therapy. New York: Harper &

Row.

Goldfried, M. R. (1980). Toward the delineation of

therapeutic change principles. American Psychologist, 35,
991-999.

Goldfried, M. R. (1995). From cognitive-behavior to

psychotherapy integration. New York: Springer.

Gordon, S. (1980). The new you. Lafayette, NY: Ed. U Press.

Hayes, S. C. (1994). Content, context and the types of

psychological acceptance. In S. C. Hayes, N. S. Jacobson, V.
M. Follette, & M. J. Dougher (Eds.), Acceptance and change:
Content and context in psychotherapy (pp. 13-32). Reno, NV:
Context Press.

Hayes, S. C., Strosahl, K., & Wilson, K. G. (1999).

Acceptance and Commitment Therapy. New York: Guilford.

368
Heidegger, M. (1962). Being and time. New York: Harper &
Row.

Herzberg, A. (1945). Active psychotherapy. New York: Grune

& Stratton.

Hoellen, B. (1993). Richard Huelsenbeck und Albert Ellis.

Zeitschrift fur Rational-Emotive Therapie und Kognitive
Verhaltens Therapie, 4, 5-37.

Hollon, S. D., & Beck, A. T. (1994). Cognitive and

cognitive-behavior therapies. In A. E. Bergin & S. L. Garfield
(Eds.), Handbook of psychotherapy and behavior change (pp.
428-466). New York: Wiley.

Horney, K. (1950). Neurosis and human growth. New York:

Norton.

Janet, P. (1898). Neuroses et idée fixes. Paris: Alcan.

Jones, M. C. (1920). The elimination of children’s fears

.Journal of Experimental Psychology, 7, 383-390.

Kelly, G. (1955). The psychology of personal constructs. New

York: Norton.

Kierkegaard, S. (1953). Fear and trembling and The sickness

unto death. New York: Doubleday.

Kwee, M. G .T., & Ellis, A. (1997). Can multimodal and

rational emotive behavior therapy be reconciled? Journal of
Rational-Emotive and Cognitive-Behavior Therapy, 15 (2).

369
Lazarus, A. A. (1989). The practice of multimodal therapy.
Baltimore, MD: Johns Hopkins.

Mahoney, M. J. (1974). Cognition and behavior modification.

Cambridge, MA: Ballinger.

Mahoney, M. J. (1991). Human change processes. New York:

Basic Books.

McGovern, T. E., & Silverman, M. S. (1984). A review of

outcome studies of rational-emotive therapy from 1977 to
1982. Journal of Rational-Emotive Therapy, 2(1), 7-18.

Meichenbaum, D. (1974). Self instructional training: A

cognitive protheses for the aged. Human development, 17,
273-280.

Meichenbaum, D. (1977). Cognitive-behavior modification.

New York: Plenum.

Neimeyer, R. A., & Mahoney, M. J. (1995). Constructivism in

psychotherapy. Washington, DC: American Psychological
Association.

Perls, F. (1969). Gestalt therapy verbatim. New York: Delta.

Richardson, F. (1977). Basic Ellis revised. Journal of

Individual Psychology, 33, 270-271.

Robie, W. F. (1925). The art of love. Ithaca, NY: Rational

Life Press.

370
Rogers, C. R. (1943). Counseling and psychotherapy. Boston:
Houghton Mifflin.

Rogers, C. R. (1951). Client-centered therapy. Boston:

Houghton Mifflin.

Salter, A. (1949). Conditioned reflex therapy. New York:

Creative Age.

Sartre, J. (1968). Being and nothingness. New York:

Washington Square.

Schutz, W. (1967). Joy. New York: Grove.

Silverman, M. S., McCarthy, M., & McGovern, T. (1992). A

review of outcome studies of rational-emotive therapy from
1982-1989. Journal of Rational-Emotive and
Cognitive-Behavior Therapy, 10(3).

Sullivan, H. S. (1953). The interpersonal theory of psychiatry.

New York: Norton.

Tillich, P. (1953). The courage to be. Cambridge: Harvard

University Press.

Watson, J. B. (1919). Psychology from the standpoint of a

behaviorist. Philadelphia: Lippincott.

Watson, J. B., & Rayner, R. (1920). Conditioned emotional

reactions. Journal of Experimental Psychology, 3, 1-14.

Wolpe, J. (1990). The practice of behavior therapy (4th. ed.).

Needham Heights, MA: Allyn and Bacon.

371
372
Chapter 10

From Psychodynamic to
Behavior Therapy:
Paradigm Shift and Personal
Perspectives
Cyril M. Franks

Distinguished Professor Emeritus, Rutgers University

“No wind blows in favor of a ship that has no direction.”

Essays, Michael Eyquem de Monte (Montaigne 1533 – 1592).

“The times they are a changin.’”

Bob Dylan, circa 1964.

Introduction
The value of an intellectual biography lies in tracing the
connection between events and thoughts. This intellectual
autobiography spans some 50 years, 3,000 miles and the
greatest adventure of all, a journey of the mind from the birth
of behavior therapy to its coming of age on the threshold of a
new millennium. My early years are likely to be of only
historic interest to the general reader and alien to most

373
American psychologists. Nevertheless, a description of these
formative years might clarify the manner in which my
thinking about behavior therapy evolved. The basic
ingredients include: a classical education stressing language
and literature; training in both the technology and
methodology of applied science; my wartime experiences; my
growing disappointment with psychoanalysis, at best a
pseudo-scientific model with a total unawareness of outcome
evaluation, accountability and the like; the cavalier treatment
of psychologists in medical facilities; the self-satisfied
medical establishment’s unawareness that any therapy other
than Freudian could be either possible or necessary. Half a
century ago, most clinical psychologists seemed to tolerate
their exclusion as therapists and see little that was wrong with
this situation. Inspired by the work of Hans Eysenck and his
associates, such factors sparked my desire to establish a new,
and eventually accepted, behavioral paradigm. To the best of
my knowledge, these matters have not been described
elsewhere.

Formative Years
I was born in 1923 in a primarily English — rather than
Welsh — speaking resort town in South Wales. Until war
came my childhood was pleasant and uneventful. As with all
children of promise, my education was fully subsidized from
kindergarten through postgraduate university levels by the
British regional department of education. I attended an elitist
high school for boys from ages 11 through 17 where selection
for a strictly academic education was determined by high
scores on a rigorous screening examination. Intellectually
qualifying girls received a similar education in a separate

374
school. My earliest recollection was of Dylan Thomas, a then
unknown one-of-the-crowd aspiring writer several years
ahead of me in school, striving to instill the elements of
Chaucer into my unappreciative head. In this rarefied setting,
typing and woodwork were out of the question and sports and
physical training were compulsory after-school activities in
addition to a demanding homework schedule. Nevertheless,
morale and teacher/student camaraderie were high despite
continuing academic pressures.

French or German, plus Latin, were compulsory and, in

addition, my family expected me to attend after-school
Hebrew classes and Jewish cultural studies three times a
week. Furthermore, all students had to go to weekly
Welsh-language classes where Welsh was of little interest to
many of us and of even less utility. All in all, I acquired an
excellent education and a lasting ability to express myself
effectively in speech and writing, both of which stood me in
good stead over the years.

In 1939 the long-expected war came to the UK But, despite

severe vicissitudes, my formal education continued
surprisingly smoothly until repeated nationwide Nazi
bombing razed our impressive 300-year-old school building
to the ground and killed several classmates. Understandably,
war curtailed what was previously an idyllic education.
Nevertheless, morale remained high as we tried to cope as
best we could. After four years of general education and
searching school examinations I opted, as was the custom in
those days, to specialize in three subjects for two more years:
applied mathematics, physics and electronics (premature
school specialization has been long since discontinued)
where, despite wartime disruptions, two years later I managed

375
to pass all school examinations. But, by then, my technical
interest in both physics and electronics had began to wane. As
the war intensified, at 18 years of age I was directed, because
of my specialization subjects, to become a full-time student at
the nearby University of Wales and complete an intensive,
accelerated four-year degree program in two years, continuing
with the same three high school subjects. It was too late to
change and, in any event, I had no say in the matter. Time
was not on the British side and, despite the Soviet Union’s
valiant war effort, we were still very much alone. Fortunately,
things changed when America entered the war. Two years
later I obtained my expected B.S. degree as planned and,
despite a still declining interest in applied physics and
electronics, once again I was drafted, this time to a top-secret
government facility for electronic “hot shots” who, for
painfully obvious reasons, soon turned out not to include me.

I was directed to work on the development of an urgently

needed infrared device for military truck driving in convoy at
night in total darkness where both sides were striving to
perfect this device prior to the imminently expected second
front. By this time, my technical deficits becoming
increasingly apparent, I was ignominiously reassigned to a
leading, but less pressured, electronics company in London.
The area turned out to be one of the most dangerous and
stressful regions in the hard-pressed British Isles. Here,
German rocket bombs, known as V2s, indiscriminately and
unpredictably rained down sudden death by day and night.
Without warning, several fellow workers were blown to
fragments in the coming months.

Fortunately, the war in Europe ended soon after and I was

free, subject to college acceptance, to study whatever and

376
wherever I chose, again at government expense. Although by
then vastly more interested in the methodology of science
than the technology of electronics, my love of literature and
writing remained. At the same time, my interest in people and
whatever made them “tick” blossomed. So, as a compromise,
I enrolled in a 16-month teacher-training program at London
University, specializing in teaching applied sciences and
elementary electronics at technical high school levels.

The Search
My first, and last, school job was teaching general science to
Merchant Service cadets at the London Nautical School.
Classes ending in the early afternoon left me ample time to
take numerous university courses in general and clinical
psychology. These under my belt, four years later I was
eligible to apply for admission both to what is now known as
a clinical internship and a Ph.D. program in psychology. Then
came the daunting task of finding an acceptable university
base. Even at this early stage I knew that “acceptable” meant
a broadly behavioral climate and a stringent questioning of
Freudian theory and practice which, even then, I regarded as a
pseudo-science. Freudian psychotherapy went unchallenged
by all mental health practitioners and necessities such as
control groups, validation, outcome evaluation, follow-up and
patient satisfaction were unheard of.

For three months I briefly visited and rejected most of the few
university departments in Western Europe offering training in
clinical psychology at that time. Another requirement, then
thought out of the question for non-physicians, was that
fully-trained clinical psychologists be permitted to serve as

377
therapists in medically controlled settings. I felt that this
situation would never change in the foreseeable future.
Fortunately, I was wrong but, at the time, I had no idea how
change might come about.

The next step was to make an extended visit to the University

of London Institute of Psychiatry, Maudsley Hospital, where
Hans Eysenck and his staff offered both an internship-type
program in clinical psychology and a related Ph.D. program
after completion of the former. So I looked no further after
making a second visit to the Maudsley, where Eysenck,
becoming well-known, endorsed my decision to apply for
admission to both programs. Eysenck’s Ph.D. program
offered a combination of clinical and experimental
psychology, which was just what I wanted. But even at the
Maudsley, clinically trained psychologists were expected and
permitted only to give psychological tests, write reports, do
some occasional vocational guidance and interviewing, and
engage in non-threatening, physician sanctioned research.
Most curious of all, or so it seemed to me, virtually all clinical
psychologists, outside the Maudsley, subscribed
unconditionally to a Freudian model.

What especially irked me was that, despite four years of

study, most clinical psychologists disregarded their training,
fully accepting Freudian dogma, and never thinking seriously
about an alternative. They also seemed to tolerate a
demeaning, hospital status. Thus, on my second Maudsley
visit, I knew where I wanted to spend the next few years as a
student. What I did not know at the time was that this
experience was to determine my career path for the rest of my
professional life.

378
Maudsley Days
Eysenck, then at the beginning of his career and already
department head, even in those early days, was a favorite
target of criticism. Prominent in psychology circles, Eysenck
had astutely gathered around him a coterie of loyal but, at
times, critical students and junior faculty periodically
reinforced by an infusion of visiting colleagues from around
the world. Our first, and long-term-goal, then regarded by the
mental health community as an impossible dream, was to
supplement the ubiquitous, and then only, model of Freudian
therapy with an, as yet, undetermined data-based learning
theory approach. In effect, this involved a paradigm shift
from a psychodynamic to a behavioral model and in which
the two paradigms were expected initially to co-exist with
gradually diminishing psychodynamic influence.

The notion of a paradigm, applied initially only to the social

sciences by Kuhn (1970), was later extended to psychology to
meet our needs. By a paradigm shift we meant a sweeping,
significant, and hopefully lasting, change in a prevailing,
comprehensive, explanatory system. For example, when
Galileo courageously announced his data-based explanation
of the locations and orbits of heavenly bodies in our solar
system, there was a gradual shift from the then literally
heretical, church-ordained dogma about the central positions
of the sun and earth, away from the old to the new
observation-based paradigm, much akin to the eventual
replacement of Freudian dogma with data.

Under Eysenck’s leadership, the second goal was to gather

enough data to validate an appropriate personality structure in

379
terms of two factorially determined orthogonal dimensions,
first neuroticism and introversion-extraversion followed,
much later, by psychoticism (the latter remaining more a hope
than a reality). To advance these goals a soundproof classical
conditioning laboratory had to be constructed, using primarily
eyeblink and GSR conditioning.

Program research, nothing new and offering many tactical

advantages, was the favored strategy. Thus, being familiar
with program research and, when my incorrectly presumed
technological skills also became known, I was invited to join
Eysenck’s department as an unpaid student member of his
research group and put in charge of the construction and
development of the new conditioning laboratory. Accepting
this invitation with alacrity I enrolled in both programs, first
the 12-month clinical internship following approval of my
carefully scrutinized psychology training and, second,
enrollment in the at least three-year full-time Ph.D. research
program.

Unfortunately, Monte Shapiro, Eysenck’s distinguished

colleague, directed the internship program. If my memories of
long-past events do not lead me too far astray, the clinical
internship presented a major dilemma. Most projective
techniques, correctly judged invalid, were not taught and
exclusive reliance was placed upon relatively more objective
assessment tests such as neuropsychology procedures,
intelligence testing, certain multiple-choice questionnaires,
the MMPI and structured interviews. Consequently, students
untrained in everyday bread-and-butter projective techniques,
such as the Rorschach, were unable to gain employment as
clinicians virtually anywhere. Fortunately, this total ban on

380
training in projective techniques was relaxed soon after even
though they were always used sparingly and with caution.

Reinforced by group enthusiasm and stimulating weekly

discussion in Eysenck’s home, I successfully, but not
enthusiastically, completed this fragile internship and began
my Ph.D. research in earnest as well as active participation in
the ongoing search for a new paradigm. The new conditioning
laboratory continued to be my responsibility by virtue of a
rashly presumed electronic expertise, laying the groundwork
for a series of published drug and personality conditioning
studies involving both normal and clinically abnormal
populations. Publication of an invited article about our new
laboratory in the prestigious journal Nature further provided a
timely and heartwarming boost to morale (Franks 1955).

After much investigation and thought various field models

were examined and found wanting in one way or another.
Eventually we settled on a combination of Hullian and
Pavlovian S-R learning theory and the methodology of
behavioral science adapted to meet our clinical requirements,
such as empirical validation, outcome evaluation, follow-up
and client satisfaction. Having found what then seemed to be
the only exclusively “behaviorally correct” formula, it took
me many years to realize that, without abrogating rigid
behavioral principles, I was as intolerant as our despised
Freudian counterparts. My thinking began to change with the
pioneering and influential text Behavior Therapy and Beyond
(Lazarus, 1971).

381
The Dawn of Behavior Therapy
Having agreed that our new approach to therapy would stem
from the notions of Pavlov and Hull, if only because there
was really no feasible alternative, we set out to develop some
behaviorally valid theory-based innovative procedures, a task,
which turned out to be more arduous than anticipated. At first,
laboriously we produced little more than a few novel, but
validated, parent-training reinforcement techniques, some
phobia extinction and anxiety reduction strategies and
unimaginative aversion conditioning training for the mentally
retarded. Not knowing better, initially we focused exclusively
on the presenting problem and none of this could even
remotely be called behavior therapy. This was around the
time that Eysenck (1959) coined the term behavior therapy in
Europe, Wolpe and Lazarus (1958) in South Africa and
Lindsley, Skinner and Solomon in the USA (1958), all
working independently. Regrettably, we overlooked the
operant conditioning of Skinner and his associates, perhaps
because we viewed Skinner as primarily an animal
psychologist and not a clinician.

My initial focus on trying to develop a conditioned aversion

to alcohol abuse seemed a feasible beginning but, as we soon
discovered, this was still not behavior therapy. All treatments
of alcoholism were notoriously unsuccessful to-date and,
consequently, to our surprise, some physicians were willing to
let lay therapists “have a try.” At first, I focused exclusively
on the development of a conditioned aversion to alcohol. In
so doing, never talking to the patient or looking into anything
meaningful in the patient’s life other than the alcohol abuse, I
never looked into relevant life circumstances. This procedure,

382
still unworthy of the name behavior therapy, or any therapy
for that matter, failed dismally, as we might have expected.
Slowly, I became more sophisticated and learned from
experience, publishing a few exploratory studies of the
behavioral treatment of alcoholism and achieving no real
success until I updated my thinking. Anticipating the
trail-blazing multimodal therapy of Lazarus (1971), I went
beyond conditioned aversion per se, began to talk to the
patient about seemingly significant concerns in addition to
presenting problems and started to explore multi-level
situations, real life settings and multi-stimulus approaches,
including the development of a conditioned aversion, trying
to modify the patient’s self-defeating life-style. Although our
short-term success rates were modestly better, but far from
spectacular, for me this was the beginning of primitive
behavior therapy (Franks, 1963), a tiny crack in the
physician’s Freudian superstructure.

When Dollard and Miller’s (1950) pioneering American text

tried to explain psychoanalytic practices and S-R learning
theory principles under one cover, I believed that things were
beginning to change for the better but, alas, I soon learned
that this was premature, merely wishful thinking! In essence,
what these two forward-looking mental health professionals
had unwittingly produced was a scholarly exercise in
translating the limited language of learning theory into
psychoanalytic gospel and vice versa in order to facilitate
communication. Both distinguished authors, one more
learning theory-oriented than the other, made the tacit
assumption that psychoanalytic theory and practice were
unassailable truths. What they had really produced was a
sophisticated dictionary. So, discouraged again, having
completed my internship, I took a year off to take an M.A.

383
degree at the University of Minnesota where primarily, I
learned that everyone “did his or her own thing.” Few had
heard of behavior therapy whereas, at the University of
Minnesota, everyone knew all about the MMPI!

A year later I returned to London with my American wife,

Violet, now a leading feminist behavior therapist, and I
completed my Ph.D. in 1954, eventually becoming a tenured
junior faculty member of Eysenck’s department. Meanwhile,
the name Eysenck had justifiably become a byword
throughout British psychology and elsewhere. Parenthetically,
Eysenck’s published autobiography was provocatively called
Rebel With a Cause! Over the years Eysenck reveled in
specious arguments and outrageous statements. Two years
after his 1998 death the debate continues – was Eysenck a
brilliant scientist or a controversial provocateur? Probably
both.

Move to the USA, Start of the AABT,

and Subsequent Developments
In 1958, Wolpe’s groundbreaking desensitization text, already
well-known to our Maudsley group in manuscript form
appeared in print, first in the USA then worldwide. For the
first time there was, at last, a viable behavioral alternative to
psychodynamic therapy which could be readily applied to
treating the sorts of patients and problems that make up the
bulk of general psychiatric practice. Thus, still disappointed
with Dollard and Miller but encouraged by Wolpe’s book, I
decided to relocate to the USA in the naïve hope that

384
American therapists would offer better opportunities in a
more friendly climate. Once again I was disappointed.

Nevertheless, in 1957 I accepted a position as Director of

Psychology at the Neuropsychiatric Institute, Princeton, NJ
where my first goal, with the help of an NIMH grant, was to
establish an American counterpart of our British conditioning
laboratory. Gradually this became a reality and the
publication flow went across both sides of the Atlantic. With
Eysenck, my second and equally important goal, facilitated by
Wolpe’s text, was to make a behavioral mental health
paradigm a meaningful reality. Unfortunately, I soon found
that the climates in both the USA and the UK, among
psychiatrists and clinical psychologists alike, were much the
same. In this frame of mind, discouraged again, I came across
few kindred spirits until a chance meeting occurred, a meeting
which would lead eventually to the creation of the
Association for Advancement of Behavior Therapy (AABT).

Dorothy Susskind, then a graduate student at New York’s

Yeshiva University, was in the final stages of a Ph.D.
dissertation examining certain aspects of Eysenck’s work.
Soon I learned not only more about Susskind’s dissertation
but also about our common dissatisfaction with mental health
training and therapy in both countries and, even more to the
point, we talked extensively about what might be done about
it. I also learned that Susskind was both a knowledgeable
clinical psychologist and an experienced administrator and
organizer. As a first step we set up several meetings with a
few carefully selected participants and soon we were
convening regularly to explore the possibilities of forming our
own organization. Eventually, about 40 seriously interested
professionals formed the nucleus of what later became

385
AABT. These individuals include such luminaries as Paul
Brady, Edward Dengrove, Andrew Salter, Arthur Staats,
Leonard Ulmann, Leonard Krasner, Arnold Lazarus, Joseph
Cautela and Joseph Wolpe in addition to Dorothy Susskind
and myself. We constituted the initial planning group, with
Stuart Agras and Eysenck offering periodic support from afar.
Sadly, Salter, Wolpe, Eysenck and Cautela are no longer with
us and we continue to miss them. On a happier note, three
founding fathers are still professionally active; Staats,
Lazarus, and myself.

Our first task was to settle on an appropriate name for our

fledgling association. Having been actively involved with the
British Association for Advancement of science for several
years, my first name suggestion was unanimously accepted
and our organization briefly became the Association for
Advancement of the Behavioral Therapies. From the start, the
word “advancement” was meant to imply scholarly rather
than personal advancement. An equally important goal was to
develop behavior therapy as a conceptual unity and, by
general consent, the name of the AABT was changed to its
present singular form and the name, Association for
Advancement of Behavior Therapy has remained this way
ever since.

In 1966, I was elected first President of the AABT, probably

because no one else was willing to take on this demanding
chore and, probably for similar reasons, I also agreed to
become the first Editor of the Association’s newly formed
Newsletter and first Program Chair for a half-day meeting
held in parallel with the Annual Convention of the American
Psychological Association. In those early months, Susskind
and I carried out much of the slog work, including such

386
mundane chores as stuffing envelopes and folding and
mailing announcements. It was around this period, 1970, that
I became Professor of Psychology at Rutgers University.

Later, for nine years, I also served as founding Editor of the

AABT’s first journal, Behavior Therapy, including choosing
and negotiating with a suitable publishing house. Throughout
this period my blinkered vision of behavior therapy gradually
matured. It took many years for me to appreciate that there is
no such thing as an interaction involving behavior alone,
making me wonder how so-called cognitive behavior therapy
could be paradigmatically different from behavior therapy.
And, though not well developed as yet, a similar argument
applies to affect.

The Times They are Indeed a

Changin’
In 2001, the AABT, now firmly launched, has a membership
approaching 6,000. Behavior therapy and its now probably
dominant offshoot, cognitive behavior therapy, exceed most
optimistic expectations as far as numbers alone (a flawed
criterion) are concerned. Perhaps some two dozen behavior
therapy, cognitive behavior therapy and behaviorally related
journals of one kind or another, mostly in English, are
scattered around the globe. Nowadays, for the most part the
behavioral and the Freudian paradigms, respectively, coexist
in mutual tolerance if not always in mutual peace and
harmony. With maturity and security the early need of some
behavior therapists to adopt an I-am-better-than-you attitude
is no more. More important, especially in the USA and UK,

387
the occasional pockets of resistance, still encountered from
some physicians, are becoming fewer and fully-qualified
nonmedical behavior therapists are now free to practice in
most psychiatric facilities on almost equal footings—a
remarkable index of progress in a comparatively short time.

While this is not the place to write a needed history of

behavior therapy, if I may be indulged at this point I would
like to mention, by way of history, a few personal
accomplishments which continue to give me gratification.
First, there is my continuing Editorship of the quarterly
journal Child and Family Behavior Therapy, now in its 25th
year. Second, with different collaborators at various times,
Terry Wilson and I produced the Annual Review of Behavior
Therapy for 12 years, gaining in size and substance and
serving as a continuing chronicle of significant behavioral
happenings from year to year. The series continued without
interruption until the growth of the behavioral literature, some
written in languages other than English, made it impossible to
maintain this pace. Reluctantly, the series had to be
discontinued. Hopefully, someone, or some group, will
eventually take on this daunting task.

Third and fourth are two earlier texts, both “firsts” in their
respective ways. The first, Conditioning Techniques in
Clinical Practice and Research (Franks, 1964), focused on
the limited behavior therapy literature of that time. The
second, an overview of the still limited behavioral literature in
1969, consists primarily of an appraisal of what I then
considered to reflect the overall status of behavior therapy,
written by others and myself. Appropriately, the title is
Behavior Therapy: Appraisal and Status (Franks, 1969).

388
Since the field changes so rapidly, both books are now of
primarily historic interest.

In terms of more recent developments, the rapid rise of

cognitive behavior therapy and its impact upon behavior
therapy at large is remarkable. With little doubt, cognitive
behavior therapy, for better or worse, probably both, is now
the dominant influence in our field. But I am still unclear
what is meant by cognitive behavior therapy, a term used in
diverse ways. In this regard, one may speculate, perhaps
uncharitably, about the needs of a few cognitive behavior
therapists to achieve professional visibility as soon as possible
even if this sometimes entails taking a few shortcuts. Other
cognitive behavior therapists sincerely perceive many of the
early, painstaking methodological exactitudes as now
unnecessary and, perhaps, never were necessary. I now
recognize that both positions have merit and Krasner’s verdict
(in this book) that cognitive behavior therapy is the
“oxymoron of the century” may be unwarranted. In any event,
regardless of validity and long-term outcome, a variety of
appealing and creative new cognitive behavior therapy
techniques emerge regularly. Part of the problem is that the
original, precise notions about the nature of a stimulus and
response sometimes become unrecognizable and, in so doing,
cognitive behavior therapy is in danger of becoming
nonexclusive, thereby blurring the once unique identity of
behavior therapy.

At first, the combination of changing times and the many new

faces of behavior therapy made me feel very uneasy, so much
so that, in 1981, I wrote an article with the self-explanatory
title “2081: Will we be many, or one—or none?” (Franks,
1981). It seemed to me, at that time, that my vision of the

389
unity of behavior therapy was rapidly falling apart. Until the
1990’s my still lingering belief in the unity of behavior
therapy seemed feasible and that, somehow, even the unity of
psychology at large was at hand, so much so that, in 1993, I
became President of Arthur Staats’ now defunct organization
SUNI: the Society for Studying Unity Issues. Now, however,
recent events force me to reconsider my position both with
respect, on the one hand, to the desirability of the notion of
unity in behavior therapy and, on the other hand, the
probability that diversity rather than unity in behavior therapy
adds vitality to our movement. After all, not surprising, if
long-established disciplines such as physics and medicine
continue to fall far short of unity how can the relatively new
discipline of behavior therapy expect to succeed? Regrettably,
Staats’ lifelong attempt to generate a comprehensive system
of clinical/behavioral psychology has not received the
recognition it merits.

As I now see it, behavior therapy is becoming, at least,

bi-modal. There is the declining minority of hard-core,
card-carrying behavioral diehards who still conform faithfully
to a strict, behavioral foundation and there is the
freewheeling, growing majority of loosely behavioral
psychologists, especially the self-styled cognitive behavior
therapists, and there is a spectrum of positions in between. I
still favor some form of behavioral orientation and some form
of recognizable S-R learning theory but I no longer regard
changes in the perceived nature of behavior therapy with
apprehension. Consequently, I am now prepared to consider
seriously most novel interpretations of behavior therapy. But
there is one proposed innovation that I find hard to stomach
and that is the proposal to combine behavior therapy and
psychodynamic therapy in one way or another. For reasons

390
obvious to me I find this curious notion totally unacceptable.
Behavior therapy and psychodynamic therapy are two
incompatible systems and both, in my book, would probably
fare better to go about their respective ways (Franks, 1984).

Behavior therapy began as an exclusively behavioral protest

against the prevailing psychodynamic regime. Originally a
hard-nosed entity, behavior therapy is now confronted with a
panoply of bewildering behaviorally-related acronyms: CT,
BT, RECBT, ACT, DBT, and more. There are probably a
dozen varieties of CBT alone. Conditioning, Dialectical,
Cognition, Behavior, and Affective are vying with each other
for their share of the limelight.

My journey has come a long way in the past 50 years, from

closed-minded notions about the nature of behavior therapy to
tolerance and cautious acceptance. As I continue my journey I
hope one day to see something which, so far, has not
occurred, the emergence of some new and fundamental
behavioral concept. Since the eras of Pavlov and Skinner, no
new generally accepted concept of significance relating to
conditioning/behavior therapy has emerged. The impressive
accomplishments and “busy work” of behavior therapists are
gratifying but I hanker for a new Pavlov or Skinner to
emerge. Maybe I am expecting too much. Meanwhile, I have
reached the present stage of my journey, a stage still in
formation, with a growing awareness of alternate behavioral
models, different than the one to which I have long been
accustomed.

Decades ago, during an extended visit to what was then

known as the U.S.S.R., my Soviet colleagues explained
politely but firmly that behavior therapy, well-developed in

391
the West, could legally be neither practiced nor studied
anywhere in the then Communist world regardless of the fact
that my brand of behavior therapy stems largely from the
work of Pavlov. In all Communist countries, behavior
therapy, said the Soviets, is a regressive product of S-R
learning theory built upon a “false and simplistic” 19th
century mechanistic materialism. Patiently, my Soviet
colleagues tried, with no success, to correct my faulty
thinking to reflect a “politically correct” dialectical
materialism rather than my “primitive” mechanistic
materialism.

It was not until very recently that I admitted that, as far as the
differences between mechanistic and dialectical materialism
are concerned, the Soviets had a point, but, of course, I never
could accept the Marxist overtones. Only recently did I begin
to appreciate contemporary developments in operant
conditioning, applied behavior analysis and their, as yet
largely unrealized clinical potentials. Gradually I became
aware of very different behavioral models couched in
functional dynamic, operant conditioning and contextual
terms involving notions still new to me, such as integrative
couples therapy, dialectical behavior therapy and more. It is
time to take into account more than an exclusively
mechanistic behavior therapy, models hitherto neglected by
most therapists, models not, as yet, well-developed clinically
in sharp contrast to their mechanistic, classical
conditioning-based predecessor.

It was not until late 1999 that Hayes and Toarmino clearly
pinpointed two very different, and rarely interacting,
traditions in behavior therapy, thereby sharpening my fuzzy
thinking about such matters, two compartmentalized

392
traditions with few points of interaction, the first popular and
well-established clinically and the second clinically not
well-developed and yet to show what it could accomplish; the
first tradition, still largely mechanistic and based upon
classical conditioning, the second less mechanistic, stemming,
in large part, from operant conditioning, and applied behavior
analysis.

For the time being, while recognizing the virtues of both

traditions and the many impressive accomplishments of the
first, I will continue to work within both traditions while
learning more about the second. As yet I am more
comfortable with, and more knowledgeable about,
mainstream traditional behavior therapy. Meanwhile, the
second tradition waits in the wings ready to move to center
stage. Meanwhile I will work with interest and curiosity
towards what the future of behavior therapy may bring.

Postscript
In 1970 I became Professor of Psychology at Rutgers
University in its clinical Ph.D. program. From time to time
my positions have included Clinical Director and Director of
post-Doctoral Training. In 1974 the Rutgers Graduate School
of Applied and Professional Psychology came into being,
offering a full-time doctorate in applied and professional
psychology (Psy.D.) for clinicians, the first of its kind
anywhere. My active involvement with both programs
continued until retirement as Distinguished Professor
Emeritus in 1991.

393
The evening of retirement festivities began with a merciless,
but lovingly crafted, “roasting” engineered by half a dozen
longtime associates in the presence of several hundred
colleagues and friends from the USA and abroad. The
evening concluded with a quite different, totally unexpected,
event, the annual “Cyril M. Franks’ Award for Excellence in
Research.” A small committee would meet annually to select
what, in the committee’s opinion, was the most outstanding
doctoral dissertation of the year at Rutgers in professional
psychology. As part of the graduation ceremony, held at the
end of each academic year, the award-winning new doctor is
presented with an appropriately inscribed plaque and his or
her name added to an annual, wall-mounted list. The recipient
also receives a modest honorarium, a procedure followed ever
since.

Following retirement, other than formal teaching and tedious

committee meetings I did almost as much as usual until, in
1997, I suffered a major stroke which still severely curtails
most professional activities other than an occasional address
on special occasions and the editorship of my journal Child
and Family Behavior Therapy. This brings my chronicle
up-to-date, a culmination of half a century of endeavor. At the
AABT’s 34th Annual Convention, in November, 2000, I
became the recipient of the Association’s Lifetime
Achievement Award, presented at a formal Awards
ceremony. I will always be receptive to new ideas as long as I
am able. As Gertrude Stein is alleged to have said on her
deathbed after a life of characteristic bombast, “I will
continue to ask questions and seek answers until better
questions and better answers come along.”

394
References
Dollard, J., & Miller, N. D. (1950). Personality and
psychotherapy. New York: McGraw-Hill.

Eysenck, H. J. (1959). Learning theory and behaviour

therapy. Journal of Mental Science, 195, 61-75.

Franks, C. M. (1955). A conditioning laboratory for the

investigation of personality and cortical functioning. Nature,
175, 984-985.

Franks, C. M. (1963). Behavior therapy, the principles of

conditioning and the treatment of the alcoholic. Quarterly
Journal of Studies of Alcohol, 24, 511-529.

Franks, C. M. (1964). Conditioning techniques in clinical

practice and research. New York: Springer Publishing
Company.

Franks, C. M. (Ed.). (1969). Behavior therapy: Appraisal and

status. New York: McGraw-Hill.

Franks, C. M. (1981). 2081: Will we be many, or one, or

none? Behavioural Psychotherapy, 9, 287-290.

Franks, C. M. (1984). On conceptual and technical integrity

in psychoanalysis and behavior therapy: Two fundamentally
incompatible systems. In H. Arkowitz & S. B. Messer (Eds.),
Psychoanalytic therapy and behavior therapy (pp. 223-247).
New York: Plenum Press.

395
Hayes, S. C., & Toarmino, D. (1999). The rise of clinical
behaviour analysis. The Psychologist, 12, 105-108.

Kuhn, T. S. (1970). The structure of scientific revolutions

(2nd ed.). Chicago: University of Chicago Press.

Lazarus, A. A. (1971). Behavior therapy and beyond. New

York: McGraw-Hill.

Lindsley, O. R., Skinner, B. F., & Solomon, H. C. (1953).

Studies in behavior therapy (Status report I). Waltham, MA:
Metropolitan State Hospital.

Staats, A. (1996). Behavior and personality: Psychological

behaviorism. New York: Springer Publishing Company.

Wolpe, J. (1958). Psychotherapy by reciprocal inhibition.

Stanford, CA: Stanford University Press.

396
Chapter 11

Cognitive Behavior
Therapy: The Oxymoron of
the Century
Leonard Krasner

Stanford University

Introduction
The topics I plan to cover are the rationale for the topic of the
paper “Cognitive Behavior Therapy – The Oxymoron of the
Century,” a history of “behavior therapy,” “behavior
modification” and the Unabomber, some views and the
joining of the label of “cognitive,” of cognitive therapy, with
“behavior therapy” and a quick touch on the future of
“behavior therapy.”

I begin my paper with a confession that I suffer from a very

serious mental disorder called “throw nothing away-itis.” So
in preparing this paper, I tried to go through the cartons of
articles, papers, and correspondence on “Behavior Therapy”
that I have in my house, as well as dozens of books, journals
and magazines I have on “Behavior Therapy.” Since there are
page limitations, I apologize in advance for omitting any
relevant references.

397
In effect, the contentions of this paper are that both behavior
therapy and cognitive therapy are helpful procedures in
alleviating health and mental health problems, although my
own bias makes me prefer behavior therapy. I am not
precluding a therapist using both procedures, cognitive
therapy and behavior therapy to help a troubled client.
However, the major theme of the paper is that the label
“cognitive behavior therapy” is an “oxymoron.”

Behavior Therapy
As for the history of behavior therapy, I will start with a paper
I wrote on “Behavior Therapy.” It was in volume 22 of the
Annual Review of Psychology, 1971, and it was the first paper
on this topic in the Annual Review.

In this 1971 article of 28 years ago, I cited 397 publications,

which included publications by almost all of the speakers in
this week’s conference: Sidney Bijou, Leo Reyna, Ogden
Lindsley, Albert Bandura, Stewart Agras, Walter Mischel,
Leonard Krasner, Cyril Franks, Donald Baer, Todd Risley,
Arnold Lazarus, Montrose Wolf, Gordon Paul, Gerald
Davison, and the father of Julie Vargas, a gentlemen named
Fred Skinner.

The first paragraph of the 1971 article:

In recent years most authors start their chapters in the Annual

Review of Psychology with a declaration that the task of
reviewing their particular field is overwhelming because of
the literature explosion. We can do no less, especially in view
of the fact that this is the first Annual Review chapter on

398
behavior therapy. In preparing the material for this chapter we
have compiled a bibliography of over 4000 items, most of
which should be included in any comprehensive review of
this field. Needless to say, space limitations in this volume
and the finiteness of the author’s life preclude full justice
being done to all.

In this article, I noted that there were 15 streams of

development that have come together in the last part of the
1960’s to form a distinctive approach to helping individuals
with behavior socially labeled as deviant. These 15 streams
were:

Perhaps the most important general stream is that of

experimental psychology and within it the concept of
“behaviorism.” A recent definition of behaviorism by Kantor
(1969) captures the spirit of this viewpoint.

What is behaviorism? … it is a renunciation of the doctrines

of soul, mind, and consciousness. Positively expressed,
behavioristic psychology is the study of the behavior of
organisms interaction with their surroundings.

The field of instrumental conditioning tracing back to

Thorndike and overwhelmingly influenced by the research
and philosophical views of Skinner (1938, 1953, 1957, 1961,
1966, 1969) has represented the most influential stream in the
development of behavior therapy as exemplified by the
volume of research and application reported in the current
literature.

A major influence on current behavior therapy has been the

research, clinical work, and writings of the psychiatrist Wolpe

399
(1958, 1968, 1969), who in turn was influenced by the
psychologists James Taylor and Leo Reyna. Wolpe
introduced the technique of reciprocal inhibition, basing it in
part on the classical conditioning research of Pavlov and Hull.

A group of psychologists and psychiatrists, working within

the framework of experimental psychology primarily
influenced by Hullian learning concepts, practiced at the
Maudsley Hospital in London under the general direction of
Eyseneck (1960, 1964).

There was a stream of investigators and practitioners in the

United States, frequently working in educational institutions,
who were applying behavioral, conditioning, and learning
concepts to various problem behaviors. Influenced by J. B.
Watson, this stream can be traced through the works of Mary
Cover Jones, Burnham, Dunlap, Mowrer (1938),
Hollingworth (1968), Guthrie, and more recently Phillips
(1968) and Pascal (1959). In effect, these investigators can be
retroactively labeled as behavior therapists.

There has been a group of investigators attempting to

interpret psychoanalysis in learning theory terms (Hullian).
The most influential work has been the Dollard & Miller
(1950) book and the papers of Shoben and Shaw of that same
period.

An important stream is that tracing directly to the applications

of the research of Pavlov. Treatment procedures in the Soviet
Union and to a lesser extent in this country have been directly
influenced by Pavlovian classical conditioning.

400
The social psychology laboratory has contributed a
psychology of social influence which is being increasingly
incorporated into a behavior therapy framework.

The field of developmental and child psychology has offered

us a stream of research exemplified by the contributions of
Sears, Miller & Dollard, Bijou, Baer, Gewirtz, N. Ellis, and
Stevenson. The current emphasis is on vicarious learning via
modeling such as the research of Bandura (1968).

Investigations of the parameters of the social influence

process have brought within its framework a series of human
interactions previously seen as unique or discrete phenomena.
This has included research on such clinical experimenter bias,
subject and patient expectancy, and the effects of non-verbal
cues in interviews.

It is now clear that there are available behavioral, social

learning alternatives to the traditional disease model of
psychopathology as developed in the mid-nineteenth century.
Every one of the current books on behavior therapy presents
such alternative behavioral models. More than any other
point, this change in the conceptualization of the target
behavior is the key to modern behavior therapy.

In tracing the streams of development of behavior therapy, it

is of importance to include one negative stream. That is, the
apparent failure of psychodynamic and psychoanalytic
psychotherapies as indicated by outside critiques and internal
dissatisfaction. It is not a question of denigrating an opposing
point of view but rather of becoming aware of the broader
paradigm clashes that are involved. Behavior therapy grew, in

401
large part, out of dissatisfaction with traditional
psychotherapy techniques.

The training of clinical psychologists has traditionally been

within the so-called “Boulder model” which conceived of the
professional role of the clinical psychologist both as a
research scientist and as a professional person who applies
that science. Behavior therapy opens up the possibility of
finally being able to successfully achieve the spirit of this
training, demonstrating that the researcher and the clinician
are not separate but are integral parts of one role.

Cognizance must be taken of a stream of influence from

psychiatrists who have stressed the importance of observation
of behavior and human interactions such as Adolph Meyer, H.
S. Sullivan, and those therapists who attempted
environmental manipulation in ward milieu programs and in
some community programs.

There is an increasingly important stream that can be labeled

utopian in its emphasis on planning the social environment to
elicit and maintain the best of man’s behavior. It includes an
ethical concern for the social implications of behavior control,
as well as offering blueprints for a better life such as
Skinner’s Walden Two (1948). This stream can of course be
traced from Plato’s Republic to the setting up of a token
economy on a psychiatric ward or in a community setting.

These historical streams of development are now converging

into the field of behavior therapy. Thus behavior therapy is
more than a series of techniques or the application of learning
theory; it is a broad conceptualization of human behavior.

402
Unless we view behavior therapy in this context, its
applications and implications will be irrelevant.

A major integrating theme of the early post-World War II

group of behavior therapy investigators was a broad model of
human behavior which emphasized social/environmental
causation. The basic theoretical framework of the emerging
“behavior therapy” movement of the 1960’s was its social
learning alternative to the then current pathology oriented
medical school (Ullmann & Krasner, 1965).

The issues in very broad terms of clashing conceptual models

of human nature, oversimplified perhaps, by calling them
“inner” and “outer” explanations of locus of causation of
behavior (the perennial “nature” vs. “nurture” controversy,
which has been with us as a society for a very long period of
time). There are theorists and investigators who conceptualize
in terms of inner concepts, variables or metaphors, such as
disease, pathology, traits, personality, mind, cognitions, and
mind-body, health-illness dichotomies. Others, primarily but
not exclusively, identified as behaviorists, focus on the outer,
environmental, social consequences, social learning emphasis,
and a “utopian” stream—the planning of social environments
to elicit and maintain the best of human behavior.

Thus, we are arguing that behavior therapy developed in the

1960’s, and that it represented a clear alternative paradigm in
the mental health industry to the then predominant paradigm
with its focus on inner processes. This is, of course, not to say
that there were no usage of inner concepts and terminology in
behavioral thinking, such as awareness, self, anxiety, phobia,
conditioning, bias, expectancy, etc. However, the major focus
was on outer environmental concepts.

403
The first use of the term behavior therapy in the literature was
in a 1953 status report by Lindsley, Skinner, and Solomon,
referring to their application of operant conditioning (of a
plunger pulling response) research with psychotic patients.
Lindsley suggested the term to Skinner, based on it simplicity
and linkage to other treatment procedures.

Independently of this early usage, Lazarus (1958) used the

term to refer to Wolpe’s application of reciprocal inhibition
techniques to neurotic patients, and Eysenck (1959) used the
term to refer to the application of what he termed “modern
learning theory” to the behavior of neurotic patients based in
large part on the procedures of a group of investigators then
working at the Maudsley Hospital in London.

Cyril Franks, in his 1969 volume on Behavior Therapy:

Appraisal and Status, pointed to the disagreements then
among self-identified behavior therapists as to the definition
of behavior therapy. He pointed out that “Responses alone are
the data available to the student of human behavior, and all
else is a matter of inference and construct.” Franks noted and
attributed theoretical importance to the base of a “common,
explicit, systematic and priori usage of learning principles to
achieve well-defined and pre-determined goals.”

During this period of the 1960’s there were considerable

publications about, discussions of, and drawing up of lists of
the ways in which behavior therapy differed from traditional
psychotherapy. These lists ranged from those having a dozen
or more clear differences (pointing to behavior therapy as a
clear alternative model) to those de-emphasizing the
differences or even calling for a bridging of the two
approaches. Whether it is possible, or even desirable to

404
combine the two approaches was a controversy of the period.
Eysenck (1969), for example, took the position that bridging
between behavior therapy and psychoanalysis was
undesirable. My own position on this issue is that the bridging
of two paradigmatic models which have historically
developed as clear alternatives to each other would generally
show a misunderstanding of the basic principles of both.

A unifying factor in behavior therapy was its basis in

derivation from experimentally established procedures and
principles. The specific experimentation varied widely but
had in common all of the attributes of scientific investigation
including control of variables, presentation of data,
replicability, and a probabilistic view of behavior.

A more encompassing framework comes from those who

viewed behavior therapy in the broader context of social
learning terms. (The authors: Bandura, 1969; Eysenck and
Rachman, 1968; Franks, 1969; Kanfer and Phillips, 1970;
Krasner and Ullmann, 1965; Staats, 1962; Ullmann and
Krasner, 1965; and Wolpe, 1969). My opinion, and I stress
the word “opinion,” was that the disciplinary matrix and
exemplars manifested in these books did indeed represent a
new paradigm that was a clear alternative to that then current
in the healing professions focusing on inner pathologies.

Another succinct descriptive statement about behavior

therapy comes from the first editorial in the first issue of a
new journal Behavior Therapy in 1970, by the first editors of
the journal Cyril Franks and John Paul Brody – “Behavior
therapy did not arise from a single source but resulted from
the confluence of several diverse streams of thought, each of

405
which entailed more or less original ways of thinking about
and approaching clinical problems.”

Behavior Modification
Next we move on to the label of “behavior modification” and
its impact on an individual who has become quite infamous in
our society, the Unabomber.

We adopted the description of behavior modification offered

by Watson (1962). In presenting a historical introduction to
Bachrach’s (1962) collection of research on the experimental
foundations of clinical psychology, Watson used the term
behavior modification to cover a multitude of theoretical
approaches:

It includes behavioral modification as shown in the structured

interview, in verbal conditioning, in the production of
experimental neuroses, and in patient-doctor relationships. In
a broader sense, the topic of behavior modification is related
to the whole field of learning (Watson, 1962, p. 19).

The field of behavior modification itself was a major

illustration of learning theory applied in the environment. In
their introduction to this field, Ullmann and Krasner (1965)
defined the then emerging field in the framework of applied
learning theory: “In defining behavior modification we follow
the work of Robert Watson…who noted that behavior
modification included many different techniques, all broadly
related to the field of learning, but learning with a particular
intent; namely; clinical treatment and change” (p. 1).

406
Bandura (1969), in a most influential and widely cited book,
placed “the principles of behavior modification” within the
conceptual framework of social learning….By requiring clear
specification of treatment conditions and objective assessment
of outcomes, the social learning approach…contains a
self-corrective feature that distinguishes it from change
enterprises in which interventions remain ill-defined and their
psychological effects are seldom objectively evaluated (p. v).

Bandura integrated the investigations, by then greatly

expanded, that were derived from the influence of Skinner,
Wolpe, and the British group (e.g., Eysenck).

On April 26, 1995, the New York Times published on the

front-page excerpts from a letter received from the
Unabomber. The first three sentences were:

We have nothing against universities or scholars as such, all

the university people whom we have attacked have been
specialists in technical fields. (We consider certain areas of
applied psychology, such as behavior modification to be
technical fields.)

Reading that article in the New York Times was disturbing

because it was 1995 and by then the term “behavior
modification” had pretty much disappeared from the literature
replaced by “behavior therapy” and the first books which had
the term “behavior modification” in the title “Case Studies in
Behavior Modification” and “Research in Behavior
Modification” were written and edited by Leonard Ullmann
and myself in 1965. My first thought on seeing that New York
Times article was – could the Unabomber have been a
graduate student of mine? In fact, I got several phone calls

407
from graduate students saying “watch your mail.” However,
within the next year, Ted Kaczynski was arrested as the
Unabomber and on the list of people he had sent bombs to, in
1985, was James McConnell, a psychology professor at the
University of Michigan, who had written articles on
“behavior modification.” McConnell did not open the
package but his secretary did and was badly hurt.

Kaczynski had been a graduate student at the University of

Michigan and obtained a doctorate degree in math there.
Thus, he may have taken a psychology course with
McConnell or read of his research. My own surmise is that
Kaczynski correctly interpreted the consequences of the
behavioral modification model, being able to train people in
changing their environment could and should lead to a better
and more ideal world as symbolized by the utopian book
Walden Two written by one of the major founders of
“behavior therapy” and “behavior modification,” a gentlemen
named Fred Skinner.

It may be that Kaczynski was determined to prevent this

utopian society since he would not be in control of such a
society. Anyway, keep alerted to any new information that
may emerge about the Unabomber.

Cognitive Behavior Therapy

Having disposed of the Unabomber, I hope we now come to
the section of this paper which elaborates on “cognitive
behavior therapy as an oxymoron.” We start with dictionary
definitions of “cognition,” “behavior,” and “oxymoron.”

408
Cognition: “The mental process or faculty by which
knowledge is acquired. That which comes to be known as
through perception, reasoning or intuition; knowledge”
(American Heritage Dictionary).

Behavior: “Manner of conducting oneself, manners, conduct,

course of action” (American Heritage Dictionary).

Oxymoron: “A rhetorical figure by which contradictory terms

are cojoined so as to give point to the statement or expression;
a contradiction in terms” (Oxford Universal Dictionary).

In presenting the dictionary definitions of “cognition” and

“behavior” we are presenting the basis of two contradictory
models. We can simplify the models by labeling them as
“inner” and “outer,” cognition is an inner and behavior is an
outer model. In terms of applying these models to working
with human beings, in systematic research or in helping with
problems in life. These two models should be labeled
“cognitive therapy” and “behavior therapy.” We are not
saying which is more effective or desirable. We are not even
contending that both models should not be applied to the
same individual. What we are contending is that the label of
“cognitive behavior therapy” is an oxymoron since the two
terms “cognitive” and “behavior” are contradictory.

As behavior therapy developed and became successful in

terms of attracting adherents to the paradigms, the very nature
of the model began to shift and there developed a merger with
the model to which it had been a genuine alternative. In
oversimplified terminology, behavior therapy was an “outer”
(social/environmental) model of human nature as against the
then predominant “inner” (personality/biological/mental/

409
cognitive/disease) model. Behavior therapy has, to a large
extent, been co-opted by and merged into the inner model,
thus it would no longer represent the paradigm in which many
of its early adherents believed. The term “cognition” has
returned to a predominant position in psychology. We view
these developments not as another paradigm shift but rather
as a paradigm lost.

Skinner’s relevance to the behavior therapy movement was

always central as one of the founders, having coined the term
“behavior therapy” to describe early work that extended
animal laboratory findings to human patients in the
Metropolitan State Hospital in Waltham, Massachusetts
(Lindsley, Skinner, & Solomon, 1953). The title of Skinner’s
paper in the August, 1987 issue of the American Psychologist:
“Whatever Happened to Psychology as the Science of
Behavior?” expressed his disenchantment with the current
scene in behavior therapy and psychology more generally.
His chief complaint was about how the entire field had been
swept away with enthusiasm for cognitive psychology.

A curve showing the appearance of the word cognitive in the

psychological literature would be interesting. A first rise
could probably be seen around 1960; the subsequent
acceleration would be exponential. Is there any field of
psychology today in which something does not seem to be
gained by adding that charming adjective to the occasional
noun? The popularity may not be hard to explain. When we
became psychologists, we learned new ways of talking about
human behavior. If they were “behavioristic,” they were not
very much like the old ways. The old terms were taboo, and
eyebrows were raised when we used them. But when certain
developments seemed to show that the old ways might be

410
right after all, everyone could relax. Mind was back.
Cognitive psychologists like to say that “the mind is what the
brain does,” but surely the rest of the body plays a part. The
mind is what the body does. It is what the person does. In
other words, it is behavior, and that is what behaviorists have
been saying for more than half a century.

In looking back over his half century of developing the

experimental analysis of behavior, Skinner was not sanguine
about how the field of psychology had developed. In his
introduction to his paper on “Why I am not a cognitive
psychologist,” Skinner (1977) nicely summarizes the outer/
environmental model:

The variables of which human behavior is a function lie in the

environment. We distinguish between (1) the selective action
of that environment during the evolution of the species, (2) its
effect in shaping and maintaining the repertoire of behavior
which converts each member of the species into a person, and
(3) its role as the occasion upon which behavior occurs.
Cognitive psychologists study these relations between
organism and environment, but they seldom deal with them
directly. Instead they invent internal surrogates which become
the subject matter of their science (p. 1).

A more encompassing framework comes from those who

viewed behavior therapy in the broader context of social
learning (Bandura, 1969). Ullmann and Krasner (1965)
described behavior therapy as “treatment deducible from the
sociopsychological model that aims to alter a person’s
behavior directly through application of general psychological
principles.” This was contrasted with “evocative
psychotherapy” which was “treatment deducible from a

411
medical or psychoanalytic model that aims to alter a person’s
behavior indirectly by first altering intrapsychic
organizations.”

As major contributor to and founder of the behavior therapy

movement, Wolpe expressed his disenchantment in a 1986
paper in Comprehensive Psychiatry entitled
“Misrepresentation and Underemployment of Behavior
Therapy.” This was also a bitter paper in which Wolpe
reviewed current psychiatry, psychology, and psychotherapy
literature to demonstrate his contention that “Despite its
well-documented record of success in the treatment of the
neuroses, behavior therapy is little taught in departments of
psychiatry because of an inaccurate image based on
misinformation” (p. 192). He documented and illustrated with
research reports and literature reviews the claim that
“Misinformation about behavior therapy has a long history.
The earliest reports elicited a great deal of scorn from the
psychiatric establishment” (p. 192). For example, Wolpe
referred to a review of his 1958 book Psychotherapy by
Reciprocal Inhibition which had many factual errors, and he
noted that “Misreporting, often with pejorative overtones, has
been the rule ever since” (p. 192). Similar to Skinner’s lament
about the rise of cognitive ideas, Wolpe noted that:

More harmful of late have been allegations by the cognitivists

that revive in a new way the idea of behavior therapy being
simple and mechanistic. In promoting a number of
idiosyncratic cognitive techniques that they claim (without
justification) to have improved the results of behavior
therapy, they also assert that standard behavior therapy
overlooks thoughts and feelings (p. 193).

412
In effect according to Wolpe, the theoretical and practical
roots of behavior therapy were deviated from, misunderstood,
and misrepresented.

Before concluding, I will cite passages from four recent, from

the late 1990’s, articles which offer succinct summaries of the
topics I have been discussing. These are by Albert Ellis,
David Reitman, Albert Bandura, and Stewart Agras:

For the past half century, traditional behavior therapy has

done a credible job of helping clients to alleviate their
dysfunctional feelings and behaviors and to maintain this
improvement. Rational Emotive Behavior Therapy (REBT),
and some other forms of cognitive behavior therapy (CBT),
have added to behavior therapy’s record of success by
including cognitive and philosophic restructuring techniques
that aim to help some clients not only to feel better, but to
become less disturbed and less disturbable. Unlike more
traditional forms of behavior therapy, REBT and CBT often
lead to profound and more lasting attitudinal change in clients
that include anti-musturbation, unconditional self-acceptance,
unconditional other-acceptance, high frustration tolerance,
anti-awfulizing, and minimal overgeneralizing. Such lasting
attitudinal changes, and understanding the processes and
mechanisms by which they are achieved, may enhance and
extend the goals of behavior therapy into the next millennium
(Ellis, 1997).

Ellis (1997) describes how Rational Emotive Behavior

Therapy (REBT) can be distinguished from Behavior Therapy
(BT) and, perhaps more significantly, “adds” to BT. Among
the issues raised by Ellis is the suggestion that traditional BT
may lead only to temporary or superficial behavior changes,

413
and that the goals of BT should be extended to include more
comprehensive and enduring “personality change.” In this
commentary, I discuss the underlying medical model adopted
by cognitive therapists, the empirical data that bear on the
issue of “adding” to BT, and an alternative framework to
evaluate clinical practice. It is suggested that there are more
similarities among therapists, and their therapies, than
differences. Thus, as Charles Ferster (1972) suggested 25
years ago, more effort should be devoted to studying what
successful therapists do, and less to arguing the merits of
therapists’ theoretically informed explanations for success
(Reitman, 1997).

The present commentary discusses the scientific legitimacy of

theories confined to correlations of observables and those that
specify the mechanisms governing the relations between
observable events. Operant analysts frame the theoretical
differences misleadingly when the operant approach is
portrayed as addressing environmental influence for affecting
change but cognitive approaches are depicted as disembodied
from environmental influences and thus can only provide
correlates with action. In point of fact, both approaches
encompass environmental influences. The major issues in
contention are whether human thinking is entirely or only
partially shaped by environmental influences (Bandura,
1996).

If the central aim of behavior therapy is to help people

overcome clinical problems in order to improve their lives,
then the field has been a remarkable success. Although there
were a few examples of controlled outcome studies in the
psychotherapy literature when behavior therapy emerged,
there was little evidence that any form of psychotherapy was

414
efficacious (Eysenck, 1952). Behavior therapy has changed
all that by adding hundreds of controlled treatment trials to
the literature in the intervening years. The major advances in
the treatment of anxiety disorders, depression, eating
disorders, alcohol dependence, and the many applications
within the field of behavioral medicine all attest to the
remarkable success of behavior therapy. The field has brought
experiment to the clinic, has refined the methodology for such
studies, and has developed the standards for determining the
efficacy and effectiveness of psychosocial treatments.
Non-behavioral therapies must now meet the same standards,
and some of them, for example, interpersonal therapy
(Weissman & Markowitz, 1994), are beginning to do so. Such
trends bring us precariously close to a purely empirical
approach to therapeutic behavior change, but that is what is
needed if our aim is to understand the mechanisms by which
all effective psychosocial treatments work in order to
understand how best to change behavior (Agras, 1997).

Behavior Therapy in the Year 2001

We have not yet completed data collection on the third
generation of behavior therapists. In our framework, this
includes those who completed degrees in the period 1976 to
1995. What has been rather remarkable about the first two
generations has been the relative consistency of the values of
behavior therapists as compared to their non-behavior
therapist contemporaries. Behavior therapy was and remains a
movement unified by some philosophical and theoretical
foundations. In this sense, behavior therapy has not lost its
identity even though one can readily identify controversies

415
and disagreements that have occurred within the behavior
therapy movement.

One such controversy between first and second-generation

behavior therapists was about whether or not to include the
word “cognitive” as a description of what was done by
behavior therapists. Although opinions on this matter differed
sharply in the 1970’s and 1980’s, our investigations suggest
that most of what made behavior therapy unique remained in
tact. Even though the oxymoron, “cognitive behavior
therapy,” was introduced into the literature, the fear that this
would lead to the abandonment of basic scientific
assumptions did not materialize. Behavior therapists, even
“cognitive” behavior therapists, did not turn to mysticism, and
they did not denounce empiricism.

We are optimistic about behavior therapy and the future, not

only because we can foresee the continuation of the influence
of the behavior therapy movement on the mental health field,
but also because being hopeful about the future has always
been a major feature of behavior therapy as a distinct
community of behavioral scientists. The optimism of behavior
therapy was given classic expression in Skinner’s Walden
Two (1948).

References
Bandura, A. (1968). Modeling approaches to the modification
of phobic disorders. In R. Poner (Ed.), The role of learning in
psychotherapy (pp. 201-16). London: Churchill.

416
Bandura, A. (1969). Principles of behavior modification. New
York: Holt, Rinehart & Winston.

Dollard, J., & Miller, N.E. (1950). Personality and

psychotherapy. New York: McGraw Hill.

Eysenck, H. J. (1959). Learning theory and behaviour

therapy. Journal of Mental Science, 195, 61-75.

Eysenck, H. J. (1960). Behaviour therapy and the neuroses.

London: Pergamon.

Eysenck, H. J. (Ed.). (1964). Experiments in behaviour

therapy. New York: Pergamon.

Eysenck, H. J. (1969). The two faces of behaviour therapy.

Association for the Advancement of Behavioral Therapy, 4,
1-2.

Eysenck, H. J. (1969). Relapse and symptom substitution

after different types of psychotherapy. Behaviour Research
and Therapy. 7, 287-88.

Franks, C. M. (Ed.). (1969). Behavior therapy: Appraisal and

status. New York: McGraw-Hill.

Kanfer, F. H., & Phillips, J. S. (1970). Learning foundations

of behavior therapy. New York: Wiley.

Kantor, J. R. (1969). The scientific evolution of psychology

(Vol. 2). Chicago: Principia.

417
Krasner, L. (1971). Behavior therapy. In P. H. Mussen (Ed.),
Annual Review of Psychology (Vol. 22). Palo Alto,
CA:Annual Reviews.

Krasner, L., & Ullmann, L. P. (Eds.). (1965). Research in

behavior modification: New developments and implications.
New York: Holt, Rinehart & Winston.

Lazarus, A. A. (1958). New methods in psychotherapy: A

case study. South African Medical Journal, 33, 660-64.

Lindsley, O. R., Skinner, B. F. & Solomon, H. C. (1953).

Studies in behavior therapy. Status report 1. Waltham, MA:
Metropolitan State Hospital.

Mowrer, O. H. (1938). Apparatus for the study and treatment

of enuresis. American Journal of Psychology, 51, 163-66.

Pascal, G. R. (1959). Behavioral change in the clinic. A

systematic approach. New York: Grune & Stratton.

Phillips, E. L. (1968). Achievement place: Token

reinforcement procedures in a home-style rehabilitation
setting for “pre-delinquent” boys. Journal of Appliued
Behavior Analysis, 1, 213-23.

Skinner, B. F. (1938). The behavior of organisms. New York:

Appleton-Century-Crofts.

Skinner, B. F. (1948). Walden two. New York: Macmillan.

Skinner, B. F. (1953). Science and human behavior. New

York: Macmillan.

418
Skinner, B. F. (1957). Verbal behavior. New York:
Appleton-Century-Crofts.

Skinner, B. F. (1961). Cumulative record. New York:

Appleton-Century-Crofts.

Skinner, B. F. (1966). Contingencies of reinforcement in the

design of a culture. Behavioral Science, 11, 159-66.

Skinner, B. F. (1968). The technology of teaching. New York:

Appleton-Century-Crofts.

Skinner, B. F. (1969). Contingencies of reinforcement: A

theoretical analysis. New York: Appleton-Century-Crofts.

Skinner, B. F. (1997). Why I am not a cognitive psychologist.

Behaviorism, 5, 1-10.

Ullmann, L. P., & Krasner, L. (Eds.). (1965). Case studies in

behavior modification. New York: Holt, Rinehart & Winston.

Watson, R. I. (1962). The experimental tradition and clinical

psychology. In A. J. Bachrach (Ed.), Experimental
foundations of clinical psychology. New York: Basic Books.

Wolpe, J. (1958). Psychotherapy by reciprocal inhibition.

Stanford: Stanford University Press.

Wolpe, J. (1968). Psychotherapy by reciprocal inhibition.

Conditional Reflex, 3, 234-40.

Wolpe, J. (1969). The practice of behavior therapy. New

York: Pergamon.

419
Wolpe, J. (1969). Basic principles and practices of behavior
therapy of neuroses. American Journal of Psychiatry, 125,
1242-47.

Wolpe, J. (1969). How can “cognitions” influence

desensitization? Behaviour Research Therapy, 7, 219.

420
Chapter 12

The Development of
Behavioral Medicine
W. Stewart Agras

Stanford University School of Medicine

Introduction to Behavior Therapy and

the Experimental Analysis of Behavior
In 1958, Wolpe published his seminal book, Psychotherapy
by Reciprocal Inhibition, while at the Center for Advanced
Study in the Behavioral Sciences at Stanford. Having read the
book, I visited Wolpe, then Professor of Psychiatry at the
University of Virginia, to observe his treatment of phobics
with systematic desensitization. He was an enthusiastic
teacher and a kind colleague, I was after all a very junior
person, just finishing my Fellowship in psychiatry at McGill.
His influence led me to formulate a model for psychotherapy
research, with phobia as a distinct and measurable disorder,
and systematic desensitization, as a simple treatment, the
procedures of which were straightforward, well documented,
and easily learned. At the same time I read much of what
Skinner had published and could see how applicable operant
conditioning might be to mental illness.

421
All this may have come to nothing had I not moved to the
University of Vermont, to a newly formed Department of
Psychiatry headed by Tom Boag, a faculty member from
McGill. Until that time, psychiatry had been a division of the
department of medicine, rather than being a department in its
own right. Here, fortuitously, Harold Leitenberg, with a Ph.D.
in experimental psychology from the operant program at
Indiana, arrived as an assistant professor in the Department of
Psychology at the same time. I was looking for an
experimental psychologist to work with, and he was looking
for an entree to the clinic to apply operant procedures.
Together we worked with two different experimental methods
to identify effective therapeutic procedures for the treatment
of phobia. Following Lang’s work using snake phobics as an
analogue of clinical phobia (Lang & Lazovik, 1963; Lang,
Lazovik, & Reynolds, 1965), we examined the hypothesized
therapeutic ingredients in desensitization in randomized
controlled studies, eventually demonstrating that none of the
hypothesized procedures such as a hierarchy or pairing
imagined feared scenes with relaxation, appeared critical to
the outcome (Agras, Leitenberg, Barlow, Curtis, Edwards, &
Wright, 1971; Barlow, Leitenberg, Agras, & Wincze, 1969).
In a complementary approach with phobic patients, we took
an operant approach to phobias of various types, examining
the effects of reinforcement, informational feedback, and
exposure to the feared situation, in a series of single case
controlled research studies. We were fortunate in being able
to admit these patients to the University of Vermont Clinical
Research Center (CRC), where they could be studied using
single case experimental designs, in a controlled environment
free of charge. We were also fortunate to attract an excellent
group of graduate students in psychology, the first of which
was David Barlow. With this group, we examined other

422
disorders in similar ways: anorexia nervosa (Agras, Barlow,
Chapin, Abel, & Leitenberg, 1974; Leitenberg, Agras, &
Thompson, 1968), tics (Agras & Marshall, 1965), hysterical
paralysis (Agras, Leitenberg, Barlow, & Thompson, 1969),
and even delusional speech (Wincze, Leitenberg, & Agras,
1972). In each case demonstrating experimental control over
these difficult to manage behaviors.

An Example of Early Research

The single case research on phobia published in a series of
papers forms an excellent example of early work in the
application of the experimental analysis of behavior to
clinically relevant behaviors. With very few patients we were
able to discover much of importance. This work was
influenced by Lindsley’s pioneering applications of Skinner’s
findings to psychiatric patients, as well as the studies of
Ayllon and Azrin in schizophrenia (Ayllon & Michael, 1959;
Ayllon & Azrin, 1965). In our first study, the participants
were three severely agoraphobic individuals, with numerous
fears including: fear of leaving home by themselves, traveling
alone, crowds, illness, and death (Agras, Leitenberg, &
Barlow, 1968). Two of the three patients were unable to leave
their homes alone, while the third was just able to manage a
five-minute drive to work. In each case the patient’s central
difficulty, walking any distance from a “safe place” by
themselves, was assessed by distance walked alone from the
clinical research center. To assess this behavior, a course was
laid out from the CRC to downtown Burlington Vermont, a
distance of about one mile. Landmarks were identified at
25-yard intervals and the patient was asked what point they
had reached when they returned. Because much of the course

423
was observable, frequent checks of the patient’s behavior
were made throughout the study, confirming the accuracy of
their reports. Following a baseline phase in which the patient
was told to walk as far as they could alone, praise for progress
in walking further was given in a shaping schedule. As can be
seen in Figure 1, distance walked increased steadily while
being reinforced, declined dramatically when no
reinforcement was given, showing first a typical extinction
burst, and then quickly recovered when reinforcement was
reinstated, with the patient being able to walk downtown
alone, something she had not done for many years.

Having had little success with psychodynamic psychotherapy

for such patients, this degree of experimental control was
remarkable to me. The patient’s maladaptive behavior
improved, relapsed, and improved again, depending on the
reinforcement schedule. The reactions of contemporaries to
these data were mixed. The paper was published in the
Archives of General Psychiatry denoting acceptance by the
more scientifically minded psychiatric community. On the
other hand during presentations to psychiatric audiences
comments such as “this type of treatment is immoral. .
.because it is superficial and does not deal with the underlying
problems” were also common. Similar objections were raised
concerning the use of psychopharmacological agents by many
psychoanalysts. Our findings were accepted with much
interest by the then small behavior therapy community, and
particularly by the growing group of individuals engaged in
the rapidly burgeoning field of Applied Behavior Analysis.

In the next study, which involved two phobic patients, we

examined the effects of feedback of progress in single case
experiments (Leitenberg, Agras, Thompson, & Wright, 1968).

424
The first patient was a severe claustrophobic whose phobia
dated back to childhood when she was locked in a small
cupboard by some friends, one of the rare occasions when a
specific trauma is reported as leading to a phobia. The fears
became increasingly incapacitating and were much worsened
when her husband died about seven years before admission to
the CRC. Briefly the experiment consisted in giving the
patient a stopwatch and asking her to record the amount of
time she was able to remain in a small dimly lit room. This
provided her with feedback as to her progress. After 22
sessions in which she was able to increase her time in the
room, the stopwatch was removed, with the excuse that it was
broken. Under this condition her progress slowed and then
picked up again when the stopwatch was reintroduced.
Similar results were found for the second case.
Parenthetically, the close relationship to animal studies is
revealed in this paper by the detailed description of the
experimental environment, e.g. “A room 4 ft. wide and 6 ft.
long, illuminated by a 100W shaded bulb, provided a
situation in which the patients claustrophobia could be
measured.”

So far then, both positive reinforcement and feedback as to

progress had been shown to be therapeutic in cases of phobia.
However, when we examined the baseline behavior of three
agoraphobics simply given the instruction to walk as far as
you can without undue anxiety, varying responses were
found, from no response, to a modest increase in distance
walked with some relapse, to steady improvement. It
appeared that externally provided positive reinforcement and/
or feedback was not always needed. The likely suspect was
exposure to the phobic situation, which led to the next series
of single case experiments.

425
Five patients took part in these experiments, although only
one of these patients who participated in two studies will be
discussed here (Leitenberg, Agras, Edwards & Thompson,
1970). The patient was a knife phobic. Her problem had
begun some seven years before the experiment. While using a
kitchen knife she suddenly thought that she might kill one of
her grandchildren who were running around the kitchen
making a noise. This very frightening intrusive thought
recurred and during the next year her obsessive thoughts
increased so that she became unable to use a knife. She was
admitted to a psychiatric unit and treated with various
medications and systematic desensitization and was
discharged improved. Eighteen months later her husband
died, precipitating a marked depression that required further
hospitalization and treatment with electroshock therapy, again
improving. However, her fear of knives slowly returned, and
once more she was unable to handle or even look at a sharp
knife, and had been unable to cook or work in her kitchen.

In the first experiment with this patient, a sharp knife was

placed in a box. On a signal the patient opened the box and
looked at the knife until she could look no longer. A timer
recorded the duration of each trial. In the first phase of the
experiment the patient’s ability to look at the knife gradually
increased. Then exposure to the knife was stopped for several
days, a period matched to the preceding exposure treatment.
No improvement occurred during this phase of the
experiment. When the patient again began practice she again
began to improve. A second no-exposure phase showed the
same result of no progress, while the final exposure phase
again resulted in continued progress. It can be argued,
however, that the lack of progress in the no exposure phase
was due to the patient perceiving that she was not being

426
treated. Hence, in the next series of trials, psychotherapy was
alternated with exposure practice, with a positive expectancy
engendered for both treatments. Again, however, no progress
occurred during the no exposure (psychotherapy) phases,
while continued progress occurred during exposure trials.
This result was confirmed in our other work using
randomized controlled experiments with snake phobic
individuals (Barlow, Leitenberg, Agras, & Wincze, 1969;
Leitenberg, Agras, Barlow, & Oliveau, 1969; Oliveau, Agras,
Leitenberg, Moore, & Wright, 1969). It was clear that
exposure to the actual phobic situation was far superior to
imagined exposure as would occur in systematic
desensitization.

I suspect that Wolpe knew this. During my visit to him I was

struck by the fact that he insisted on his patients practicing in
the phobic situation following each desensitization session.
As time has gone on, exposure therapies have become the
standard approach to the treatment of phobia, displacing
systematic desensitization. This does not detract from
Wolpe’s seminal contribution to the field. Systematic
desensitization provided an enormous stimulus to
psychotherapy research, led to the analog psychotherapy
experiment with snake and spider phobics, which in turn
allowed for a dissection of those aspects of desensitization
that worked in reducing phobic behavior.

It may be of interest to note that the University of Vermont

CRC was located one floor above the psychiatric unit. But
these were very different milieu’s and it was difficult to
transfer the results of our research one floor down. This was
an early introduction to the enormous problem of

427
disseminating new psychotherapeutic procedures, a problem
that remains with us today.

The Underpinnings of Behavioral

Medicine
In 1969, I moved to the University of Mississippi Medical
Center as Chair of the Department of Psychiatry,
accompanied by David Barlow who became Director of
Psychology training. The Dean of the Medical School was
interested in improving the Department and giving it a new
direction and was, therefore, most supportive of a behavioral
approach to psychiatric disorders. The idea behind the move
was to integrate the new psychology and psychiatry within an
academic medical center giving equal power to both
disciplines. In line with this, Barlow used the name
Psychology Residency for the psychology internship training
program. Fortunately, others shared this vision and we were
joined by Michel Hersen, Edward Blanchard, Peter Miller,
Leonard Epstein, Tom Sajwaj, Gene Abel, and Matig
Mavissakalian (the latter two being psychiatrists) amongst
others, all of whom went on to become well known
researchers. Moreover, the psychology residency attracted
many outstanding individuals. Within this new environment it
was easier to transfer effective behavior change procedures to
the clinic. However, the intellectual basis of the Department
was the investigation of disordered human behavior based on
theories and procedures derived from psychology. Soon we
had three research rounds a week, yet were unable to present
all the ongoing research.

428
It should be noted that the introduction of behavioral science
teaching in medical schools began in 1958 at the University
of Kentucky. At the time of my move to Mississippi there
were some 17 Departments of Behavioral Science in schools
of medicine. Hence the need to introduce the findings of
modern behavioral science to medical students and residents
in various disciplines was well recognized. Yet much of this
teaching was not viewed as relevant by medical students and
other physicians in training. This was probably because
nothing was taught about the application of behavioral
science to the treatment of medical conditions. Mainly, of
course, because very little was known about such applications
at the time.

I also became more deeply involved in the Society for the

Experimental Analysis of Behavior, first as Associate Editor,
and then as Editor of the Journal of Applied Behavior
Analysis. Our research enterprise was deeply influenced by
the field of applied behavior analysis, and particularly by the
findings of researchers such as Montrose Wolfe, Ted Allyon,
Nathan Azrin, and John Paul Brady, to name just a few. Yet it
was also integrated with the broader field of behavior therapy,
and was becoming integrated into psychiatric practice.

The scientist practitioner model proved very appealing within

the medical school environment, not only within psychiatry
but also with other disciplines. For here were researchers who
could apply behavior change procedures that were known to
work and effectively care for patients with difficult problems.
It is my view that the integration of the psychologist scientist
practitioner into medical school departments where they were
treated as equals, with the opportunity to apply behavior
change procedures first to psychiatric problems, and then to

429
medical problems, was one of the important prerequisites for
the later development of behavioral medicine.

The Rise of Behavioral Medicine

What were the other prerequisites for the development of
behavioral medicine? The first, in terms of historical
sequence, was the development of Psychosomatic Medicine.
This field had begun to grow in the thirties, and the journal
Psychosomatic Medicine was begun in 1939. One of the
primary hypotheses underlying psychosomatic medicine was
that many physical ailments were caused by intrapsychic
conflict interacting with organ predisposition, presumably
stemming from either genetic or environmental causes, and
specific personality patterns, for example, dependent
personality. Understanding these patterns would lead, it was
thought, to better treatment of patients with disorders such as
asthma, peptic ulcer, and ulcerative colitis. It was regarded as
a field of much promise for medicine. Yet, in 1979 the
President of the Psychosomatic Society noted in his
Presidential address that “Psychosomatic medicine has by no
means had the influence. . .that was predicted for it 30 years
ago” (Graham, 1979). I have previously argued that the main
reason for this lack of influence on the practice of medicine
was due to the lack of intervention research (Agras, 1982).
Comparing a random sample of papers published in the
journals Psychosomatic Medicine and Behavior Therapy, an
equal number of articles (about half in each journal) were
clinically focused. Hence, psychosomatic medicine was
clearly a clinical field. However, 15% to 20% of all articles in
Behavior Therapy reported controlled intervention research
compared with only 3% of the papers published in

430
Psychosomatic Medicine during the same period.
Nonetheless, the basic psychological research stemming from
the psychosomatic medicine field, as it moved back to the
laboratory away from psychoanalytic theory, was an
important prerequisite for the development of behavioral
medicine.

A further influence was the realization that many of the risk

factors for chronic disabling disease, for example
cardiovascular disease and some cancers, were behaviors.
These behaviors included: inactivity, overeating, high fat
consumption, high alcohol consumption, cigarette smoking,
and risk taking in various forms. Moreover, procedures
deriving from behavior therapy could be used to modify some
of these behaviors and enhance self-control. At the same time
there was a growing realization that disease prevention and
health promotion were needed to delay, ameliorate, or even
prevent the onset of various diseases, to improve the nations
health and reduce the growing health care costs. These
influences, combined with pertinent biologic research, came
together in the mid-70’s, a catalytic confluence, from which
behavioral medicine emerged. There are various definitions of
the field, some focusing on intervention or prevention, others
on the etiology or risk factors for medical disorders, in which
case the interaction between environment and biologic
variables becomes the focus of interest.

In 1973, I moved to Stanford and began to consider these

issues and develop a new research program influenced not
only by my previous research background but also by the
pioneering work of Jack Farquar at Stanford who was taking a
community behavior change approach to cardiovascular
disease prevention (Farquar, et al., 1977; Leventhal, Safer,

431
Cleary, & Gutman, 1980). In 1975 the Laboratory for the
Study of Behavioral Medicine was founded at Stanford under
my direction. Simultaneously, demonstrating that the notion
of behavioral medicine was in the air at the time, Ovid
Pomerleau and Paul Brady opened a behavioral medicine
program at the University of Pennsylvania (Pomerleau, 1975;
Pomerleau, 1979). Both programs were focused on clinical
research using theories and procedures from the experimental
analysis of behavior and from behavior therapy. At this time
Bandura’s formulation of social learning theory formed the
theoretical background to our work (Bandura, 1965).
Researchers from the Stanford program became consultants to
the Coronary Primary Prevention Project, bringing their
expertise to the recruitment of the large number of
participants needed for this study, and later to the problem of
providing help to the participants in adhering to the
medication regimen in this long-term study. Hence, adherence
research became one focus of the Stanford program, followed
in the next few years by studies on the treatment of obesity,
and the use of non-pharmacological methods to reduce blood
pressure. The program produced basic psychological research
pertinent to the clinical problems being investigated,
identified behavioral risk factors for these disorders, and
carried out both preventive and intervention research. In
addition to the research focus, an outpatient clinic for the
treatment of stress disorders, obesity, and other medical
problems for which behavior change procedures seemed
justified, was opened, followed by an inpatient unit
specializing largely in eating disorders, pain problems, and
other medical disorders with accompanying psychological
problems. The close linkage between research and the clinic
was important because it allowed reciprocal feedback from

432
the two different worlds, on the one hand informing research,
and on the other informing clinical practice.

In 1977, the Yale Conference on behavioral medicine

sponsored by the National Heart Lung and Blood Institute
(NHLBI) brought together an interdisciplinary audience to
consider theoretical models, definitions for the field, the
elements constituting the field, and the possible futures of
such a field (Schwartz & Weiss, 1977). The NHLBI at the
National Institutes of Health, with the prompting of Steven
Weiss, became one of the first Institutes to provide continuing
support to the field of behavioral medicine, leading eventually
to a behavioral medicine grants review committee at the
National Institutes of Health, which in turn provided a
funding mechanism for much of the research in behavioral
medicine.

By 1978, it had become clear that a more specialized

academic society than the Association for the Advancement
of Behavior (AABT) was needed. Until that time AABT had
been home to the burgeoning field of behavioral medicine.
The Academy for Behavioral Medicine Research was
founded at a meeting at the Institute of Medicine in 1978.
However, that society was limited to senior researchers in the
field. It was clear that a society structured similarly to AABT
was needed, a society that would attract researchers and
clinicians, to keep the field moving forward. At the 1978
meeting of AABT various individuals and groups interested
in such a society were brought together, and with the help of
AABT, the Society for Behavioral Medicine (SBM) was
founded, and I was elected the first President of this
interesting new organization. In my view the key purpose of
the Society was to facilitate communication among the

433
various disciplines involved in behavioral medicine. To this
end the Abstracts of Behavioral Medicine was begun, a
journal that provided selected abstracts of pertinent studies.
The notion was that primary research papers should be
published in the scientific journal most relevant to that
research and not in a specialized behavioral medicine
publication with limited readership. This step was aimed at
better disseminating the results of research from this new
field. The Abstracts was aimed at cross-disciplinary
communication.

Both the Academy and SBM successfully pursued their

somewhat different paths, eventually forming a linkage
between the two societies, such that Fellows of SBM are
members of the Academy. Most medical schools have
divisions or groups focused on behavioral medicine research
and clinical work, and similar developments have occurred in
other countries. Apparently, the time was right for such a
development, one that appears to have been accepted by other
medical disciplines.

From Single Case to Multisite Studies

In the late 1970’s the remarkable increase in the number of
cases of eating disorders, particularly bulimia nervosa,
presenting in our clinic led me to return to an earlier interest
in the eating disorders. Previously we had used single case
experiments to isolate some of the therapeutic procedures
leading to weight gain in patients with anorexia nervosa.
Among these were positive and negative reinforcement,
informational feedback, and the serving of large meals.
(Agras et al., 1974; Leitenberg, Agras, & Thompson, 1968).

434
Each of these procedures led to increased caloric intake and
weight gain in patients with anorexia nervosa and were used
in the treatment of such patients (Agras & Werne, 1977). The
research with bulimia nervosa began somewhat differently,
eventually following the research flow described in more
detail elsewhere (Agras & Berkowitz, 1980). The first step
was to treat a series of cases of bulimia nervosa treated with a
version of cognitive-behavior therapy (CBT) reported by
Fairburn, (1981). The results reported by Fairburn were
replicated with a reasonable proportion of bulimics
recovering by the end of treatment (Schneider & Agras,
1985). The treatment was based on the hypothesis, deriving
from clinical observation, that extreme dieting caused by
severe weight and shape concerns, was the driving force
behind binge eating and purging. Later studies carried out in
our laboratory demonstrated that when bulimics were
deprived of food by not serving them breakfast and lunch, and
were then served a buffet, they would eat more at the buffet
than a non-deprived control group of bulimics. However, the
number of calories eaten over the whole laboratory day were
not statistically significantly different between groups (Telch
& Agras, 1996). This suggests that the caloric regulation of
bulimics is normal. However, because they restrict food
intake and then eat a large amount of food at one sitting, they
perceive this to be a binge. Later studies suggested that mood
was likely to alter the perception of a feeding episode. In a
negative mood bulimics were more likely to classify an eating
episode as a binge (Agras & Telch, 1998).

The primary goal of cognitive-behavioral therapy was to help

the patient attain a regular eating pattern with lessened dietary
restriction. Self-monitoring became the primary tool to
determine the ongoing eating, binge eating, and purging

435
patterns, at to guide the therapist and patient in selecting goals
for treatment, as well as providing monitoring of progress.
Weight and shape concerns and other cognitive and emotional
distortions were addressed with behavioral experiments and
cognitive therapy in the second half of treatment.

The next step in the research program was to compare the

new treatment with a psychotherapy that did not contain any
of the key procedures comprising cognitive-behavioral
therapy, but that engendered an equal expectancy of success
in the patient. Cognitive-behavioral therapy was more
effective than a placebo (non-directive) psychotherapy
(Kirkley, Schneider, Agras, & Bachman, 1985). This led to a
series of controlled outcome studies both with bulimics and
the later defined syndrome “binge eating disorder” further
documenting the effectiveness of CBT and defining some of
the critical elements comprising the treatment package
(Agras, 1995; Wilson & Fairburn, 1997). In addition, CBT
was compared to antidepressant medication, the latter having
been found more effective than placebo in a series of
controlled outcome studies (Agras et al., 1992). CBT
appeared to be more effective than antidepressant medication,
and adding medication to CBT was only marginally
beneficial.

When a research field accrues sufficient single site controlled

studies it may become necessary to conduct larger scale
studies to increase study power. Two such studies appeared
important to pursue. The background to the first study was the
finding that a form of interpersonal therapy (IPT) although
slower to achieve its effects was as effective at follow-up as
CBT (Fairburn, Jones, Peveler, Hope, & O’Connor, 1993). In
order to compare two active treatments, a large sample size

436
was necessary, leading to the first multisite study. The
background to the second study was the question as to what
treatment to offer patients who fail CBT. The two potential
candidates were antidepressant medication and IPT both of
which had proved useful in controlled trials. In order to have
sufficient subjects for such a comparison, i.e. treating only the
CBT failures (about 45% of those treated) a large sample size
was again needed, leading to the second multisite trial. The
conduct of multicenter trials is of necessity more complex
than the conduct of a single site trial. There are, however,
many advantages besides the large sample size that such trials
are able to accrue. The involvement of several investigators
leads to improvements in all aspects of the study including:
assessment procedures, therapist training, data acquisition and
analysis, as well as writing up the results of the study.

Some Lessons for the Future

What can we learn from the development of behavior therapy
and behavioral medicine? It is clear that one cannot forecast
the development of a scientific field because developments in
science interact with the changing needs of society, in turn
altering the course of the scientific field. For a clinically
oriented field to gain acceptance in medicine it must develop
effective therapeutic procedures based on well designed
studies. Unlike behavioral medicine, psychosomatic medicine
failed to do this, and gained little general acceptance as a
clinical field. Hence, the first lesson for a clinical field is that
it must develop treatments that have been shown to be
effective in controlled trials. To do this the field must attract a
cadre of researchers. Apparently, a new theoretical approach
such as behavior therapy that begins to demonstrate efficacy

437
in its therapeutic procedures will attract innovative
researchers who will move the field forward. It may often be
the case that the early findings, seminal to the development of
the field, will later be found wanting. But such studies
provoke new research which is best served if it can utilize
findings and theories from basic science. After all, Wolpe
described animal experiments in his book, but did not provide
any controlled studies with human behavior problems.

Another important lesson has been the flexibility of the field

of behavior therapy and the Association for the Advancement
of Behavior Therapy in accepting and tolerating a range of
viewpoints useful to the advancement of the field. Arguments
have been solved by the data accrued by proponents of one
viewpoint or another, rather than leading to schisms within
the field. For example, it is to the credit of many of the early
workers in applied behavior analysis that they joined AABT
rather than having a separate society, adding a great deal to
the scientific discourse. The ability to include and tolerate
different views is of extraordinary importance to the
development of a new field. Behavior therapy, and following
in its footsteps, behavioral medicine, were able to do this, and
both fields prospered because of this ability.

But there have been some relative failures along the way. The
notion of prevention has not penetrated clinical medicine as
much as might have been hoped. It is true that the prevalence
of cigarette smoking has dropped as have the rates of lung
cancer. But the prevalence of overweight and obesity has
steadily increased over the years, even among children,
forecasting a rise in the diseases comorbid with obesity. For
behavioral medicine to advance prevention efforts it is clear
that scientists must ally themselves with political activists so

438
that changes can be made at the societal level. Altering the
ways in which cigarettes can be advertised can do more than
individual physicians can do in their efforts to help someone
stop smoking.

A field with much to offer the public has to position itself

correctly. Behavior modification was not a term acceptable to
the public, with its overtones of George Orwell’s 1984. Other
words had to be used. Similarly, the remarkable popularity of
“alternative medicine,” even though many of the procedures
used have not been shown to be effective, should remind us
that what is attractive to science is not necessarily attractive to
the public. Luckily, science is being injected into alternative
medicine, and I would regard it as another frontier for
behavioral medicine.

Finally, there are often long delays in translating the findings

of applied research to the clinical domain. Behavior change
procedures are often complex and need to be sensitively
applied to each individual. Unlike new medications, for which
large budgets for disseminating the use of the drug exist,
psychosocial research has no such money. Although
workshops can be helpful in introducing a new therapeutic
procedure, more is required for therapists to attain a
reasonable level of expertise. For example, as is done in
controlled clinical trials, ongoing supervision of a number of
cases is often necessary for therapists to gain the necessary
expertise. Even more problematic is the fact that the practice
of many psychotherapists is comprised of a mix of patients
with different behavior problems, not allowing the therapist to
gain sufficient expertise through continued practice with
particular diagnostic entities. One solution to this would be to
introduce different levels of therapy for use in different

439
locales, simpler modes for the solo practitioner, more
complex modes for the specialty clinic, where a particular
kind of case would form a substantial proportion of therapists
time. An example of a simpler therapeutic mode is the use of
self-help manuals with brief therapist support. More
sophisticated self-help treatments are also available in
computerized form, and would be easily accessible via the
internet.

The future of behavioral medicine is uncertain. Enormous

structural changes are occurring in the health system and will
likely continue for some time to come. These changes will
eventually shape the direction of behavioral medicine, from
science to applications. Hopefully the changes will be rational
with a greater emphasis on healthful living from infancy
through old age.

References
Agras, W. S. (1982). Behavioral medicine in the 1980’s:
Nonrandom connections. Journal of Consulting and Clinical
Psychology, 50, 797-803.

Agras, W. S. (1995). Treatment of the obese binge eater. In

K. D. Brownell & C. G. Fairburn (Eds.), Eating disorders and
obesity: A comprehensive handbook (pp. 531-535). New
York, NY: The Guilford Press.

Agras, W. S., Barlow, D. H., Chapin, H. N., Abel, G. G., &

Leitenberg, H. (1974). Behavior modification of anorexia
nervosa. Archives of General Psychiatry, 30, 2799-286.

440
Agras, W. S., & Berkowitz, R. I. (1980). Clinical research in
behavior therapy: Halfway there? Behavior Therapy, 11,
472-487.

Agras, W. S., Leitenberg, H., & Barlow, D. H. (1968). Social

reinforcement in the modification of agoraphobia. Archives of
General Psychiatry, 19, 423-427.

Agras, W. S., Leitenberg, H., Barlow, D. H., Curtis, N.,

Edwards, J., & Wright, D. (1971). The role of relaxation in
systematic desensitization. Archives of General Psychiatry,
225, 511-514.

Agras, W. S., Leitenberg, H., Barlow, D. H., & Thompson, L.

E. (1969). Instructions and reinforcement in modification of
neurotic behavior. American Journal of Psychiatry, 125(10),
1435-1439.

Agras, W. S., & Marshall, C. (1965). The application of

negative practice to spasmodic torticollis. American Journal
of Psychiatry, 122(5), 579-582.

Agras, W. S., Rossiter, E. M., Arnow, B., Scheider, J. A.,

Telch, C. F., Raeburn, S. D., Bruce, B., Perl, M., & Koran, L.
M. (1992). Pharmacologic and cognitive-behavioral treatment
for bulimia nervosa: A controlled comparison. American
Journal of Psychiatry, 149, 82-87.

Agras, W. S., & Telch, C. F. (1998). The effects of caloric

deprivation and negative affect on binge eating in obese
binge-eating-disordered women. Behavior Therapy, 29,
491-503.

441
Agras, W. S., & Werne, J. (1997). Behavior modification in
anorexia nervosa: Research foundation. In R. Vigersky (Ed.),
Anorexia nervosa. A monograph of the National Institute of
Child Health and Development (pp. 181-195). New York,
NY: Raven Press.

Ayllon, T., & Azrin, N. H. (1965). The measurement and

reinforcement of behavior of psychotics. Journal of the
Experimental Analysis of Behavior, 8, 357-369.

Ayllon, T., & Michael, J. (1959). The psychiatric nurse as a

behavioral engineer. Journal of the Experimental Analysis of
Behavior, 2, 323-334.

Bandura, A. (1965). Principles of behavior modification. New

York, NY: Holt, Rhinehart, & Winston.

Barlow, D. H., Leitenberg, D., Agras, W. S., & Wincze, J. P.

(1969). The transfer gap in systematic desensitization: An
analogue study. Behavior Research and Therapy, 7, 191-197.

Eysenck, H. J. (1952). The effects of psychotherapy: An

evaluation. Journal of Consulting and Clinical Psychology,
16, 319-324.

Fairburn, C. G. (1981). A cognitive-behavioral approach to

the management of bulimia. Psychological Medicine, 11,
707-711.

Fairburn, C. G., Jones, R., Peveler, R. C., Hope, R. A., &

O’Connor, M. (1993). Psychotherapy and bulimia nervosa.
Longer-term effects of interpersonal psychotherapy, behavior

442
therapy, and cognitive-behavior therapy. Archives of General
Psychiatry, 50, 419-428.

Farquar, J. W., et al. (1977). Community education for

cardiovascular health. Lancet, 1, 1192-1195.

Graham, D. T. (1979). What place in medicine for

psychosomatic medicine? Psychosomatic Medicine, 41,
357-362.

Kirkley, B. G., Schneider, J. A., Agras, W. S., & Bachman, J.

A. (1985). A comparison of two group treatments for bulimia.
Journal of Consulting and Clinical Psychology, 53, 43-48.

Lang, P. J., & Lazovik, A. D. (1963). Experimental

desensitization of a phobia. Journal of Abnormal and Social
Psychology, 519-525.

Lang, P. J., Lazovik, A. D., & Reynolds, D. J. (1965).

Desensitization, suggestibility, and pseudotherapy. Journal of
Abnormal Psychology, 70, 395-402.

Leitenberg, H., Agras, W. S., Barlow, D. H., & Oliveau, D. C.

(1969). The contribution of selective positive reinforcement
and therapeutic instructions to systematic desensitization
therapy. Journal of Abnormal Psychology, 74(1), 113-118.

Leitenberg, H., Agras, W. S., Edwards, J. A., & Thompson,

L. E. (1970). Practice as a psychotherapeutic variable: An
experimental analysis within single cases. Journal of
Psychiatric Research, 7, 215-225.

443
Leitenberg, H., Agras, W. S., & Thompson, L. (1968). A
sequential analysis of the effect of selective positive
reinforcement in modifying anorexia nervosa. Behavior
Research and Therapy, 6, 211-218.

Leitenberg, H., Agras, W. S., Thompson, L. E., & Wright, D.

E. (1968). Feedback in behavior modification: An
experimental analysis in two phobic cases. Journal of Applied
Behavior Analysis, 1, 131-137.

Leventhal, H., Safer, M. A., Cleary, P. D., & Gutmann, N.

(1980). Cardiovascular risk modification by community based
programs for life-style change. Comments on the Stanford
study. Journal of Consulting and Clinical Psychology, 48,
150-158.

Oliveau, D. C., Agras, W. S., Leitenberg, H., Moore, R. C., &

Wright, D. E. (1969). Systematic desensitization,
therapeutically oriented instructions, and selective positive
reinforcement. Behavior Research and Therapy, 7(1), 27-35.

Pomerleau, O. F. (1975). Role of behavior modification in

preventive medicine. New England Journal of Medicine, 292,
1277-1282.

Pomerleau, O. F. (1979). Behavioral medicine: The

contribution of the experimental analysis of behavior to
medical care. American Psychologist, 34, 654-663.

Schneider, J. A., & Agras, W. S. (1985) A cognitive

behavioral group treatment of bulimia. British Journal of
Psychiatry, 146, 66-69.

444
Schwartz, G. E., & Weiss, S. M. (1977). Proceedings of the
Yale conference on behavioral medicine (NIH Publication
No. 78-1424). Washington, DC: U.S. Department of Health,
Education, & Welfare.

Telch, C. F., & Agras, W. S. (1996). The effects of short-term

food deprivation on caloric intake in eating disordered
subjects. Appetite, 26, 221-234.

Wilson, G. T., & Fairburn, C. G. (1998). Treatments for

eating disorders. In P. E. Nathan & J. M. Gorman (Eds.),
Guide to treatments that work (pp. 501-530). New York, NY:
Oxford University Press.

Wincze, J. P., Leitenberg, H., & Agras, W. S. (1972). The

effects of token reinforcement and feedback on the delusional
verbal behavior of chronic paranoid schizophrenics. Journal
of Applied Behavior Analysis, 5, 247-262.

445
Chapter 13

Toward a Cumulative
Science of Persons: Past,
Present, and Prospects
Walter Mischel

Columbia University

1968 Revisited: The Paradigm

Challenge
The first of these issues took shape for me slowly over the
first ten years of my career. It led me to fundamentally
question and ultimately reject the reigning paradigm of
personality and clinical psychology that defined the
mainstream at the time, and indeed since the beginnings of
the field of personality. My concerns about that paradigm
grew as I tried to find my bearings in psychology, first
teaching at the University of Colorado from 1956 to 1958,
and then in four years on the faculty of Harvard University’s
Department of Social Relations. In 1962 I moved to Stanford
University’s Psychology Department which provided an ideal
scholarly context. In that supportive intellectual setting I
could concentrate on trying to make sense of a number of
paradoxes that became increasingly disconcerting as I looked

446
more closely at the data of the field and at my own research
findings.

Early Hopes
This was a time when, beginning in the mid-1950’s, the field
of clinical psychology and personality had a giant growth
spurt, both as a profession and as a science. It was stimulated
on the one hand by demands for clinical psychologists in new
health care facilities for millions of World War II veterans,
and on the other by the opening of the Sputnik-launched
space race, which fueled federal support for science in the
United States. In this context, a new generation of
clinically-schooled psychologists was trained to dedicate
themselves to two-pronged careers as “scientist/practitioners.”
Most of us tried to straddle both roles; scientists devoted to
invalidating our favorite hypotheses in research, and
clinicians eager to apply the findings of what promised to
become the new psychological science to improve the human
condition.

A vast research literature sprouted rapidly, which tried to

demonstrate the reliability and validity of an array of
techniques that claimed to be analogous to X-rays of the
human mind — from Rorshachs and Thematic Apperception
Tests to Minnesota Multiphasic Personality Inventories and
interviews — inherited from an earlier generation of
practitioners. It was widely hoped, and often claimed, that
these methods at a minimum would yield psychometrically
sophisticated, reliable ways to capture the essentials of what
the person “is really like,” stably over time and across
situations. This mission was guided by the “classic” view of

447
human dispositions. Shared both by trait and psychodynamic
approaches, these traditional assumptions about personality
arose from a self-evidently true observation. On practically
any dimension of human behavior, there are substantial,
distinctive differences in the response of different persons
within the same social situation: Obviously, within the same
objective stimulus situation, there often are also large
differences between individuals. Second it is assumed that
individuals are characterized by stable and broadly
generalized dispositions that endure over long periods of time
and that generate consistencies in their social behavior across
a wide range of situations. With this belief, assessors tried to
predict behavior in many domains and contexts from a variety
of personality indicators or “signs” from which they inferred
these dispositions.

As an example, working as an assessment consultant at the

beginnings of the Peace Corps in Washington DC, while still
teaching at Harvard, the goal was to predict the probable
success of Peace Corps teachers in Nigeria on the basis of a
battery of measures while they were still in training. The
project, as well as the finding, was representative of the
expectations and assessment strategies of that time (Mischel,
1965). What was learned highlighted both the typical data
obtained and the concerns that grew from them. Briefly,
global ratings of the trainees made by the faculty, by the
assessment board for the project, and by an interviewer were
significantly intercorrelated. For example, the assessment
board and the interview ratings correlated .72, showing that
the assessors had similar impressions of the candidates’
personalities in training. Independently, field performance of
the Peace Corps teachers when they were on their
assignments in Nigeria was assessed on six criterion subscales

448
(which also were highly intercorrelated), and these were
aggregated into a multiple scale criterion. To illustrate the
major findings, simple self-reports and self-ratings, as on
anxiety-relevant items and attitudes to authority, yielded
modest but often statistically significant correlations with
outcome criteria, accounting for small but significant amounts
of variance. On the other hand, more global and indirect
measures failed even to reach significance. For example,
behavior in the interviews, one of the favorite global methods,
predicted total criterion performance in Nigeria with a
correlation of .13, accounting for a trivial percentage of the
variance.

A three-person subcommittee of the assessment board based

its pooled recommendations of each candidate on their
discussion and review of all data from all observer sources
during training. They thus integrated information from faculty
evaluations, academic records, peer ratings, and interviews. A
larger final assessment review board discussed these
recommendations, considering each candidate individually.
None of the separate predictions made by the training staff
correlated significantly with criterion performance, to the
dismay of those who were so confident in the power of their
procedures. Especially notable is that even when the data
were aggregated to enhance reliability, the resulting combined
evaluations of each candidate by the total assessment board
predicted aggregated performance outcomes in the field with
a nonsignificant correlation of .20.

449
1968 Challenges to Classic
Dispositional Assumptions
Findings like these were startling at a time when even small
samples of behavior had seemed to promise a diagnostic
X-ray to illuminate the core of personality — to allow rapid
inferences from a few subtle signs observed by experts to
broad generalizations about what the individual was like “on
the whole,” and then from these inferences about generalized
global dispositions to predict specific outcomes. Were such
predictive failures anomalies?

By the late 1960’s I had scrutinized and sifted through the

findings emerging from the voluminous investigations of the
preceding years. These data — which seemed consistently to
undermine and contradict the most central beliefs about the
nature of personality consistency and coherence — led me to
a thorough reexamination of the traditional global trait and
psychodynamic approaches to personality, challenging its
core assumptions (Mischel, 1968). Probably neither the
findings nor the challenge were surprising to Skinnerians,
given that Skinner had dismissed the mainstream of
personality and clinical psychology altogether thirty years
earlier. But whereas Skinner’s critique could be rejected
within personality psychology as arbitrary and from an
outsider who had not taken the data of the field seriously, my
1968 challenge traumatized the established paradigm and its
guardians.

To recapitulate the essentials, considerable evidence was

found that cognitive constructions about oneself and the
world, including other people, are often extremely stable and

450
highly resistant to change. Self-concepts, and the impressions
of other people including clinical judgments — the theories
that we have about ourselves and each other — these
phenomena and many more of the same type were found to
have consistency and even tenacious continuity (Mischel,
1968). Indeed, our constructions about other people are often
built quickly and on the basis of little information (e.g.,
Bruner, Olver, & Greenfield, 1966), and soon become
difficult to disconfirm. An impressive degree of continuity
and consistency also was found for another aspect of
cognition: namely, cognitive or information-processing styles.
These styles are often closely related to measures of
intelligence and cognitive competence and, like “intelligence”
itself, they tend to have higher consistency relative to more
social dimensions of behavior (e.g., “conscientiousness,”
“honesty,” “friendliness”).

Apart from cognitive and intellective dimensions, the domain

of social dispositions and interpersonal behavioral
consistency proved much harder to document. In fact, a
surprisingly reliable degree of behavioral specificity or
“discriminativeness” (Mischel, 1973) was found regularly in
the behavioral referents for such character traits as rigidity,
social conformity, dependency, and aggression; for attitudes
to authority; and for virtually any other nonintellective
personality. In some readings of the literature, noncognitive
personality dispositions began to seem much less global than
traditional psychodynamic and trait positions had assumed
them to be, with response patterns even in highly similar
situations typically failing to be strongly related.

Individuals show far less cross-situational consistency in their

behavior than has been assumed by trait-state theories. The

451
more dissimilar the evoking situations, the less likely they are
to produce similar or consistent responses from the same
individual. Even seemingly trivial situational differences may
reduce correlations to zero. Response consistency tends to be
greatest within the same response medium, within self-reports
to paper-and-pencil tests, for example, or within directly
observed nonverbal behavior. Intraindividual consistency is
reduced drastically when dissimilar response modes are
employed. Activities that are substantially associated with
aspects of intelligence and with problem solving
behavior—like achievement behaviors, cognitive styles,
response speed—tend to be most consistent (Mischel, 1968,
p. 177).

Evidence of this sort, in smaller amounts, had been noted for

many years (e.g., Hartshorne & May, 1928; Newcomb, 1929)
indicating instability and lack of consistency across situations
in domains of behavior expected to reflect generalized and
stable traits. In the past, however, such data were interpreted
to reflect the imperfections of tests and tools and the resulting
unreliability and errors of measurements; the fallibility of
clinical judges; and other similar methodological problems.
The new criticisms also noted that these methodological
sources constituted serious constraints, but took another step,
suggesting that the observed inconsistency so often found in
studies of noncognitive personality dimensions may reflect
the state of nature and not just the noise of measurement. Of
course, this need not imply a capriciously haphazard world,
but it did suggest a world in which personality consistencies
seem greater than they are and in which the organization of
behavior seems simpler than it is (Mischel, 1969, pp.
1014-1015).

452
The Psychodynamic-Clinical
Alternative
At the same time that global traits were challenged in their
core assumptions, I — along with many others (e.g., Peterson,
1968) — also questioned, but on different grounds, the
dispositional paradigm of psychodynamic approaches. In
contrast to the neglect of situational variables for which
classic trait approaches were being criticized, psychodynamic
approaches to personality dispositions had long recognized
both the specificity and complexity of behavior, rejecting the
idea of broad, overt behavioral consistencies across situations
at the “surface” level. They believed that the observed
inconsistencies in the individual’s overt behavior could be
understood as merely superficial diversities that masked the
fundamentally consistent, underlying dispositions and
dynamics (Mischel, 1971a, p. 153). Thus, psychodynamic
theories at an abstract level could readily deal with the facts
of inconsistencies in the person’s behavior. But they were
subject to another problem. The embarrassment for them was
in the failure to provide compelling empirical evidence that
the inferences they generated about the underlying or
genotypic dispositions were useful either for the prediction of
behavior or for its therapeutic modification, especially when
compared to simpler, less inferential, and less costly
alternatives.

Perhaps the most serious challenges to classic dispositional

approaches, both of the trait and of the psychodynamic type,
arose primarily from the clinical experiences of the 1950’s
and early 1960’s with clients seeking help. It was in that
clinical context, not in the laboratory, that many clinicians

453
came to doubt the value both of the psychodynamic and of the
trait-dispositional portraits to which they were devoting most
of their effort (e.g., Peterson, 1968; Vernon, 1964).
Skepticism about the utility of such global assessments arose
not from any lack of interest in the client’s dispositions nor
from a neglect of individual differences. Instead, it arose from
a growing anxiety that psychodynamic and trait “personality
diagnostics,” too often generated without close attention to
the clients’ own views of their lives and specific behaviors,
might be exercises in stereotyping that missed the uniqueness
of individuals and pinned the persons instead on a continuum
of clinician-supplied labels, as George Kelly (1955) had
charged years earlier.

Many empirical studies had investigated the utility of

clinicians’ efforts to infer broad dispositions indirectly from
specific symptomatic signs and to unravel disguises in order
to uncover the hypothetical dispositions that might be their
roots. The results on the whole threw doubt on the utility of
clinical judgments even when the judges were well-trained,
expert psychodynamicists, working with clients in clinical
contexts and using their own preferred techniques (Bandura,
1969; Mischel, 1968; Peterson, 1968). Clinicians guided by
concepts about underlying genotypic dispositions did not
seem better able to predict behavior than the persons’ own
direct and simple self-report, demographic variables, or in
some cases the clinicians’ secretaries (e.g., Mischel, 1968).
The disappointments of expert clinical judgment were
especially disconcerting when contrasted with evidence for
the predictions possible from indices of directly relevant past
behavior, such as an individual’s past record of
maladjustment and hospitalization (Lasky, et al., 1959). A
correlation of .61 was found, for example, between the weight

454
of a patient’s file folder and the incidence of rehospitalization.
As results like these illustrate, even a simple measure of a
person’s past can sometimes powerfully predict relevant
aspects of the future, in sharp contrast to more complex,
indirect, costly efforts. It should be equally evident that the
simple fact that one cannot predict well from some previously
favored measures and strategies denies neither the importance
of individual differences nor the potential value of all sorts of
assessments for all sorts of purposes (e.g., Mischel, 1983). It
does not mean that individual differences are necessarily
unpredictable, but it does indicate that the nature and locus of
that predictability may be quite different from what had been
assumed.

When concerns with clinical practice were combined with the

evidence from empirical studies of global traits, the challenge
in 1968 both to trait and to psychodynamic approaches to
personality became considerable. After review of the utility of
psychometrically measured traits, as well as of
psychodynamic inferences about states and traits, I was led to
the following conclusion:

Responses have not served very usefully as indirect signs of

internal predispositions. Actuarial methods of data
combination are generally better than clinical-theoretical
inferences. Base rates, direct self-reports, self-predictions, and
especially indices of relevant past behavior typically provide
the best as well as the cheapest predictions. Moreover, these
predictions hold their own against, and usually exceed, those
generated either clinically or statistically from complex
inferences about underlying traits and states. In general, the
predictive efficiency of simple, straightforward self-ratings
and measures of directly relevant past performance has not

455
been exceeded by more psychometrically sophisticated
personality tests, by combining tests into batteries, by
assigning differential weights to them, or by employing more
complex statistical analyses involving multiple-regression
equations. The conclusions for personality measures apply, on
the whole, to diverse content areas including the prediction of
college achievement, job and professional success, treatment
outcomes, rehospitalization for psychiatric patients, parole
violations for delinquent children, and so on. In light of these
findings it is not surprising that large-scale applied efforts to
predict behavior from personality inferences have been
strikingly and consistently unsuccessful. . . . (Mischel, 1968,
p. 145-146).

Finally, beyond the empirical challenge questioning the utility

of global traits was the practice of endowing such dispositions
with causal powers in theoretical explanations of behavior.
Allport (1937) had most articulately argued that behind the
confusion of trait terms, the disagreement of judges, and the
errors of empirical observation, trait terms ultimately refer to
“bona fide mental structures” (p. 289) that generate (i.e.,
produce) consistencies in behavior not only over time, but
also across situations. To the degree that traits also were
commonly used as causal entities in explanation of the
determinants of behavior, they became vulnerable to
criticisms of circularity (Mischel, 1968; Peterson, 1968;
Skinner, 1953; Vernon, 1964). If descriptions of behavior are
used to invoke traits, which in turn are offered as explanations
of the same behavior from which they were inferred in the
first place, the circularity of the reasoning becomes
embarrassing.

456
The critiques of traditional dispositional approaches that
emerged at this juncture were first read by many as
“situationist” attacks on personality itself and as unjustified
denials of the importance of individual differences. Reactions
of this type were understandable, given a long tradition of
dichotomizing the person and the situation and contrasting the
relative importance of the two sources of variance, rather than
clarifying how they interact psychologically. But while the
challenge to traditional dispositional paradigms called
attention to the significance of situations or contexts in the
study of persons, my basic message was not a negation either
of personality as a field or of individual differences as a
phenomenon. On the contrary, the focus was on the
idiographic nature of each person interacting with the specific
contexts of his or her life and on the need to revise some
favorite assumptions of traditional personality theories to take
those unique interactions into account seriously. Far from
denying individual differences in personality, the criticisms
were largely motivated to defend individuality and the
uniqueness of each person against the tendency, prevalent in
1960’s clinical and diagnostic efforts, to use a few ratings or
few behavioral signs to categorize people into categories on
an assessor’s favorite nomothetic trait dimensions. It was
common practice to assume in the 1960’s that such
assessments were useful to predict not just “average” levels of
individual differences, but a person’s specific behavior on
specific criteria as well as “in general.” It was not uncommon
to undertake decision making about a person’s life and future
on the basis of a relatively limited sampling of personological
“signs” or “trait indicators.” It was this type of practice that I
challenged:

457
Global traits and states are excessively crude, gross units to
encompass adequately the extraordinary complexity and
subtlety of the discriminations that people constantly make.
Traditional trait-state conceptions of man have depicted him
as victimized by his infantile history, as possessed by
unchanging rigid trait attributes, and as driven inexorably by
unconscious irrational forces. This conceptualization of man,
besides being philosophically unappetizing, is contradicted by
massive experimental data. The traditional trait-state
conceptualizations of personality, while often paying lip
service to man’s complexity and to the uniqueness of each
person, in fact lead to a grossly oversimplified view that
misses both the richness and the uniqueness of individual
lives (Mischel, 1968, p. 301)

In sum, the dissatisfactions that crystallized two decades ago

were wide in range, reflecting many concerns. Global
dispositional approaches were faulted as not useful for the
planning of specific individual treatment programs, for the
design of social change programs, or for the prediction of the
specific behavior of individuals in specific contexts. Perhaps
most troubling theoretically, they also were criticized for not
yielding evidence of cross-situational consistency and for
failing to provide a theoretically compelling analysis of the
basic psychological processes that underlie the individual’s
cognition, affect, and actions (e.g., Mischel, 1973).

458
Aftermath of the Challenge: Toward a
Theoretical Integration
These challenges fueled a period of prolonged and heated
controversy about personality dispositions and the construct
of personality itself that dominated thinking in the area
throughout the 1970’s and early 1980’s (e.g., reviewed in
Magnusson & Endler, 1977). The debate was multifaceted —
engaging many segments of the field, spilling into adjacent
areas, and spanning from one extreme that exaggerated the
dilemma to another that trivialized it. The claims ranged from
contentions that personality was a largely fictitious
construction in the mind of the perceiver (e.g., Shweder,
1975), to counterarguments intended to prove that global
dispositions as traditionally conceptualized were “alive and
well” if one simply employed a more reliable measurement
strategy to find them (e.g., Epstein, 1979). In the same period,
social psychologists amassed evidence for the power of
situational variables, and proposed that humans have a
persistent tendency to invoke dispositions as favorite (albeit
erroneous) explanations of social behavior (e.g., Nisbett &
Ross, 1980; Ross & Nisbett,1991). In that sense, Skinner’s
focus on the importance of the situation and of stimulus
control in the regulation of social behavior, although never
acknowledged within social psychology, was at last echoed
within it.

In the abstract and as a general framework, “person/situation

interactionism” was easily and widely given lip service and
even embraced in the 1970’s within social and personality
psychology. Indeed it was prematurely hailed as yielding a
solution to the long-standing controversy and the growing

459
confusion. The continuing challenge, however, still awaited
answers: how to reconceptualize dispositions to take such
interactions into account incisively, a priori in the form of
specific predictions, and not just in post hoc attempts to deal
with unpredicted and perhaps basically unpredictable
higher-order interactions after they are found in the data. My
efforts to address this challenge and to provide a theoretical
framework for understanding and predicting individual
differences in the interactions of persons and situations have
been spelled out in detail over the years elsewhere (e.g.,
Mischel, 1973, Mischel & Shoda, 1995, 1998).

Briefly, a set of person variables was proposed in the early

1970’s (Mischel, 1973), based on theoretical developments in
the fields of social learning and cognition that had been
bypassed or ignored by personality theory at that juncture. In
light of the complexity of the interactions between the
individual and the situation that was emphasized by the critics
of global trait assumptions, the focus in the search for person
variables shifted. This shift called attention away from
inferences about what broad traits a person has to focus
instead on what the person does in particular conditions in the
coping process. Of course, what people do encompasses not
just motor acts, but what they do cognitively and affectively,
including the constructs they generate, the projects they plan
and pursue, and the self-regulatory efforts they attempt in
light of long-term goals (to illustrate from the vast array).
Moreover, it was argued that these descriptions of what
persons do must include the specific psychological conditions
in which they do it, thus providing more condition-qualified,
“local,” contingent, and specific characterizations of persons
in contexts, in contrast to context-free traits.

460
The cognitive-social learning approach to personality
(Mischel, 1973) shifted the unit of study from global traits
inferred from behavioral signs to the person’s cognitions,
affect, and action assessed in relation to the particular
psychological conditions in which they occur. The focus thus
changed from describing situation-free people with broad trait
adjectives to analyzing the interactions between conditions
and the cognitions and behaviors of interest. In the 1960’s
much personality research on social behavior was undertaken
to study the processes of cognition and social learning
through which potential behaviors are acquired, evoked,
maintained, and modified (e.g., as reviewed in Bandura,
1969; Mischel, 1968). Less attention had been devoted to the
psychological products within the person of these processes in
the course of development. The cognitive-social learning
reconceptualization of personality was intended to identify a
set of interrelated person variables that capture these
“products” of the individual’s psychological history and that
in turn mediate the manner in which new situations are
interpreted.

The person variables that were proposed consisted of such

constructs as the person’s expectations, goals, values, and
self-regulatory competencies. Although the proposed
variables overlap and interact, each yields distinctive
information about the probable specific interaction between
the individual and any given psychological situation. Each
may be assessed objectively. Most important theoretically,
each is also amenable to study in two distinct but
complementary ways. Each may be conceptualized as a
person variable that is the product of the individual’s
social-cognitive development and on which individuals differ.
Each also may be conceptualized in terms of the

461
psychological processes relevant to understanding the
operations of that variable and its psychological meaning.
Thus, each variable has both a structural and a functional
aspect in an emerging theory of personality. More recent
versions of this approach address the specific nature of the
organization of the person variables, their interactions within
the mediating system (the Cognitive Affective Personality
System, or CAPS), and the stimulus features in the social
environment that activate them. These interactions have been
analyzed and modeled in detail, including as computer
simulations (Mischel & Shoda, 1995; Shoda & Mischel,
1998).

A Conditional Approach to
Dispositions
Historically, the failure to find strong support for
cross-situational consistency at the behavioral level, given the
widespread assumption that personality consists of traits,
expressed across many different situations as generalized,
global behavior tendencies, was read originally as a basic
threat to the construct of personality itself. As noted above,
the effect was an unfortunate and prolonged “person versus
situation” debate, and a paradigm crisis in the area.

Aggregating over situations to remove them. After years of

debate, consensus was reached about the state of the data: The
average cross-situational consistency coefficient is nonzero
but not by much (Bem, 1983; Epstein, 1983; Mischel &
Peake, 1982a, b). But there was and still is deep disagreement
about how to interpret the data and proceed in the study of

462
personality (e.g., as discussed in Mischel & Shoda, 1995,
1998; Pervin, 1994). The most widely accepted strategy used
by the classic dispositional or trait approach currently
acknowledges the low cross-situational consistency in
behavior found from situation to situation: It then
systematically removes the situation by aggregating the
individual’s behavior on a given dimension (e.g.,
“conscientiousness”) over many different situations (e.g.,
home, school, work) to estimate an overall “true score,”
treating the variability across situations as “error.”

An alternative conception of stability: Incorporating the

situation into the search for consistency. In the traditional
approach to behavioral dispositions, the observed variability
within each person on a dimension is seen as “error” and
averaged out to get the best approximation of the underlying
stable “true score,” so the question simply becomes: Is person
A different overall in the level of helpfulness than person B?
This question is important, and perhaps the best first question
to ask in the analysis of personality invariance. But it may
also be its premature end if we ignore the profile information
about when and where A and B differ in their unique pattern
with regard to the particular dimension of behavior. What if
person A’s helping behavior occurs mostly with people from
whom he can expect a “return,” while person B’s helping
behavior is correlated with the perceived level of need for
help, irrespective of the possibility of future returns? These
differences in their pattern of variability in relation to
situations may be a possible key to understanding
individuality and personality coherence and their underlying
motivations and personality systems. In that case, these
patterns are potential “signatures” of personality that need to
be identified and harnessed rather than deliberately removed.

463
Evidence for the conditional (contextualized) expression of
dispositions. Many of the results of my research program on
the structure of consistency in social behavior were based on
behavior observed intensively and extensively at a
well-controlled field laboratory site. The long-term, intensive
observational field laboratory developed in this research
program was located within a 6-week summer residential
camp setting and treatment program for troubled children,
called Wediko (e.g., Mischel, 1990; Shoda, et al., 1993a,
1993b, 1994). This setting provided an exceptional
opportunity to examine behavior in vivo as it unfolds across
situations and over time under unusually well-controlled
research conditions that assured the reliability and density of
measurement.

We found a type of stability that was seemingly contradictory

to earlier assumptions about the consistency and structure of
dispositions and that was systematically bypassed, rather than
harnessed. Consistent with — and predicted by — the
Cognitive Affective Personality System (CAPS) model
developed with Yuichi Shoda (Mischel & Shoda, 1995, 1998;
Shoda & Mischel, 1998), individuals may be characterized by
distinctive and stable patterns of variability in their prototypic
behavior in relation to different significant psychological
situations, as will be described below. Thus, although
findings showing the variability of behavior and the
importance of the situation in the past have been interpreted
as evidence against the utility of the personality construct, our
work has shown that this need not be the case (Mischel &
Shoda, in press; Shoda, et al., 1994). On the contrary, at least
some of this variability is intra-individually stable and
meaningful: Indeed it seems to reflect some of the essence of
personality coherence and to provide a potential window into

464
the underlying social cognitive and affective processes and
system that generate it (e.g., the individual’s construals, goals,
and motivations).

Distinguishing nominal and psychological situations: Finding

the “active ingredients” (psychological features) within
situations. To test for the existence and meaningfulness of the
hypothesized, stable if/then situation-behavior relations,
everyday social behavior was observed as it unfolded over the
course of each summer in the Wediko residential camp
setting, yielding an exceptionally large and comprehensive set
of systematic observations for each participant (Shoda, et al.,
1989, 1993a, b, 1994; Shoda, 1990). The first requirement in
the Wediko field study was to identify the situations in which
the behavior occurred (Shoda, et al., 1994). But which
situations? In studies of the consistency of behavior across
situations, the situations usually have been defined in nominal
terms, as places and activities in the setting, for example, as
woodworking activities, arithmetic tests, dining halls, or
school playgrounds (e.g., Hartshorne & May, 1928;
Newcomb, 1929). Individual differences in relation to such
specific nominal situations, even if highly stable, necessarily
would be of limited generalizability. On the other hand, if
situations are redefined to capture their basic psychological
features then information about a person’s behavior
tendencies specific to those situation features (Kelly, 1955;
Mischel, 1973) might be used to predict behavior across a
broad range of contexts that contain the same psychological
features (Shoda, et al., 1993a). For example, situations that
include criticism or lack of attention from a partner might be
those in which individuals hypersensitive to rejection in
intimate relations become consistently more upset than others
(e.g., Downey & Feldman, 1996; Mischel & Shoda, 1995).

465
The key for achieving generalizability, therefore, is to identify
psychological features of situations that (1) play a functional
role in the generation of behaviors, and (2) are contained in a
wide range of nominal situations.

It was thus important to identify the relevant psychological

features — the “active ingredients” (if we make the analogy
with the potent aspects of a chemical substance) as opposed to
the inactive, filler elements that exert a significant impact on
the behavior of the person. These active ingredients may
occur within many different nominal situations (Shoda, et al.,
1994). Individual differences in response to nominal
situations, such as the daily activities within a camp, then may
be understood in terms of the person’s stable responses to the
encoded active psychological features within the nominal
situations. These psychological features, in turn, may consist
of combinations of even more specific features and may be
analyzed in terms of their overlap and similarity.

As predicted, we found that individual differences in behavior

were relatively inconsistent across different types of
psychological situations. However, they were significantly
more consistent across the same types of psychological
situations, i.e., those that contain the same active ingredients
even when they were embedded within different nominal
situations (Shoda, et al., 1994). Further, cross-situational
consistency increased substantially in the individual’s
behavior as the number of shared features increased. In short,
the situation specificity of individual differences in behavior
found repeatedly over the years has often been seen as
limiting the generalizability and utility of personality
descriptions, prompting repeated paradigm crises in the field.
Our results suggest a very different possibility: If the

466
distinctive and stable if/then contingencies for the individual
are defined in relation to the basic psychological ingredients
or features that occur in many different nominal situations,
then it may be possible to understand and predict behaviors in
a novel situation given that its psychological ingredients are
known. Moreover, to the degree that particular sets of such
active ingredients or psychological features for an individual
(or for a personality type) are imbedded in diverse nominal
situations (e.g., at woodworking in camp, on playground at
school, at mealtime at home), it may become possible to
predict behavior across those seemingly different situations
and contexts, allowing much broader predictability even for
quite specific behavioral manifestations (Mischel & Shoda,
1995, 1998; Shoda, et al., 1994).

Resolving the Consistency Paradox

It was in the 1970’s that the so-called “consistency paradox”
(e.g., Bem & Allen, 1974) was articulated, and it proposed a
possible resolution to the paradigm crisis produced by the
evidence for the lack of cross-situational consistency within
the individual’s behavior. This now classic paradox refers to
the notion that while intuition seems to support the belief that
people are characterized by broad dispositions resulting in
extensive cross-situational consistency, the research in the
area (as discussed above) has persistently failed to support
this intuition. To resolve this paradox, Daryl Bem predicted
that behavioral cross-situational consistency would be
demonstrable at the level of specific situations, but only for
that subset of people who view themselves as consistent on
the particular dimension and for whom that dimension is thus
personally relevant (Bem & Allen, 1974).

467
To test this proposition, my studies of the consistency
paradox (Mischel & Peake, 1982, 1983) were conducted with
a sample of Carleton College undergraduates who
volunteered to participate in extensive self-assessments
relevant to their conscientiousness and friendliness. Directly
contradicting Bem’s prediction, we found that the students’
perceptions of their own overall consistency or variability on
conscientiousness were not related closely to the observed
cross-situational consistency of their actual behavior directly
observed as it occurred in vivo across diverse daily situations.
Although inter-judge agreement was greater for those
students who saw themselves overall as consistent in
conscientiousness, their average behavioral consistency
across the measures was not significantly greater than that of
students who saw themselves as variable. In short, in this
study, people who saw themselves as consistent on a
dimension were seen with greater interjudge agreement by
others, but their overall behavior was not necessarily more
consistent cross-situationally.

If, as the Carleton data suggest, the self-perception of

consistency is unrelated to the level of cross-situational
consistency in the referent behaviors, in what is it rooted? To
try to answer this question, we proposed that consistency
judgments rely heavily on the observation of central
(prototypic) features, so that the impression of consistency
will derive not from average levels of consistency across all
the possible features of the category, but rather from the
observation that some central features are reliably (stably)
present. This perspective suggests that extensive
cross-situational consistency may not be a basic ingredient for
either the organization or the perception of personal
consistency in a domain.

468
We hypothesized that the impression that a person is
consistent with regard to a trait is not based mostly on the
observation of average cross-situational consistency in all the
potentially relevant behaviors (e.g., punctuality for classes,
punctuality for appointments, desk neatness, etc.). Instead, we
proposed that when people try to assess their variability (vs.
consistency) with regard to a category of behavior, they scan
the temporal stability of a limited number of behaviors that
for them are most relevant (prototypic) to that category. That
is, it was hypothesized that the impression of consistency is
based extensively on the observation of temporal stability in
those behaviors that are most relevant to the prototype. No
relationship was expected between the impression of high
consistency versus variability and overall cross-situational
consistency. The results supported these expectations
(Mischel & Peake, 1982). Those students who saw
themselves as highly consistent in conscientiousness were
significantly more temporally stable on these prototypic
behaviors than were those who viewed themselves as more
variable from situation to situation — an effect that was
replicated in the domain of friendliness by Peake within the
same sample of students (Peake, 1982). In contrast to the
clear and consistent differences in temporal stability for the
prototypic behaviors, the self-perceived low and high
variability groups did not differ in mean temporal stability for
the less prototypic behaviors. Finally, also as expected,
self-perceived consistency and behavioral cross-situational
consistency were unrelated in Mischel and Peake’s (1982)
study.

These findings suggested that peoples’ intuitions of their

cross-situational consistency are not illusory: they are based
on data, but these data are not highly generalized

469
cross-situational consistencies in their behaviors on the
whole. Intuitions about one’s consistency seem to arise,
instead, from the observation of temporal stability in
prototypical behaviors. This would certainly not be a
fictitious construction of consistency. The “error” simply
would be to confuse the temporal stability of key behaviors
with pervasive cross-situational consistency, and then to
overestimate the latter.

Condition-Behavior Stabilities: In
Search of Local Predictability
The classic trait strategy essentially treats situations as if they
were error, and seeks to cancel their effects by aggregating
across them to eliminate their role and to demonstrate stable
individual differences. The approach my associates and I
favor, in contrast to the classic route, seeks consistencies by
linking the behavior of interest to a circumscribed set of
contexts, thus pursuing consistency on a more local,
condition-bound, contingent, and specific (rather than global)
level. Both approaches seem to accept the fact that average
levels of consistency in behavior from situation to situation
tend to be modest, even after aggregation over multiple
occasions. Advocates of the traditional trait strategy propose
circumventing this constraint by abandoning attempts to
predict behavior from situation to situation altogether.
Instead, they confine their predictive efforts to aggregates
over multiple situations (e.g. Epstein, 1979). Such a strategy
can enhance the resulting coefficients dramatically (as the
Spearman-Brown formula has long recognized). But it
bypasses rather than resolves the classic problems found in

470
the search for coherences from situation to situation by
“averaging out” the situation rather than predicting behavior
in it. And, of course, it places a low ceiling on the accuracy
possible for predicting behavior in specific situations.

We have continued to explore the view that personality

coherences involve prototypic features of behavior that are
cross-situationally discriminative but meaningful, temporally
stable dispositional indicators when they occur in certain
diagnostic contexts. Instead of pursuing high levels of overall
consistency from situation to situation for many aggregated
behaviors in a wide range of aggregated contexts, the goal is
to identify the distinctive “bundles” or sets of temporally
stable prototypic condition-behavior relations that
characterize the individual under predictable circumstances
(e.g., Mischel, 1973; Wright & Mischel, 1987). Although
these if/then condition-behavior relations may occur only
some of the time, they may figure crucially both in the
perception of personality and in its organization (e.g., Mischel
& Peake, 1982; Wright & Mischel, 1987, 1988).

In this vein, we developed a conditional approach to

dispositions in which dispositional constructs are viewed as
clusters of if/then propositions. Rather than construing
dispositions as generalized response tendencies aggregated
over diverse situations, we view them as propositions
summarizing contingencies between categories of conditions
and categories of behavior. A basic unit in the analysis of
dispositions, then, becomes the conditional frequency of acts
that are central to a particular behavior category in
circumscribed, “diagnostic” conditions (Wright & Mischel,
1987, 1988). This type of if/then proposition contrasts with
the traditional focus on the overall frequency of

471
dispositionally relevant behaviors aggregated across a wide
range of situations. It calls explicit attention to inextricable
specific links between conditions and actions in determining
the implications of people’s behavior for dispositional
judgments about them (e.g., Mischel & Shoda, 1995, 1998,
1999; Shoda, Mischel, & Wright, 1989). It equally highlights
the interactive nature of the person-situation relations that
characterize social behavior.

To summarize, although context sensitivity and

discriminativeness across situations may be the rule rather
than the exception for most social behavior, it is also possible
to find specific coherences that differentiate individuals and
that can be identified under predictable contingencies for at
least some people and behaviors, suggesting “local” areas of
relative predictability without resorting to aggregation across
situations. Such local coherences seem to occur at least in
situations requiring cognitive and self-regulatory
competencies that make high demands and strain people’s
available competence, at least with regard to some categories
of disadvantageous behavior (aggression, withdrawal).
Moreover, we have found significant links between ratings on
these dimensions and individuals’ actual behavior in difficult
situations at relatively molecular episodic levels of
observation, even without benefit of aggregating across
different behavioral features.

472
Beyond Stimulus Control: Why I
Became a Cognitive Social
Psychologist
The fact that the situation plays an enormously powerful role
in the often automatic activation and regulation of complex
human social behavior was of course central to my critique of
traditional approaches to clinical and personality psychology.
I saw the power of the situation convincingly in my own work
on the willingness and ability of young children to delay
gratification by continuing to wait for two little treats later as
opposed to settling for one right now. We found that such a
seemingly trivial change in the situation as whether the
rewards remain exposed on the plate facing the preschool
child or are placed under it can change the average delay time
from less than a minute to more than ten (e.g., Mischel, Shoda
& Rodriguez, 1989). Thus whether or not the young child
finds delay of gratification excruciatingly difficult or easily
achievable hinges on the subtleties of the situation, and when
these IFs are properly understood and introduced the THENs
that follow can become highly predictable.

Overcoming Stimulus Control

Through Self-Regulation
The same set of studies also generated a second set of
findings, however, that led me to join the cognitive revolution
but without abandoning my focus on behavior and its
determinants (e.g., Mischel, 1973). Most compelling for me
was the finding that regardless of the objective stimulus

473
facing the subject, it was its mental representation, as primed
by suggestions on how to think about the rewards, that
controlled the delay of gratification behavior, regardless of
the actual rewards present in the situation. Namely, when
these representations focused on the “hot” consummatory
features of the stimulus (e.g., “while you’re waiting you can
think about the pretzel’s salty, crunchy taste”) the frustration
of continued delay of gratification became unbearable for
most children even when the external stimulus facing the
subject was completely controlled. Conversely, when the
mental representation focused on the “cool,” informative cue
properties (e.g., “you can think about the pretzels as if they
were little sticks”) , sustained, goal-directed delay of
gratification and “willpower” became manageable, again
regardless of the external stimulus in the situation (Mischel,
1974, 1996; Mischel, Shoda & Rodriguez, 1989). Thus the
way the child represented the stimuli cognitively during the
delay period profoundly transformed their impact: the power
thus resides in the head, not in the external stimulus.

Likewise, even in the presence of the rewards, which makes

delay difficult for children, delay was easily sustained when
the children self-distracted (e.g., by imagining that they were
playing with a toy, or by thinking about pleasant activities
such as swinging on a swing at the birthday party). If the
crucial process is in the head, individuals should be able to
influence it through self-generated strategies to influence the
mental representations of the reward objects, just as it is
influenced by the strategies that the situation suggests.

In fact, in the same research program we found impressively

stable individual differences in the ability to overcome
stimulus control pressures in the purposeful pursuit of

474
long-term goals (e.g. Mischel, Shoda, & Peake, 1988; Shoda,
Mischel, & Peake, 1990). For example, in laboratory
situations in which individual differences in such strategies
were activated, those 4-year-old children who delayed longer
became more socially and cognitively competent young
adults, also achieving higher levels of scholastic performance,
as reflected in their SAT verbal and quantitative scores (e.g.,
Shoda, Mischel, & Peake, 1990). In short, behavior in the
delay situation was a function not only of the characteristics
of the situation but also of the individuals in it. The fact that
these distinctively human efforts also require the strategic
utilization and support of situations undermines neither the
role of the person nor of the situation, but requires attention to
their reciprocal interaction. These interactions have been
conceptualized more recently in a theoretical two-system
“hot/cool” framework that takes account both of the
automatic, stimulus-response aspects of functioning and of
the more reflective, cognitive mediating system, focusing on
the specifics of their interplay (Metcalfe & Mischel, 1999).

The findings from this research also make the common

distinction between the power of the situation and of the
person fuzzy. For example, does the fact that attention is
focused away from the rewards in the delay situation (e.g., by
covering the rewards or avoiding them cognitively)
demonstrate the power of situational variables in self-control?
It does in the sense that they show how specific changes in
the situation can make delay either very difficult or very easy.
But the same results also show that even young children can
and do increase their own mastery and personal ability to
control the effects of stimuli on them by modifying how they
think about those stimuli, by “reframing” them cognitively, or

475
by distracting themselves and focusing on other aspects of the
situation while continuing in their goal-directed behavior.

Empirically, it by now seems undeniable: in everyday life, as

well as in the laboratory, people can and do modify and
transform the power and impact of the stimuli that they
encounter and create, persisting in pursuit of long-term
difficult goals even in the face of potent barriers and
temptations along the route (e.g., Mischel, Cantor & Feldman,
1996). These phenomena (as when the habitual smoker gives
up tobacco, and the difficult new year’s resolution to exercise
is actually executed) may be the rare events, but it is their
importance for being human, not their frequency, that is at
issue, and surely it is part of what psychologists must explain
in a comprehensive account of what is significant in everyday
life.

In sum, the prevalence and significance of automaticity of if/

then links in everyday life and of the power of the situation is
not diminished (it may be enhanced) by the concurrent
recognition of the field’s other major conclusion: regardless
of its frequency, humans do engage at least some of the time
and under some circumstances in self-regulatory behavior in
pursuit of their long-term goals and values. In these moments
they manage to purposely modify, transform, and even
overcome the power of the immediate stimulus, interjecting
their own personal agendas between the external IF and the
observable external THEN, in ways that reveal their
distinctive personality signatures.

476
References
Allport, G. W. (1937). Personality: A psychological
interpretation. New York: Holt, Rinehart, & Winston.

Bandura, A. (1969). Principles of behavior modification. New

York: Holt, Rinehart, & Winston.

Bem, D. J. (1983). Further Déjà vu in the search for

cross-situational consistencies in behavior: A response to
Mischel and Peake. Psychological Review, 90, 390-393.

Bem, D. J., & Allen, A. (1974). On predicting some of the

people some of the time: the search for cross-situational
consistencies in behavior. Psychological Review, 81, 506-520.

Bruner, J. S., Olver, R., & Greenfield, P. (1966). Studies in

cognitive growth. New York: Wiley.

Downey, G., & Feldman, S. (1996). Implications of rejection

sensitivity for intimate relationships. Journal of Personality
and Social Psychology, 70, 1327-1343.

Epstein, S. (1979). The stability of behavior: On predicting

most of the people much of the time. Journal of Personality
and Social Psychology, 37, 1097-1126.

Epstein, S. (1983). Aggregation and beyond: Some basic

issues on the prediction of behavior. Journal of Personality,
51, 360-392.

477
Hartshorne, H., & May, A. (1928). Studies in the nature of
character: Studies in deceit (Vol. 1). New York: MacMillan.

Kelly, G. A. (1955). The psychology of personal constructs

(Vols. 1-2). New York: Norton.

Lasky, J. J., Hover, G. L., Smith, P. A., Bostian, D. W.,

Duffendack, S. C., & Nord, C. L. (1959). Post-hospital
adjustment as predicted by psychiatric patients and their staff.
Journal of Consulting Psychology, 23, 213-218.

Magnusson, D., & Endler, N. S. (Eds.). (1977). Personality at

the crossroads: Current issues in interactional psychology.
Hillsdale, NJ: Erlbaum.

Metcalfe, J., & Mischel, W. (1999). A hot/cool system

analysis of delay of gratification: Dynamics of willpower.
Psychological Review, 106, 3-19.

Mischel, W. (1965). Predicting the success of Peace Corps

volunteers in Nigeria. Journal of Personality and Social
Psychology, 1, 510-517.

Mischel, W. (1968). Personality and assessment. New York:

Wiley.

Mischel, W. (1973). Toward a cognitive social learning

reconceptualization of personality. Psychological Review, 80,
252-283.

Mischel, W. (1974). Cognitive appraisals and transformations

in self-control. In B. Weiner (Ed.), Cognitive views of human
motivation (pp. 33-49). New York: Academic Press.

478
Mischel, W., Shoda, Y., & Peake, P. K. (1988). The nature of
adolescent competencies predicted by preschool delay of
gratification. Journal of Personality and Social Psychology,
54, 687-699.

Mischel, W. (1969). Continuity and change in personality.

American Psychologist, 24, 1012-1018.

Mischel, W. (1971). Introduction to personality. New York:

Holt, Rinehart, & Winston.

Mischel, W. (1983). Alternatives in the pursuit of the

predictability and consistency of persons: Stable data that
yield unstable interpretations. Journal of Personality [special
issue on prediction], 51, 578-604.

Mischel, W. (1990). Personality dispositions revisited and

revised: A view after three decades. In. L. Pervin (Ed.),
Handbook of personality psychology (pp. 111-134). New
York: The Guilford Press.

Mischel, W. (1996). From good intentions to willpower. In P.

M. Gollwitzer & J. A. Bargh (Eds.), The psychology of
action: Linking cognition and motivation to behavior (pp.
197-218). New York: Guilford Press.

Mischel, W., Cantor, N., & Feldman, S. (1996). Principles of

self-regulation: The nature of willpower and self-control. In
E. T. Higgins and A. W. Kruglanski (Eds.), Social
psychology: Handbook of basic principles (pp. 329-360).
New York: Guilford.

479
Mischel, W., & Peake, P. K. (1982). Beyond Déjà vu in the
search for cross-situational consistency. Psychological
Review, 89, 730-755.

Mischel, W., & Peake, P. (1982). In search of consistency:

Measure for measure. In M. P. Zanna, E. T. Higgins, & C. P.
Herman (Eds.), Consistency in social behavior: The Ontario
Symposium (Vol. 2, pp. 187-207). Hillsdale, NJ: Erlbaum.

Mischel, W., & Peake, P. K. (1983). Analyzing the

construction of consistency in personality. In M. M. Page
(Ed.), Nebraska Symposium on Motivation, 1982:
Personality-Current theory and research (Vol. 30, pp.
233-262). Lincoln: University of Nebraska Press.

Mischel, W., & Shoda, Y. (1995). A cognitive-affective

system theory of personality: Reconceptualizing situations,
dispositions, dynamics, and invariance in personality
structure. Psychological Review, 102(2), 246-268.

Mischel, W., & Shoda, Y. (1998). Reconciling processing

dynamics and personality dispositions. Annual Review of
Psychology, 49, 229-258.

Mischel, W., & Shoda, Y. (1999). Integrating dispositions and

processing dynamics within a unified theory of personality:
The cognitive affective personality system (CAPS). In L.
Pervin & O. John (Eds.), Handbook of personality: Theory
and research (pp. 197-218). New York: Guilford Press.

Mischel, W., Shoda, Y., & Rodriguez, M. L. (1989). Delay of

gratification in children. Science, 244, 933-938.

480
Newcomb, T. N. (1929). Consistency of certain
extrovert-introvert behavior patterns in 51 problem boys.
New York: Columbia University, Teachers College, Bureau
of Publications.

Nisbett, R. E., & Ross, L. D. (1980). Human inference:

Strategies and shortcomings of social judgment. Englewood
Cliffs, NJ: Prentice Hall

Peake, P. K. (1982). Searching for consistency: The Carleton

Student Behavior Study. Doctoral Dissertation, Stanford
University. Dissertation Abstracts, 43, Pt. B, Section 8, p.
2746. (University Microfilms No. AAD 83-01259).

Peterson, D. R. (1968). The clinical study of social behavior.

New York: Appleton-Century-Crofts.

Pervin, L. A. (1994). A critical analysis of current trait theory.

Psychological Inquiry, 5, 103-113.

Ross, L., & Nisbett, R. E. (1991). The person and the

situation: Perspectives of social psychology. New York:
McGraw-Hill.

Shoda, Y., Mischel, W., & Peake, P. K. (1990). Predicting

adolescent cognitive and self-regulatory competencies from
preschool delay of gratification: Identifying diagnostic
conditions. Developmental Psychology, 26, 978-986.

Shoda, Y., Mischel, W., & Wright, J. C. (1989). Intuitive

interactionism in person perception: Effects of
situation-behavior relations on dispositional judgments.
Journal of Personality and Social Psychology, 56, 41-59.

481
Shoda, Y., Mischel, W., & Wright, J. C. (1993). The role of
situational demands and cognitive competencies in behavior
organization and personality coherence. Journal of
Personality and Social Psychology, 65, 1023-1035.

Shoda, Y., Mischel, W., & Wright, J. C. (1993). Links

between personality judgments and contextualized behavior
patterns: Situation-behavior profiles of personality prototypes.
Social Cognition, 11, 399-429.

Shoda, Y., Mischel, W., & Wright, J. C. (1994).

Intraindividual stability in the organization and patterning of
behavior: Incorporating psychological situations into the
idiographic analysis of personality. Journal of Personality
and Social Psychology, 67, 674-687.

Shweder, R. A. (1975). How relevant is an individual

difference theory of personality? Journal of Personality, 43,
455-485.

Skinner, B. F. (1953). Science and human behavior. New

York: Macmillan.

Vernon, P. E. (1964). Personality Assessment: A critical

survey. New York: Wiley.

Wright, J. C., & Mischel, W. (1987). A conditional approach

to dispositional constructs: The local predictability of social
behavior. Journal of Personality and Social Psychology, 53,
1159-1177.

482
Wright, J. C., & Mischel, W. (1988). Conditional hedges and
the intuitive psychology of traits. Journal of Personality and
Social Psychology, 55, 454-469.

483
Chapter 14

A Small Matter of Proof

Donald M. Baer

University of Kansas

Intellectual Biography
My parents were immigrants who had to work rather than
finish high-school educations. Appraising their new country,
they saw two ways up for their children: business and
education. They sent me to school with very clear
instructions: It was the most important part of my life; I
should learn everything that was taught; and I should cause no
trouble. I was a compliant child; it would be a long time
before I disobeyed any of those instructions.

Massachusetts and Pennsylvania public schools made me

literate, grammatical, and able to calculate. They also taught
me how a few parts of the world worked. They called that
Science. I loved knowing how any part of the world worked; I
do not know why. Kantor might have labeled it maximizing
an ecological reinforcer (Kantor, 1924/1926), but a label is
not an explanation.

Science was not taught in the public schools then as if it were

interesting, glamorous, all-explanatory, or crucial to the future
of our society. Instead, it was taught as if this was what we

484
could prove about how some part of the world worked. I
found that something called “understanding” or “explanation”
was already one of my most potent reinforcers, and I was
taught implicitly that it always came wrapped in experimental
proof.

When I studied at the University of Chicago, my teachers

promptly displayed the diversity of activities people call
Science. I found that proof was supremely important only in
what they called Natural Science. Other paradigms of Science
depended much more heavily on something called theory.

Theory came in at least two kinds. One kind guessed that

what had been proven true in many particular cases, almost
without exception, would remain true in all or most future
cases. For example, if the three-term contingency had proven
demonstrably analytic for a long list of behaviors and stimuli,
why not assume, as an act of theory, that it always would?
Especially if subsequent experience required us only to
change the three-term contingency to the N-term
contingency? This, I was taught, was a legitimate but minor
use of theory, and quite typical of Natural Science.

The second kind of theory was a way of seeming to

understand, explain, or predict what you could not prove. For
example, if people behaved in a remarkable variety of ways,
why not invent an Id to motivate all those ways, an Ego to
learn how to let the real world regulate the maximal overall
satisfaction of the Id, and a Superego to explain exceptions to
apparent maximal satisfaction? This, I was taught, was a
legitimate and major form of theory. It was the kind of theory
that was prized, honored, battled over, disseminated, and
turned into show business and profit. In its best forms, its

485
major terms were verbal inventions; by definition, they were
immune to direct proof or disproof. That put proof in its
properly small place, which was to confirm or disconfirm any
observable deductions from the inventions. Confirmations
meant the inventions were good; disconfirmations meant the
inventions were even better, because disconfirmation was
much too harsh a term for this event. It really was a rich
encounter, one that allowed the theorist to “discover” certain
details in the inventions not appreciated before. Given an
apparent disconfirmation, theorists did not drop their
disconfirmed theory and build a better one; instead, they
“discovered” that there was more to the original theory than
they had appreciated before. Explaining away
disconfirmations let them claim an even greater
“understanding” than before.

The only abstainers to that kind of theory were people who

defined understanding as something both explanatory and
proven. For them, their major reinforcer, understanding,
required the conditioned establishing operation (Michael,
1995), proof.

I thanked the University of Chicago for showing me all the

wonderful ways people could behave and call it Science.
Again, but this time by choice, I endorsed Natural Science.
My best teachers then, Leo Nedelsky, Jacob Gewirtz, and
Howard Hunt, affirmed that choice, as would my next best
teacher, Sidney Bijou. I was content to do the minor theory
that characterized Natural Science, and to use inference only
with foreboding. The dictionary definition of foreboding is an
apprehension of coming misfortune.

486
An epiphany led me to behaviorism. I was studying
mathematics and physics in graduate school, and not liking
them. One day a friend in psychology asked me to explain an
equation in one of his textbooks. The equation was Weber’s
Law. I was astounded that psychology could use differential
calculus. My undergraduate brush with psychology had
offered it as psychodynamic, and the media I had consumed
equated it to the rather bizarre forms of insight that allowed a
happy ending.

I borrowed my friend’s textbook, which was on experimental

psychology. It presented a natural science of behavior: It
offered an experimental proof of almost everything it
asserted, and labeled any unproven assertions as possibly
wrong. I could not imagine anything more fascinating. I
stayed up all night to read it through. With the perfect
symbolism of the dawn, I decided to transfer to psychology
and study the natural science of behavior, especially human
behavior.

Of everything psychology subsequently showed me, the

approach that most depended on natural science was that form
of behaviorism called operant conditioning then, and behavior
analysis later. Of all I had seen so far, operant psychology and
psychophysics were the only ones constantly wedded to
proof, and the subsequent behavior analysis had ambition —
it aimed for scope. It tried for scope by proceeding from what
it had proven to what it had not yet proven, but might. The
rest of psychology did the reverse, as if the greatest honor was
to arrive at something unprovable.

So to speak, I had not been raised to be a behaviorist; I had

been raised to be a proof consumer. Behavior analysis

487
apparently valued proof more than any other approach, so I
joined.

A related intellectual epiphany occurred much later. In a

faculty bar after a colloquium, a professor of psychology told
me, with the intensity that only a very drunk academic can
muster, that Hullian theory had never been disproven. I
realized immediately that he was correct: it had not been
disproven, because it could not be disproven. It had merely
extinguished, and had left behind some betrayed subscribers.
It was not their theory that had betrayed them. When they
were embarrassed momentarily by some fact, they could
always “discover” another fractional antedating goal response
in the theory, and so explain away the fact. The betrayal had
been by a shallow Science audience that simply could not
commit to an ever more complex relationship. I began then to
suspect that major theorists would ultimately want not
research but marital therapy.

A Case Study of Two Articles by Baer,

Wolf, and Risley
In 1968, Montrose Wolf, Todd Risley, and I published an
article in the new Journal of Applied Behavior Analysis. I had
the privilege of writing the first draft, and the pleasure of
discovering that my coauthors largely agreed with it. The
article was entitled “Some current dimensions of applied
behavior analysis.” It proposed and elaborated seven
generative dimensions of this new discipline: The discipline
should be applied, behavioral, analytic, technological,

488
conceptual, effective, and capable of producing appropriately
generalized behavior changes.

In 1987, in response to an invitation by Jon Bailey, then

editor of that journal, we published another version of it.
Bailey’s argument was that 20 years of relevant experience
might or might not change some of its arguments, and that
either case would be interesting. Again, I had the privilege of
writing the first draft, and the pleasure of discovering that my
coauthors largely agreed with it. We postulated the same
seven generative dimensions. As Bailey had supposed, their
basic meanings remained the same as in 1968, but their
ramifications were more elaborate and more realistic in 1987.
(Bailey’s proposal was wise and prudent. Any discipline that
can ever state its generative dimensions should do so, and
review them every 20 years to see if they have changed, and
how, and argue whether that is good or bad.)

In my opinion, then and now, there were not really seven

generative principles of applied behavior analysis; basically,
there were only three. Those three were that the discipline be
capable of producing generalized behavior changes, be
capable of explaining how it did that, and be capable of
proving its power and its explanations. The most urgent of
these capabilities was proof. In support of that thesis, consider
the seven dimensions more or less in order, to see their
underlying allegiances.

Applied

The surface meaning of “applied” was that the target behavior

changes would have considerable social significance and
importance. Subsequent experience elaborated that surface

489
meaning into two kinds of reasons for changing behavior: one
was to find out how it could be changed; the other was that
someone did not like the behavior the way it was, and did like
some other way it could be. The first connoted basic science;
the second connoted application, depending on the social
importance of that someone.

But “applied” could have an underlying meaning as well. It

could be an acid test of proof. Behavior analysis had proven
the power of its principles to change convenient behaviors of
convenient organisms in well controlled laboratories. Could it
prove that those principles were similarly powerful for the
complicated behaviors called problems in people in real
trouble in real-world settings? While some applied behavior
analysts were trying to be useful to their society, others were
asking for a better proof of these principles. Analogous to the
conceit that if you can make it in New York you can make it
anywhere was the premise that if these principles work in
application they work anywhere. Whatever else application
was, it was a proof technique.

Behavioral

One reason for a science of behavior is that behavior exists,

therefore deserves analysis, and under analysis proves lawful.
But another reason is that “behavior” usually denotes
something observable, and hence something in which changes
are amenable to direct proof rather than inference. Proof
consumers will of course prefer a psychology of directly
measurable behaviors to a psychology of necessarily inferred
behavioral “constructs.” Whatever else behaviorism is, it is a
proof technique.

490
Analytic and Conceptual

In 1968, “analytic” meant experimental analysis: We could

prove what had caused the behavior change. “Conceptual”
meant that we not only could prove what had changed the
behavior, we also could explain why that technique should
have changed that behavior in that way. As Skinner had
daringly proved by his book, Verbal Behavior (1958),
behavior analysis could pursue conceptual analysis quite
extensively without a shred of direct proof. However, the
applied behavior analysis proposed by Baer, Wolf, and Risley
could not. In it, conceptual analysis might follow a proof of
what had changed behavior, or might precede a proof of what
would change behavior, but it never was divorced from proof.
Yet proof could be divorced from it, and still be proof.

For many psychology audiences, this wedding of conceptual

analysis and proof was the worst disadvantage of applied
behavior analysis; for those audiences, flights of conceptual
analysis should never be seriously restrained by anything as
mundane as proof. But for proof consumers, this constraint
was of course the greatest recommendation of this kind of
applied behavior analysis. It also made clear that the proper
question was about our reinforcers. For me, the question was,
if I could have a natural science of behavior, why would I
want an unnatural one?

Technological

A technology is a set of recipes for reliably accomplishing

some outcome. The term “technological” implies forcefully
that these recipes always work. Technology is therefore a
body of completely described procedures accompanied by

491
complete descriptions of the parameters necessary and
sufficient for them to work reliably. That requires a lot of
proof. Whatever else the development of a technology is, it is
the result of a great deal of proof. Applied behavior analysis
tried to develop and accumulate a behavioral technology, so it
had to prove what was necessary to yield a recipe that should
always work, and then prove the extent to which it did.

Effective

In 1968, the basic way to be effective was to state a

measurable goal and then demonstrate how closely the
procedures applied to a problem attained that goal. That was
essentially a problem in proof. We usually chose the goal to
be a behavior change other than a client’s statement that
things were better now. We knew that the relation between
what people said and how they behaved otherwise could be
anything; we simply did not trust their statements about
behavior changes to be true, unless we had programmed them
to be true — and we were not sure we could do that in a
thoroughly, permanently generalized way. It seemed better to
measure goal attainments directly, and to prove that they were
attainments of the procedures that had been applied.

By 1987, effectiveness also meant that the goal was worth the
cost of attaining it the way we had attained it. Cost meant the
costs of the procedures that accomplished the goal, and any
costs of everything else those procedures caused, if anything.
Both were problems in proof, because it was safer to prove
what were and what were not the systematic consequences of
the procedures, than to infer that things must be better now
and that nothing could have gone wrong.

492
But 20 years had shown a use for clients’ statements that
things were or were not better now. Sometimes behavior was
changed in ways the clients or other audiences did not like.
That could result in countercontrol, which severely
diminished anything that could be called effectiveness, and
was bad for our reputation. We began asking all relevant
audiences if they liked the procedures, their outcomes, and
the personnel; we meant to assess something called the social
validity (Kazdin, 1977; Wolf, 1978) of our programs. We did
that not just for the pleasure of positive answers, but also
because negative answers might predict countercontrol. The
old problem of proof arose even more severely: If we allowed
unhappy audiences to lie to us about liking our procedures,
we would fail to predict the subsequent countercontrol, and
effectiveness and reputation would be lost. Experience
suggests that audiences easily lie about what has just
happened to them. Suddenly we were in the ironic position of
needing truthful talk about behavior from the very people we
earlier had not trusted to talk truthfully about behavior.

Thus, to be effective had become an urgent problem in

increasingly complex and difficult proofs. Those proofs
required measuring everything that happened, attributing it
correctly to our procedures or elsewhere, and knowing the
relation between what our audiences said and what they
would do later.

Capable of Appropriately Generalized Behavior Changes

Whatever else the production of appropriately generalized

behavior changes may be, it is, if the term “appropriately” is
deserved, a technology. So it must be based on a large body
of proof, which is the only way to produce any technology. In

493
1968, it was sufficient to prove how we accomplished
appropriate generalizations; in 1987, it was imperative to
prove as well that our choices of “appropriate” were in fact
appropriate. It also seemed reasonable to acknowledge that
“generalization” was a pragmatic term, and “stimulus control”
was its analytic obverse.

Overview

Of the seven generative dimensions proposed to define

applied behavior analysis, six of them are essentially proof
techniques or proof demanders. The 1968 “Current
Dimensions” paper had been written to formalize what we
were teaching our students then. The simple part of what we
taught them was whatever was known then about how to
change behavior. That was a mixture of procedures, a
conceptualization that gave the procedures meaning, and a
conceptualization about how to match procedures to
problems. The complex and difficult parts of what we taught
were what we and our students had to prove in the process, if
our adventures were to be part of a scientific discipline. The
essence of the argument was the primacy of proof.

That argument seemed worth publishing in the first issue of

the Journal of Applied Behavior Analysis. Twenty years later,
it seemed worth reaffirmation and elaboration, but mainly
reaffirmation.

494
Important Developments in the Rise of
the Behavior Therapies
For those of us who saw applied behavior analysis primarily
as the acid test of the power and generality of behavior
analysis, the decades after 1968 were very informative.
Behavior analysis passed instance after instance of this test,
and continues to pass more instances of it today. Most of us
saw that some problems of people and society arose from the
behavior of those people and their society. The final
significance of applied behavior analysis might well be that it
would show how to solve some of those problems.

Thus applied behavior analysis took a place among the

already existing behavior therapies. The behavior therapies
had been doing very nicely with some problems that seemed
suitable for the logic of classical conditioning and its
extinction, inhibition, and disinhibition. The targets of those
procedures were sometimes directly observable behaviors,
and sometimes were inferred internal behaviors. The inferred,
internal behaviors were considered the explanation of the
success seen with the external, directly observable behaviors.
In that sense, these behavior therapies hardly seemed different
from the psychodynamic therapies against which they were a
rebellion. But because the procedures applied were
conditioning-relevant procedures, and since those procedures
had in the past always been applied to behaviors, using them
in this clinical way must be “behavior therapy.”

Sometimes those behavior therapists measured how well they

had done with the measurable parts of their approach, and
sometimes they proved how well they had done with those

495
parts. They did not prove anything about their inferred
internal behaviors, because they could not; they merely
inferred that these internal behaviors mediated their
observable success, and thereby explained it.

That practice of course attracted criticism from

natural-science adherents. Rather than give up those internal
explanatory events, these practitioners coined the term,
“cognitive-behavior therapy.” That denoted that they
considered the inferred “cognitive” construct (1) explanatory;
(2) either necessary or at least useful; (3) real, in that we all
know that we do behave privately at times; and (4) still
essentially behavioral, because the inferred behavior was
presumably being controlled by procedures that traditionally
controlled observable behaviors.

To solve more of the clinical problems presented, the

cognitive-behavior therapists needed not that kind of
explanation, but the addition of operant logic. It is easier to
generate applications, especially skill-building applications,
from the N-term contingency than to strain Pavlovian
mechanisms, observable or not, into those pursuits. That
argument is perhaps only pragmatic. But behavior analysis
also offered a world view — if anyone wanted one.

So, cognitive-behavior therapies came increasingly to use

techniques from the operant or subsequent behavior-analytic
realm. These cognitive-behavior therapies paid applied
behavior analysis one immense compliment and one casual
insult. The compliment was to adopt some of its procedures.
The insult was to impose an internal cognitive argument on
the behavior-analytic thesis. The behavior-analytic thesis was
that a complete analysis of the external environment would

496
show when those procedures would be effective and when
they would not. When the cognitive-behavior therapies gave
primacy to cognitions instead, they allowed, and perhaps
encouraged, a neglect of the crucial details of environmental
control.

An example may be instructive: time out. Time out is

probably a widely used and largely misused procedure in our
society today. It is often prescribed as an acceptable way to
reduce undesirable behavior, and the behavior therapies often
package it with an attempt at cognitive reorganization. But
behavior-analytic logic teaches that time out has no necessary
or fixed function for behavior. If the reinforcement and
punishment schedules of the time-out environment are worse
than those of the time-in environment, time out will weaken
the behavior on which it is systematically contingent, but only
if the contingency is managed well. If the reinforcement and
punishment schedules of the time-out environment are equal
to those of the time-in environment, time out will not change
the behavior on which it is systematically contingent, other
than continue to detract from the time available for good
programming. If the reinforcement and punishment
contingencies of the time-out environment are better than
those of the time-in environment, time out will strengthen the
behavior on which it is systematically contingent, but only if
the contingency is managed well.

A booklet is needed to explain thoroughly what those three

statements mean and why they are unavoidably true. Another
booklet is needed to describe how to manage a time-out
contingency, or any contingency, well or badly. Another
booklet is needed to describe how to assess the reinforcement
and punishment schedules naturally operative in the two

497
environments. Another booklet is needed to describe how to
change those schedules to make and keep the time-out
contingency effective. Another booklet is needed to describe
how to record everything relevant and graph it, so as to reveal
whether progress is being made, and how much; and if no
progress is being made, to explain how to best choose the
next procedures. Another booklet is needed to explain
convincingly to the potential user of time out that the user’s
behavior of using it, even ineffectively, is subject to quite
powerful reinforcement, especially when time out frees the
user for a while from living with the behavior of the
timed-out person. Another booklet is needed to explain that
no behavior problem is solved merely by eliminating an
undesirable behavior: The behaver needs desirable ways of
gaining the reinforcers and avoiding the punishers that the
undesirable behavior gained and avoided. Those desirable
ways may not already exist in the behaver’s repertoire. Then
they will have to be taught. Many booklets are needed to
describe the relevant teaching processes well enough to get
them done effectively.

Usually, when time out is stolen from applied behavior

analysis, its many booklets are not stolen as well. Instead,
hapless parents are told that whenever their child misbehaves,
they should shut the child in a room alone until the child
behaves better. They are often told what lessons their child
will learn from this, as if that explained why the child would
behave better.

Sometimes, that painfully incomplete recommendation

actually reduces the undesirable behavior; sometimes it does
not. Most often, we will never know which is true, because
most often, the people who recommend the technique rely on

498
the parents’ report of how well it worked. Like all behaviors,
parent report is subject to control by many contingencies.
Sometimes one of those contingencies is to reassure the
therapist that the problem has been solved, not because it has,
but because the parents want a graceful way to leave this
therapist and find a different one. When you give up proof,
you give up knowing.

In general, the cognitive-behavior therapies find the causation

or mediation of behavior change in the cognitions of the
behaver. Consequently, they target those cognitions when
change is necessary, on the premise that if the cognitions
change, so will the relevant behaviors. They give little
attention to a second possibility, which is that if only the
behaviors are targeted for change, and are changed,
cognitions about them will change as well, and exactly as the
therapist would like. And the usual cognitive-behavior
therapy gives very little attention to a third possibility, which
is, in metaphor, that behaviors and cognitions about those
behaviors can lead perfectly independent lives. In only one of
those three cases is the cognition a reasonable target for
intervention, because only in that case is it the explanatory
mediation of the change.

The need to explain what is true creates most of the

arguments among the cognitive-behavior therapists and the
applied behavior analysts. Their difference seems to lie
primarily in where they seek explanation. Perhaps nothing is
more destructive of peace among scientists than
disagreements on how to explain what they know is true.
Behavior analysts will look for “why” in the organism’s
environment; cognitive scientists will look for it inside the

499
organism (even as they agree that its now far-distant early
origins probably were in the external environment).

Some cognitive-behavior therapies use a few putative

behavior-change techniques, minus their booklets, for a
secondary role. They use whatever putative behavior-change
techniques the therapist knows or likes for the primary
behavior-change role; but they still target changes in
cognitions first. They do not see the cognitive changes as
behavior-change techniques; they see them as what will make
the behavior-change techniques powerful. It is as if clients
must first (in one notable example) accept themselves and
their world as they are, and then commit themselves to some
goal, because only after that will the differential
reinforcement available to a therapist change their behavior in
a useful, dependable way.

When applied behavior analysis is informed by the principles

of radical behaviorism, it is not hostile to the notion of private
behaviors, which may be all that is meant by “cognition.” It
is, however, congenitally cautious about inference, and
extremely skeptical of the premise that inferred cognitions
have any autonomy — that they are first causes. In radical
behaviorism, cognitions are seen as behaviors — private
behaviors, for the most part, meaning only that they are not
directly measurable. But private behaviors, like observable
behaviors, may play any of three roles in a problem-solving
chain: (1) One chain is that a problem is presented and evokes
the unobservable behaviors called cognitive processing,
which in turn evoke a problem solution. In such chains,
teaching or repairing the cognition should prove quite
effective. (2) A second chain is that a problem is presented
and evokes a solution, which in turn evokes cognitive

500
processing of the solution and its problem. In this chain, there
is very little to be gained by targeting the cognition; it is not
the cause of the solution, but the result of the solution. (3) A
third chain is that a problem is presented, and it
simultaneously evokes a solution and cognitive processing of
the problem. In this chain, the cognitive processing and the
solution have no relation to each other; each is an independent
response to the problem. Then targeting the cognition will
again have little value in promoting a solution.

Most cognitive-behavior therapies package what their

practitioners will call cognition-change techniques with what
applied behavior analysts will call behavior-change
techniques. When those packages succeed, the presumed
cognition changes are seen as crucial, or at least functional.
Yet there is no proof of that. True, in the world of practice,
packages are seen as highly desirable: an addition of one kind
of strength to another kind of strength, which presumably will
most benefit the maximal number of clients. But in the world
of proof, packages are seen as confoundings: There may well
be a proof that the package is effective, but there probably is
not a proof of which of its components contributed what, if
anything, to that effectiveness. The experimental analysis of
therapy packages is rarely done, probably because it is
forbiddingly expensive and probably not cost-effective, as
will be seen below.

If the three chains postulated above all can be real, then

applied behavior analysts cannot logically dismiss the
cognitive-behavior therapies for wasting time and effort on
cognitions of unproven function in the effective package. One
of those three causal chains is almost exactly what the
cognitive-behavior therapists postulate is the general case. So,

501
the question is, in the problems with which we all work, how
prevalent is each of these three causal chains?

And the next question is, predictably, is the first question

amenable to proof?

Perhaps many applied behavior analysts who consider that

question will be reminded of an unfinished bit of proof the
field still requires, namely, the possibility of a placebo effect.
The problem is crystal-clear in medical research. In the
evaluation of new medical therapies, a placebo control is now
mandatory. Any medical procedure usually is delivered to
patients with a good bit of theater: the distinctiveness of the
clinic or hospital setting; the formidable equipment on every
side; the white coat; the societal role; the past personal cures
remembered; the promise, belief, or hope by patients,
clinicians, and observers alike that this will work; the general
successes steadily celebrated in the media, especially the
entertainment media; the authoritative manner of medical
personnel; and the extraordinary costs. Once medical
researchers discovered that good theater could “cure” certain
ailments by itself, they had to distinguish forevermore
between medical cures on the one hand, and drama cures on
the other — i.e., placebo cures that deliver all the theater but
without any medical agent.

The alternative is to recognize good theater as not a placebo

but as simply another medical agent, useful in some cases but
not in others.

The parallel question remains unanswered for the behavior

therapies and for applied behavior analysis: How much of our
effectiveness is attributable to good theater, and how much to

502
the procedures our science said should have been the effective
agents? Perhaps this question has been neglected by applied
behavior analysis so long because we quietly suspect that we
are very bad theater. We put causation in the environment,
when almost every client knows it surely must well up from
within. Thereby, we are bad theater. In that context, what
cognitive-behavior therapies add to applied behavior analysis
looks, at least, exactly like much better theater: Cognitions
are processes welling up from within! There is a therapy
worth the ticket price.

So, one proof problem is to ask if the explicit targeting of

cognitions is the appropriate placebo control for applied
behavior-analytic interventions. And if, as in medical
research, this presumed placebo turns out to be effective for
the behavior changes we target, then once again, an
alternative is to consider it not a placebo but a
behavior-analytic agent, specifically, a bit of
conditional-stimulus control or a conditioned establishing
operation (Michael, 1995).

The question still begging for a proof is the separate

contributions, if any, of what are called the cognitive and
what are called the behavioral components of the typical
cognitive-behavior therapy package. If that is the question, we
can of course, as always, answer it at the level of theory,
ideology, business, or rhetoric. But we might also consider it
a problem for proof.

What would a proof require? First, it needs a null hypothesis.

That hypothesis could assign priority to the cognitive
procedures of the package, and ask if the behavior-change
techniques add anything to them; or it could give priority to

503
the behavior-change procedures of the package, and ask if the
cognitive techniques add anything to them. I argue here that
the procedures of applied behavior analysis are well known to
me as behavior-change procedures, whereas the procedures of
cognitive reorganization seem to me to be much less
dependable, as I look back on my own education. Thus I will
for myself phrase the null hypothesis as if applied behavior
analysis had the priority: my null hypothesis is that the
addition of cognitive procedures to behavior-analytic
procedures does not alter the effectiveness of the
behavior-analytic procedures. Anyone else may recast the null
hypothesis to give priority to the cognitive procedures: the
alternative null hypothesis is that the addition of behavioral
procedures adds nothing to the effectiveness of cognitive
procedures. Either way, the problem for proof is much the
same; we shall either compare cognitive-only to
cognitive-plus-behavioral, or compare behavioral-only to
behavioral-plus-cognitive. (If we are well funded, which we
never are, we can compare cognitive-only, behavioral-only,
and cognitive-plus-behavioral.)

This question is about the comparative effectiveness of some

therapies for a population of potential clients. That makes it
an actuarial question. Actuarial questions require actuarial
designs, i.e., well sampled group designs. In the most
rudimentary design, we need to compare a group of clients
receiving a package of behavior-change and cognition-change
procedures to a group of clients receiving only the
behavior-change procedures (or only the cognition-change
procedures).

Fortunately, or perhaps unfortunately, many research

professionals have for several decades considered the proof

504
requirements for the comparison of any Therapy A to any
Therapy B. Here is a summary of what they have shown the
problem requires:

• Agreement on common outcome measures, which

should always include social-validity assessments
made by the clients and the therapists, and
procedural-fidelity assessments made of the
therapists’ behavior.
• A very large sample of cases representative of the
populations of interest.
• Random assignment to Therapy A and Therapy B of
very many cases from a homogeneous sample.
• Alternatively, matched assignment to Therapy A and
Therapy B of very many cases from a heterogeneous
sample, according to the variables that define them as
heterogeneous.
• Assessment of each case’s pleasure or resentment
with the assignment, so as to analyze the degree to
which any unequal distribution of these reactions
across the two groups could bias the outcome.
• Random assignment to Therapy A and Therapy B of
very many therapists from a homogeneously skilled
population of such therapists.
• Alternatively, matched assignment to Therapy A and
Therapy B of very many therapists from a
heterogeneously skilled population of such therapists,
according to the variables that define them as
heterogeneously skilled.
• Assessment of each therapist’s pleasure, resentment,
confidence, or hesitation with the assignment, so as to
analyze the degree to which any unequal distribution

505
 of these reactions across the two groups could bias
the outcome.
• Assessment of what therapies, if any, each case has
had previously, and with what effectiveness, so as to
analyze the possibility that Therapy A or Therapy B
does best if preceded by some Therapy X or worst if
preceded by some Therapy Y, and so as to analyze
the degree to which any unequal distribution of these
histories across the two groups could bias the
outcome.
• Assessment of the correlates of attrition of cases from
each group, so as to analyze the degree to which an
unequal distribution of attrition, or of reasons for
attrition, across the two groups could bias the
outcome.
This list of the requirements for an accurate comparison of
some Therapy A to some Therapy B is not complete; it is only
a good beginning.

Object Lessons for the Future

Until this kind of extraordinarily expensive proof is in hand,

which would seem to be never, there is small good reason to
debate the relative merits of applied behavior analysis and the
cognitive-behavior therapies. There is even less reason to
attempt literature-based meta-analyses of what each is best
and worst at doing. Such analyses are almost inevitably full
of, or susceptible to, the seriously misleading biases just
listed, which a proper proof would either prevent or assess for
some form of covariance analysis.

506
We might better simply note that each discipline has found an
evolutionary niche in which it prospers and reproduces, a
little, and that each probably will continue to do so for many
years to come. What we cannot afford to prove about our
variety can always be observed, not to see what is true, but to
see what survives. And if survival (rather than correctness) is
of interest, then we might well remember an old rule of
evolution: A population with some diversity has a better
chance of surviving some sudden change in the survival
contingencies than a population with little diversity.

Are the behavior therapists and the applied behavior analysts

a population? To decide that, we should first compare our
disciplines to the theories, size, power, and entrenchment in
our society of all the disciplines and therapies we are not.
That comparison will show that, despite our diversity in
where we seek explanation, we are, compared to our
alternatives, very much alike, and very different from them.
Similarly, simple politics will show that if we behave as one
population, we will survive much better than if each of our
subdivisions secedes because it considers itself the lone
fraction of the population that is doing our discipline
correctly.

Being scientists as well as practitioners and survivors, we

probably will continue seeing small differences in where to
seek explanation as crucial differences, until proof or
extinction decides the matter. But to continue being survivors,
we should remember to forget those differences at all
survival-relevant moments.

507
References
Baer, D. M., Wolf, M. M., & Risley, T. R. (1968). Some
current dimensions of applied behavior analysis. Journal of
Applied Behavior Analysis, 1, 91-97.

Baer, D. M., Wolf, M. M., & Risley, T. R. (1968). Some

still-current dimensions of applied behavior analysis. Journal
of Applied Behavior Analysis, 20, 313-327.

Kantor, J. R. (1924/26). Principles of psychology (Vol. 1-2).

Chicago: Principia Press.

Kazdin, A. E. (1977). Assessing the clinical or applied

significance of behavior change through social validation.
Behavior Modification, 1, 427-452.

Michael, J. (1995). What every student of behavior analysis

ought to learn: A system for classifying the multiple effects of
behavioral variables. The Behavior Analyst, 18, 273-284.

Skinner, B. F. (1957). Verbal behavior. New York:

Appleton-Century-Crofts.

Wolf, M. M. (1978). Social validity: The case for subjective

measurement, or how applied behavior analysis is finding its
heart. Journal of Applied Behavior Analysis, 11, 203-214.

508
Chapter 15

Do Good, Take Data

Todd R. Risley

University of Alaska

The Origins of Applied Behavior

Analysis
In the late 1950’s, Jack Michael1, a bright but irritating young
psychology instructor, moved from the Universities of Kansas
to Houston to Arizona State. Along the way he befriended
two nontraditional students, protected them through their Ph.
D. programs, and turned them loose on the world: Teodoro
Ayllon, who produced a series of unprecedented field studies
at a mental hospital in Saskatchewan for his dissertation in the
late 1950’s (Ayllon, 1959); and Montrose Wolf who
proceeded to set the parameters of Applied Behavior Analysis
at the Institute for Child Development at the University of
Washington in 1962-64. The methods they modeled are now
so universal in Applied Behavior Analysis that their origins
go unnoticed:

• Systematic observation and recording by people;

• with checks for the bias and drift that are likely when
people are involved.
• Repeated observations within sessions and across
days, weeks and months;

509
 • with experimental designs that go with observation
and recording over time.
But their most important “breakthrough” contribution was the
demonstration that naively simple immediate things were
actually powerfully important in the real lives of people. You
see, at the time we were all talking about the principles of
learning and behavior but we thought they would be actually
expressed only in complex, multiply-interactive combinations
in the ongoing actions of people in real life. (In this respect,
we were all “Hullians.”) We assumed that their role could
only be isolated and analyzed after carefully designed
histories in specially arranged settings — in other words, in
laboratories. And laboratory work, both ‘neohullian’ and
‘human operant,’ was flourishing in the early 1960’s. It was
certainly flourishing at the University of Washington. (My
master’s thesis research [Risley, 1964] was an example.)

(At this point, we will leave Ayllon to his own affairs, with
the reminder that he was recruited by Nathan Azrin, an
established leader in experimental psychology at the time and
the most prolific contributor of good research on many topics
to the experimental analysis of behavior. At Anna State
Hospital Azrin and Ayllon started [Ayllon & Azrin, 1964],
and Azrin developed, arguably the most productive and
important program of problem-solving field research that has
yet existed [cf. Azrin, 1977]).

Montrose Wolf arrived at Washington as a post-doc in the

summer of 1962. There he found a flourishing setting that had
been deliberately and bravely created by Sidney Bijou. Bijou
was a full professor of psychology and was responsible for
the child clinic, the preschools and the experimental-child

510
laboratory at the Institute of Child Development, (ICD) and
had established a research lab at Rainier State School. He had
recruited Donald Baer and Jay Birnbrauer as new assistant
professors in developmental psychology and filled the
post-doc position of child clinic director with successively,
Ivar Lovaas, Ralph Wetzel, and Robert Wahler. On his
research and training grants, Bijou had recruited and
supported a cadre of graduate students including me. (I had
finished my undergraduate work at San Diego State, where
Virginia Voeks, who had studied with both Hull and Guthrie,
taught the learning and child psychology courses. She sent me
to the University of Washington in 1960 where I was offered
an assistantship at ICD.)

When he arrived, Wolf was assigned to teach the preschool

teachers an introductory course in learning principles. The
four class projects designed by Wolf and carried out by the
teachers constituted the discovery of the power of social
attention in real life. We had never seen or imagined such
power! We were all amazed at the speed and magnitude of the
effects of simple adjustments of such a ubiquitous variable as
adult attention. Thirty-five years later, positive attention,
praise, and “catch them being good” have become the
foundation of most American advice and training for parents
and teachers — making this the most influential discovery of
modern psychology.

We were also enthused by the methodology that was evolving

as the studies progressed: direct observation with interval
recording, interobserver reliability, reversals of conditions,
and concurrent multiple measures — this at a time when the
only real-time data being collected were from automatic
recorders in laboratory settings, and the few real-world efforts

511
were being documented only with field notes (excepting the
hospital studies of Ayllon and associates, 1959, 1962, 1963).

Among Wolf’s other duties at Washington was the task of

building an experimental classroom for children with mental
retardation at Rainier State School, 40 miles away. He had
brought from Arizona State the first seed of a system of
symbolic reinforcers that he cultivated into a durable
motivational system to maintain academic behavior, a system
now called a point system or token economy. (He also helped
design an ingenious curriculum of reading comprehension.)
(See Risley, 1997 for a list of the publications from Wolf’s
1962-64 work at Washington.)

Wolf also modeled ease and enthusiasm in directly interacting

with the preschool children at ICD and the institutionalized
people at Rainier School. There-to-fore everyone seemed to
avoid such direct contact — ostensibly “to avoid research
contamination,” but also because no one was very good at
talking to children. (I have seen many, many academics who
are so obviously uncomfortable and clumsy at direct contact
with the people they study that they only deal with them
through graduate student proxies.)

At the same time, Montrose Wolf accepted the task of getting

a vision-impaired 3-year-old boy with autism, who displayed
tantrums and self-injury and who resided in a psychiatric
hospital 50 miles distant, to wear his glasses. I assisted him
with this little chore. Thus began a year of weekly drives that
culminated in the premier study of behavior modification
(Wolf, Risley, & Mees, 1964). That study introduced the
procedure of contingent social isolation, labeled it time-out
(to note the vague similarity to the laboratory procedure), and

512
demonstrated its power with a reversal design. Thirty-five
years later, and now recommended by the American
Academy of Pediatrics (1998), half the parents and teachers in
the United States use this nonviolent practice and call it
“time-out,” which makes it a social invention unmatched in
modern psychology.

That work and related work over the next year (Risley &
Wolf, 1964, 1967; Wolf, Risley, Johnston, Harris, & Allen,
1967) are also noteworthy in that they introduced the direct
reinforcement of verbal imitation and the shaping of
meaningful speech. It is amusing to recollect the exclamations
of amazement from professionals — even behavioral
professionals working with children with autism — at the
sight of Wolf or me matter-of-factly using bites of food to
shape functional speech with children with autism and mental
retardation. (Ferster gave us credit for our procedures, but did
not use them. Lovaas used them.)

In 1963, I had decided to pursue the deliberate development

of functional speech in children with severe speech deficits as
my dissertation project and, with Bijou’s support, started
working with a dozen children with various diagnoses
(autism, retardation, aphasia). In Bijou’s laboratory, I had
been using the Wisconsin General Test Apparatus (a tray with
choices that was pushed through a curtain when the subject
was sitting with quiet hands and looking at the tray front).
This I adapted into the now familiar face-to-face “discrete
trial” procedure — after much reluctance because it was a
“restricted operant” and not a “pure” free-operant procedure.
Reflecting this preference for free-operant (initiated)
behavior, I alternated periods of offering (or “baiting” the
environment with) attractive things or activities and waiting

513
for the child to initiate, before prompting a more elaborate
request — the germ of the procedure Betty Hart and I later
elaborated into “incidental teaching.”

With Montrose Wolf’s demonstrations that the things we

were studying and discussing could actually be powerfully
influential in simple ways in real life, the Institute of Child
Development became caught up in a remarkable time of
discovery and excitement. The Institute’s research meeting
was the high point of our week. People came from miles
around to listen, to question and to present their work. It
seemed that everything anyone innovated was immediately
adapted by everyone else.

A steady stream of visitors came from around the world to see

Bijou and the work at the Institute. Wolf and I would take
them on our twice-a-week rounds to the classrooms (and labs)
at Rainier State School, then to the mental hospital near
Olympia and then back to Seattle. All the while we would be
“double teaming” them with descriptions, argument, and
excited discussions. After another day at ICD — observing
my speech shaping sessions, preschool teachers taking data
and using adult attention, and Wahler’s clinic staff exploring
functional analyses of children’s problem behaviors; and after
attending one of the Institute’s research meetings — most
visitors departed as enthusiasts and many became active
colleagues.

514
The Journal of Applied Behavior
Analysis
Now, let us skip ahead a few years to 1966, after Baer and
Richard Schiefelbusch had recruited some of us together
again to the University of Kansas. The combination of field
research methods and problem-solving strategies that Ayllon
and Wolf modeled for us had now evolved in sophistication
and example and had proliferated across the country (except
peculiarly in those places where operant laboratory research
was strongest). By this time, books of readings (such as
Ullman & Krasner, 1965, or Ulrich, Stachnik, & Mabry,
1966) were not enough to handle the studies being generated.
We needed a journal. Wolf campaigned for a journal and Sage
and Academic Press each responded favorably. But by then
we had convinced the Society for the Experimental Analysis
of Behavior to sponsor an applied companion to the Journal
of the Experimental Analysis of Behavior (JEAB). With the
wise guidance of Azrin, the Society selected Montrose Wolf
as the first editor, and he designed and named the Journal of
Applied Behavior Analysis (JABA).

The Baer, Wolf and Risley (1968) article was written,

primarily by Baer, as an attempt to differentially prompt
certain types of submissions. Wolf and I intended that article
to be heuristic (“Some Current”) rather than definitive
(“Dimensions of Applied Behavior Analysis”). He and I
assumed that the enterprise of Applied Behavior Analysis
would evolve — that findings would condense into
knowledge and technology, and that new problems and
opportunities would require and beget new research

515
methodologies. Underlying all the suggestions in that article
were the dimensions that most concerned Wolf and me:

The encouragement of field research;

• the insistence that you should seek lawfulness in the

everyday activities of people; and
• the pursuit of the invention (and documentation) of
new behavioral technology.
We wished to devote JABA’s space and reviewing resources
to the display and shaping of the analysis of those variables
that actually influence what people actually do — not analogs
to what people do, and not derived variables that may or may
not account for what people do. Laboratory research had other
outlets and audiences, such as JEAB.

We did not think that laboratory research findings are

unimportant to human affairs — quite the contrary. It is just
that one must study and directly analyze human affairs to
know what derived principles and findings might, in fact, be
relevant in any instance. We saw too many examples of
behavioral researchers behaving like other psychologists and
casually extrapolating their findings to account for things they
actually knew little or nothing about. To quote the wisdom of
Sidney Bijou, “Before you try to explain something be sure
that it, in fact, exists.” As a matter of intellectual honesty or
scientific integrity, if you are going to claim a relationship
between laboratory findings and human affairs you need to
know a great deal about both of them. And if you are going to
propose some new and better way of doing something, you
need to demonstrate the utility of your proposal. We saw too
many studies, even behavioral studies, whose importance to

516
human affairs were highlighted in their introduction and
discussion sections but absent in their procedures and results
sections.

And Wolf and I saw the exciting possibility of experimentally

analyzing human affairs rather than simply passively
observing them as other applied psychologists were doing.
We were sure that a new era of natural science would emerge
that would directly explore more and more of human affairs
— and through field (and laboratory) analysis, account for
that reality in progressively fewer terms. And we were sure
that a new era of behavioral and social invention and
technology would flourish — and through cultural insertion,
continually improve human kindness and productivity. I leave
it to you to judge whether we are succeeding. In the field or in
the laboratory, let me remind you of our common grounds in
Behavior Analysis:

Realism: There is a reality that exists independent of our

perspective of it.

Natural History: The reliable description of that reality.

Natural Science: The description of reality in progressively

fewer terms.

Experimental Analysis: You can understand best when you

intervene (“to carve reality at its joints”).

Empiricism: It is best to derive our concepts from description

and analysis of reality. To paraphrase the author of Walden
(one): “How much virtue there is to simply see, and to fasten
words again to visible things.”

517
My Version of Applied Behavior
Analysis
Within Applied Behavior Analysis there was (and is) a
diversity of emphases from religious to philosophical to
empirical: for some, B. F. Skinner was the final word and
their task was to defend and interpret his writings; for others,
Radical Behaviorism was a philosophical worldview for
logical analysis, explanation and debate; for many of us,
Methodological Behaviorism was an attempt to import the
simplest precepts of natural history and experimental analysis
into the slow, incremental process of observing, accounting
for, and improving human affairs. If allowed to oversimplify
it might be said that Bijou and Baer were mostly interested in
explaining the world, and Wolf was mostly interested in
fixing the world. I think I was mostly interested in exploring
the world.

I was most influenced by Skinner’s urgings for the

development of behavioral and social technology to overcome
our genetic predilections and our cultural superstitions. In the
first three chapters of Science and Human Behavior (1953),
Skinner had outlined an agenda for an inductive, empirical
approach to a science of human behavior. (Which was
followed by 26 chapters of a deductive, logical explanation of
uninvestigated human behavior.) In Walden Two (1948),
Skinner had envisioned an experimenting community,
constantly monitoring and redesigning itself. (Without,
however, envisioning much effort to document and
demonstrate the resultant technology and inventions.)

518
Murray Sidman’s Tactics of Scientific Research (1960) gave
me an experimental, inductive model of science that assumed
that one should strive to discover things and invent
technology in the largely uncharted world of behavior. And
Donald Campbell (1957, 1963, 1969) gave me both the
mission and the methodology to pursue reforms as
experiments.

John F. Kennedy and the civil rights movement convinced me

that it was not only acceptable to act on social problems — it
was imperative to act. And act I have.

Although speech and language development has been a

predominant theme in my career, a parallel theme has been to
be a “social entrepreneur” — pursuing “openings of
opportunity” to find out how the world really works by taking
responsibility and trying to intervene in human problems that
presented themselves to me.

After applying my molecular technology for shaping and

strengthening and generalizing behaviors, I usually found
myself complaining about and blaming teachers, parents, and
attendants who were not doing all day long what I advised
them to do. I soon found that even those who were skillful
and willing often could not divert the time or attention from
their other tasks — and that those other tasks, often badly
designed and orchestrated, consumed more time and attention
than they should. I therefore became most interested in human
organizations, human settings — and human crises.

Organizations are arrangements of people which have

activities and goals that persist though the participants may
change. Organizations provide predictable establishing

519
operations and contingencies of social and material
reinforcers across all participants, and provide predictable
curricula of models, prompts, instruction and reinforcement
criteria for each participant. The organizations in which a
person participates provide the infrastructure of predictable
schedules, discriminative stimuli, contingencies and
reinforcers of most of her or his daily life. People suffer when
their family, social, community and work organizations are
disrupted or incompetent.

Settings are the physical environments in, on, and around

which human activities occur. Settings facilitate or impede
human actions across every occupant by the antecedent,
response effort or consequence effects of their designs.
Everyone passes through many settings each day and they are
at risk of failure or even injury when the settings that make up
their homes, neighborhoods, streets, markets, schools,
playgrounds, and work or recreation places are badly
designed for their activities.

Crises are opportunities. Although many human organizations

and human settings are poorly suited to the activities intended
in them, there is little opportunity to explore and intervene
and understand them, and to invent appropriate organizational
technologies and physical designs, until a crisis comes. A
crisis improves the social acceptance of any change. (Change
is stressful and is usually resisted by affected people.) A crisis
improves the political and resource support for large changes.
(Without crisis, only small aspects of organizations and
settings can be explored, probed, understood, and changed.)
And a crisis provides the ethical basis for taking personal and
professional responsibility and for making assertive
interventions.

520
The following is an account of the human organizations and
settings where I have taken responsibility for people in some
significant aspects of their lives — with the intent of making
their lives better and contributing knowledge and technology
for use for others. I have omitted several projects that are too
complicated or sensitive to describe, and many more where
aspects of peoples’ lives were not at stake (see endnote 1 for
some of these), or where someone else was primarily
responsible (endnote 2 names most of those people). In each
of the following projects I was personally at social, political
and professional risk for the conduct of the project, and
ethically responsible to ensure benefits to the people being
served. Although there were several other participants in each
project (partially reflected in the authorship of resultant
publications), I have named only those persons who shared
the risks and responsibilities of starting a project. The term
‘created’ indicates a service where none previously existed;
and ‘begun’ indicates the conversion of an existing service. In
the references, the publications are listed by project.

The Child Speech-Shaping and Behavior Lab at Florida

State University where parents (and graduate students and
colleagues) were taught to shape and reinforce skills of
severely impaired children. Created in 1964, it served a dozen
children and their mothers, and was turned over to Bill
Hopkins and Jack May in 1965. It was funded out-of-pocket
except for space. No publications were generated although
speech-shaping refinements made here were included in
Risley & Wolf, 1967.

The Turner House Language Development Preschool,

where welfare children attended a half-day preschool to
enhance their language development. Created with Betty Hart

521
in 1965, it served 200-250 children and closed in 1981 when
our longitudinal study of children’s everyday language
experience (Hart & Risley, 1992, 1995, 1999) began. It was
funded by NICHD grants. Seventeen publications were
generated.

The Parent Cooperative Preschool where welfare mothers

came to learn to teach their own and each other’s children.
Created in 1966, it served 50-60 mothers and their children
and was turned over to Rodney Hammond and Don Bushell in
1969. It was funded by OEO and NICHD grants. Three
publications were generated.

The Turner House Urban Recreation Project where we

assumed program responsibility for an existing very
problematic evening recreation center serving older children
and adolescents in a high-crime neighborhood. This
recreation center occupied the same building as our preschool
and we took it over more or less in self-defense. Begun with
Charles Pierce in 1970, it served several hundred children and
youth (about 40 any evening) and was relocated in 1971. It
was funded by the Episcopal Church and a NICHD graduate
student stipend. Two publications were generated.

The Day Care Environments Project where we assumed

management responsibility for the operation of two existing,
typically problematic day care centers serving children of the
working poor. Begun in 1970, it served about 150 children
and ended in 1973. It was funded by NICHD grants and
parent fees. Seven publications were generated.

The Juniper Gardens Tenant’s Association Project where

residents of a low-income, high-crime housing project were

522
organized to decide on community rules and operate a
community security patrol. Created in 1970 with Edward
Christophersen, it served about 1500 residents living in 420
apartments and ended in 1974. It was funded by NIMH
grants. Three publications were generated.

The Lawrence Infant Day Care Center, which provided

full day care for babies, 6-weeks-old to walking. Created in
1970, it served over 300 babies, 20 at any one time, and
closed in 1984. I started this program with my own son and
10 other babies and was the supervisory caregiver for much of
the first summer to design the program. It was funded by
USOE grants and parent fees. Five publications were
generated.

The Juniper Gardens Community Recreation Center that

was open each weekday after school for children from the
housing project and surrounding low-income, high-crime
neighborhood. Created in 1971 with Robert Quilitch, it served
over 500 3-16 year old children, about 50 on any one day,
across two years. It was funded by an NIMH grant and a
USOE graduate student stipend. Two publications were
generated.

The Nursing Home Activities Project that provided daily

leisure activities for the aged residents of a nursing home.
Begun with Lynn McClannahan in 1971, it served all of the
ambulatory residents in a 100-bed private nursing home for 2
years. It was funded by an NIMH grant and a USOE graduate
student stipend. Six publications were generated.

The Roadrunner Project that adapted the Infant Day Care

operations manual into the active-treatment day program in an

523
institution, for a group of people who were profoundly
retarded and non-ambulatory. Begun with Jim Favell in 1971
with a 2-week working retreat by me with the team of
graduate students and post-docs from my Kansas projects, the
project continued to serve the 16 residents of the Roadrunner
‘cottage’ until 1982. With Judy Favell, portions of the
Toddler Day Care manual were also adapted and
institution-wide quality assurance systems were developed to
improve services to the other 300 institutionalized residents
during that time. It was funded by the State of North Carolina
and NICHD grants. Ten publications were generated.

Guidelines for Behavioral Procedures in Programs for

Persons with Mental Retardation was a multi-year effort to
clarify behavioral procedures and their use, and suggest client
protection systems to enable State programs to confidently
use behavioral programming. This project started as a rational
response by a State Director of Mental Retardation to real and
rumored abuses in Florida institutions. In 1972, I participated
in a resident abuse investigation; then throughout 1973 I site
visited all Florida MR institutions; I then co-chaired and
orchestrated a blue-ribbon task force of behavioral and legal
experts to construct guidelines in 1974; and finally I designed
professional procedures for human rights and peer review
panels and coordinated their implementation from 1975-1980.
This effort was variously funded by the State of Florida, the
National Association of Retarded Citizens, and the
Association for the Advancement of Behavior Therapy. Eight
publications were generated.

The Lawrence Toddler Day Care Center that provided full

day care for children from walking to 30 months old. Created
with Mike Cataldo in 1972, it served about 150 toddlers, 20 at

524
any one time, and closed in 1986. It was funded by USOE and
Maternal and Child Health grants and parent fees. Eight
publications were generated.

The Johnny Cake Child Study Center; a residential

treatment program for dependent-neglected children with 3
group homes, a school, a recreation center, an office building,
four houses and 10 apartments for staff (and our own fire
engine) around a private lake in the mountains of Arkansas. A
private philanthropy of a wealthy man, I converted the
center’s primary mission into developing, testing and
disseminating child-rearing technology for the problems
normal middle-class families would be facing in the future. In
1973, Mike Cataldo, Rusty Clark, 2 other post-docs, 7
graduate students, and 11 employees were recruited to create
a large research program while serving 18 children in
residential care. When business reverses ended the
philanthropy of the sponsor, the Center was closed in 1976.
Four publications were generated.

The Nashville Police Operations Project that developed and

tested new procedures for problematic areas in the daily
duties of patrol officers. Begun with Robert Kirchner in 1974
as an adaptation of the mechanical monitoring developed for
the Juniper Gardens security patrol, the project expanded to
include six graduate students (including 2 police captains in
key positions) and one post-doc. The supervision, patrol, and
case preparation routines of about 500 police officers serving
a city of a half-million people were rationalized, refined,
manualized and implemented when the project ended in 1979.
It was funded by the Nashville Metropolitan police
Department and DOJ and DOT grants. Five publications were
generated.

525
The Nursing Home Operations Project where we assumed
responsibility for the daily care routines of a 100-bed private
nursing home and the meal service of a 100-bed county
nursing home; adapting the engagement measures and the
staff management and nutrition systems from our child day
care centers. Begun in 1975 with a one-week working retreat
by me and my students working as nursing aides, it served
about 300 fragile elderly people over four years. It was
funded by USPHS Health Services Research grants. Three
publications were generated.

Individualized services for People with Mental Health or

Developmental Disabilities where severely disturbed
children and people with developmental disabilities in a State
were provided access to care and treatment services designed
for each individual. The effort began in 1984 when I arranged
for some new DD funds to be targeted to serve difficult
clients in individual homes — to begin building Alaska’s
capacity to do without a MR institution. It advanced in 1985
with Karen Ward and Theda Ellis with a statewide effort to
convert all sheltered workshops into supported employment
programs. It advanced in 1986 with John VanDenBerg and
the Alaska Youth Initiative to bring back from out-of-state
institutions all Alaskan children and youth and one-by-one
create heavily supported homes and community lives for each
of them. In 1988, to protect and extend these initiatives I
temporarily became the Director of Alaska’s Division of
Mental Health and Developmental Disabilities. With Mike
Renfro approximately 300 people had been provided complex
services in individual arrangements when I returned to the
University in 1991 — and the State’s MR institution was
closed, on schedule, in 1996. This effort was funded with

526
NIMH, ADD, and VR grants and State of Alaska general
funds. Two publications were generated.

Alaska’s Autism Intensive Early Intervention Project

where young children with autism throughout Alaska are
provided state-of-the-art treatment in their homes by their
parents and teams of family volunteers. Created in 1993, the
project has trained the parents and volunteer teams, and
orchestrated the treatment of over 50 children. With Cheryl
Risley the program is now, finally, able to reliably ensure
full-dosage treatment (pervasive assertive parenting, plus 40
hours/week of direct instruction and incidental teaching for
two years, and a final year of self-control training). It is now
starting 6 new children per year into a 36-month treatment
regimen. Pilot work for national field trials are scheduled to
begin next year. (This will probably be the last service project
I will start.) It is funded by State of Alaska DD services,
USDOE and ADD grants. One publication has been generated
so far.

I live at ‘Risley Mountain,’ the homestead of four generations

of my family where I was born in 1937. My strongest
impression from childhood is of my mother’s long hours of
energetic toil at the tasks of homestead living without electric
lights, indoor plumbing and central heating. To wash a load of
clothes, she had to have hauled about 40 gallons of water
from the spring 200 yards behind the cabin — 10 gallons at a
time with a child’s wagon in the summer and a sled in the
winter. She had to have hand-sawn about 10 rounds of wood
off of logs skidded in by horse — and split them into
hand-size stove wood with a double-bitted axe. She heated the
water in copper boilers on the wood stove, and dipped it into
a washing machine that was powered by a gasoline engine

527
that crankstarted and ran somewhat dependably. With a stick,
she would fish out the washed clothes, garment by garment,
and feed each through the motor-driven wringer into a first
and then into a second tub of rinse water beside the washing
machine. After a final wring, she would clothespin each
garment on an outdoor clothesline and take it down when it
was dry or when it began to rain, whichever came first. In the
winter the clothes would freeze stiff as boards, but would
slowly dry anyway even if it snowed. (We used cloth diapers
then so I was probably toilet trained early.) My mother
weighed 90 pounds so her approach to the heavy labors of
homestead living was of small loads and many trips all day
long. (But mother also read for self-improvement every
evening by lamplight and talked to me of ‘bookish’ things.
With me in tow, she would stop by to see her “school marm”
friends whenever she arranged a ride into town for groceries,
or to the Sunday afternoon gathering of school teachers and
other educated people at the government doctor’s house to
listen to classical music on his victrola. And I would hear talk
of foreign things like wine or sidewalks and foreign lands like
France or California.)

Fortunately, when I became old enough to use a handsaw and

axe, we moved closer to town and had oil for heat and
propane for cooking; and I was liberated from hour-a-day
firewood chores. Next we acquired a small diesel-powered
‘light plant’ and had electric lights to read by; thus liberating
us from the daily chore of cleaning and servicing kerosene
lamps and Coleman lanterns. Then rural electrification came,
enabling indoor plumbing which liberated us from trips to a
cold outhouse; and liberated me from my duties of hand
pumping and carrying from the neighborhood well all the
water used by two households. (The neighborhood joke was

528
that we had ‘walking water’ that was slowly carried by a boy
with his nose in a book and his head in the clouds. My fourth
grade teacher wrote to me last year: “What a dreamer you
were. When we had your attention you did excellent work.”)

From direct experience I have a favorable view of invention

and technology and am less enthusiastic than most about the
benefits of getting back to the ‘simple’ life. The simple life
was simple because waking hours were consumed by hard,
thoughtless, simple chores. Since I was a child on a primitive
homestead, the world population has tripled — and more
people than ever live lives that are harder, more stupefying,
and more time demanding than my family’s homestead life.
And as populations grow, as people migrate and become
strangers, as economies shift and work skills become
obsolete, as physical inventions and technologies change the
nature of daily life, the need for new behavioral and social
inventions and technologies is accelerating.

In reality, we don’t know much about what people actually do

in their everyday lives. And in theory, we only know a little
about why they do what they do. But, we do have some tools
for finding out what people do and some tools to influence
what people do — and so we can invent some ways to help.
In the process we will learn and pass on some things that
apply to other people with other problems in other
circumstances — and thus add to our knowledge about what
people do, why they do it, and how to help. Do you have
enough tools to see reality clearly and change it for the better?
You will never know unless you try.

529
References
Turner House Language Development Preschool

Greenwood, C. R., Hart, B., Walker, D., & Risley, T. R.

(1994). The opportunity to respond revisited: A behavioral
theory of developmental retardation and its prevention. In R.
Gardner, III, D. M. Sainato, J. O. Cooper, T. E. Heron, W. L.
Howard, J. W. Eshleman, & T. A. Grossi (Eds.), Behavior
analysis in education: Focus on measurably superior
instruction. Pacific Grove, CA: Brooks Cole.

Hart, B. M., & Risley, T. R. (1968). Establishing use of

descriptive adjectives in the spontaneous speech of
disadvantaged preschool children. Journal of Applied
Behavior Analysis, 1, 109-120.

Hart, B. M., & Risley, T. R. (1974). Using preschool

materials to modify the language of disadvantaged children.
Journal of Applied Behavior Analysis, 7, 243-256.

Hart, B. M., & Risley, T. R. (1975). Incidental teaching of

language in the preschool. Journal of Applied Behavior
Analysis, 4, 411-420.

Hart, B. M., & Risley, T. R. (1976). Community-based

language training. In T. D. Tjossem (Ed.), Intervention
strategies for high risk infants and young children. Baltimore,
MD: University Park Press.

530
Hart, B. M., & Risley, T. R. (1978). Promoting productive
language through incidental teaching. Education and Urban
Society, 10, 407-429.

Hart, B. M., & Risley, T. R. (1980). In vivo language

intervention: Unanticipated general effects. Journal of
Applied Behavior Analysis, 13, 407-432.

Hart, B., & Risley, T. R. (1980). Incidental teaching of

language. Austin, Texas: PRO-ED.

Hart, B., & Risley, T. R. (1981). Grammatical and conceptual

growth in the language of psychosocially disadvantaged
children. In M. Begab, R. Barber, & C. Haywood (Eds.),
Psychosocial influences in retarded performance
(pp.181-198). Baltimore, MD: University Park Press.

Hart, B., & Risley, T. R. (1983). Incidental strategies. In R. L.

Schiefelbuch (Ed.), Communicative competence: Acquisition
and intervention. Baltimore, MD: University Park Press.

Reynolds, N. J., & Risley, T. R. (1968). The role of social and

material reinforcers in increasing talking of a disadvantaged
preschool child. Journal of Applied Behavior Analysis, 1,
253-262.

Risley, T. R. (1972). Spontaneous language and the preschool

environment. In J. C. Stanley (Ed.), Preschool programs for
the disadvantaged: Five experimental approaches to early
childhood education. Baltimore, MD: John Hopkins
University Press.

531
Risley, T. R. (1977). The social context of self-control. In R.
B. Stuart (Ed.), Behavioral self-management: strategies,
techniques and outcome. New York: Brunner/Mazel.

Risley, T. R. (1978). Behavior modification perspective and

bilingual/bicultural education models. Bilingual Resources, 1,
8-10.

Risley, T. R., & Hart, B. (1968). Developing correspondence

between the non-verbal and verbal behavior of preschool
children. Journal of Applied Behavior Analysis, 1, 267-281.

Risley, T. R., Hart, B., & Doke, L. A. (1971). Operant

language development: The outline of a therapeutic
technology. In R. L. Schiefelbusch (Ed.), The language of the
mentally retarded. Baltimore, MD: University Park Press.

Risley, T. R., & Reynolds, N. J. (1970). Emphasis as a

prompt for verbal imitation. Journal of Applied Behavior
Analysis, 3, 185-190.

Parent Cooperative Preschool

Jacobson, J. M., Bushell, D. B., Jr., & Risley, T. R. (1969).

Switching requirements in a Head Start classroom. Journal of
Applied Behavior Analysis, 2, 43-47.

Risley, T. R. (1968, January). Jenny Lee: Learning and

lollipops. Psychology Today, 25.

Risley, T. R., Reynolds, N. J., & Hart, B. (1970). The

disadvantaged: Behavior modification with disadvantaged
preschool children. In R. Bradfield (Ed.), Behavior

532
modification: The human effort. Palo Alto: Science and
Behavior Books.

Turner House Urban Recreation Project

Pierce, C. H., & Risley, T. R. (1974). Recreation as a

reinforcer: Increasing membership and decreasing disruptions
in an urban recreation center. Journal of Applied Behavior
Analysis, 7, 403-411.

Pierce, C. H., & Risley, T. R. (1974). Improving job

performance of neighborhood youth corps aides in an urban
recreation program. Journal of Applied Behavior Analysis, 7,
204-215.

Day Care Environments Project

Doke, L. A., & Risley, T. R. (1972). The organization of day

care environments: Required versus optional activities.
Journal of Applied Behavior Analysis, 5, 405-420.

Krantz, P., & Risley, T. R. (1977). Behavioral ecology in the

classroom. In K. D. O’Leary & S. G. O’Leary (Eds.),
Classroom management: The successful use of behavior
modification (2nd ed.). New York: Pergamon Press.

LeLaurin, K., & Risley, T. R. (1972). The organization of day

care environments: “Zone” versus “man-to-man” staff
assignments. Journal of Applied Behavior Analysis, 5,
225-232.

533
Montes, F., & Risley, T. R. (1975). Evaluating traditional day
care practices: An empirical approach. Child Care Quarterly,
4, 208-215.

Risley, T. R., & Twardoz, S. (1976). The preschool as a

setting for behavioral intervention. In H. Leitenberg (Ed.),
Handbook of behavior modification and therapy). Englewood
Cliffs, NJ: Prentice Hall.

Risley, T. R. (1977). The ecology of applied behavior

analysis. In A. Rogers-Warren & S. Warren (Eds.),
Ecological perspectives in behavior analysis. Baltimore, MD:
University Park Press.

Twardosz, S., & Risley, T. R. (1982). Behavioral-ecological

consultation to day care centers. In A. Jager & R. Slotnick
(Eds.), Community mental health and behavioral ecology.
New York: Plenum Press

Juniper Gardens Tenant’s Association Project

Chapman, C., & Risley, T. R. (1974). Anti-litter procedures in

an urban high-density neighborhood. Journal of Applied
Behavior Analysis, 7, 377-383.

Christophersen, E. R., Doke, L. A., Messmer, D. O., &

Risley, T. R. (1975). Measuring urban problems: A brief
report on rating grass coverage. Journal of Applied Behavior
Analysis, 8, 230.

Kloss, J. D., Christophersen, E. R., & Risley, T. R. (1976). A

behavioral approach to supervision. Security Management,
20, 48-49.

534
Lawrence Infant Day Care Center

Cataldo, M. D., & Risley, T. R. (1974). Infant day care. In R.

Ulrich, T. Stachnik & J. Mabry (Eds), Control of human
behavior (Vol. 3). Glenview IL: Scott Foresman.

Herbert-Jackson, E., O’Brien, M., Porterfield, J., & Risley, T.

R. (1977). The infant center: A complete guide to organizing
and managing infant day care. Baltimore, MD: University
Park Press.

Risley, T. R. (1975). Day care as a strategy for social

intervention [Introduction]. In E. Ramp & G. Semb (Eds.),
Behavior analysis: Areas of research and application.
Englewood Cliffs, NJ: Prentice Hall.

Twardosz, S., Cataldo, M. F., & Risley, T. R. (1974). Infants’

use of crib toys. Young Children, 29, 271-276.

Twardosz, S., Cataldo, M. F., & Risley, T. R. (1974). Open

environment design for infant and toddler day care. Journal of
Applied Behavior Analysis, 7, 529-546

Juniper Gardens Community Recreation Center

Quilitch, H. R., & Risley, T. R. (1973). The effects of play

materials on social play. Journal of Applied Behavior
Analysis, 6, 573-578.

Quilitch, H. R., Christophersen, E. R., & Risley, T. R. (1977).

The evaluation of children’s play materials. Journal of
Applied Behavior Analysis, 10, 401-502.

535
Nursing Home Activities Project

McClannahan, L. E. (1973). Therapeutic and prosthetic living

environments for nursing home residents. Gerontologist, 1,
33-41.

McClannahan, L. E. (1973). Recreation programs for nursing

home residents: The importance of patient characteristics and
environmental arrangements. Therapeutic Recreation, 2,
26-31.

McClannahan, L. E., & Risley, T. R. (1973, June). A store for

nursing home residents. Nursing Homes, pp. 8-12.

McClannahan, L. E., & Risley, T. R. (1974). Design of living

environments for nursing home residents: Recruiting
attendance at activities. Gerontologist, 14, 236-240.

McClannahan, L. E., & Risley, T. R. (1975). Design of living

environments for nursing home residents: Increasing
participation in recreational activities. Journal of Applied
Behavior Analysis, 8, 261-268.

McClannahan, L. E., & Risley, T. R. (1975). Activities and

materials for severely disabled geriatric patients. Nursing
Homes, 24, 10-13.

Roadrunner Project

Cataldo, M. F., & Risley, T. R. (1974). Evaluation of living

environments: the MANIFEST description of ward activities.
In P. O. Davidson, F. W. Clark, & L. A. Hamerlynck (Eds.),

536
Evaluation of social programs in community, residential and
school settings. Champaign IL: Research Press.

Favell, J. E., Risley, T. R., Wolfe, A. F., Riddle, J. I., &

Rasmussen, P. R. (1981). The limits of habilitation: How can
we identify them and how can we change them? Analysis and
Intervention in Developmental Disabilities, 1(1), 37-43.

Favell, J. E., & Cannon, P. R. (1976). Evaluation of

entertainment materials for severely retarded persons.
American Journal of Mental Deficiency, 81, 357-361.

Favell, J. E., Favell, J., Riddle, J. L., & Risley, T. R. (1983).

Promoting change in mental retardation facilities: Getting
services from the paper to the people. In W. P. Christian, J.
Hannah & T. J. Glahn (Eds.), Programming effective human
services. New York: Plenum

Hart, B., & Risley, T. R. (1976). Environmental

programming: Implications for the severely handicapped. In
H. J. Prehm & S. J. Deitz (Eds.), Early intervention for the
severely handicapped: Programming and accountability
[Monograph No. 2]. University of Oregon: Severely
Handicapped Learner Program.

Jones, M. L., Favell, J. E., & Risley, T. R. (1983).

Socioecological programming of the mentally retarded. In J.
L. Matsen & F. Andrasik (Eds.), Treatment issues and
innovations in mental retardation. New York: Plenum Press.

Jones, M. L., Risley, T. R., Favell, J. E. (1983). Ecological

patterns. In J. L. Matson, & S. E. Breuning (Eds.), Assessing
the mentally retarded. New York: Grune & Stratten.

537
Jones, M. L., Favell, J. E., Lattimore, J., & Risley, T. R.
(1984). Improving independent engagements of
nonambulatory multihandicapped persons through systematic
analysis of leisure materials. Analysis and Intervention in
Developmental Disabilities, 4, 313-332.

Jones, M. L., Lattimore, J., Ulicny, G., & Risley, T. R.

(1985). Programming for engagement. Environmental design.
In R. P. Barret (Ed.), Treatment of severe behavioral
disorders: Contemporary approaches with the mentally
retarded. New York: Plenum.

Lattimore, J., Stephens, T. E., Favell, J. E., & Risley, T. R.

(1984, April). Increasing direct care staff compliance to
individualized physical therapy body positioning
prescriptions: Prescriptive checklists. Mental Retardation,
22(2), 79-84.

Guidelines for Behavioral Procedures in State Programs for

Persons with Mental Retardation

Azrin, N. H., Risley, T. R., Stuart, R. B., & Stolz, S. B.

(1977). Ethical issues for human services. Behavior Therapy,
8, v-vi.

Favell, J., Favell, J. E., Riddle, J. I., & Risley, T. R. (1981). A

quality-assurance system for ensuring client rights in mental
retardation facilities. In G. T. Hannah, W. P. Christian, & H.
B. Clark (Eds.), Preservation of client rights: A handbook for
practitioners providing therapeutic, educational, and
rehabilitative services (pp. 345-346). New York: The Free
Press Publishing Co.

538
Favell, J. E., & Risley, T. R., et al. (1982). The treatment of
self-injurious behavior (AABT Task Force Report.) Behavior
Therapy, 13, 529-554.

May, J. G., Jr., Risley, T. R. Twardosz, S., Friedman, P.,

Bijou, S. W., Wexler, D., et al. (1976). Guidelines for the use
of behavioral procedures in state programs for retarded
persons. Arlington Texas: National Association for Retarded
Citizens.

Risley, T. R. (1975). Certify procedures not people. In W. S.

Wood (Ed.), Issues in evaluating behavior modification.
Champaign, IL: Research Press.

Risley, T. R., & Sheldon-Wildgen, J. (1980). Suggested

procedures for human rights committees of potentially
controversial treatment programs. The Behavior Therapist,
3(2), 9-10.

Risley, T. R., & Sheldon-Wildgen, J. (1982). Invited peer

review: The AABT experience. Professional Psychology.

Sheldon-Wildgen, J., & Risley, T. R. (1983). Balancing

clients’ rights: Establishing human rights and peer review
committees. In A. Bellack, M. Herson, & A. Kazdin (Eds.),
International handbook of behavior modification. New York:
Plenum Press.

Lawrence Toddler Day Care

Herbert-Jackson, E., Cross, M. Z., & Risley, T. R. (1977).

Milk types and temperatures — what will young children
drink? Journal of Nutrition Education, 9, 76-79.

539
Herbert-Jackson, E., & Risley, T. R. (1977). Behavioral
nutrition: consumption of foods of the future by toddlers.
Journal of Applied Behavior Analysis, 10, 407-413.

O’Brien, M., Porterfield, J., Herbert-Jackson, E., & Risley, T.

R. (1979). The toddler center: A practical guide to day care
for one and two-year olds. Baltimore, MD: University Park
Press.

O’Brien, M., Herbert-Jackson, E., & Risley, T. R. (1978;

1979). Menus for toddlers in day care: A toddler taste test
turns up nutritious foods suited to young preschoolers [Four
part series]. Day Care and Early Education, 6(1), 48-53; 6(2),
49-54; 6(3), 49-54; 6(4), 48-53.

O’Brien, M., Houston, A. C., & Risley, T. R. (1983). Sex

typed play of toddlers in a day care center. Journal of Applied
Developmental Psychology, 4, 19.

O’Brien, M., & Risley, T. R. (1983). Infant-toddler day care:

practical considerations and applications to children with
special needs. In E. M. Goetz & K. E. Allen (Eds.), Early
childhood education. Rockville, MD: Aspen Systems.

Porterfield, J. K., Herbert-Jackson, E., & Risley, T. R. (1976).

Contingent observation: An effective and acceptable
procedure for reducing disruptive behaviors of young children
in group settings. Journal of Applied Behavior Analysis 9,
55-64.

Twardosz, S., Cataldo, M. F., & Risley, T. R. (1975). Menus

for toddler day care: Food preference and spoon use. Young
Children, 30, 129-144.

540
Johnny Cake Child Study Center

Clark, H. B., Green, B. F., Macrae, J. W., McNees, M. P.,

Davis, J. L., & Risley, T. R. (1977). A parent advice package
for family shopping trips: Development and evaluation.
Journal of Applied Behavior Analysis, 10, 605-624.

Dineen, J. P., Clark, H. B., & Risley, T. R. (1977). Peer

tutoring in elementary students: Educational benefits to the
tutor. Journal of Applied Behavior Analysis, 10, 231-238.

Greene, B. F., Clark, H. W., & Risley, T. R. (1978). Shopping

with children: Advice for parents. San Rafael, CA: Academic
Therapy Publications.

Risley, T. R., Clark, H. B., & Cataldo, M. F. (1976).

Behavioral technology for the normal middle-class family. In
E. J. Mash, L. C. Handy, & L. A. Hamerlynck (Eds.),
Behavior modification and families (pp. 34-60). New York:
Brunner/Mazel.

Nashville Police Operations Project

Carr, A. F., Larson, L. D., Schnelle, J. F., Kirchner, R. E., &

Risley, T. R. (1980). Effective police field supervision: A
report writing evaluation program. Journal of Police Science
and Administration, 8(2), 212-219.

Currey, G. H., Carr, A. F., & Schnelle, J. (1979). Juvenile

warning citations: A diversion from Juvenile Court. FBI Law
Enforcement Bulletin, 48 (12).

541
Domash, M. A., Schnelle, J. F., Stromatt, E. L., Carr, A. F.,
Larson, D., Kirchner, R. R., & Risley, T. R. (1980). Police
and prosecution systems: An evaluation of a police criminal
case preparation program. Journal of Applied Behavior
Analysis, 13, 397-406.

Larson, L. D., Schnelle, J. F., Kirchner, R. E., Carr, A. F.,

Domash, M. A., & Risley, T. R. (1980). Reduction of police
vehicle accidents through mechanically-aided supervision.
Journal of Applied Behavior Analysis, 13, 571-582.

McNees, M. P., Egli, D. S., Marshall, R. S., Schnelle, J. F., &

Risley, T. R. (1976). Shoplifting prevention: Providing
information through signs. Journal of Applied Behavior
Analysis, 9, 399-405.

Nursing Home Operations Project

Risley, T. R. (1978). Toward a system of nursing home

organization and management. In T. Glynn & S. McNaughton
(Eds.), Behavior Analysis in New Zealand (pp. 1-25).
Auckland, New Zealand: University of Auckland.

Spangler, P. F., Risley, T. R., & Bigelow, D. D. (1984). The

management of dehydration and incontinence in
nonambulatory geriatric patients. Journal of Applied
Behavioral Analysis, 17, 397-401.

Traughber, B., Erwin, K. E., Schnelle, J. F., & Risley, T. R.

(1983). Behavioral Nutrition: An evaluation of a simple
system for measuring food and nutrient consumption.
Behavioral Assessment, 5, 263-280.

542
Individualized Services for People with Mental Health or
Developmental Disabilities

MacFarquhar., L. W., Dowrick, P. W., & Risley, T. R.

(1993). Individualizing services for seriously emotionally
disturbed youth: A nationwide survey. Administration and
Policy in Mental Health, 20(3), 165-174.

Risley, T. R., (1996). Get a Life! Positive behavioral

intervention for challenging behavior through life
arrangement and life coaching. In L. K. Koegel, R. L. Koegel,
& G. Dunlap (Eds.), Positive behavioral support: Including
people with difficult behavior in the community. Baltimore,
MD: Paul Brookes.

Alaska’s Autism Intensive Early Intervention Project

Risley, T. R. (1997). Family preservation for children with

autism. Journal of Early Intervention, 21, 15-16.

Other References Cited

American Academy of Pediatrics, Committee on

Psychosocial Aspects of Child and Family Health (1998).
Guidance for effective discipline. Pediatrics, 101, 723-728.

Ayllon, T. (1959). The application of reinforcement theory

toward behavior problems. Unpublished doctoral dissertation,
University of Houston.

Ayllon, T. (1963). Intensive treatment of psychotic behavior

by stimulus satiation and food reinforcement. Behavior
Research and Therapy, 1,53-61.

543
Ayllon, T., & Azrin, N. H. (1964). Reinforcement and
instructions with mental patients. Journal of the Experimental
Analysis of Behavior, 7, 327-331.

Ayllon, T., & Haughton, E. (1962). Control of the behavior of

schizophrenic patients by food. Journal of the Experimental
Analysis of Behavior, 5, 343-352.

Ayllon, T., & Michael, J. (1959). The psychiatric nurse as a

behavioral engineer. Journal of the Experimental Analysis of
Behavior, 2, 323-334.

Azrin, N. H. (1977). A strategy for applied research: Learning

based but outcome oriented. American Psychologist, 32,
140-149.

Baer, D. M., Wolf, M. M., & Risley, T. R. (1968). Some

current dimensions of applied behavior analysis. Journal of
Applied Behavior Analysis, 1, 91-97.

Campbell, D. T. (1957). Factors relevant to the validity of

experiments in social settings. Psychological Bulletin, 54,
297-312.

Campbell, D. T. (1969). Reforms as experiments. American

Psychologist, 24, 409-429.

Campbell, D. T., & Stanley, J. C. (1963). Experimental and

quasi-experimental designs for research. Chicago: Rand
McNally.

Hart, B., & Risley, T. R. (1992). American parenting of

language-learning children: Persisting differences in

544
family-child interactions observed in natural home
environments. Developmental Psychology, 28(6), 1096-1105.

Hart, B., & Risley, T. R. (1995). Meaningful differences in

the everyday experiences of young American children.
Baltimore, MD: Paul Brookes.

Hart, B., & Risley, T. R. (1999). The social world of children

learning to talk. Baltimore, MD: Paul Brookes.

Kazdin, A. (1992). Personal communication.

Risley, T. R. (1964). Generalization gradients following

two-response discrimination training. Journal of the
Experimental Analysis of Behavior, 7, 199-204.

Risley, T. R. (1997). Montrose M. Wolf: The origin of the

dimensions of Applied Behavior Analysis. Journal of Applied
Behavior Analysis, 30, 377-381.

Risley, T. R., & Wolf, M. M. (1964). Experimental

manipulation of autistic behaviors and generalization into the
home. Paper read at American Psychological Association,
Los Angeles.

Risley, T. R., & Wolf, M. M. (1967). Establishing functional

speech in echolalic children. Behaviour Research and
Therapy, 5, 73-88.

Sidman, M. (1960). Tactics of scientific research. New York:

Basic Books.

Skinner, B. F. (1948). Walden two. New York: Macmillan.

545
Skinner, B. F. (1953). Science and human behavior. New
York: Macmillan.

Wacker, D. P. (1998). Editorial. Journal of Applied Behavior

Analysis, 31, 511.

Wolf, M. M., Risley, T. R., Johnston, M., Harris, F., & Allen,
E. (1967). Applications of operant conditioning procedures to
the behavior problems of an autistic child: A follow-up and
extension. Behaviour Research and Therapy, 5, 103-111.

Wolf, M. M., Risley, T. R., & Mees, H. L. (1964).

Applications of operant conditioning procedures to the
behavior problems of an autistic child. Behaviour Research
and Therapy, 1, 305-312.

Ullman, L. P., & Krasner, L. (Eds). (1965). Case studies in

behavior modification. New York: Holt, Rinehart & Winston.

Ulrich, R., Stachnik, T., & Mabry, J. (Eds.). (1966). Control

of human behavior. Glenview, IL: Scott Foresman.

546
Chapter 16

Application of Operant
Conditioning Procedures to
the Behavior Problems of an
Autistic Child: A 25-Year
Follow-Up and the
Development of the Teaching
Family Model
Montrose M. Wolf

University of Kansas

The first thing that I would like to describe is how I first got
into the field of behavioral psychology. Jack Michael arrived
at the University of Houston in 1957. The first description of
Jack by a fellow student was, “He obviously can’t be a very
good teacher because he dresses so badly.” I had also heard
the student rumor that Jack had been asked to leave Kansas
University because he was teaching Skinner. Then Jack
arrived in his strangely colored short pants and flip-flop
sandals. And, to make matters worse, Jack announced that the
course would concentrate on animal research! I began looking
through my timetable for a course more in line with my

547
neo-Freudian interests. But, Learning Theory was a required
course for psychology majors. However, I did arm myself
with drop slips in case the course became too painful. I was
already thinking about changing my major to philosophy,
anyway.

Jack, rather than choosing something interesting like Carl

Jung, assigned Keller and Schoenfeld’s Principles of
Psychology (1950) as our textbook. The first few chapters did
turn out to be fascinating. By the third chapter I was
converted. I converted not because they had all the answers.
But, because their natural science approach seemed to hold
the greatest promise among the psychological systems for
achieving the answers. In addition, Jack was also excited
about the application of behavioral principles to important
human problems. Jack described Ted Ayllon’s pioneering
research at Saskatchewan Hospital. At Jack’s suggestion, I
read Skinner’s newest book Science and Human Behavior
(1953) which made a strong case for the application of
behavioral principles to human problems.

I described my conversion to my girlfriend, Sandra, who was

also a psychology major. She was horrified! But, she sat in on
some lectures and became a convert, too.

Others became Jack’s students. Ted Ayllon published with

Jack the pioneering applied behavioral study: The Psychiatric
Nurse as a Behavioral Engineer (JEAB, 1959). That paper
was the model for applied research. It showed the use of
mediators, in this case the Psychiatric Nurses, who were
trained to apply the behavioral principles, working with the
behavior analyst in the psychiatric institution. And, how
mediators could be used in other settings like schools and

548
institutions for other populations. Finally, others also became
members of the group which met in Jack’s living room and
planned how to save the world through behaviorism,
including John Mabry, Pat Corke, Leland Johnson, Jerry
Short, Lee Meyerson, Nancy Kerr, Lloyd and Polly Brooks
(both deceased), and Sam Toombs (deceased).

Jack shared his library with us, so we began reading more

Skinner and the early issues of JEAB. We helped to celebrate
the approval by APA of Division 25.

Jack also generously shared his research equipment. I built a

rat box to study the effects of combined discriminative
stimuli. Jack let us use the equipment at night. We were
responsible for rewiring the equipment so it would be ready
for Jack’s research the next day. Unfortunately, there were
times when we didn’t rewire the equipment correctly. Jack
would make cracks about our long line of serendipitous
experiments.

One day Jack got a call from friends at Arizona State

University. Their plan was to establish ASU as the first
psychology department specializing in behavioral research,
theory, and application. Some of us went to ASU as graduate
students and helped to set up Fort Skinner in the desert.
Unfortunately, the new chairperson began receiving criticism
from other departmental heads about the ASU department
being too narrow. The new chairperson began referring to us
students as poorly trained relay raconteurs.

At ASU, during the good days, we had lots of exciting

courses. For example, we got to read neat stuff like Bijou and
Baer’s Child Development (1961).

549
We also read a very interesting theoretical analysis by Charles
Ferster (1961). He discussed how the behavior problems of
childhood autism might be due to subtle positive
reinforcement, punishment, and extinction contingencies.
Ferster and DeMyer (1961) also published a very important
laboratory study about autistic children. They showed that
neutral stimuli could be conditioned in the laboratory.

We also read about the fascinating human operant research

taking place in institutions. Few had made the leap from the
lab to the other side of the one-way glass or to schools or to
homes. In fact, some were of the opinion that such a leap was
premature and unwise because we didn’t know enough, that
we needed to wait for more basic human operant research.

In 1962, Sidney Bijou visited ASU. He described to us his

interest in taking this leap. Sid described his exciting
programs and opportunities. For example, to teach a
behavioral course to pre-school teachers who were asked to
convert from a psycho-dynamic to a behavioral orientation.
And, a research opportunity to work with Jay Birnbrauer on a
new behavioral classroom at Rainier State School for the
developmentally disabled. We described to Sid the token
economy research that we had carried out with pre-school
children (Staats, Staats, Schultz, & Wolf, 1962).

So, in 1962, Sid hired me on his NIMH grant. I arrived at the

University of Washington on the first day of July. I had been
there less than two weeks, when Sid called Todd Risley, a
graduate student, and me into his office. Sid had just
completed a conversation with Jerman Rose, the Director of
the Childrens’ Psychiatric Hospital at Fort Steilacoom,

550
Washington. Jerman Rose wanted us to work with a child
with the following characteristics.

A list of Dicky’s Characteristics:

• He was 3 1/2 years old.

• He had developed normally until about nine months
old.
• He was diagnosed as autistic.
• He developed a high rate of self-destructive temper
tantrums (“He was a mess, all black and blue and
bleeding.” His parents reported after a typical
tantrum.)
• No normal language (Echolalic).
• Not toilet trained.
• Did not go to bed at night. (Tantrumed unless a
parent stayed up with him).
• Did not eat at the table (Grazed).
• Because his problems were so severe, permanent
institutionalization had been recommended to the
parents.
• Cataracts had been removed from the lenses of both
eyes.
• The ophthalmologist predicted that unless he begins
wearing glasses within the next six months he would
permanently lose his macular vision.
• He had been hospitalized for three months and they
had not taught him to wear his glasses.
No one had used behavioral clinical procedures before with
autistic children. So, this looked like a tough case. Especially
in the hostile environment of a psychoanalytic childrens’
hospital. Todd and I pointed out these concerns to Sid. But,

551
Sid kept replying that we should wait and see what was up
before making a final decision.

And Sid was right, of course (Wolf, Risley, & Mees, 1964;
Wolf, Risley, Johnston, Harris, & Allen, 1967).

You may be wondering what happened to the pre-school

teachers in the behavioral course that I taught? Well, it went
better than you might expect. Todd and I began our weekly
trips to Dicky’s hospital that summer. So, we took the
pre-school teachers and others on our weekly trips and
presented Dicky’s data at every class session. The teachers
began looking for opportunities to replicate the Dicky study
by behaviorally analyzing the normal behavior problems of
the pre-school children in their classrooms. They began with
“regressed” crawling behavior (Harris, Johnston, Kelly, &
Wolf, 1964). And, then they moved to isolate behavior
(Allen, Hart, Buell, Harris, & Wolf, 1964), operant crying
(Hart, Allen, Buell, Harris, & Wolf, 1964), and motor skills
(Johnston, Kelly, Harris, & Wolf, 1966).

Because the teachers made the conversion we were able to

provide Dicky with a strong behavioral treatment program for
the next two years which made him ready for special
education classes at school. And, we were later informed that
he went into regular classes. He graduated from high school
as well.

His IQ also changed. During his early years he was

untestable. A Washington psychologist sent us a copy of a
report in 1985 when Dicky was 26 years old that included an
IQ test. According to that psychologist Dicky’s WAIS Verbal
IQ score was 98.

552
Todd has visited with Dicky in person and on the phone.
Since high school he has lived independently and has had a
series of jobs.

You may be wondering what happened at the new behavioral

classroom at Rainier State School for the developmentally
disabled with Jay Birnbrauer? It was also exciting. We were
able to set up a powerful demonstration program (Birnbrauer,
Bijou, Wolf, & Kidder, 1965; Birnbrauer, Wolf, Kidder, &
Tague, 1965).

We can draw some conclusions about the impact of these

studies. First, time-out has been widely disseminated.
According to Hart and Risley (personal communication),
about half the teachers and parents in their study use time-out
as a non-violent disciplinary procedure. Furthermore, the
American Academy of Pediatrics published an article in the
journal Pediatrics encouraging pediatricians to recommend
that parents use ‘time outs’ and positive reinforcement instead
of spanking when children misbehave (Pediatrics, 1998).

Finally, the behavioral autism treatment program has been

widely adapted by others.

The Development of the

Teaching-Family Model
As one of the founders of the Teaching-Family model, I was
honored for our team’s contribution to the success of Boys
Town’s programs with the Boys Town’s Father Flanagan
Award for Service to Youth, 1996. We were also honored by
our recent awards from The Society for the Advancement of

553
Behavior Analysis Award for Distinguished Service to
Behavior Analysis for 1998, and the American Psychological
Association, Division 25, Award For Outstanding Applied
Research — For innovative and important research on
applications of behavioral principles to address socially
significant human behavior, 1998.

We have described the history of the model’s development in

a couple of recent articles. One description appears in Wolf,
Kirigin, Fixsen, Blase, & Braukmann (1995):

A case study in program development and refinement is

presented. We describe the Teaching-Family model and its
history, the original research goal of developing a
community-based program that was more humane, more
effective in teaching community-living skills, and less
expensive than the traditional large state institutions prevalent
when we began. We present the research on the components
of the model and the outcome research on the complete
model.

We share the serious problems that occurred when we

attempted to replicate the program in other communities. We
argue that the subjective consumer feedback questionnaire
(and the other components of the comprehensive quality
refinement system that have evolved over the past 25 years)
have played an important role in the survival and success of
the model. We recommend that people interested in
increasing the quality and survival rates of their human
services programs may want to consider developing a similar
technology driven by systematic reciprocal feedback from
consumers and line staff instead of relying on unplanned
consumer and staff feedback, as many programs do now.

554
Such feedback helps us to continue improving the quality of
the always evolving Teaching-Family model (p. 11-12).

References
Allen, K. E., Hart, B., Buell, J. S., Harris, F. R., & Wolf, M.
M. (1964). A study of the use of reinforcement principles in a
case of “isolate” behavior. Child Development, 35, 511-518.

Ayllon, T., & Michael, J. (1959). The psychiatric nurse as a

behavioral engineer. Journal of the Experimental Analysis of
Behavior, 2, 323-334.

Bijou, S. W., & Baer, D. M. (1961). Child development: A

Systematic and Empirical Theory (Vol. 1). New York:
Appleton-Century-Crofts.

Birnbrauer, J. S., Bijou, S. W., Wolf, M. M., & Kidder, J. D.

(1965). Programmed instruction in the classroom. In L. P.
Ullmann & L. Krasner (Eds.), Case studies in behavior
modification (pp. 358-363). New York: Holt, Rinehart, &
Winston.

Birnbrauer, J. S., Wolf, M. M., Kidder, J. B., & Tague, C. E.

(1965). Classroom behavior of retarded pupils with token
reinforcement. Journal of Experimental Child Psychology, 2,
219-235.

Ferster, C. B. (1961). Positive reinforcement and the

behavioral deficits of autistic children. Child Development,
32, 437-456.

555
Ferster, C. B., & DeMyer, M. K. (1961). The development of
performances in autistic children in an automatically
controlled environment. Journal of Chronic Diseases, 13,
312-345.

Hart, B. M., Allen, K. E., Buell, J. S., Harris, F. R., & Wolf,
M. M. (1964). Effects of social reinforcement on operant
crying. Journal of Experimental Child Psychology, 1,
145-153.

Harris, F. R., Johnston, M. K., Kelly, C. S., & Wolf, M. M.

(1964). Effects of positive social reinforcement on regressed
crawling of a nursery school child. Journal of Educational
Psychology, 55, 35-41.

Johnston, M. K., Kelly, C. S., Harris, F. R., & Wolf, M. M.

(1966). An application of reinforcement principles to the
development of motor skills of a young child. Child
Development, 37, 370-387.

Keller, F. S., & Schoenfeld, W. N. (1950). Principles of

psychology: A systemic text in the science of behavior. New
York: Appleton-Century-Crofts.

Skinner, B. F. (1953). Science and human behavior. New

York: Macmillan.

Staats, A. W., Staats, C. K., Schultz, R. E., & Wolf, M. M.

1961). The conditioning of textual responses using “extrinsic”
reinforcers. Journal of the Experimental Analysis of Behavior,
5, 33-40.

556
Wolf, M. M., Kirigin, K. A., Fixsen, D. L., Blase, K. A., &
Braukmann, C. J. (1995). The Teaching-Family model: A
case study in data-based program development and
refinement (and dragon wrestling). Journal of Organizational
Behavior Management, 15 (11-68).

Wolf, M. M., & Risley, T. R. (1965). Application of operant

conditioning procedures to the behavior problems of an
autistic child. Behavior Research and Therapy, 1, 302-312.

Wolf, M. M., Risley, T. R., & Mees, H. (1964). Application

of operant conditioning procedures to the behaviour problems
of an autistic child. Behavior Research and Therapy, 1,
305-312.

Wolf, M. M., Risley, T. R., Johnston, M. K., Harris, F. R., &

Allen, K. E. (1967). Application of operant conditioning
procedures to the behavior problems of an autistic child: A
follow-up and extension. Behavior Research and Therapy, 5,
103-111.

557
Chapter 17

The Active Unconscious,

Symptom Substitution, &
Other Things That Went
‘Bump’ in the Night
Gordon L. Paul

University of Houston

Introduction and Overview

The title of this chapter signifies theoretical constructs that
were relevant as my thinking gradually shifted from a
psychoanalytic orientation to a behavioral point of view. The
“active unconscious,” “symptom substitution,” and “other
things that went ‘bump’ in the night” all refer to concepts and
principles that I once believed were the way to understand
human behavior and the best ways to go about changing it.
They also reflect beliefs that had to be painfully abandoned in
the face of firm evidence of their spuriousness and lack of
utility.

Even though I now use behavioral conceptualizations and

behavior therapy techniques, I do not call myself a behavior
therapist. The functional analysis of behavior is a major part

558
of what I now do, but I do not call myself a behavior analyst.
The fact is that I abhor the categorical restrictions, frequent
cult-like trappings, and pigeonholing of any “school”
approach. This position has been personally costly in many
ways. For example, I belong to neither the Association for the
Advancement of Behavior Therapy (AABT) nor the
Association for Behavior Analysis (ABA)—and some of my
most respected friends have been key players in these
organizations. Beyond missing the camaraderie, my failure to
regularly network through such groups has resulted in fewer
people being familiar with my work and that of my coworkers
than otherwise would have been the case.

On the other hand, the lack of affiliation with any doctrinaire

school has been professionally valuable for both research and
clinical practice. It has allowed freedom to easily move
between the classical and instrumental sides of the behavioral
fence. It has allowed me to rationally select or combine
principles and procedures with the greatest utility for the
problems I address at any given time, including those
problems that are called cognitive ones.

This chapter fulfills the editors’ request for coverage of four

topics: (1) an intellectual autobiography; (2) observations of
developments during the “behavioral revolution;” (3) an
historical case study of one of my publications; and (4)
reflections on possible object lessons for the future. Rather
than a linear sequence, the topics are blended as I describe a
personal odyssey that combines reconstruction of my
developmental autobiography with observations of my own
paradigm shift.

559
For more than 30 years my work has emphasized
institutionalized populations, observational assessment, and
treatment of psychoses (e.g., Mariotto, Paul, & Licht, 1995;
Paul, 1987b; Paul, 2000; Paul & Lentz, 1977, 2001). These
efforts have drawn more heavily on the work of people
identified with applied behavior analysis than did my earliest
work on anxiety-related problems.

Rather than focus on more recent work, I selected for the case
study the publication reporting an examination of insight
versus desensitization in psychotherapy (Paul, 1966). The
investigation was completed for my Ph.D. dissertation in
1964. This publication was chosen because of its
developmental importance for me as well as for the field.
Following the case study of the selected publication, I return
to more autobiography with developmental significance
before offering reflections and recommendations based on
these experiences.

Intellectual Autobiography

Official Academic Genealogy

My official academic genealogy is presented in Figure 1. This
is a lineage of major professors of which anyone could be
proud — tracing back to William James. Surprisingly, these
academic ancestors all demonstrate the approach to science
that Nickles (2001) ascribes to James. I view Nickles’ five
points of contrast as continua of relative focus rather than
mutually exclusive categories, but the group does share the
following characteristics. All emphasized problem solving

560
over abstract truth seeking as a goal. All emphasized
pragmatic over foundational-epistemological selection of
problems for study. All made heavy use of heuristic rather
than just epistemic appraisal of knowledge. All showed a
forward-looking prospective focus rather than a
backward-looking historical one. Science for this line of my
academic ancestry was clearly an enterprise that should make
a difference to societal problems and to the lives of real
people. It is for me as well.

Figure 1. Official academic genealogy.

For his fifth point of contrast, Nickles (2001) also ascribes to

James an emphasis on case-based thought and practice over
thoroughgoing rule-based inquiry. (Usage of “case-based
thought” and “case study” in this context does not include

561
single-subject behavior analysis, which would be considered
“rule-based inquiry.”) This portrayal may characterize the
majority of James’s own work. However, the remainder of the
group appears to have used both case-based and rule-based
inquiry, depending on whether the task involved the context
of discovery or the context of justification (Reichenbach,
1938). Within the context of discovery, ideas and hypotheses
were derived from many sources, including traditional case
studies. Such case studies have been used for generating
novel concepts, for teaching, for clarification, and for casting
doubt on widely accepted practices. (Ullmann & Krasner,
1965, p. 44-49, provide an excellent exposition on such use of
case studies.) However, within the context of justification, my
academic ancestors and I seem to be in agreement that data
are an absolute requirement. Evidence from rule-based
inquiry is necessary to separate facts from opinions and
speculations.

Developmental Prelude
My academic genealogy shows a line of influence that is part
of the public record. The more important aspects of an
intellectual autobiography, of course, are not yet in the public
domain. Newly entering professionals often seem to believe
that the senior people in any discipline were born with a
grand plan for accomplishment. That grand plan is further
presumed to be one that was based on the senior person’s
current conceptual approach or research program. Senior
people may be viewed as having always been as they
currently appear. Obviously such notions are not true.

562
A broad range of developmental experience influences each
person’s intellectual approach, accomplishments, and ultimate
status. These include nonacademic life experiences, chance
events, choices based on irrelevant factors, and lucky or
unlucky timing as well as formal academic influences.
Ultimately, consequences interactively shape the approach
that becomes a basis for some rational professional action.
The following excursion is an attempt to summarize a number
of these experiences that now seem important to my
professional development. 1

1940-1954: Public schools and community college in

Marshalltown, IA. Elementary through high-school years and
1 year of community college were spent in the public schools
of Marshalltown, IA — a town with less than 30,000
population. I had no notion that there were competing
orientations and schools of psychology during this period. In
fact, I was hardly aware that the discipline of psychology
existed. I majored in life, love, and music — attending school
full time while working part-time and summer jobs. Jobs
included work as a paperboy, baby sitter, caddy, butcher’s
apprentice, warehouse and field laborer, construction worker,
theatre usher and doorman, mechanic, truck driver, salesman,
and musician.

Always taller than my peers, I achieved my adult height (6 ft.

4 in.) by the time I was 13 years old. This allowed me to pass
as much older and work in travelling dance bands and jazz
groups. That work provided many maturing experiences. I
organized 5-piece and 7-piece combos and a 12-piece dance
band during my last 2 years of high school and first year of
college — booking whatever group the employers could

563
afford. Family and friends were the most important
developmental influences during this period.

My mother, Ione Hickman (Perry), taught piano and was the

single teacher in a one-room school for 1st through 8th
grades. She taught many years with a 2-year associate’s
certificate. While continuing to teach full time, she earned the
B.A. degree in 1961, followed by advanced study,
certification, and acclaim in special education and sex
education. Her reluctant retirement from “going out” to work
came in 1991, following a total of 54 years in the classroom
and reaching the forced-retirement age in two separate school
systems. Since her last formal retirement, she has continued
to teach piano and tutor special-education students at home —
being 89 years young at the time of this writing. This special
lady provided a model of excellence and compassion for
others that has been a major influence in everything I have
tried to do.

My father, Leon D. Paul, ended formal schooling in the 6th

grade when he was expelled for striking the principal during
an argument about driving a car to class. His life was a
troubled one involving hard work, fast cars, chain smoking,
and heavy drinking as well as extremes of both hypermanic
activity and severe depression. After losing the family farm in
1939, he worked group piecework in a local factory until his
sudden death, under questionable circumstances, at age 40.
He had difficulty expressing affection and was a strict
disciplinarian who, under the influence of alcohol, often
became physically abusive. He was fond of saying, “Do as I
say, not as I do.” When sober, he demonstrated a brilliant
mind, natural leadership, charm, and a genuine concern for
others. At those times, family and friends adored my father

564
and it was fun to be with him. Friendships were far more
important to him than career advancement or financial gain.
He was an anti-snob for whom “being one of the guys” meant
refusing elected offices and promotions to management
positions. Explicit advice from my father came in a talk at the
time of my parents’ divorce. I was a sophomore in high
school. “Always remember that your word is your bond,” and,
“Do whatever you want to do, but be the best damned one in
town.” I have tried to follow his advice in those regards.
Unfortunately, we only became close after I could relate to
him as an adult — just before his death in 1954.

Although an only child, I was raised in a large extended

family of cousins, aunts, uncles, and maternal grandparents.
Nobody had much money. Most were employed as
tradesmen, factory workers, sales people, clerical workers, or
sharecroppers. One uncle was a military pilot. Many were
involved with music, some professionally. The family
provided multiple models for hard work, assumption of
personal responsibility, and helping one another. These ethics
were so strongly shared that they seldom needed
verbalization.

There were also extremes of values and behavior on many

dimensions. Some family members were ministers, including
a “hellfire and damnation” evangelical preacher. Some were
outspoken atheists. Most were quiet Quakers, Christians, or
agnostics. Politically they spanned the gamut from far
left-wing liberals to arch right-wing conservatives. My
grandmother was president of the Women’s Christian
Temperance Union while most in the following generations
were heavy users of alcohol. Some banned playing cards from
their homes while others were inveterate gamblers. The

565
family had its share of folks who could be called
“philanderers” and those who could be called “prudes.” Large
family reunions were enjoyable events that provided
demonstrations of tolerance as well as multiple examples of
conflict resolution. They also showed selective avoidance of
problem topics, based on respect for others.

Several close family members endured chronic and painful

physical conditions (e.g., cancer, arthritis, heart disease,
surgeries). Many of us had to deal with premature deaths of
family members from disease, accidents, and a suicide. There
were a few known instances of serious child abuse, several
failed marriages, and both an aunt and a close cousin whose
emotional problems were severe enough that they were
hospitalized with diagnoses of schizophrenia. The beneficial
role of social support from family and friends was much in
evidence here.

My experience with behavioral and emotional problems

during this early developmental period resulted in a strong
desire. I wanted to understand the distressing actions,
thoughts, and emotions of troubled and troubling loved ones,
in particular, as well as those of people in general. However,
entering a profession to try to gain this understanding and to
do something about it had not yet occurred to me. Rather, I
was committed to a career as a professional musician.

1954-1958: U. S. Navy, music school, bands, and San Diego

City College. A stint of military service was required in those
days. After discovering that my height excluded me from
becoming a fighter pilot, I auditioned for and gained
admission to the U.S. Naval School of Music in Washington
DC. This provided a way of jointly meeting my military

566
obligation and furthering my career in music. I majored in
music, minored in business, and discovered psychology as an
avocation during my 4-year enlistment as a contract musician
in the U. S. Navy. My last tour of duty helped to change
career focus from music to psychology.

During my first leave, in-route from basic training to the

music school, I renewed contact with a young woman in
Marshalltown whom I briefly had met before entering the
navy—Joan M. Wyatt. Following a whirlwind courtship, she
became my wife on Christmas Eve, 1954, and returned with
me to Washington. We lived off base in a one-room
apartment, sharing the lone bathroom with no less than 15
other people. She worked in the post-exchange office,
attended most of the duty jobs that I played, and made me the
envy of every sailor on the station. For 46 years, Jo Paul has
been my lover, partner, and best friend — participating in
most of the decisions that influenced my personal and
professional development.

Familiarity with performance anxiety is the most memorable

among many experiences in music school that had impact on
my later work as a psychologist. At the music school,
instruction, practice, rehearsals, and duty jobs with navy
bands (concerts, parades, shows, and dances) entailed a
complete immersion in music, 18 hours a day, 7 days a week
(not unlike the commitment required of graduate school in
psychology). A performance exam on each student’s major
instrument was required prior to graduation (tenor sax for
me). This exam determined the quality of the next assignment
and whether or not any choice was allowed among
alternatives. It was conducted in a room that was accessible
only by a spiral staircase to the third floor of the music

567
school. Everyone called it the “ivory tower.” The student
performed prepared pieces to demonstrate virtuosity,
followed by demonstration of whatever the examiners might
request — such as transpositions, sight-reading, or riffs.
Examiners included the commandant of the school and heads
of the section (e.g., reeds) and groups (jazz, symphonic) as
well as each student’s individual instructor.

The context of this exam already maximized its threat value

— especially when an outcome could be 2 years separation
from my new bride in some foreign country with a foul
climate. To make matters worse, a recent oral infection had
been followed by extraction of four impacted wisdom teeth. I
had been conducting and announcing instead of playing the
saxophone for some time when my exam was scheduled —
with only one week to prepare. I experienced the full rush of
debilitating overarousal that I had seen others suffer. The
week’s preparation was a waste as I unsuccessfully tried to
gain control with prescription drugs. The trip up the spiral
staircase, trailing behind my examiners, produced an
excruciating anxiety spiral, resulting in a total inability to
perform. The damn horn would not make a sound!

Fortunately, my reed instructor was familiar with the

circumstances. His “state-versus-trait” explanation for my
failure was successful in getting the exam extended. I was
able to practice, drug-free, and regain a level of instrumental
proficiency sufficient to pass the exam and obtain my choice
of duty stations. Ever after, I have had an empathic
appreciation of the experience of anxiety and of its effects on
performance. The experience also strengthened my desire for
understanding these powerful emotions.

568
I was able to obtain one of the best assignments for anyone
planning the musical career that Jo and I had envisioned. It
was with the band assigned to the commander of cruisers and
destroyers, Pacific fleet. This band was known for its
top-quality jazz musicians and for its great USO show
schedule. During my tenure with the group, musicians were
recruited on discharge directly to the Lawrence Welk
Orchestra (considered a sell-out) and to the Stan Kenton
Orchestra (the epitome for big band jazz). Even though the
assignment counted as sea duty, the admiral, his staff, and the
band transferred among flagships such that sea duty involved
a trip to Hawaii every 6 months. Otherwise, everyone lived on
dry land in San Diego, CA. Jo and I lived in navy housing
where there was an active social life. Both of our sons were
born there.

Because we worked shows and dance jobs only a couple of

nights a week, I was able to attend night school at San Diego
City College. It was there that I took introductory psychology
and became fascinated with the subject matter. Psychology,
especially psychoanalytic theory, became an avocation.
Playing in jazz clubs, observing drunken sailors, and other
wild parties increased my interest in knowing about the
unconscious bases of human motivations.

Additional time for self-directed study of psychology

inadvertently came from an attempt to advance my future
career in music. I was offered the opportunity to organize and
direct a band on the cruiser that was to be the American
flagship for the Australian Olympics. I was reluctant to accept
such an offer as Jo had recently given birth to our first son,
Dennis, and was pregnant with our second son, Dana.
However, we decided that the publicity from leading the

569
official American band at the Olympics — just a year before I
was to be discharged — was such a great opportunity for our
future in the music business that I could not refuse. I violated
the well-known military taboo and volunteered for temporary
duty on that flagship.

Unlike duty in San Diego, this ship regularly went to sea —

typically 3 weeks out, followed by a weekend in port. The
ship was based in Long Beach, CA. This required more than a
100-mile drive each way to join my family in San Diego. As
the only rated musician on the flagship, I declined jobs other
than working with the band, while the band members also had
other shipboard duties. They could work with me only 3
hours per day. After about 3 days at sea, I had exhausted
everything of interest on the ship (excluding gambling in the
anchor locker). This left 14 or 15 hours a day for me to
entertain myself.

I pursued my avocational interest in psychology as a way to

escape excruciating boredom — pouring through the
available psychology books. These were mostly by or about
Freud. This self-directed education further strengthened my
interest in Freudian theory. I became enamored with the deep
and mysterious unconscious and with the hydraulic operation
of libidinal energies.

After all the effort, I never did get to the Olympics. The duty
radioman, a friend whose shipboard hammock was in the
same area as mine, awaked me one night with some
disturbing news. A military crisis had resulted in the cruiser’s
reassignment to the Suez Canal theatre of operations. Rather
than serving as the flagship for the Australian Olympics, the
cruiser was to depart for the Suez hotspot within 48 hours. I

570
quickly arranged to terminate my temporary duty and
returned to the admiral’s band in San Diego.

The last year in the navy was spent at Great Lakes with a
show band whose schedule regularly involved 3-week
traveling stints. We lived in an off-base apartment in a “Little
America” ghetto in Waukegan, IL. The band was great, but
the travel schedule was not conducive to a good family life
with Jo and our two sons. As this was a typical schedule for
civilian bands as well, I again volunteered to try something
different — the job of bandmaster for the Service School
Command. In addition to training officers in parade
procedures, this job entailed organizing 100 non-professional
musicians to perform concerts and parades. I imagined that
this would be very similar to teaching high school and college
bands. That experience made it clear that a satisfying career in
music for me was not to be found outside of big band jazz.
Ten years as a professional musician was thus put on the
shelf.

Undergraduate Major in Psychology

— University of Iowa: 1958-1960
My decision to major in psychology was one of expedience.
Avocational interests had resulted in more credits in
psychology than in anything other than music. Psychology
would take the least amount of time in which to complete a
bachelor’s degree. Jo was willing to work and take primary
responsibility for the boys while I attended classes.

Because we were residents of the state, the University of

Iowa, Iowa City, offered low-cost tuition and even lower-cost

571
married student housing in “tin-hut” converted barracks.
Tuition, rent, and utilities were cheap enough to be covered
by the GI Bill. Jo would need to earn only enough to cover
incidentals, such as food, clothing, and transportation. Iowa
City was also close to our extended families. On these
grounds — with no clue as to the nature of the psychology
department — I applied to the University of Iowa and was
accepted with junior-level standing.

Things are not always as anticipated. I majored in psychology

and minored in mathematics at the University of Iowa. The
impact of education at this institution on my development was
recognized only later. The 2 years in Iowa City were not at all
what we had planned — personally or academically.

On the personal side, we discovered that Jo was pregnant with

our daughter, Joni, just before my discharge from the navy.
Our family soon included three children. Jo continued with
child rearing and emotional support as her primary jobs; for
money, she did contract typing in our tin hut. I worked 30-40
hours weekly as a fitter and salesman at a local shoe store
while attending classes full time. We did not do much for
entertainment, but we were all together.

On the academic side, conflict among schools of psychology

was evident from the first set of classes. Conflicts among
competing theories of learning and among theories of
personality seemed to drive most of the faculty’s interests. In
retrospect, the greatest genuine conflict was one of
contrasting epistemologies. At the time, I mistakenly viewed
it as a conflict of “basic research with rats” versus “clinical
applications with people.” The strong push for theory-driven
laboratory research and apparent derogation of clinical work

572
by the senior professors at Iowa had a paradoxical effect on
my intellectual development. It solidified my commitment to
the predominant psychodynamic approaches of the time and
to a clinical career, in psychiatry or clinical psychology,
rather than research.

Instructors and academic influences. Kenneth Spence, head of

the psychology department, appeared to rule with an iron
hand that left most faculty as well as students quivering. I was
never enrolled in a course with him. He seldom spoke to
undergraduates. However, the influence of “Spence-Hull”
theory — as it was known there — was ubiquitous. An
anti-clinical bias appeared prevalent among the senior faculty.

As an undergraduate, my classes exposed me to only two

clinicians in the psychology department — young assistant
professors, Charles VanBuskirk and Ron Wilson. I also
received instruction from well-known senior faculty,
including Jud Brown, I. E. Farber, Dee Norton, and Gustav
Bergmann, the last member of the Vienna Circle. More
applied courses were smuggled from the faculty at the Child
Welfare Research Station, including Boyd McCandless,
Charles Spiker, and Charles Truax, who was visiting for a
year. John Knott ran the EEG lab at the medical school and
taught my physiological psychology and anatomy courses —
with both cats and human cadavers as resources. All were
strong instructors.

Research methods, psychometrics, and statistics courses were

required of psychology majors. I hated those courses. Rather
than run rats to examine some theory-relevant principle for
my undergraduate research project, I investigated acoustics —
specifically, the overtones produced by different fingerings on

573
the saxophone. It was not much related to psychological
theory, but the findings had some practical use (musical
arrangers should be careful where they score trills).

I liked the courses in abnormal psychology, personality

theory, motivation, and physiology. Even the study of
sensation and perception was interesting. This was the meat
of my avocational interests, packaged in a formal discipline.
Translating Freud’s letters to Wilhelm Fliess to fulfill the
foreign language requirement further enhanced my interest in
Freudian theory. Ron Wilson semi-secretly guided additional
clinical studies as special topics. From that, I found Anna
Freud, Harry Stack Sullivan, and Eric Fromm all added
embellishments to psychodynamic thought that seemed to
make sense. Wilson also arranged for me to work briefly as
an aide at the state mental hospital. That was my first
exposure to institutionalized people with severe psychoses.

Learning theory was all right, particularly after I discovered

ego-analytic theorists, such as Heinz Hartman, Ernst Kris, and
Rudolph Lowenstein. Their notion of conflict-free spheres of
functioning allowed understanding of unconscious
motivations and of the mysterious active unconscious without
stereotypically invoking libido theory. I always had a little
trouble accepting the ubiquity of that. Hullian theory as
applied by Dollard and Miller (1950) provided a translation of
psychoanalytic principles that removed some of the
mysticism from Freud while showing the applicability of
learning theory to complex clinical problems.

Ideological commitment and career choice. I became

committed, prematurely, to a neo-Freudian, ego-analytic
approach. I also became committed to clinical practice — not

574
research. The decision became, should it be psychiatry or
clinical psychology? Both could take as long as 7 years before
earning a living above the poverty level. We now had three
children and only the GI Bill as a firm financial resource.

Jo again demonstrated her customary support. Either

psychiatry or psychology was fine with her. I should focus,
however, on things that really were of interest if we were
going to endure the hard work and lengthy period of austerity
entailed by either choice. On careful study, it was clear that I
did not want to spend 3 years learning about cuts and bruises.
Why spend years on material that appeared only tangentially
related to understanding and ameliorating the distressing
actions, thoughts, and emotions of troubled and troubling
people? I wanted immediately to get to the real stuff of
learning to understand the mysterious unconscious bases of
human behavior. Clinical psychology seemed the route to
follow.

We needed to find a university that would provide real

clinical training. As money to move our family was in short
supply, the distance from Iowa City was an important factor
for consideration of any graduate program. Availability of
financial support through a fellowship or assistantship,
including a tuition waiver, was also of critical importance.

The University of Illinois at Urbana-Champaign and Ohio

State University, Columbus, both advertised large and active
clinical programs. I did some careful library work to verify
the advertisements, the facilities available for clinical
training, quality ratings, and the nature of faculty
publications. It seemed that research training was a necessary
evil to obtain clinical training in psychology. At least the

575
research in these two departments had an applied focus with
people. Ron Wilson concurred that both had good reputations.
I applied to those two programs and was accepted at both
places. Both also offered support with a tuition waiver for the
first year. Continuing support would be dependent on my
performance.

Before I responded to either offer, Jud Brown called me into

his office to say that the Iowa faculty thought I would be
throwing away my career by pursuing this “clinical business.”
Although I hadn’t applied there, he said the faculty would like
me to stay at Iowa for graduate school. He offered admission
with a tuition waiver and guaranteed financial support. I was
flattered by the Iowa offer. Guaranteed financial support in
our situation was also very tempting. However, the University
of Iowa did not provide the clinical emphasis that I was
seeking.

Graduate School — University of

Illinois at Urbana-Champaign:
1960-1964
Jo and I selected the doctoral program in clinical psychology
at the University of Illinois at Urbana-Champaign for
graduate training. This choice over Ohio State was based on
factors that were as unrelated to professional issues as the
earlier selection of the University of Iowa. The major reason
for selecting Illinois rather than Ohio State was, again, one of
expediency. Moving to Champaign-Urbana from Iowa City
and visits to relatives in Iowa would be cheaper than
corresponding travel to and from Columbus. Another factor

576
that determined this choice was the quality of inexpensive
married-student housing. Although still renovated barracks,
instead of the Iowa-style tin huts with 6-foot high window
sills, Illinois housing had frame walls with windows through
which Jo and the kids could actually look. After Iowa City,
ordinary windows and doors had become important.

Some things are as anticipated — and more! I majored in

clinical psychology and minored in physiological psychology
and educational psychology (group therapy) during our 4
years of graduate school at the University of Illinois.
Homemaking continued as Jo’s major job — rearing our three
children and providing emotional support to all four of us.
She added preschool daycare to her home typing to help with
finances after our sons started school. My earnings now came
from efforts in psychology rather than music or sales. I first
worked as a half-time research assistant. Following initial
clinical training, I also worked as an assistant in the
department’s training and research clinic, with added summer
work at both the Student Counseling Bureau and the
Veteran’s Administration (VA) Hospital, Danville, IL. I
obtained a predoctoral fellowship from the National Institute
of Mental Health (NIMH) for support during the final year.
Social activities and entertainment centered on the
psychology department and the student housing project. Jo
and the kids were regularly in close contact with other
spouses and children who were in similar circumstances.

Our entering class at Illinois numbered about 60 graduate

students, half of whom were clinical majors. We all shared
proseminar, quantitative methods, and research design
courses during the first year. The welcome by Lloyd
Humphreys, head of the department, set the stage for a very

577
stressful year. He noted a few “empirical facts.” (1) Graduate
students should become immersed in psychology as a
discipline — successful students typically devoted at least 80
hours per week to classes, work, and study in psychology. (2)
It had been documented that people could survive on 4 hours
of sleep per night for the period of time we should plan to be
in graduate school. (3) Two-thirds of the entering class would
likely flunk or leave before earning doctorates. (4) Students
were unlikely to be successful in both graduate school and in
marriage. Clearly, this was going to take real effort in the
interpersonal as well as academic arena.

Avoidance schedules produce a lot of behavior, but they do

not feel good. The amount of work required was not much
different from what Jo and I had done for the previous 2 years
at Iowa, but it was more intense. The threat of flunking also
made it much more anxiety laden. To add to the stress, we all
had been used to being the top students in our undergraduate
courses. Now we were told our competition came from the
highest 5-10 % of the undergraduate population. Clinical
students learned that only 6-8 of 30 entering students each
year typically made it through qualifying exams for doctoral
candidacy. We really thought it was a plan to weed-out those
who could not work under pressure — a version of “trial by
fire.”

Now I tell my students that I not only had to walk 2 miles in

the snow to get to graduate school, but it was uphill both
ways! The proseminar course, as conducted in our first year,
was a brutal experience. I generally favor the concept of a
proseminar, in which faculty members each spend a week or
two with the entire group of first-year students.
Unfortunately, for our entering class, there were as yet no

578
guidelines for the amount of material to be assigned. Each
faculty member attempted to cover everything that was
important in his own specialty. These assignments often
exceeded 2000 pages of reading a week, with testing so
comprehensive as to require recall of specific footnotes. The
expected attrition in the first year class did occur, either from
failure or from people’s unwillingness to tolerate the required
life style. Two students left to be admitted to acute psychiatric
units. Unnecessarily stressful, indeed!

Graduate students organized into study groups as a way to

deal with the massive amount of reading and to maintain
some degree of interpersonal support. After a few abortive
trials with other students, Tom D’Zurilla, Ron Krug, Rick
Schulte, and I settled in as an effective working group. We
continued to meet twice a week for study and review of all
common course material throughout graduate school,
including preparation for the doctoral qualifying exams. We
became close friends who socialized with each other’s
families and friends, in the limited time available for those
activities. The reviews, arguments, discussions, and mutual
support within our study group were a major influence on my
intellectual development during this period.

Instructors and academic influences. Those who successfully

endured the “trial by fire” were rewarded by such intense
education and training experiences that our horizons had to
undergo major expansion. By the time I undertook my
doctoral dissertation, I had received classroom or seminar
instruction from what I later realized was an incredibly
talented list of figures in the psychology department and
affiliated labs or institutes. These influential instructors
included Jack Adams, Wes Becker, Ray Cattell, Lee

579
Cronbach, Don Dulany, Charles Eriksen, Fred Fiedler, Marty
Fishbein, Ray Frankman, Bob Grice, Harry Hake, Lloyd
Humphreys, Joe Hunt, Will Kappauf, Sam Kirk, Joe
McGrath, Bill McGuire, Hobart Mowrer, Larry O’Kelly,
Charles Osgood, Don Peterson, Hal Rosen, Don Shannon,
Ivan Steiner, Larry Stolurow, Garth Thomas, Harry Triandis,
Ledyard Tucker, Mort Weir, and Jerry Wiggins. Merle
Ohlsen, Cecil Patterson, and Fred Proff taught courses in the
educational psychology counseling program, where I
completed a minor in group counseling and therapy.

I initially viewed research experience only as a way to earn

money and as a necessary evil to obtain the doctoral degree.
Incorporation of research and scientific method as natural
problem-solving strategies developed only gradually.
Practical training in applied research was continuous, through
paid employment as a research assistant, class work, and
required thesis and dissertation projects.

Based on my undergraduate work in acoustics, my first

research assistantship was with Grant Fairbanks in the Speech
Research Laboratory. My next assistantship was in the
psychology department with Wes Becker. He was a “factor
analyst” at that time, before his conversion to a “behavior
analyst.” That position included helping to build his garage as
well as work on grant-supported projects. Roy Hamlin
supervised research work at the VA Hospital. Charles Eriksen
was advisor for both my master’s and doctoral studies. He
also encouraged incidental research to answer other questions
of interest. Other members of my doctoral committee were
Wes Becker, Lloyd Humphreys, Merle Ohlsen, Don Shannon,
and Jerry Wiggins. All were devoted to the discipline and
served as strong models for professional work.

580
The clinical facilities, supervision, and training were as rich
as advertised. I overbooked to maximize my clinical
experience. This included extra practica, special topics, and
formal supervision for the clinical aspects of research
assistantships as well as paid clinical service jobs. By
tradition, students continued to carry clients in the
departmental clinic throughout their tenure in the clinical
program — even while collecting data and writing
dissertations.

I received clinical supervision as part of course work and

formal practica in assessment and psychotherapy from a
number of well-known clinicians in several different
facilities. Supervision through the department’s Psychological
Clinic came from Joe Becker, Wes Becker, Nate Eisen, Don
Shannon, Jerry Wiggins, and Arnie Miller. Don Shannon also
supervised my work in a paid position there as well as
assessments in the public school system. Larry O’Kelly ran
the EEG lab. O’Kelly and Wes Becker, along with Angela
Folsom, from the Danville VA Hospital, supervised
neuropsychological assessments in both clinic and hospital
settings. Helaine Moody and Bill Ward supervised other
clinical work at the VA Hospital, where they were full-time
clinical staff. Bill Gilbert and Alice Jonietz were supervisors
for clinical practica through the Student Counseling Bureau,
while Tom Ewing and Mort Wagman supervised my paid
clinical work there. Merle Ohlsen supervised my conduct of
therapy groups through the Guidance and Counseling Center
of the College of Education.

By the time other requirements for the Ph.D. degree were

completed, I had accumulated more than 2000 hours of
supervised clinical work beyond the minimum specified for

581
graduation in the program. This was sufficient to successfully
petition to bypass the predoctoral internship requirement and
go directly to postdoctoral internship training.

Len Ullmann joined the clinical faculty at Illinois after my

dissertation study was in progress. I informally audited some
of his classes. He helped to reinforce changes in thinking that
were already underway as my dissertation findings unfolded.
He was especially influential in expanding my interest in the
application of operant principles. However, his influence
came after the data had already been collected for the focal
publication. He guided me to Stanford University Press as a
publisher for the monograph to report the study and its
findings. He also introduced me to Len Krasner, who was
instrumental in helping to arrange a postdoctoral internship at
the VA Hospital in Palo Alto/Menlo Park, CA.

Upheaval and attempts at recovery. Most clinical training was

based on Freudian, Rogerian, neo-Freudian or ego-analytic
theories and associated insight-oriented approaches to
psychotherapy. Sullivan, Rogers, Fromm, Dollard and Miller,
and eclectic ego-analysts were the predominant affiliations of
my clinical supervisors — and I loved it! However, the
evidence-based content and research courses regularly failed
to support the fundamental principles underlying the clinical
training. I hated that! I became determined to demonstrate the
basic truth of psychodynamic concepts.

As we all had to complete a master’s thesis, I planned to use

that project to document the operation of unconscious
learning, through use of heat stress with cool-air
reinforcement. Even though I had not enjoyed acquisition of
the knowledge, my previous course work in quantitative

582
methods and research design, from both Iowa and Illinois,
had prepared me to critically evaluate the literature. Previous
studies, I believed, had simply not involved variables or
designs that were strong enough to demonstrate the
phenomena of learning without awareness in a laboratory
setting. I wanted to ensure that the quality of the research
design was so strong that no alternative explanations could
sully the conclusion that unconscious learning occurred.

Following his presentations in proseminar, I asked Charles

(Erik) Eriksen to supervise my master’s study. Beyond the
fact that he was the expert in the topic area, Erik had a
reputation as a strict methodologist. He also had been
debunking many cherished principles of psychodynamic
theory. In fact, the title of this chapter is fashioned after a
presentation of his, entitled, “Subliminal perception,
preconscious thought, and other things that go ‘bump’ in the
night.”

My master’s research found quite the opposite of what I

hoped to demonstrate. The study did not simply fail to
document the operation of unconscious learning. Findings
actually showed how reinforcement effects could produce
different classes of “crazy behavior” that were explainable by
correlated hypotheses regarding stimulus control — not
psychodynamic principles (Paul, Eriksen, & Humphreys,
1962). This outcome — of my own investigation — created a
major chink in my beliefs.

The next explicit effort to support psychodynamic concepts

came in a literature review for a physiological course with
Larry O’Kelly. I had tinkered some with hypnosis. Surely
hypnotic phenomena demonstrated the active unconscious.

583
The production of non-herpetic skin blisters by hypnotic
suggestion should be a natural demonstration. I did a critical
review of that literature, again drawing on my previous forced
methodological training. Once more, my own efforts found
better explanations for the phenomena, not support for the
psychodynamic principles (Paul, 1963). This added another
chink in my beliefs.

I was not the only one who was disturbed by the lack of
evidence to support our practical clinical training. Several of
the faculty and students were as well. Much ferment was
present among advanced clinical students as well as among
the remaining members of my class. A group of us convinced
Erik to offer a seminar in which we could seriously examine
Eysenck’s (1952, 1961) outrageous claims that
insight-oriented psychotherapy did not work and that this new
approach called “behavior therapy” did. The behavior therapy
literature was sparse enough at that time to be covered in a
single semester. We concluded that Bandura’s (1961) review
article was essentially correct regarding the promise for that
approach. However, we did not believe that the literature
demonstrated ineffectiveness of the psychodynamic approach.
We concluded that there was simply no good evidence
regarding the effectiveness of insight-oriented psychotherapy
— only a lot of flawed studies. Nevertheless, the absence of
evidence to support my preferred approach added another
disquieting chink in beliefs.

Concerted clinical efforts. Concurrently, I decided to

personally treat a few cases with Wolpe’s (1958) systematic
desensitization to demonstrate to myself that the results were
simply transference cures. Surely, anxiety was a symptomatic,
distorted discharge of accumulated tension — a derivative of

584
unconscious conflicts between contradictory impulses and
defensive forces. Once I personally observed the superficial
nature of client response to desensitization, I thought, I could
comfortably discard Wolpe’s uncomplicated
counter-conditioning model for the more comprehensive
psychodynamic approach.

The definitions in Wolpe’s book and other publications were,

unfortunately, not precise enough for me to replicate his
procedures. Hal Johnson was a research assistant in Erik’s lab
at the time. Together, we used our knowledge of learning
principles and of physiological responses to imaginal stimuli
to develop procedures that should be followed, if Wolpe’s
hypotheses were true. We ran several subjects with polygraph
monitoring to establish the timing parameters for progressive
relaxation training and for presentation of hierarchically
ordered stimuli for imaging.

Following our development of timing parameters, I arranged

for clinical coverage and personally treated 10 or 11 clients
with anxiety related problems. I was careful to limit my
interventions to systematic desensitization — with no
psychodynamic interpretations. This was especially difficult
as some of the cases involved perfectly obvious symbolism.
The reason for treating more than a handful of clients was that
every one demonstrated clinical improvement! I continued to
add more clients on the presumption that the next one would
be unsuccessful. I knew enough to realize that these results
could be more apparent than real, based as they were on
uncontrolled case studies. Still, this created another chink in
my beliefs.

585
I arranged to do a small comparison with 11 students who
requested treatment for test anxiety, following their
participation in a validity study of an anxiety scale (Paul &
Eriksen, 1964). I treated five with individual systematic
desensitization and compared their outcomes to six equated
but untreated classmates. In contrast to the untreated group,
the entire treated group showed improvement on self-reported
anxiety and on course exam performance. We never
published this study because of the obvious within-class
confounding of therapist characteristics and lack of control
for nonspecific treatment effects. Nevertheless, the findings
were disquieting. They added another chink in my beliefs.

The Focal Study of Insight versus

Desensitization (Paul, 1966)
The lack of adequate controls was the only reassuring aspect
of my apparent and unanticipated successes with systematic
desensitization. It was probable, I thought, that placebo
responses and the nonspecific effects of psychological
support, simply from undergoing treatment, had been
operating. My application of systematic desensitization still
could be merely producing transference cures. I had not
followed these clients long enough to check on the occurrence
of early relapse or symptom substitution. Yes, I thought. That
was it! It really would take a well-controlled study to
demonstrate these effects and the superiority of
insight-oriented psychotherapy.

586
Background and Preparation
I had to complete a dissertation study anyway. If that amount
of work was to be done, it ought to provide some answers to
questions that were personally important. My thoughts were
as follows. It was clear that previous real-life treatment
studies were badly flawed. Both the “known,” but
undemonstrated, effectiveness of insight-oriented
psychotherapy and the early relapse and symptom substitution
expected from systematic desensitization should appear in an
investigation with really good research design. An ideal
design was needed.

I decided to undertake such a comparative study and, again,

asked Erik to be my research advisor. Even though he was not
actively involved in treatment research, if Erik approved a
design, I was confident that the findings would be solid. He
and I, together, identified other desirable faculty for my
doctoral committee. I wanted critical expertise in
measurement, quantitative methods, learning theory, and
psychodynamic theory — including coverage by one or more
of the participating insight-oriented therapists. This would
ensure that the study surpassed the highest standards of all
stakeholders. After I had outlined my overall plan and
obtained agreement of potential therapists, I asked Wes
Becker, Lloyd Humphreys, Merle Ohlsen, Don Shannon, and
Jerry Wiggins to serve on the committee. All agreed.

Design principles. Before selecting the committee, I scoured

the literature to determine the domains and classes of
variables that needed to be measured, manipulated, or
controlled in order to provide cause-effect evidence of

587
effectiveness for any psychotherapeutic approach. The
resulting principles and concepts were later expanded in two
publications, where the “ultimate clinical question” was
explicated. A paper on the strategy of outcome research in
psychotherapy was published in the Journal of Consulting
Psychology (Paul, 1967b). The recommendations were further
elaborated in a chapter on design and tactics in Cyril Franks’
edited book (Paul, 1969a). Before detailing this material in
print, however, I had formulated what was needed to design a
comparative study of treatment effectiveness that would allow
unambiguous conclusions. My forced training in research
methods at Iowa and Illinois again proved useful.

Target problem. I identified interpersonal-performance

anxiety as a timely target for treatment research. Anxiety was
the major component in most, if not all, theories of neurosis
and the reduction of anxiety was a goal of nearly every
psychotherapeutic approach. My experience in music school
provided first-hand knowledge of the potential devastating
effects of anxiety on performance. Epidemiological data
showed interpersonal-performance anxiety to be a frequent
emotional problem with a serious impact on functioning.
Further, this was an anxiety problem that appeared delimited
enough to allow rigorous experimental methodology, yet
significant enough for broad generalization of findings —
generalization that would not be possible with mere
laboratory analogues of psychological difficulties. Public
performance also provided prototypic stress conditions for the
direct measurement of improvement or worsening. These
situations would similarly allow assessment of the predicted
effects from the conflicting theories on which competing
treatments were based — generalization of improvement
versus symptom substitution.

588
At that time, a course in public speaking was a graduation
requirement for liberal arts students. Many found
performance anxiety to be a major impediment to their
successful completion of this requirement. People often
delayed the course until their senior year. Large numbers
sought treatment and a very large percentage was reported to
simply drop out without graduating. Here was a population of
young adults who should be essentially psychiatrically
normal, but with a serious emotional problem that they were
highly motivated to overcome. An ideal group, it seemed,
with which to arrange multiple treatment and control
conditions within the necessary partial-factorial research
design.

Performance anxiety was a serious concern to the faculty and

administration of the speech department as well. So much so
that the faculty were eager to announce the availability of
treatment and allow me to collect assessment data within their
classes.

Insight-oriented therapists. Experienced practitioners were

crucial if the superiority of psychodynamic principles and
procedures were to be demonstrated. I solicited participation
of only those insight-oriented therapists with the best
reputations. Although I intended to obtain grant funding to
pay therapists, the hourly rate would be nominal compared to
their usual charges. Therapists also needed to be confident
enough in their usual treatment approach to pit it against
systematic desensitization and a stylized attention-placebo
treatment. Further, to control for therapist attributes,
nonspecific treatment effects, and placebo responses, these
therapists had to be willing to learn and deliver the competing
treatment and control procedures themselves — and be

589
monitored in their performance. Five of the area’s most
highly regarded insight-oriented psychotherapists agreed to
participate: Joe Becker, Alice Jonietz, Merle Ohlsen, Fred
Proff, and Don Shannon. All were doctoral practitioners who
had been previous supervisors of mine.

Commitments and funding. I obtained commitments from the

numerous players, wrote the proposal, and had it approved by
my doctoral committee. By concurrently submitting a
proposal to the U. S. Office of Education, I received fast-track
contract funds to purchase supplies and equipment, pay
therapists, and support data analyses. Of course, notification
of the award arrived only after the latest date that everyone
had to commit to work on the project. Jo volunteered to
handle the typing and supervise the many duplicating and
scoring tasks.

As a sidelight on scientific progress in the clinical area, Erik

and I had earlier approached the NIMH regarding possible
grant support. We were told the proposed study could not be
funded because symptom substitution was a highly probable
outcome for two of the planned treatment conditions. By
policy, the NIMH could not support treatments that would
have negative outcomes on participants. My hypotheses
obviously had support within the government agency that was
responsible for looking after the mental health of the nation!

Design and Procedures

Some writers have mistakenly referred to this investigation as
an “analogue study” of psychotherapy. That faulty inference
apparently resulted from the fact that the clients were also

590
students, concurrently enrolled in a course in public speaking,
and because actual treatment contacts were time-limited to
five sessions within a 6-week period.

Table 1. Design of the original study of insight vs.

desensitization in psychotherapy (Reproduced from Paul,
1966).

As detailed elsewhere (Paul, 1966, 1967a, 1969e), I do not

consider this study to be an analogue. Clients were “real
clients” who requested treatment for personal difficulties
following announcement of its availability. They sought
treatment for serious “real-life problems,” with no incentive
provided beyond their own potential improvement in
functioning. Therapists were “real therapists” with years of
experience. The experienced therapists selected the number of

591
sessions, based on their consensus of the usual number
needed to treat the focal problem (given that identification
and assessment of the problem occurred in advance).
Treatment sessions took place in the therapists’ own clinical
offices, in one of three “real clinical-service facilities,” rather
than in a laboratory setting. Finally, treatment procedures
were applied as they would have been in ordinary practice,
except for being monitored by audiotapes to ensure fidelity.

Treatment and control groups. The overall design and

procedures for the focal study are summarized in Table 1.
There were five groups in the study. Three were treatment
groups — systematic desensitization (group D),
insight-oriented psychotherapy (group I), and an
attention-placebo control treatment (group AP). The same
five therapists conducted all three of the individual treatments
with one female and two male clients from each group. The
attention-placebo condition involved therapists performing a
believable set of actions and attending to clients. The clients,
meanwhile, engaged in an irrelevant task that was presented
as being helpful for anxiety reduction. The task also
prevented therapists and clients from talking to each other for
45 minutes of each 50-minute session. This was an excellent
control for the nonspecific effects of receiving treatment and
attention from the same therapists. As shown later, it works!

The remaining two groups were control conditions. The

treatment control group (group TC) was a wait-list control
that received all assessments and classroom procedures, but
no individual treatment. The no-contact control group (group
CC) consisted of people who had requested treatment and met
all other selection criteria, but received no personal contact.
They merely continued in the speech course and completed

592
the pretreatment and follow-up batteries with the entire class
population. People in this group were unaware of their
participation in a treatment study — providing a base for
evaluation of possible improvement resulting from the
additional attention, practice, and anticipation of treatment in
group TC.

Pretreatment battery and subject selection. The pretreatment

battery, noted in Table 1, was administered to the entire
population at the beginning of the required speech course.
This battery included self-report scales to assess anxiety
levels in both the focal performance situation and in other
theoretically relevant interpersonal-evaluative contexts.
Nonfocal anxiety scales were included for later testing of
hypotheses regarding generalization versus symptom
substitution. Scales assessing general anxiety, extroversion,
emotionality, and falsification were included for the latter
purpose as well. These scales also served to describe the
sample and to evaluate possible mediators of clients’ response
to treatment.

A cover letter in the pretreatment packet explained that

treatment would be available free of charge to a few people. It
said the psychology department was conducting a study of the
way personality characteristics might interact with alternative
treatments. What people benefit most from different
approaches? Further, although treatment was to be paid by a
federal grant, in return, anyone receiving services needed to
commit 2-3 hours more for additional assessment.
Demographic data were collected on a data sheet, which
provided a place for people to request treatment and rate their
degree of motivation.

593
A total of 380 people requested treatment following
administration of the pretreatment battery. I selected the 96
who were the most debilitated by anxiety for participation in
the study (68 males, 28 females). People in the resulting
sample were “good bets” for psychotherapy, being young,
intelligent, middle-class, highly motivated, and with strong to
severe interpersonal-performance anxiety of 2-20 years
standing. In most cases, anxiety was reported in nearly any
social, evaluative, or interpersonal context — being most
severe in the public speaking situation. Their scores on the
focal anxiety scales were, in fact, higher than those of most
people applying to community clinics with similar problems.
The selected sample also scored significantly higher than the
broad student population on both general anxiety and
emotionality as well as scoring significantly lower on
extroversion. The severity of the problem was further evident
among those not selected for participation, although they had
requested treatment. Even though the latter group scored
lower on anxiety scales than those selected for participation,
32% dropped out without completing the speech course. None
of the treated students dropped out.

Pretreatment test speech, treatment assignment, and

interview. Of the 96 people in the selected sample, 74
underwent additional pretreatment assessment under stress
conditions. This involved delivery of a test speech before an
unfamiliar audience of 10-17 people. The audience included
four trained observers (graduate students), introduced as
“clinical psychologists and speech people who will be helping
us to evaluate your reactions.” During each presentation, the
observers sat in the center front row, coding the presence or
absence of 20 specific manifestations of anxiety on a
timed-behavior checklist — every 30 seconds for 4 minutes

594
(with reliabilities exceeding r = .95). Additional
stress-condition measures were obtained just before each
person presented his or her test speech. These included a
self-report Anxiety Differential and two measures of
physiological arousal — pulse rate and palmar sweat. At the
end of the pretreatment test speech, each participant was
scheduled for an interview with me.

The 74 “contact” people were then randomly assigned to one

of the three treatment or control groups from stratified blocks
based on pooled behavior checklist scores. Assignments were
stratified within gender so that each therapist was randomly
assigned one female and two male clients for each of the three
treatment conditions. Each treatment group thus included 15
people while the no-treatment control group had 29 and the
no-contact control group had 22 people. The five groups were
equated on all relevant variables at the pretreatment
assessment.

I met personally with each person who underwent

pretreatment stress condition assessments to provide a
common motivational induction and final screening. Each one
assigned to a treatment group was given a standard rationale
for the treatment and a description of the procedures. The
intention had been to dismiss or reassign anyone whose
expectations were opposed to the assigned treatment. This
was necessary in only two instances. After checking
schedules to assign the first session with a specific therapist, I
then attempted to further induce common expectations by
saying, “Oh, you’ll be seeing Dr. X. (s)he is very good with
problems of this sort. (s)he’s had a great deal of experience,
and I think you’ll find working with (her)him to be not only
quite helpful, but interesting as well.”

595
People assigned to group TC were told that time was not
available for experienced therapists to see everyone that
semester. Rather than ask anyone to work with a less
competent therapist, names had been “picked from a hat” to
determine who would have to wait until the following
semester to receive treatment. I assured them that treatment
would be provided for those who wanted it (and it was). It
still would be necessary, however, for them to return for
another evaluation speech in a few weeks, as our measures
would be meaningless without them.

Treatment period. Treatments were conducted concurrently

by the five therapists, rescheduling all missed appointments
within a week. Written manuals were used prior to the
treatment period to train therapists in their conduct of
desensitization and attention-placebo treatments. Therapists
described their own approach on a standardized instrument
constructed for that purpose. I personally monitored
audiotapes of every session and met weekly with each
therapist to ensure fidelity and absence of cross-treatment
contamination. Therapists provided self-ratings of their
confidence in effecting change with each treatment approach.
At the end of the treatment period therapists rated each of
their client’s improvement, prognosis, likability, and a variety
of other features that might be of relevance.

People assigned to the no-treatment and no-contact control

groups continued in the speech classes, without other contact
during the treatment period.

Posttreatment test speech and follow-up battery. Within a

week of treatment termination, clients in the three treatment
groups and the wait-list controls (group TC) were brought

596
back for posttreatment stress-condition assessments. All
treated clients returned for this evaluation, but 7 of 29 in the
wait-list condition were “no shows.” These no shows were
among the most anxious of that group at pretreatment. The
same measures obtained before and during the pretreatment
test speech were repeated for the posttreatment test speech. I
also added blinding and other procedures to avoid carryover
biases from the pretreatment test speech.

The follow-up battery was administered 6 weeks following

termination of treatments. It was administered to the entire
course population to protect the anonymity of the people
within the five groups of the focal study. The follow-up
battery consisted of the same self-report scales that had been
administered in the pretreatment battery. It also included
scales for each treated client to rate his or her therapist on
likability and competence as well as self-ratings of
improvement and satisfaction. A cover letter reemphasized
the confidentiality of information and informed everyone of
the opportunity to receive interpretation of his or her scores.

Results
Group and individual differences on all data were evaluated
by analyses of variance, multiple comparisons of the
difference between differences, regression analyses, and
individually significant changes on each measure for each
client. The results were surprisingly clear and consistent, with
converging evidence from all information sources. The
superiority I had hoped to demonstrate for insight-oriented
psychotherapy was no where to be found. The
attention-placebo treatment was, in fact, as effective as

597
insight-oriented psychotherapy! Clients treated by systematic
desensitization were, unmistakably, the most improved of all
groups at treatment termination. In fact, systematic
desensitization was remarkably effective on an absolute level.
The 6-week follow-up demonstrated a similar pattern of focal
treatment effects. For treated clients, only a few additional
changes, beyond the focal effects, appeared at the 6-week
follow-up. There was no evidence of symptom substitution. A
few trends even suggested generalization of positive effects.
The following highlights the major findings.

Stress condition results at treatment termination. Changes

observed under stress conditions are the most stringent test of
treatment effects. The percentages of significantly improved
cases within each group under stress-condition assessments
are presented in Figure 2. The dark bars at the top of each set
in Figure 2 show that clients treated by systematic
desensitization demonstrated significant improvement on all
three classes of measurement — a much greater proportion of
improved clients than either insight-oriented psychotherapy or
attention-placebo treatments. In fact, 100% of clients treated
by systematic desensitization showed improvement on both
overt performance and self-report measures, with 87% also
showing improvement on physiological measures.

The second and third bars of each set in Figure 2 reflect the
percentages of improved cases treated, respectively, by
insight-oriented psychotherapy and attention-placebo
treatment. Both of these treatments produced greater rates of
improvement than the no-treatment controls, but the insight
and attention-placebo groups did not differ in effectiveness
from one another on a single measure! The bottom two bars
of each set represent improvement rates for the no-treatment

598
control group. The differing numbers reflect the “no shows”
on the second assessment for those in the wait-list condition.
Improvement rates that include the entire no-treatment control
group, with no-shows treated as “no change,” are in the
lowest bars. The next to the lowest bar of each set shows
improvement rates that include only those with complete data.

Figure 2. Percentage of cases significantly improved under

stress conditions (Adapted from Paul, 1966).

Six-week follow-up results. As noted earlier, the 6-week

follow-up data generally showed a pattern among the four
contact groups that was consistent with the stress-condition
results at treatment termination. The no-treatment control
group did show slight, but significant, gains compared to the
no-contact control group — indicating that the additional

599
attention, stress-condition assessments, and promise of
treatment had produced some minimal benefits. However, the
strength and clarity of results among the treatment groups left
no doubt of the superiority of systematic desensitization over
competing approaches, or of the essential equivalence of the
insight and attention-placebo treatments.

No client or therapist characteristic interacted with response

to treatment. Systematic desensitization was, without fail,
more effective in the hands of all therapists, without
moderating effects by any client attribute. Therapist and client
ratings of improvement and prognosis in the focal problem
area correlated with standardized improvement data from the
self-report instruments and with improvements shown in the
objective behavioral data. The only, even suggestive,
superiority for the insight-oriented group came in a single
rating by therapists — prognosis in areas other than the focal
one — and this rating was unrelated to any other data. It was,
however, consistent with the theory underlying the
insight-oriented approach.

Two-year follow-up results (Paul, 1967a). The data just

summarized are those that were included in my dissertation
study (Paul, 1966). However, the 6-week follow-up in that
study was too short to show the long-term effects predicted by
psychodynamic theory. The early relapse and symptom
substitution predicted for both systematic desensitization and
attention-placebo clients, and the consolidation and spread of
gains, for insight-oriented clients, all required the passage of
time. Learning theory, in contrast, predicted no differential
relapse in the focal area. Additional change in untreated areas
was predicted by learning theory only to the extent of

600
generalization of focal improvements, with naturally
occurring reinforcement in the posttreatment environment.

A 2-year follow-up was undertaken to evaluate these

contrasting predictions (Paul, 1967a). Other than archival
records, all of this information was collected by mail as 64%
of the total sample had moved from the area. Data collection
included a third administration of the scales from the
pretreatment and posttreatment batteries. The battery was
supplemented by requests for information on the
posttreatment frequency of stressful events, speech
performances, and activities that might reflect predicted
symptom-substitution effects (increased dependency, anxiety,
or introversion). Information was also requested on receipt of
either drug or psychological treatment during the follow-up
period.

All but three of the earlier participants were located within

25-27 months following the end of the original treatment
period. All treated clients returned requested data. Only 70%
of controls returned data and 38% of those were excluded
from analyses because they had been treated for the focal
problem after the original study ended. The retained untreated
controls were a positively biased subsample of the original
control groups. Even so, the long-term follow-up results,
again, demonstrated clear, convincing, and consistent findings
in support of the learning theory interpretation. Not even
suggestive support appeared for psychodynamic theory.
Figure 3 presents the percentage of people in each group who
showed continued improvement, generalized improvement,
symptom substitution, or relapse on the 2-year follow-up.

601
Focal improvement among treated clients over the 2-year
follow-up period was relatively reliable (r = .78) and
predictable from improvements on stress condition
assessments 2 years earlier (r = .61). The set of bars at the
bottom of Figure 3 shows that systematic desensitization
maintained its higher rate of significant improvement over the
other two treatment groups. Insight and attention-placebo
treatment continued to show no differences in improvement
rates from one another. As noted at the top of Figure 3, there
was no evidence of relapse for any treated client. No matter
what treatment they had received, none showed an increase in
self-reported anxiety in situations that involved public
speaking. Of the positively biased group of untreated controls,
11% showed significant increases in focal anxiety that would
qualify as relapse and 22% showed decreases that qualified as
focal improvement.

Figure 3. Relapse, symptom substitution, generalization of

improvement, and continued focal improvement at the 2-year
follow-up (Adapted from Paul, 1967a).

602
The middle set of bars in Figure 3 shows evidence of
generalized improvement, with the desensitization group
showing more generalization of positive effects than other
groups. The insight and attention-placebo groups failed to
differ from each other, once again. As reflected in the top set
of bars in Figure 3, changes that could provide evidence of
possible symptom substitution occurred at exceptionally low
rates. None of the groups, including controls, differed from
chance-level changes and there were no differences among
groups.

The results of the 2-year follow-up clearly substantiated the

cause-effect relationships found earlier in my dissertation
study. It also extended the evidence to show that the produced
effects were lasting ones. Systematic desensitization not only
worked but it worked best. No relapse. No symptom
substitution. Insight-oriented psychotherapy produced no
greater benefits than the nonspecific attention-placebo effects
achieved by skilled therapists. Learning principles rather than
psychodynamic theory accounted for the pattern of findings
for all treatments.

Consequences of the Investigation

The impact of the focal study was substantial even before
publication of the 1966 monograph. Emerging effects were
apparent on the participating therapists and on me. I listened
to audiotapes of every session to monitor the fidelity of
treatment applications. Clients’ in-session reports of progress
in systematic desensitization were routinely positive and
consistent with the underlying theory. Participating
insight-oriented therapists began giving demonstrations of

603
desensitization and using it in their own practices well before
analyses of the objective data. Informal knowledge of the
findings spread rapidly through local psychology, speech, and
counseling departments as I analyzed the results.
Word-of-mouth proliferation quickly extended beyond the
university once committee members read the completed
dissertation.

Adoption of a utility criterion to guide clinical work. The

combination of my previously summarized personal
experiences, accumulating evidence from the literature, and
the findings of my own research finally forced me to discard
the unsupported beliefs in the active unconscious and its
clinical corollaries. I explicitly adopted a utility criterion for
guidance in the clinical arena. That is, among sets of
principles that can explain any given phenomenon, first use
those that have firm evidential support. From among those
sets of principles with empirical support, select the simplest
set that can explain both the phenomenon and provide
direction for how to change it. On that basis, I concluded that
laboratory derived principles of learning and performance,
including biological and social contexts, provide the best
working hypotheses for understanding and ameliorating the
great majority of clinical problems involving psychosocial
functioning, be it motoric, emotional, or ideational.

This paradigm shift resulted in clinical practices that require

more work on my part. I could occasionally see as many as
8-10 clients per day when I engaged in 50-minute
insight-oriented psychotherapy sessions. After the shift, my
limit for individual adult assessment and treatment became
4-5 hours per day — often using 2-hour sessions. Conducting
functional analyses, designing and maintaining treatment

604
programs, and engaging clients and their physical-social
environments in the ongoing process of reeducation and
change is simply harder work. The increased effort is justified
by routine improvements in clients’ functioning.

Additional consequences. The informal dissemination of

information about the focal study by committee members and
therapists led to my rapid entry to the talk circuit. Papers at
regional conferences, invited addresses, and job-candidate
colloquia all spread knowledge of the study and its findings
before the monograph was published. This, in turn, resulted in
networking with others who were engaged in what we now
see as the behavioral revolution. For a period of 5 or 6 years, I
participated several times each year in invited symposia with
rosters that included some combination of the contributors to
this volume and other notable players (e.g., Ted Ayllon, Nate
Azrin, Izzy Goldiamond, Fred Kanfer, Peter Lang, Gerry
Patterson, Bob Peterson, Len Ullmann, and Joe Wolpe).
These symposia were so well attended that audiences often
had standing room only. Job offers, invited chapters, and
positions of influence on review panels and policy
committees were early career consequences of the
word-of-mouth dissemination.

The field was in ferment and clearly ready for change. Most
people were interested in the clinical outcomes. Methodology
became of interest as stakeholders in competing theoretical
camps sought to embrace or dismiss the findings. The
timeliness of work in this area is further reflected by the
interest in the focal publication (Paul, 1966) and in the
chapters in Cyril Franks’ book that reviewed systematic
desensitization studies (Paul, 1969d, 1969e) and

605
methodological design and tactics (Paul, 1969a) — all of
which became citation classics.

Intellectual Autobiography:
Afterwards
My incorporation of the utility criterion resulted in
consolidation and expansion of the new paradigm following
completion of the focal study. The approach was strengthened
during a postdoctoral year in California and through
interactions with colleagues and further research at the
University of Illinois, after I returned as faculty. These
experiences also contributed to my later focus on psychoses,
mental hospitals, and mental health systems in Illinois and
Texas.

Consolidation and Expansion

Most professors at Illinois advised me to join the faculty of a
major research university immediately upon graduation.
However, a university appointment was not what I had in
mind as a career goal. University faculty, I thought, had too
much pressure to simply generate grants and publications. It
seemed that accumulation of numbers often became the
uppermost goal rather than using research to answer questions
that really made a contribution. I still intended to make
clinical practice my primary career focus. My interests had
broadened to include supervision of clinical work and access
to resources for researching questions that arose in clinical
practice. I also wanted more direct experience with psychoses
and hospitals before leaving the trainee role.

606
The willingness of my major professors to nominate me for
faculty positions at top universities was very much
appreciated. It was with some trepidation that I chose not to
follow their advice. But, there were many reasons to pursue
postdoctoral clinical training instead.

1964-65: VA postdoctoral training in Palo Alto/Menlo Park,

CA. Our earlier experience in San Diego was so positive that
Jo and I planned to return to California on completion of my
doctorate. Selection of the VA Hospital in Palo Alto/Menlo
Park for postdoctoral training was jointly determined by the
desire to live there and by the remarkable intellectual activity
in the area at that time. It was also close to Stanford
University Press, where I had submitted the manuscript for
the insight-versus-desensitization monograph. As noted
earlier, Len Ullmann introduced me to Len Krasner, who
helped arrange the postdoctoral position.

Still lacking finances, we replaced our old station wagon with

a 48-passenger school bus to transport furniture and our other
worldly goods. The bus became a moving van by removing
all but four double front seats — one each for Jo and our three
children. We licensed it as a camper. The bus trip involved
many exciting events, but those are stories for another place.
It was a bit embarrassing, after arriving in Palo Alto, when
the bus was our “family car” for 6 weeks before we were able
to replace it. Nevertheless, for the first time in our marriage,
we lived in an ordinary single-family house and participated
as regular community residents. The climate was great, the
kids loved it, and we socialized a lot.

Professionally, the postdoctoral year in Palo Alto was even

better than I anticipated. In addition to two formal 6-month

607
rotations at the VA hospital, I also did some paid consulting
and spent a few hours each week with private clients. As
noted earlier, I was also active on the talk circuit during this
period. VA trainees were treated to weekly seminars with
Stanford University faculty, Mental Research Institute staff,
and visiting professionals as well as some excellent VA staff.
Influential instructors, beyond my direct supervisors, included
Al Bandura, Walt Mischel, Ernest Hilgard, John Vitali, Don
Jackson, Paul McReynolds, and Bob Weiss. Jerry Davison
and I overlapped on the same ward for the last 6 weeks.

My first rotation was with chronically hospitalized folks on

the Menlo Park campus. This placement provided direct
experience in ward administration and staff training as well as
expanding my intervention skills. My assignment, by prior
arrangement, was in the experimental token-economy
program run by Jack Atthowe and Len Krasner (Atthowe &
Krasner, 1968). Bill Fairweather’s social-milieu unit
(Fairweather, 1964) was on the same campus. Visits there and
to traditional programs allowed immediate contrasts of
unit-wide organization and structures.

The Atthowe/Krasner unit was the first full-fledged token

economy for adults in the VA, after Ted Ayllon and Nate
Azrin (1965) reported on their Illinois program at Anna State
Hospital. Consequently, a stream of behaviorally oriented
psychologists and psychiatrists visited the ward. Those visits
and the planning involved in daily operations provided many
opportunities for discussing applications of learning theory —
and reinforcement for my shift in orientation.

My second 6-month rotation was with acutely hospitalized

folks on the Palo Alto campus. Unlike Menlo Park, explicit

608
behavioral practices were rare there. Many psychologists
were content to restrict their activities to testing.
Chemotherapy predominated. Most professionals endorsed
psychoanalytic or related psychodynamic orientations,
providing the opportunity to argue and clarify differing
positions. This helped consolidate my conversion.

For example, a series of psychological assessments were

required on this rotation. The supervising psychologist
insisted that I administer the Rorschach. Although I had been
thoroughly trained in projective techniques, I found the
literature failed to support their utility for most uses. I asked
the supervisor to please specify the problems for which
information was needed in a given case and let me select the
best means for obtaining it. When I did not use any
projectives, he asked where I would find the Rorschach more
useful than other approaches? I suggested that it might
provide information on subclinical cognitive slippage that had
not otherwise impacted functioning, if anyone were really
interested in that question outside of a research project. He
assigned a clinical case with that very question the following
week. I administered the Rorschach and wrote a short report
for him. He stopped pressing for projectives once he knew
that I could administer and interpret them. That was my last
Rorschach — ever!

My primary assignment during the Palo Alto rotation was on

a milieu unit that was one of four wards in a building run by
the Stanford University Medical School. Clientele included
high-frequency elopers, whom I often had to seek out in the
San Francisco Bay area. Experimental studies with
hallucinogens were conducted there. Jack Shelton, a Szaszian
psychiatrist, ran the ward. He was an active practitioner of

609
hallucinogens as well as an investigator of their use in
treatment. To our surprise, Jack and I often gave identical
advice to folks on the ward, although we arrived there from
different conceptualizations. His non-disease, “myth of
mental illness” approach (Szasz, 1961) was quite compatible
with my own.

Weekly grand rounds with the combined staff of all four

wards provided opportunities to contrast approaches. One of
the other wards was organized as a therapeutic community. It
was run by Rudy Moos, a psychologist who became well
known for the study of ward atmosphere. A third-generation
psychoanalyst, a psychiatrist, ran the other two wards. As the
highest ranking VA professional in the building, the analyst
felt obliged to give summaries at the end of case
presentations. These often included self-styled insights, such
as, “What we see here is a problem in impulse control
deriving from unconscious anal conflict, for which massive
doses of Thorazine are required to detoxify the id.”

Such nonsensical discourse provided many occasions to

consider contrasting conceptual principles. Open discussion
of differing interpretations became even more active after
Jerry Davison was assigned to our ward. Jack Shelton, Jerry,
and I then regularly took seats in opposing corners of
grand-rounds audiences to ask challenging questions when
unsupported opinions were asserted as facts. It was great fun
and further reinforced use of the utility criterion.

Another change in plans. Following postdoctoral training, the

original plan had been to obtain a position in California that
would allow time for private practice and perhaps a day per
week for applied research. Once again, plans changed as a

610
result of new experiences. The California climate was great. I
enjoyed living on the edge of the “hip” culture, while
maintaining more traditional values. However, Jo and I did
not like the ideals that our children were developing. We
decided that it would be better to raise the children in a
community with more traditional mores than we observed in
the bay area, especially with its high rate of broken homes.

I also found that I had come to equate time with money

during the year of full-time applied work, to the extent that
another client contact always won out over reading another
journal article. In fact, I had fallen behind in the literature and
I did not like that at all! I also came to a disturbing conclusion
after presentations at several universities, public and private
hospitals, and other mental health agencies. Namely,
universities might be the only practical settings to allow all
three of the major activities that I had come to value —
clinical practice, training, and research.

Although I was ambivalent about committing to a faculty

position, a university that devoted equal resources and
reinforcement to clinical training and applied research seemed
to be worth a try. I received job offers from several top
universities, but few of them placed the strong emphasis on
clinical training that I wanted. An offer from Stanford
University was tempting, but I declined it as they had
terminated their clinical training program in psychology —
and, paradoxically, the location had become undesirable as
long as our children were young.

It was a surprise when Lloyd Humphreys inquired about my

interest in returning to Illinois. He contacted me late in the
year for recruiting. He was clearly ambivalent about violating

611
the policy against hiring the department’s own graduates. I
was also ambivalent about returning. The department
supported the three activities that I sought, but engaging in all
of them would entail 75-80 hour workweeks. Did I want to
continue working that hard? Not really! I was also a bit
concerned about switching from student to faculty roles.
However, discussions with the major players convinced me
that this would not be a problem. The Illinois climate was
lousy but the community was nearly ideal for child rearing
and socializing in ways that we all enjoyed. Jo and the kids
thought returning to Champaign-Urbana would be going
home.

1965-1980: University of Illinois at Urbana-Champaign. I had

planned to conduct long-term follow-ups of the subjects from
my dissertation study. That would be easier from Illinois. We
decided to return for a 2-year trial. That period of time should
allow completion of the follow-up studies and a good test of
job satisfaction as a faculty member. We remained at the
University of Illinois at Urbana-Champaign for 15 years. I
continued a private practice with individual clients and
became active in hospital consulting.

Teaching graduate courses in behavior disorders and

interventions and supervising clinical practica involved
immersion in the empirical and theoretical literature. The
accumulating publications continued to undercut the basic
assumptions of psychodynamic theory and to reinforce the
utility of the alternate paradigm. The 1960’s literature was
rich in this regard. It included key articles (e.g., Baer, Wolf,
& Risley, 1968; Eriksen & Pierce, 1968; Wilson, 1963) as
well as books that became classics (e.g., Bandura, 1969;
Bandura & Walters, 1963; Bijou & Baer, 1961; Franks, 1969;

612
Kanfer & Phillips, 1970; Krasner & Ullmann, 1965; Patterson
& Gullion, 1968; Ullmann & Krasner, 1965; Wolpe &
Lazarus, 1966).

An exceptionally strong faculty in the Illinois department and

affiliated institutes provided collegial support for the
expanded application and evaluation of laboratory based
principles. Wes Becker (who converted to a radical
behaviorist position while I was in Palo Alto), Charles
Eriksen, Hobart Mowrer, and Len Ullmann were all
continuing members of the faculty whose work and personal
relationships directly reinforced my change in orientation.
Several additional behavioral clinicians joined the faculty as
well, including Doug Bernstein, Sid Bijou, John Gottman,
Fred Kanfer, Bob Peterson, Bob Nay, Bill Redd, Bob
Sprague, and Warren Steinman.

Given the size of the Illinois faculty, those who identified

with behavior therapy or behavior modification were still a
minority. However, the thread that bound the entire faculty
together was a commitment to interactional models of human
functioning and the primacy of empirical evidence for guiding
practices. Those values supported my expanding efforts.

In this context, my own studies of desensitization and related

techniques continued to document the utility of a behavioral
or social-learning approach. Besides the 2-year follow-up of
subjects from the focal investigation (Paul, 1967a), these
studies included evaluation of systematic desensitization in
groups (Paul & Shannon, 1966) and long-term follow-up of
those clients (Paul, 1968). A series of investigations on
relaxation training and hypnosis further clarified the
components of effective treatment procedures for anxiety

613
related problems (Evans & Paul, 1970; Paul, 1969b, 1969c,
1969f; Paul & Trimble, 1970).

Many of these articles were reprinted over the years, long

after my research interests had turned elsewhere. My last
publications to specifically focus on systematic
desensitization and related techniques appeared in the early
1970s. Co-authored with Doug Bernstein, these included a
critique of analogue treatment studies (Bernstein & Paul,
1971) and a monograph on treatment of anxiety related
problems (Paul & Bernstein, 1973). Both of these have been
reprinted as well.

I continue to use these empirically based procedures in my

own clinical work and to train practicum students in their
application. At last count, I had personally used or supervised
application of systematic desensitization and/or progressive
relaxation training with more than 4000 clients. I try to keep
up with the literature on treatment of sexual and other anxiety
related problems. However, my interest waned for conducting
further research on these topics, once I found the
conceptualizations and technology to be routinely effective in
clinical practice.

Psychoses, Mental Hospitals, and

Mental Health Systems
When I joined the faculty at the University of Illinois,
concurrent consulting work put emphasis on the absence of
knowledge to support effective clinical practices in mental
hospitals. My conversion was complete regarding the
spuriousness of the basic assumptions underlying

614
psychoanalytic practices. However, the utility criterion still
called for empirical answers in those areas of practice where
firm evidence was lacking. Traditional practices were clearly
inadequate in the treatment of psychoses, especially for
people who were chronically institutionalized. This attracted
and maintained my interest. In fact, my work has primarily
focused on institutionalized populations, observational
assessment, and treatment of psychoses for more than 30
years.

1968-1984: Adolf Meyer Mental Health Center, Decatur, IL.

For 5 years, I directed the Psychosocial Rehabilitation Unit at
the Adolf Meyer Mental Health Center. The same staff on
parallel residential treatment wards conducted Milieu Therapy
and Social-Learning Programs. The unit was responsible for
aftercare of discharged clientele as well. The research
component was funded by a federal grant that I obtained to
operationalize and evaluate the most promising approaches
for treatment of people who were mental patients with the
most severe disabilities. Comparison programs using
traditional treatment approaches were located in a separate
hospital.

After our treatment wards were terminated due to political

changes in the state, I continued as director of the
Clinical-Research Unit at the Meyer Center for another 11
years. This unit served as a base for continued follow-up of
discharged clients as well as for grant-supported research and
development of new observational assessment technologies.
The latter work entailed statewide data collection and
examination of mental health system operations.

615
The expansion to inpatient populations drew heavily on the
work of colleagues who shared their ideas and innovations
through personal interactions. Len Ullmann, Wes Becker,
Don Peterson, Bob Peterson, Jerry Wiggins, Lloyd
Humphreys, Sid Bijou, Lew Kurke, Bernie Wagner, Joe
Williams, and John Nolte all contributed ideas to the
development of inpatient assessment and treatment
procedures. My coworkers and I sought guidance from Elaine
Cumming and Alan Kraft on milieu practices. I had direct
experience with the inpatient programs developed by Ted
Ayllon and Nate Azrin, Jack Atthowe and Len Krasner, and
Bill Fairweather. In the design of treatment programs, my
coworkers and I incorporated procedural innovations from the
work of all of these investigators as well as those developed
by Ogden Lindsley, summarized in this volume. I am pleased
to acknowledge other contributors to this volume — Don
Baer, Monte Wolf, and Todd Risley — for providing the
seminal technologies and templates with children that we
adapted for use in developing effective treatment programs
with adults.

I was the principle investigator and director of the

clinical-research group that undertook the expanded work
through the Meyer Center. That work, however, was a
collaborative effort with a host of coworkers. Many were
graduate students and interns who concurrently completed
theses or dissertations as part of the ongoing operations.
Senior clinical staff, who shared in the development and
conduct of the treatment programs over the years, included:
Titus McInnis, Beverly Holly, Dick Hagen, Ed Craighead,
Kay Davidson, Jim Calhoun, Jim Curran, Al Litrownik, Dave
Doty, Chris Power, Howard Himmelstein, Bill Kohen, Dale

616
Theobald, Bob Paden, Carolyn Paden, Paula Griffith, Peggy
Maynard, and Ralph Trimble.

Bob Lentz was a member of the clinical senior staff and

supervisor of research personnel through completion of the
residential treatment evaluations. Graduate research
assistants, who contributed to the development of instruments
and collection of assessment data, included: Dean Orris,
Lester Tobias, George Montgomery, Roger Knudsen, Rich
Edelson, Pat Vogel, Connie Duncan-Johnson, and Al
Porterfield. Marco Mariotto, Joel Redfield, and Mark Licht
were graduate research assistants who also took on the
statistical analyses of the 6-year long comparative treatment
evaluations (Paul & Lentz, 1977, 2001).

Mark Licht, Chris Power, Kathryn Engel, and Marco Mariotto

continued as coworkers and collaborators through the
Clinical-Research Unit and beyond. These former students
and current colleagues were intimately involved with the
practical development of the new observational assessment
system. They were also responsible for the collection of data
in the multi-institution samples to evaluate the feasibility and
generalizability of the component instruments.

Many methodological and practical contributions emerged

from the work at the Meyer Center. I found the
laboratory-derived principles and theoretical formulations
with greatest utility for anxiety-related problems were also the
best for understanding and treating people’s problems that
were classified under the rubric of “schizophrenia” (Paul,
1974). My earlier formulation of the ultimate clinical question
and necessary domains and classes of variables for research
with outpatient treatments was extended to inpatient

617
psychosocial and biomedical procedures and, ultimately, to
the operation of treatment facilities and entire systems of
service (Paul, 1986a, 1986b; Paul & Lentz, 1977, 2001).

Perhaps the most notable contribution from our Meyer Center

work is a psychosocial treatment program that works for
people who are the most severely disabled and chronically
institutionalized of all mental patients (Paul, Stuve, &
Menditto, 1997). This program — the Social-Learning
Program — is the result of incorporating the principles and
procedures from the work of previous researchers into a
comprehensive, unit-wide treatment program. It is established
as the treatment-of-choice on both absolute and comparative
grounds, with superior cost-efficiency as well as effectiveness
(Paul & Menditto, 1992). This program is ready for adoption
in ongoing services as soon as the supporting assessment
systems are available.

The basic developmental work on the observational

assessment technology was also completed at the Meyer
Center and other facilities in the state of Illinois. The quality
of instruments had been demonstrated in earlier publications
(e.g., ω ²s in the high .90s for reliabilities and prediction of
discharge-readiness; rs in the .50s to .70s for prediction of
postdischarge functioning in the community). However,
analyses of the massive data set for evaluation of the
instruments, prior to release for use elsewhere, was still in
progress. Also, we needed to complete the materials for
training others to implement an assessment and monitoring
system that has been characterized as “revolutionary.”

1980-present: University of Houston, Houston, TX. Jo and I

moved to Texas in 1980. The University of Houston received

618
funding to build a few prominent departments, and
psychology was targeted as one of them. More than half of
the faculty was newly recruited into a department that was to
explicitly focus on solving applied problems. A Cullen
Distinguished Chair and a group of stellar colleagues were
part of the offer. Early on, the clinical program counted Dale
Johnson, John Vincent, Roger Maley, Marco Mariotto, Lynn
Rehm, and Len Ullmann among the faculty working on adult
problems, with interests in the seriously mentally ill. The
strongest incentive was the promise of establishing a
combined service-research-training-demonstration center at
Austin State Hospital in which to continue our work and
disseminate findings (see Paul, 1990). Jo was willing to
assume the secretarial duties of our clinical-research project.
Houston had no snow. The kids were grown. It all looked
very promising.

Unfortunately, the Austin center was trashed after 2 years of

development. The Commissioner of the Texas Department of
Mental Health and Mental Retardation (TDMHMR), with
whom we had established agreements, fell into political
disfavor. His replacement viewed “a proper DSM diagnosis
and the right drug” as all that was needed for mental patients
— certainly not a psychosocial unit run by people who were
not even employees of his department. He terminated our
agreements with Austin State Hospital and TDMHMR. He
also refused to sign collaborative agreements with NIMH.
That refusal killed grant support for our continuing operations
in Illinois as well as planned expansions in Florida and Texas.

Following that disappointment, my co-authors and I

committed our time to completing the assessment materials
that would allow others to implement the Social-Learning

619
Program. Marco Mariotto, Joel Redfield, Mark Licht, Chris
Power, and I completed the theoretical analyses that underlies
our science-based assessment approach (Paul, 1986b). We
continued data analyses and development of the observational
assessment technology with the assistance of our graduate
students. The data set contains information on more than 1200
inpatients and 800 clinical staff in 35 different treatment
units. It has served as a rich source of data for theses and
dissertations.

Graduate seminars on our work have been preparing young

psychologists to assist with dissemination, as the new
technology becomes available. In addition to the
developmental group noted below, several former students
and current colleagues have remained affiliated with our
clinical-research group. These include Bob Lentz, Chris
Power, Kathryn Engel, Tony Menditto, Mark Schade, Gail
Brothers Braun, Jan Cross, and Julian Salinas. With
prepublication access to the assessment materials, Mark
Licht, Mark Schade, and Tony Menditto have all spearheaded
early implementations of the Social-Learning Program in
different states. Under the direction of Tony Menditto,
programs in Missouri are the most advanced at the time of
this writing.

The new assessment technology was developed into the

Computerized TSBC/SRIC Planned-Access Observational
Information System (summarized in Mariotto, Paul, & Licht,
1995; Paul, 1987b). This assessment and monitoring system
goes beyond support for the Social-Learning Program. It also
stands, independently, as the best way to help residential
treatment operations approach the status of an applied
science. Three parts of a five-part series on the TSBC/SRIC

620
System have been published (Paul, 1986b, 1987a, 1988). At
the time of this writing, the developmental group led by Mark
Licht is upgrading the TSBC/SRIC System computer
programs. Coworkers in the group include Paul Stuve, James
Coleman, Will Newbill, and Susan Hall. They are converting
programs to a more powerful and user-friendly database
system before we finish the final version of implementation
materials (Paul, 2001a, 2001b). The TSBC/SRIC System has
even more potential to improve ongoing practices than just
having an effective treatment program for previously
untreatable clientele. It will allow treatment facilities to offer
ongoing services that are not only “new and improved” but
“ever improving” (Paul, Stuve, & Cross, 1997).

Reflections and Recommendations

It has been quite a trip so far. What does it all mean? Can
there be “take-home” messages that go beyond a list of
platitudes and proverbs? The following includes a few things
that struck me as I reflected on both the journey and the
current state of the field. After reflections, I offer some
recommendations derived from my in-route experiences.
Others have said most of these things. I simply offer them as
my own observations and beliefs, without attempting
scholarly references to evidence, original sources, or to others
who have expressed similar notions.

Reflections
I am struck by the degree to which my nonacademic life
influenced my professional work. Family values, emotional

621
experiences, and good and bad models of deportment all play
the expected role. However, chance events and decisions
based on irrelevant factors appear more influential than I
imagined. Long-range planning is risky. Things change. It
seems best to keep options open for as long as possible.

The role of political factors is noteworthy, especially for work

involving residential units and mental health systems.
Multiple stakeholders all want to protect their interests. Most
political decisions are based on factors that should be
peripheral. Gerry Klerman, former director of the Alcohol,
Drug Abuse, and Mental Health Administration, suggested
that clinical researchers need one trait if they hoped to bring
about real change—tenacity in the face of adversity. I use the
analogy of a broken-field runner in football to illustrate how
therapists can direct a clinical session without being directive.
Sometimes he zigs, sometimes he zags, sometimes he even
runs in the opposite direction — but he is always focused on
reaching the goal. The broken-field-runner also appears to be
an apt analogy for our attempts to bring about change in the
public mental health system.

The dissemination of behavioral interventions that could be

imported to the lone clinician’s office stands in sharp contrast
to the political barriers just noted. These procedures were
often adopted more rapidly than justified by the scientific
evidence. In other cases, valid techniques were adapted and
applied without the necessary training and rigor. It was a
heady experience to be part of the early symposia, with
packed rooms and active questioning. Unfortunately, the new
techniques described by the careful research of those years
were often adapted as a “bag of tricks.” The necessary change
in concepts and principles for understanding clinical problems

622
— the real behavioral revolution — did not happen in those
instances.

Things that still go “bump” in the night. Psychoanalytic

theory no longer dominates psychiatry and psychology but it
continues to flourish in the humanities and theatre arts.
Psychoanalytically oriented treatment is clearly alive, if not
well. Many still fail to attend to the empirical literature. This
failure allows them to maintain a religious commitment to the
active unconscious and its corollaries — and continue to
argue articles of faith. They clearly have not shared the
corrective emotional experiences of my history.

Unfortunately, the remedicalization of psychiatry seeks to

supplant the disease analogies of psychodynamic thought
with return to another religion — the presumption that all
human problems result from defects that are based on real
genetic or physical diseases. Criteria from the most recent
round of voting are codified in the Diagnostic and Statistical
Manual (DSM) and called a “nosology.” Things people do are
used to assign categories that refer to things they purportedly
have — providing a pseudo-explanation, where naming
activities supposedly accounts for their existence. Does the
public notice this? Yes! Even in cartoons — Leroy Lockhorn
says that Loretta cooks poorly because she “has” a “cooking
disorder.” Remedicalized professionals consider the disorders
so named to be putative biological disease entities, for which
drug treatment is assumed to be the proper intervention.
“Biochemical imbalance” of the brain has become the
panchreston, or explain-all-that-explains-nothing, of modern
psychiatry.

623
Evidence-based practices. I endorse the past decade’s
initiatives to identify and promote empirically validated
treatments as well as the fledgling work on empirically
validated assessments. More generally, science-based practice
and policy guidelines and evidence-based practices should be
the best corrective to “things that still go ‘bump’ in the night.”

My work has been identified among those that contributed to

the development of these movements. I am pleased to offer
Paul (1966) and Paul and Lentz (1977, 2001) and their
follow-ups as practical models of the kind and scope of
research that is needed to establish the comparative
effectiveness of psychosocial and biomedical treatments. The
manuals provide operational definitions of therapeutic
procedures. They do so at a level of specification that allows
training in the artful application of established principles and
techniques, without being oversimplified “cookbooks.” The
assessment procedures provide converging evidence of
reliably documented phenomena. They do so at a level of
specificity that allows identification of aspects of functioning
that are change-worthy as well as detection of the presence or
absence of change.

As a note of caution, however, everyone should carefully

examine the criteria used by professional groups to identify
and sanction clinical procedures. Rather than undertaking a
construct-validation approach, which fits the task, committees
often develop rigid categorical checklists to ease their work. It
is with some chagrin that neither of the award-winning
monographs referenced above was included in the database of
official psychiatric or psychological task forces that produced
the first lists of effective treatments. Both groups defined the
domain of evidence to include only articles in peer-reviewed

624
journals. The psychology task force later included journal
follow-ups and summaries of these studies as valid evidence.
However, psychiatric groups continue to exclude these works,
often on the basis that treatments were not specific to a single
DSM diagnosis. Psychiatric guidelines are mostly for
psychotropic drugs. It would be better if science-based rules
of evidence played a greater role in these undertakings.

Recommendations
My major recommendation for all workers in the area is
captured in the old adage, “Anything worth doing is worth
doing well.” Whether the activity involves learning new
things, conducting research to solve problems, or engaging in
clinical work to help others — do it right. Be responsible. Be
thorough. Be rigorous. Do not take the easy way if a hard way
is required.

Try to establish an environment that is rich in personal

satisfactions and current reinforcers. External reinforcement,
such as money, publications, promotions, or awards, is far too
delayed to maintain good work let alone improve it.
Avoidance schedules produce behavior but positive
reinforcement feels better. Apply knowledge from life
experiences to your discipline and vice versa. Learn what
your reinforcers are and try to build new ones. Then use them
to achieve balance. Take care of the important personal things
— health, family, and friends — without which any degree of
professional success will seem trivial.

Suggestions for students and young professionals. Beyond the

recommendations just noted, my practical experiences

625
provide some bases for suggestions to those who are just
entering the field — students, young investigators, and young
practitioners.

Take advantage of all possible educational experiences during

your limited time in the trainee role. The best way to
influence other professionals as well as clients is to
understand their conceptual framework and the reinforcement
system within which they operate. Do not let premature
commitments to a career path or orientation limit your
exposure to new ideas and practices. Always evaluate the
evidence. Study, learn, and understand materials whether or
not you find them intrinsically interesting or immediately
applicable. They may become useful in ways you never
imagined. They may even contribute to changes in direction
of your own path.

Select problems for research about which you really care.

Although passion has been described as the source of major
achievements, luck also plays a role. You need not be
passionate about the undertaking, but you should have a
genuine interest in obtaining answers to the questions
addressed by your research. This will help carry you through
the inevitable hassles and pure drudgery that careful scientific
work entails, especially for theses and dissertations. Of
course, the problems should be ones that others care about as
well. This is particularly true if you need outside funding to
support your work. The greater the number of stakeholders
concerned about the outcomes, the greater the degree of
interest your research is likely to attract.

Make sure you have phenomena that can be reliably identified

and measured. Do not become enamored with constructs that

626
vanish under careful scrutiny. In most areas, a series of
programmatic studies will probably make a greater
contribution than isolated ones. However, investigations
undertaken as part of degree requirements must be completed
in a timely fashion. It is generally better to begin sequential
investigations, where findings build on one another, than to
try to answer all questions in a single study.

Suggestions for established investigators and practitioners. I

do not presume to give advice to established investigators and
practitioners unless they ask me. However, I encourage those
who are well established in their discipline to consider some
of the following suggestions that I give myself on occasion.
Others will determine whether or not I follow my own advice.

Society should benefit from policies and practices that are

informed by empirically based knowledge. Disseminate the
products of your research. Just publishing findings is not
enough. Collaborative negotiation is usually the best way to
work with others. “Shape, don’t rape,” is a worthy proverb
that I first heard from Len Ullmann. When rational discourse
fails, however, senior people in a discipline can risk offending
others in ways that junior people cannot. There is enough
absurdity in the world that care must be taken to avoid simply
becoming a chronic irritant. But do not just
go-along-to-get-along when the issues are critical. Actively
speak out on important things when you see “the emperor has
no clothes.” Provide leadership by example.

Surround yourself with energetic young people. They can

help you from stagnating, personally and professionally. Be
sure that those just entering the discipline become aware of
common values and practices that underlie current procedures

627
as well as those that are historically important — not just the
new findings and technologies from recent publications. Do
not rest on your laurels. Continue to evaluate the evidence.
Continue to learn. Never fully retire.

Share what you know but recognize others for their

knowledge and contributions. Try to be aware of when and
where colleagues should take the major responsibility for
continuing work that you may have started. The guru role,
without the burden of day-to-day responsibility for activities,
has many attractive features. In the words of my intellectual
forefather, William James, “The great use of life is to spend it
for something that will outlast it.”

The “Ultimate” Answer

I originally formulated the “ultimate clinical question(s)” to
summarize the domains and classes of variables needing
description, measurement, or control — if firm evidence were
to be obtained and accumulated across studies of
psychotherapies and/or psychotropic drugs. The question,
“What treatment, by whom, is most effective for this
individual, with that specific problem, under which set of
circumstances, and how does it come about?” could, of
course, never be entirely answered. It was intended to guide
investigators and practitioners. The extension of the question
and organizational scheme of variables to inpatient and
biomedical treatments as well as entire facilities and systems
of service has also proven to be useful (see Hayes, 1991).
Some instructors, I am told, even treat the ultimate question
as a near mantra, requiring their students to memorize the
words.

628
My final recommendation is this. Everyone should use the
ultimate question for guidance, but add the “ultimate” answer
as well. That should help to maintain focus on the
interactional complexities of clinical phenomena. What is the
ultimate answer? “It depends!”

References
Atthowe, J. M., & Krasner, L. (1968). Preliminary report on
the application of contingent reinforcement procedures (token
economy) on a “chronic” psychiatric ward. Journal of
Abnormal Psychology, 73, 37-43.

Ayllon, T., & Azrin, N. H. (1965). The measurement and

reinforcement of behavior of psychotics. Journal of the
Experimental Analysis of Behavior, 8, 357-383.

Baer, D. M., Wolf, M. M., & Risley, T. R. (1968). Some

current dimensions of applied behavior analysis. Journal of
Applied Behavior Analysis, 1, 91-97.

Bandura, A. (1961). Psychotherapy as a learning process.

Psychological Bulletin, 58, 143-159.

Bandura, A. (1969). Principles of behavior modification. New

York: Holt.

Bandura, A., & Walters, R. H. (1963). Social learning and

personality development. New York: Holt.

Bernstein, D. A., & Paul, G. L. (1971). Some comments on

therapy analogue research with small animal “phobias.”

629
Journal of Behavior Therapy and Experimental Psychiatry, 2,
225-237.

Bijou, S. W., & Baer, D. (1961). Child development (Vol. 1).

New York: Appleton-Century-Crofts.

Dollard, J., & Miller, N. E. (1950). Personality and

psychotherapy. New York: McGraw-Hill.

Eriksen, C. W., & Pierce, J. (1968). Defense mechanisms. In

E. F. Borgotta & W. N. Lambert (Eds.), Handbook of
personality theory (pp.1007-1040). New York:
Rand-McNally.

Evans, M. B., & Paul, G. L. (1970). Effects of hypnotically

suggested analgesia on physiological and subjective responses
to cold stress. Journal of Consulting and Clinical Psychology,
35, 362-372.

Eysenck, H. J. (1952). The effects of psychotherapy: An

evaluation. Journal of Consulting Psychology, 16, 319-324.

Eysenck, H. J. (1961). The effects of psychotherapy. In H. J.

Eysenck (Ed.), Handbook of abnormal psychology (pp.
697-725). New York: Basic Books.

Fairweather, G. W. (Ed.). (1964). Social psychology in

treating mental illness: An experimental approach. New
York: Wiley.

Franks, C. M. (Ed.). (1969). Behavior therapy: Appraisal &

status. New York: McGraw-Hill.

630
Hayes, S. C. (1991). Pursuing the ultimate clinical question:
An interview with Gordon L. Paul. The Scientist-Practitioner,
1(3), 6-16.

Kanfer, F. H., & Phillips, J. S. (1970). Learning foundations

of behavior therapy. New York: Wiley.

Krasner, L., & Ullmann, L. P. (Eds.). (1965). Research in

behavior modification. New York: Holt.

Marriotto, M. J., Paul, G. L., & Licht, M. H. (1995).

Assessment in inpatient and residential settings. In J. N.
Butcher (Ed.), Clinical personality assessment: Practical
approaches (pp. 435-459). New York: Oxford University
Press.

Nichols, T. (2001). The importance of case studies to

methodology of science. In W. T. O’Donohue, D. Henderson,
S. C. Hayes, J. Fisher, & L. J. Hayes (Eds.), A history of the
behavioral therapies: Founders’ personal theories. Reno,
NV: Context Press.

Patterson, G. R., & Gullion, M. E. (1968). Living with

children. Champaign, IL: Research Press.

Paul, G. L. (1963). Production of blisters by hypnotic

suggestion: Another look: Psychosomatic Medicine, 25,
233-244.

Paul, G. L. (1966). Insight versus desensitization in

psychotherapy: An experiment in anxiety reduction. Stanford,
CA: Stanford University Press.

631
Paul, G. L. (1967a). Insight versus desensitization in
psychotherapy two years after termination. Journal of
Consulting Psychology, 31, 333-348.

Paul, G. L. (1967b). The strategy of outcome research in

psychotherapy. Journal of Consulting Psychology, 31,
109-118.

Paul, G. L. (1968). Two-year follow-up of systematic

desensitization in therapy groups. Journal of Abnormal
Psychology, 73, 119-130.

Paul, G. L. (1969a). Behavior modification research: Design

and tactics. In C. M. Franks (Ed.), Behavior therapy:
Appraisal and status (pp. 29-62). New York: McGraw-Hill.

Paul, G. L. (1969b). Extroversion, emotionality, and

physiological response to relaxation training and hypnotic
suggestion. International Journal of Clinical and
Experimental Hypnosis, 17, 89-98.

Paul, G. L. (1969c). Inhibition of physiological response to

stressful imagery by relaxation training and hypnotically
suggested relaxation. Behavior Research and Therapy, 7,
249-256.

Paul, G. L. (1969d). Outcome of systematic desensitization I:

Background, procedures and uncontrolled reports of
individual treatment. In C. M. Franks (Ed.), Behavior
therapy: Appraisal and status (pp. 63-104). New York:
McGraw-Hill.

632
Paul, G. L. (1969e). Outcome of systematic desensitization II:
Controlled investigations of individual treatment, technique
variations, and current status. In C. M. Franks (Ed.), Behavior
therapy: Appraisal and status (pp. 105-159). New York:
McGraw-Hill.

Paul, G. L. (1969f). Physiological effects of relaxation

training and hypnotic suggestion. Journal of Abnormal
Psychology, 74, 425-437.

Paul, G. L. (1974). Experimental-behavioral approaches to

“schizophrenia.” In R. Cancro, N. Fox, & L. Shapiro (Eds.),
Strategic intervention in schizophrenia: Current
developments in treatment (pp. 187-200). New York:
Behavioral Publications.

Paul, G. L. (1986a). Can pregnancy be a placebo effect?:

Terminology, designs, and conclusions in the study of
psychosocial and pharmacological treatments of behavior
disorders. Journal of Behavior Therapy and Experimental
Psychiatry, 17, 524-544.

Paul, G. L. (Ed.). (1986b). Principles and methods to support

cost-effective quality operations: Assessment in residential
treatment settings, Part I. Champaign, IL: Research Press.

Paul, G. L. (Ed.). (1987a). Observational assessment

instrumentation for service and research — The Time-Sample
Behavioral Checklist: Assessment in residential treatment
settings, Part 2. Champaign, IL: Research Press.

Paul, G. L. (1987b). Rational operations in residential

treatment settings through ongoing assessment of client and

633
staff functioning. In D. R. Peterson & D. B. Fishman (Eds.),
Assessment for decision (pp. 145-203). New Brunswick, NJ:
Rutgers University Press.

Paul, G. L. (Ed.). (1988). Observational assessment

instrumentation for service and research — The
Staff-Resident Interaction Chronograph: Assessment in
residential treatment settings, Part 3. Champaign, IL:
Research Press.

Paul, G. L. (1990). The role of the National Institute of

Mental Health in attracting doctoral-level talent and
improving training, research, and services: A proposal. In D.
L. Johnson (Ed.), Service needs of the seriously mentally ill:
Training implications for psychologists (pp. 45-50).
Washington, DC: American Psychological Association Press.

Paul, G. L. (2000). Evidence-based practices in inpatient and

residential facilities. The Clinical Psychologist, 53, 3-16.

Paul, G. L. (Ed.). (2001a). Observational assessment

instrumentation for service and research — The
Computerized TSBC/SRIC Planned-Access Observational
Information System: Assessment in residential treatment
settings [Part 4]. Manuscript in preparation, University of
Houston.

Paul, G. L. (Ed.). (2000b). Observational assessment

instrumentation for service and research — The TSBC/SRIC
System implementation package: Assessment in residential
treatment settings, Part 5. Manuscript in preparation,
University of Houston.

634
Paul, G. L., & Bernstein, D. A. (1973). Anxiety and clinical
problems: Treatment by systematic desensitization and
related techniques. Morristown, NJ: General Learning Press.

Paul, G. L., & Eriksen, C. W. (1964). Effect of anxiety on

“real-life” examinations. Journal of Personality, 32, 480-494.

Paul, G. L., Eriksen, C. W., & Humphreys, L. G. (1962). Use

of temperature stress with cool air reinforcement for human
operant conditioning. Journal of Experimental Psychology,
64, 329-335.

Paul, G. L., & Lentz, R. J. (1977). Psychosocial treatment of

chronic mental patients: Milieu versus social-learning
programs. Cambridge: Harvard University Press.

Paul, G. L., & Lentz, R. J. (2001). Psychosocial treatment of

chronic mental patients; Milieu versus social-learning
programs [2nd ed.]. Manuscript in preparation, University of
Houston.

Paul, G. L., & Menditto, A. A. (1992). Effectiveness of

inpatient treatment programs for mentally ill adults in public
psychiatric facilities. Applied and Preventive Psychology:
Current Scientific Perspectives, 1, 41-63.

Paul, G. L., & Shannon, D.T. (1966). Treatment of anxiety

through systematic desensitization in therapy groups. Journal
of Abnormal Psychology, 71, 123-135.

Paul, G. L., Stuve, P., & Cross, J. V. (1997). Real-world

inpatient programs: Shedding some light — A critique.

635
Applied and Preventive Psychology: Current Scientific
Perspectives, 6, 193-204.

Paul, G. L., Stuve, P., & Menditto, A. A. (1997).

Social-learning program (with token economy) for adult
psychiatric inpatients. The Clinical Psychologist, 50, 14-17.

Paul, G. L., & Trimble, R. W. (1970). Recorded versus “live”

relaxation training and hypnotic suggestion: Comparative
effectiveness for reducing physiological arousal and
inhibiting stress response. Behavior Therapy, 1, 285-302.

Reichenbach, H. (1938). Experience and prediction. Chicago:

University of Chicago Press.

Szasz, T. S. (1961). The myth of mental illness. New York:

Hoeber-Harper.

Ullmann, L. P., & Krasner, L. (Eds.). (1965). Case studies in

behavior modification. New York: Holt.

Wilson, R. S. (1963). On behavior pathology. Psychological

Bulletin, 60, 130-146.

Wolpe, J. (1958). Psychotherapy by reciprocal inhibition.

Stanford, CA: Stanford University Press.

Wolpe, J., & Lazarus, A. A. (1966). Behavior therapy

techniques. New York: Pergamon.

636
Footnote
1
Preparing the prelude to my intellectual autobiography
involved revisiting aspects of my youth that are laden with
emotions. Some involve people and experiences that I simply
had not recently thought about — at least, not regarding their
impact on my own development. Others involved painful
experiences that I had diligently worked to suppress. The
redintegration of strong emotional reactions, both positive and
negative, that I experienced in considering these aspects of
my history — even while writing this chapter — suggests that
I have, indeed, selected relevant material for inclusion.

637
Chapter 18

Values and Constructionism

in Clinical Assessment: Some
Historical and Personal
Perspectives on Behavior
Therapy1
Gerald C. Davison

University of Southern California

It’s been said that key decisions in life are determined or at

least strongly influenced by unforeseen chance events. A
social gathering we decide only at the last minute to attend
turns out to be the place we meet our future spouse. A
careless moment while driving leads to a terrible accident that
affects our health and our family’s well-being for the rest of
our lives. Though the idea that chance events play a major
role in shaping our existence may not fully satisfy our
existential needs, I’ve been struck many times by how
germane this perspective is in reflecting upon the careers of
many of my friends and colleagues. It certainly applies to
mine.

The task set for participants in the Reno Conference on the

History of Behavior Therapy was to present what we see as

638
the formative influences in our professional lives, discuss a
publication that we believe has had some importance in
behavior therapy, and reflect on the nature of that influence
on the field. This paper is an effort to fulfill this unusual and
intriguing assignment.

My High School and College Years

I spent grades 7 through 12 at Boston Latin School, at the
time not a particularly reinforcing or supportive secondary
school and known for a number of notable graduates like
several signers of the Declaration of Independence, among
them Samuel Adams and John Hancock. Another signer was
Benjamin Franklin, who enrolled in the school in 1714 and
was doing very well when his father withdrew him after just
one year. It seemed that Josiah Franklin did not consider his
son pious enough for the ministry, which was the profession
that most of the boys were oriented towards after they
graduated and went on to Harvard. Other well-known
graduates were Cotton Mather, George Santyana, Ralph
Waldo Emerson, Arthur Fiedler, and Leonard Bernstein. Not
too much pressure on contemporary students! For generations
this school has been the way out of several of the Boston
ghettoes for the children of parents and grandparents who
immigrated from Europe and who saw a rigorous education as
the most reliable way for the kids to make it into the
mainstream of American society.

Nearly all of my 240 classmates of the class of 1957 went to

college, and about a fifth of these entered Harvard as I did.
Like many other Jewish boys, I was supposed to become
either a physician or a lawyer. I spent the first year studying

639
German, Russian, political science, biology, western
civilization, and other general education topics that were
supposed to enable me to declare a major (or “concentration,”
the term favored by Harvardians). By the beginning of my
sophomore year, however, I had managed only to reject
political science as a major and to locate myself in the
German department, where I found myself intrigued more by
the characters in the novels I was reading than by the
language or whatever else it was that a literature major was
supposed to find interesting.

Then something unexpected happened during the first day of

classes of my sophomore year. Having read during the
previous summer Freud’s Clark University lectures and being
both drawn to and annoyed by Freud’s ingenious speculations
about people’s putative unconscious motivations, I found
myself deciding at the last minute to drop into a class that was
sandwiched between my 9:00 and 11:00 AM lectures. It was
called Social Relations 10, the first of two semesters of a
massive introductory course in a department that had been
created after the second world war as a combination of
anthropology, sociology, and psychology. There was also a
department called Psychology, where Skinner was situated,
but there was little more than animosity and mutual suspicion
between the pigeon and rat-runners of Psychology and those
more interested in complex human interactions in Social
Relations.

So, with a long-standing curiosity about why people —

especially myself — behaved as they did, and with the
summer’s reading of Freud still knocking around in my head,
I veered off my intended path and entered Emerson Hall to

640
listen to the first lecture of the introductory Soc. Rel. course.
And my life changed.

The lecture was by Robert White, a courtly New Englander

and, in what would be an irony for me fifteen years later,
author of what was at the time one of the leading abnormal
psychology textbooks. What White did in this opening lecture
was place psychology in context, as an approach to
understanding the human condition that straddled biology,
sociology, anthropology, political science and other social
sciences, and even philosophy and theology. While my
reaction may have fallen short of being an epiphany, I
nonetheless made the decision to alter my fall schedule so that
I could enroll in the course — while continuing as a German
major for the nonce.

Throughout the two semesters that year, I had as lecturers in

addition to White the following senior professors: Clyde
Kluckhohn, Talcott Parsons, and Jerome Bruner. Not too
shabby for a poor Jewish kid from Dorchester. We read
widely in the several domains of what was called Social
Relations, but there was one set of readings and experiences
that were pivotal for me. In discussion section, we were
examining a clinical case history of Benjamin Feingold, a
young man with lots of insecurities and anxieties. One day the
topic of discussion was a dream he had of sitting at the wheel
of a car and then seeing to his right his brother-in-law coming
to a stop next to him so close to his car that the two cars
scraped together. What did the dream mean, the teaching
fellow asked us. A lively conversation ensued during which
everyone in the section except myself saw very clearly that
the dream was a disguised expression of Benjamin’s
homosexuality: the two cars scraping together obviously

641
signified a wish on his part to rub up against the body of his
brother-in-law. The Radcliffe students in the class —
“Cliffies,” we called them with a mixture of envy and
resentment since most of them were smarter and more verbal
than the Harvard students — were especially vocal in this
rendition of the dream.

Now it turns out — as I have learned at the several reunions

I’ve attended over the past thirty years — that most of us
were somewhat intimidated by our peers, all of whom
appeared brighter and better read than we. I was certainly no
exception. I recall looking around the room and deciding that
even if I were not downright stupid, I was certainly poorly
suited to any specialization of psychology that had to do with
trying to understand and help people in emotional distress. No
matter. There were other areas of psychology that intrigued
me, and I spent the next three years working with two faculty,
Richard Alpert (aka Baba Ram Dass) and Jerome Bruner on
topics relating to motivation and cognition. I ended up doing
an honors thesis with Bruner on perceptual problem-solving
under conditions of degraded but improving stimulus input.
This early interest in cognition diminished drastically at
Stanford, as seen below, but returned soon thereafter with my
involvement in cognitive behavior therapy. Towards the end
of my senior year — after a dalliance with applying to law
schools — I found myself with an acceptance to Stanford, to
study cognitive dissonance with Leon Festinger.

But the uncertainty earlier in my senior year — whether to go

to law school or to graduate school in psychology — had led
me to seek ways to postpone a firm decision. I applied for
several foreign study fellowships and was awarded a

642
Fulbright Scholarship to study for a year in Germany
following graduation in June 1961.

My Graduate School Years

The year abroad was, well, broadening. I immersed myself in
the culture, language, and wine of the southwestern part of
Germany, took courses at the University of Freiburg in dream
analysis, handwriting analysis, the Colored Pyramids Test,
and psychoanalysis, and sang in Freiburg’s Russian Chorus.
By the time June came around, I was ready to trek out to
Stanford and begin a new life.

Eager to become a Festinger-type social psychologist at

Stanford beginning in Fall 1962, I was dismayed to learn that
he had switched into eye movement research, both forcing
and freeing me to explore a bit that first year. Having an NSF
fellowship, I was able to move pretty much as I wanted and
decided to do some research in something I had never had
contact with or even given any thought to, physiological
psychology. This found me learning about brain stimulation
of the rat brain from J.A. Deutsch, who had recently
published an unusual book on what he called “a structural
theory of behavior.” Deutsch had studied at Oxford, where
much of the teaching is done via individual tutorials, so we
had innumerable seemingly discursive conversations in his
lab, with me watching him run rats while he peppered me
with questions on what I had been reading. I found it very
intellectually stimulating but somehow constraining, because
as complex and challenging as rat’s appetitive behavior was
as understood by Deutsch’s ingenious theorizing, I continued
to feel unfulfilled.

643
And now another unexpected event. Because my mother was
worried that I would not eat well in California, she’d
prevailed on me to join a meal plan for at least my first
quarter at Stanford. This found me eating dinner each evening
in the graduate dining hall with other first year students in
psychology. Some of these were clinical students (Stanford
had an APA-approved clinical program at the time), and they
often talked about a professor named Bandura and something
called “behavior therapy.” The basic notion was that all the
stuff I’d been learning as an experimental psychologist had
relevance for understanding and treating abnormal behavior.
This was a new notion for me. None of my professors at
Harvard had ever mentioned this viewpoint — and recall that
Wolpe’s classic book had been published in 1958, while I was
a sophomore. Recall also that Skinner was at Harvard and had
published a couple of papers with Ogden Lindsley on operant
conditioning of regressed schizophrenics. But these
new-fangled notions had not found their way into the
ego-analytic, psychodynamic stronghold of Emerson Hall.

These dinner conversations bounced around in my mind

during the fall and winter quarters of my first year, and then
another unexpected thing happened. Visiting that year from
the University of Illinois was a young associate professor
named Perry London, whose courses I had of course been
avoiding because they had to do with clinical. Somehow we
found ourselves playing tennis, and during a break, he asked
me what I wanted to do when I grew up. When I confessed
that I had little idea except that I thought I’d like to be an
academic like him, he took me back to his office and showed
me a few vitas of colleagues of his. He asked me whether
anything struck me about the publications. They were all very
different from each other, and each vita was, within itself,

644
very heterogeneous. Precisely, he said, but one thing they had
in common was that all the people were clinical
psychologists. Clinical psychology, he said with obvious
relish, is a bastard discipline. And that’s what makes it
exciting and promising. That conversation with London was
pivotal.

Soon I found myself in Bandura’s office, doing a song and

dance about why he should let me switch into clinical. I
interpreted his mm-hmms and nods as signs that I should
continue my persuasion attempts, but after a while he
interrupted me and said “OK.” “OK, what?” I asked. “OK,
you can switch into clinical,” he said with some bemusement.
And that was it. Stanford was a remarkably flexible place, and
I am forever grateful for that. For I knew then and there that I
would not have found myself now in a clinical program had I
gone to either of the other two places that had been options
for me, Berkeley and Michigan. I had finally found a true
intellectual and professional home in a department whose
earlier appeal had had nothing to do with clinical psychology.
Indeed, I hadn’t known what area of psychology I would
specialize in, only that it would not be clinical.

Dumb luck continued. I took Bandura’s course that spring,

worked with some autistic children at a nearby daycare
center, read virtually everything that had ever been published
in behavior therapy, and then found myself in my second year
taking an assessment course with Walter Mischel and a
behavior therapy course with a visitor from South Africa
named Arnold Lazarus.

This triumvirate — unbeknownst to them, I think —

presented me with a marvelous apprenticeship in theory,

645
research, and practice in what we then called social learning
approaches to psychotherapy, or sometimes just behavior
therapy. Lazarus began to see private patients at a greatly
reduced fee and permitted a few of the clinical students to sit
in with him. I must have spent at least 10 hours a week during
my second year, from September 1963 to June 1964,
watching Lazarus work with patients. Somewhere along the
line the conceptual introduction I had received from Bandura
and Mischel came to life in my sessions with Lazarus, such
that behavior therapy had a completely different meaning for
me at the end of that incredible year than it had had in the
beginning.2

Lazarus returned to South Africa in the summer of 1964 and I

globbed onto a dissertation project on systematic
desensitization (which I would have been very reluctant to
undertake had I not learned the procedure and related things
from Lazarus) that went well and enabled me to complete my
degree by July of the following year. In those days one could
do an internship postdoctorally, and that is what happened. I
spent a good internship year at the Veterans Administration
(VA) Hospital in Palo Alto.

One last tidbit from my formative Stanford years is in the

form of another chance event. I began my internship in July
1965 and was assigned to a ward on which Gordon Paul had
been working as an intern the previous several months. It
turned out that he was not leaving till August, so we had one
month’s overlap. During that time I did pretty much what I’d
done with Lazarus — I followed him around and sat in on
practically every meeting and session he had. (My advice to
graduate students has for some years been: Find someone
good and follow that person around.) I hate to think what my

646
internship year would have been like had it not been for
Paul’s calm and skillful introduction into the sometimes
surreal world of the VA mental hospital.

My Stony Brook Years

After completing my internship in August 1966, I migrated to
SUNY-Stony Brook to join a dedicated, sometimes
hypomanic group of behavior therapy enthusiasts in an
avowedly behavioral Ph.D. program set up by Len Krasner
(director of clinical training) and Harry Kalish (department
chair). I arrived there along with the first cohort of graduate
students as well as the first group of postdoctoral fellows in
what was the very first postdoctoral training program in
behavior therapy.

It was a heady time. Here was a program that, well before the
empirically supported treatments movement, elected to focus
on assessment and intervention that enjoyed some measure of
empirical support, eschewing unvalidated approaches and
procedures without a concern that our students would be
unable later on to obtain clinical internships. True, we were
narrow, but the Krasner-Kalish vision was to specialize in
something that we all believed had more promise than the
traditional clinical fare.

I worked most closely with three colleagues during my Stony

Brook years, and these collaborations enriched my
professional life immeasurably. First there was my attribution
research in the late 1960s with Stu Valins, a Schachter-trained
social psychologist. Together we published the first
experiment showing that attribution of behavior changes to

647
oneself rather than to a drug leads to greater maintenance of
therapeutic change (Davison & Valins, 1969). Second was
getting together with John Neale to write our abnormal
psychology textbook, the first edition being published in
1974. It was an instant success, and we recently completed
our 8th edition (Davison & Neale, 2001). And finally there
was Clinical Behavior Therapy with Marv Goldfried,
published in 1976 and reissued in 1994 in an expanded form
(Goldfried & Davison, 1976, 1994). This book helped bring
research and theoretical abstractions to life, contributed to the
cognitive trend in behavior therapy, and made a case for
trying to integrate ideas and procedures from the
non-behavioral psychotherapies. I was very fortunate to have
had such talented colleagues as these as well as other Stony
Brook faculty.

During the 1970s a number of Stony Brook faculty were

doing sex research. I was spending a lot of time with two
colleagues in particular, Jim Geer in Psychology and John
Gagnon in Sociology. We planned and we plotted, and at one
point Gagnon and I co-taught a graduate seminar in human
sexuality that attracted a lot of interesting and occasionally
unconventional students and colleagues.

Around 1971, I began doing a good deal of reading in what

was then known as the radical gay literature, books like
Lesbian/Woman by Martin and Lyon (1972) and Homosexual
Behavior Among Males by Churchill (1967). Between 1971
and 1973 I taught two advanced graduate seminars on
homosexuality, which were well attended by students in
Stony Brook’s clinical program, postdoctoral fellows in the
behavior therapy program I was directing, and a few selected
undergraduates and graduate students from other departments.

648
It was my impression that some of the students had a very
personal interest in the subject matter, but most of the seminar
members were involved in the subject more from an
intellectual than from a personal or political point of view. A
couple of colleagues mentioned to me a few years afterwards
that they wondered if these might have been the first courses
taught in a psychology department with the focus primarily
on homosexuality.

Much of what we read and discussed in seminar was new to

us, and some of it was disturbing. The disturbing part came
from the anger expressed in many of the books and articles
towards scientists who were investigating the causes of
homosexuality and towards practitioners who were engaged
in sexual reorientation programs. It took me a while to
understand the source of that anger.

Why focus on the etiology of homosexuality, this literature

asked, rather than on the etiology of heterosexuality? The
reason, it was asserted, was that the latter was viewed as the
universal norm and that the only thing worth looking into
were aberrations from that norm, i.e., homosexuality. A
not-always-articulated agenda was at work, therefore: By
focusing on the causes of homosexuality more than on the
causes of heterosexuality, the message was being conveyed
that the former was intrinsically abnormal and needed special
scrutiny. Psychoanalytic theorizing of course viewed
homosexuality as deviant, and while the rhetoric of behavior
therapy disclaimed intrinsic abnormality in favor of a socially
relative view — it depends on the culture one is in — the
reality was that some of the earliest work in behavior therapy
entailed efforts to eliminate homosexuality in favor of
heterosexuality.

649
Of more interest to me, though, was the brief against sexual
reorientation treatment. Simply stated, why spend so much
time and effort developing, evaluating, and providing
change-of-orientation therapies when they are aimed at a
“problem” that is socially defined?

This was a toughie for me. I had already written on sexual

reorientation in a paper with my student Terry Wilson
(Wilson & Davison, 1974) and had even made a training film
with Bob Liebert, Behavior Therapy for Homosexuality
(Davison & Liebert, 1971) that demonstrated the “orgasmic
reorientation” technique I had published on several years
earlier (Davison, 1968). The basic theme of the article with
Wilson was that there were more sophisticated ways to
analyze and change homosexuality than were prevalent in the
behavior therapy literature, and we offered an analysis that we
believed would eventuate in more effective and more humane
— little or no aversion therapy — ways to alter sexual
orientation from the homosexual to the heterosexual. In fact,
it was the material in this paper that formed the basis for the
workshop I gave at the AABT convention in October 1972, a
pivotal event for the contribution that I’ve selected to focus
on as important in my professional development and of some
significance as well for the field.

The 1972 AABT Convention

Chance rears its head once again, but in this case the
foundations had already been laid.

One of the people attending my 1972 AABT workshop on

better ways to change homosexuals into heterosexuals was

650
Charles Silverstein, a recent Ph.D. from the Rutgers clinical
program. I recall Chuck sitting in the meeting room with an
interested and fairly friendly expression on his face,
occasionally asking questions about why I was involved in
this sort of scholarship and application. My answer, which
was the standard response of behavior therapists at the time,
was that I would never impose such conversion treatment on
an unwilling homosexual patient, but that I saw it as
appropriate and, indeed, inherent to my professional role to
make such reorientation interventions available to gay and
lesbian patients who asked for sexual reorientation. He never
seemed quite satisfied with my answer but he didn’t push the
issue. Not during the workshop.

During a break, he came up to me and asked if he could

circulate some flyers for a symposium he had organized for
the last day of the convention. He showed me the flyer, and it
looked like one of those radical political diatribes that were
prevalent in the early 1970s on a variety of social issues. I
thought to myself that I would certainly not attend the
symposium but felt it would be imprudent and uncollegial not
to permit him to distribute it to the members of my workshop
(assuming that I actually had a choice in the matter, since he
could easily have handed them to people as they left the
workshop).

Now here’s where chance enters the picture again. I had

intended to leave the New York Hilton at a time on Sunday
that would enable me to catch a late-morning train from Penn
Station out to Port Jefferson, a town I lived in just east of
Stony Brook. But I fell into unplanned conversations with
some friends as I was trying to leave the hotel and then
realized that I would never make my intended train. I found

651
myself with a couple of extra hours, and then Silverstein’s
symposium came to mind. With no one in particular to talk to
and deciding it would be more interesting to spend the extra
time at the convention as it was ending rather than cooling my
heels elsewhere, I found the room where the radicals were to
hold forth.

It was a boisterous affair. Silverstein enraged me. He accused

people like me of strengthening the unjustifiable bias against
and discrimination towards homosexuals by virtue of even
making conversion programs available. Others spoke in a
similar vein, but what I remember most was that Silverstein
was not radical enough for some members of the audience
(some of whom may not have been actual convention
registrants — but that’s how it was in those days). So this
man, whose views I reacted to with a mixture of outrage,
curiosity, and a nescent respect, was accused in angry tones of
selling out to the oppressing establishment, to the behavior
therapy fascists, by the very fact of his participating in the
convention.

I returned to Stony Brook and over the next several weeks

began discussing these events and my reactions to them with
several friends and colleagues, and with my homosexuality
seminar. I wish I could remember how my ideas evolved after
that, but it could not have been more than a few weeks before
I concluded that Silverstein and the radical therapists were
right.

652
Being President of AABT in 1973-1974
During my presidential year, I initiated a motion in the AABT
Board of Directors, which passed the following resolution at
its meeting of May 11-12, 1974. It was supported by an
overwhelming vote of the membership later that spring:

The AABT believes that homosexuality is not in itself a sign

of behavioral pathology. The Association urges all mental
health professionals to take the lead in removing the stigma of
mental illness that has long been attributed to these patterns of
emotion and behavior. While we recognize that this
long-standing prejudice will not be easily changed, there is no
justification for a delay in formally according these people the
basic civil and human rights that other citizens enjoy.

But this position statement, as forward-looking as it was,

went only just so far. The implications (as I saw them) had
yet to be drawn out. This would be the theme of my 1974
presidential address, which I entitled “Homosexuality: The
Ethical Challenge.”

The Context of My AABT Address

The AABT convention in Chicago in November 1974 was a
tense affair. Behavior therapy had been lambasted in the
media the preceding year by several groups — by the ACLU
and by Senator Sam Ervin’s committee that was looking into
behavior modification for denying people their civil liberties
(especially prisoners in the federal penitentiary in Springfield,
Missouri)3. Also critical of behavior therapy was an unruly
and violence-threatening group of self-appointed guardians of

653
The American Dream who saw fit to engage in such activities
as circulating the home addresses of fascists like myself,
Israel Goldiamond, and other putative enemies of the people.
Since Goldiamond was a featured presenter at the 1974
convention, we had to arrange for plainclothes as well as
uniformed police to ensure the orderliness and safety of the
proceedings.

It was in this context that I presented the arguments

summarized below. For reasons that I believe will be evident,
I was fairly nervous. But with the support of a number of
friends who sat down front in the large ballroom and gave me
encouraging nonverbals — even though most of them
disagreed with the substance of my remarks — I got through
the address.

Homosexuality: The Ethical Challenge

Below is a brief rendition of my AABT address (Davison,
1974/1976), expanded in recent years to encompass more
general issues of the constructive nature of clinical
assessment (Davison, 1991).

We Only Want to Help

API (Apocryphal Press International). The governor recently
signed into law a bill prohibiting discrimination in housing
and job opportunities on the basis of membership in a
Protestant Church. This new law is the result of efforts by
militant Protestants, who have lobbied extensively during the
past ten years for relief from institutionalized discrimination.

654
In an unusual statement accompanying the signing of the bill,
the governor expressed the hope that this legislation would
contribute to greater social acceptance of Protestantism as a
legitimate, albeit unconventional, religion.

At the same time, the governor authorized funding in the

amount of twenty million dollars for the upcoming fiscal year
to be used to set up within existing mental health centers
special units devoted to research into the causes of people’s
adoption of Protestantism as their religion and into the most
humane and effective procedures for helping Protestants
convert to Catholicism or Judaism. The governor was quick to
point out, however, that these efforts, and the therapy services
that will derive from and accompany them, are not be
imposed on Protestants, rather are only to be made available
to those who express the voluntary wish to change. “We are
not in the business of forcing anything on these people. We
only want to help,” he said.

The Myth of Therapeutic Neutrality

Therapists never make ethically or politically neutral
decisions. “Any type of psychiatric [psychological]
intervention, even when treating a voluntary patient, will have
an impact upon the distribution of power within the various
social systems in which the patient moves. The radical
therapists are absolutely right when they insist that psychiatric
neutrality is a myth” (Halleck, 1971).

This is the thesis of Seymour Halleck’s noted — and too little

read — book, The Politics of Therapy, and it plays a major
role in my argument about sexual conversion therapy. Most of

655
the time the very naturalness of and familiarity with our
therapeutic practices blind us to the nonempirical biases that
affect how we construe the patient’s problems and the goals
we regard as acceptable to work towards. Better to be aware
of and own up to our biases than to pretend that we have
none.

Differences Do Not Imply Pathogens

Sometimes those who have argued in favor of sexual
conversion therapies for gays and lesbians seek to justify their
position by asserting that homosexuality is pathological and
that, as doctors of the mind, it is our duty and right to set
things straight (pun intended). One form that the argument
takes is that homosexuals differ from heterosexuals in how
they were raised, and that this difference indicates something
pathogenic. The classic study in this vein is by Bieber, Dain,
Dince, Drellich, Grand, Gunlach, Kremer, Rifkin, Wilbur, &
Bieber (1962), a survey so flawed both conceptually and
methodologically that it is hard to believe that it has been
taken seriously by anyone. The logic of the findings takes the
following form: the parents of male homosexuals more often
reflect a pattern of a “close-binding intimate mother” and a
cold and detached father. Ergo, homosexuality is a mental
illness.

A moment’s reflection reveals the absurdity of the argument.

Simply put, what is wrong with such child-rearing unless one
has decided before the fact that homosexuality is an illness?
Post hoc ergo propter hoc. Weak reasoning indeed.

656
No Cure Without a Disease
Clinicians devote effort to developing and analyzing
therapeutic procedures only if they are concerned about a
problem. Until the 1980s behavior therapists spent a good
deal of time and effort reducing homosexual attraction and
increasing heterosexual attraction in homosexuals (and for the
most part, the target population was men only). Again, until
recently little if any time — and none at all when I first made
my remarks — was spent by mainstream therapists
encouraging health professionals to change their biases
against homosexuality and foster gay-affirmative attitudes
and behavior in patients who happened to be homosexual.
The question for me was and still is the following: How can
therapists honestly speak of nonprejudice when they
participate in or tacitly support therapy regimens that by their
very existence and regardless of their effectiveness condone
the societal prejudice and perhaps also impede social change?
As Begelman pointed out many years ago (1975), sexual
reorientation therapies

. . .by their very existence constitute a significant causal

element in reinforcing the social doctrine that homosexuality
is bad. Indeed, the point of the activist protest is that behavior
therapists [and other therapists] contribute significantly to
preventing the exercise of any real option in decision making
about sexual identity by further strengthening the prejudice
that homosexuality is a “problem behavior” since treatment
may be offered for it. . .homosexuals tend to seek treatment
for being homosexuals. . .contrary to the disclaimer that
behavioral therapy is “not a system of ethics” (Bandura, 1969,
p. 87), the very act of providing therapeutic services for

657
homosexual “problems” indicates otherwise (p. 180, emphasis
in original).

I would add that the availability of a technique encourages its

use. For example, Wolpe’s (1958) systematic desensitization
ushered in a period in which behavior therapists looked
vigorously for antecedent cues that could be arranged on an
anxiety hierarchy and be paired in imagery with deep muscle
relaxation. Thus, a problem like social isolation might be
viewed at least in part as a consequence of unnecessary
anxiety that could be translated into an anxiety hierarchy and
then desensitized. The therapist’s assessment and
problem-solving efforts are shaped by the availability of
therapeutic techniques that are believed to be effective. This
is not a bad thing! But it does skew what the therapist sees
and finds out about a patient, a topic we turn to next.

Clinical Problems as Clinicians’

Constructions
As I have argued elsewhere (e.g., Davison & Neale, 2001;
Davison & Lazarus, 1995; Goldfried & Davison, 1994),
clients seldom come to mental health clinicians with problems
as clearly delineated and independently verifiable as what
patients often bring to a physician. A client usually goes to a
psychologist or psychiatrist in the way described by Halleck
(1971). That is, the person is unhappy; life is going badly;
nothing is meaningful; sadness and despair are out of
proportion to life circumstances; the mind wanders and
unwanted thoughts intrude, etc. The clinician transforms
these often vague and complex complaints into a diagnosis or

658
functional analysis, a set of ideas of what is wrong, what the
controlling variables are, and what might be done to alleviate
the suffering and maladaptation. My argument, then, is that
psychological problems are for the most part constructions of
the clinician. Clients comes to us in pain, and they leave with
a more clearly defined problem or set of problems that we
assign to them.

In the case of homosexuality, I argue that when a person with

such attractions/behavior goes to a therapist, whatever
psychological woes they have are generally construed as
caused entirely or primarily by their sexual orientation. This
happens because (a) their sexual orientation is usually the
most salient part of their personhood, to the clinician and
usually to the clients themselves because of the negative
salience homosexuality has been accorded by society; and (b)
it is regarded as abnormal, regardless of the liberal stance the
clinician may take overtly. Even with the changes in the DSM
over the past 25 years, but especially when I first articulated
this position in 1974, the clinician’s perceptions and
problem-solving are skewed in a direction that implicates
homosexuality — no matter what the actual presenting
problems are (cf. Davison & Friedman, 1981) — and, most
importantly, imply the desirability of a change in sexual
orientation.

None of this is to gainsay that being homosexual in our

society is difficult psychologically and that it can occasion
people considerable distress — particularly a generation ago
but even now, given the disproportionate exposure to hate
crimes and simple everyday prejudice that homosexuals are
subject to (discussed in next section). The moral point I tried
to make in 1974 and, despite incredulousness on the part of

659
some of my friends and colleagues I still hold to, is that
mental health professionals have a responsibility not to be
co-opted by the societal pressures that, sometimes subtly,
channel our clinical problem-solving and decision-making
into narrowly defined domains that result in a maintenance of
a status quo that, in official pronouncements, we say we do
not support.

Discrimination, Hate Crimes, and the

“Voluntary” Desire to Change Sexual
Orientation
I’d like to expand in this section on a theme that was not fully
developed in my original presentation and that may provide
the context not only for my holding to my position against
sexual conversion therapies but also for the importance I
attach to applying the analysis to other psychological issues
that come to the attention of health professionals.

Although most states have dropped their sodomy laws, which

used to be enforced selectively against homosexual acts, some
legal pressure against homosexuality remains. A 1986 U.S.
Supreme Court decision (Bowers v. Hardwick, 106 S.Ct. 284
[1986]), still valid, refused to find constitutional protection of
the right to privacy for consensual adult homosexual activity
and thereby upheld a Georgia law that prohibits oral —
genital and anal — genital acts, even in private and between
consenting adults. (Such laws can be applied to heterosexual
sex as well, but straight people don’t have to worry about that
as much as do gays and lesbians.)

660
But legal pressures are not the whole story. Research supports
the view that gays and lesbians are discriminated against in all
kinds of ways and that this discrimination takes a particularly
heavy toll on their emotional well-being. So-called “hate
crimes” highlight this problem. A hate crime (sometimes
referred to as a bias crime) is an assault that is based primarily
or solely on a person’s (perceived) membership in a group
against which the perpetrator is prejudiced. The ultimate
modern-day hate crime was, of course, the Holocaust in
Germany and other parts of Europe prior to and during World
War II. The Nazis sought out for imprisonment and execution
millions of Jews and hundreds of thousands of gypsies,
Communists, and homosexuals. The more recent “ethnic
cleansing” in Bosnia and Kosovo and in many other parts of
the world shows us that humankind has not learned much
from the Holocaust experience. But hate crimes as well as
hurtful discrimination are carried out every day in less
organized and less dramatic fashion.

Recent research shows that as many as 92 percent of gays and

lesbians have been subjected to verbal abuse and threats —
often from members of their own family — and that as many
as 24 percent have been physically attacked because of their
sexual orientation (Herek, 1989; Herek, Gillis, Kogan, &
Glunt, 1996). A quarter of gay youth are ejected from their
homes when they come out to their families, and as many as
half of the homeless young people in New York City are gay.
The lifetime risk of suicide and suicidal behaviors is much
higher among homosexual men than among heterosexuals
(Herrell, Goldberg, True, Ramakrishnan, Lyons, Eisen, &
Tsuang, 1999).

661
As compared to non-hate crimes, bias crimes and verbal
assaults may create more psychological distress, perhaps
because they are an attack not just against the person as a
physical being but against the person’s very identity (Garnets
Herek, & Levy, 1990). Furthermore, such crimes may impart
to the victim a pervasive sense of danger and even loathing of
an aspect of the self that might otherwise be a source of
pleasure and pride.

In addition to violence from strangers and acquaintances,

lesbians and gay men experience “invisibility, isolation, lack
of information, lack of role models, negative attitudes from
others, lack of family and social support, uninformed or
biased helping professionals, religious prohibitions,
workplace discriminations, lack of legal supports, and
internalized homophobia” (Fassinger & Richie, 1997, p. 90).
Fassinger (1991) concluded that, while growing up, most gays
and lesbians acquire the same negative attitudes towards gays
as heterosexuals do, and this internalized homophobia makes
it all the more difficult for them to confront their sexual
orientation and to consider it in a positive light.

Anti-gay attitudes are strong, sometimes virulent, with many

people believing that homosexuals are sick and their behavior
disgusting (Herek, 1994). These negative attitudes can take
the form of open heterosexism — as when people directly
insult a gay person with epithets like faggot or dyke — or a
more subtle, indirect kind of anti-homosexual stance — as
when people tell jokes that deride homosexuality without
knowing (or caring) if a gay person is present. This prejudice
creates what has been termed “minority stress”, a source of
pressure and tension that is a special burden of those in
despised or feared minorities (Meyer, 1995) and no doubt is

662
the major factor in gay and lesbian people suffering
particularly high levels of depression (Herek et al., 1996).

In light of all this, is it surprising that gays may seek out

sexual reorientation treatment? Being subjected to verbal and
physical assault for being gay is not likely to enhance one’s
sense of comfort with and acceptance of one’s sexual
orientation. Little wonder, then, that questions have been
raised about how voluntary is the desire of some gays to
change their sexual orientation.

A Proposal Regarding Sexual

Reorientation Therapy
These several considerations led me to make a proposal that
surprised no one more than myself, an idea that was present
for several years in some of the gay activist literature (see
especially Silverstein, 1977): Therapists should stop engaging
in change-of-orientation programs, whether the client makes
the request or someone else does. The social pressures,
discrimination, and in some cases violent hatred directed
toward people with homosexual inclinations make it highly
doubtful that client-requests for conversion therapy approach
what we regard as voluntary. In a sense, by attending to the
reasons for a “voluntary” request for change, we are, I
believe, doing nothing less than remaining true to our
deterministic stance. And without entering the free
will-determinism morass, we can, I believe, consider more
carefully than we have the societal pressures that would seem
to underlie “voluntary” requests for conversion therapy.

663
Long ago, Perry London (1969) warned of an unappreciated
danger in behavior control technology, namely clinicians’
increasing ability to engineer what we have tended to regard
as free will on the part of our patients. In his view, therapists
are capable of making patients want what is available and
what they believe their patients should want. Moreover, even
if therapists assert that they do not work against the will of
their patients, this does not free them from the responsibility
of examining those factors that determine what is considered
free expression of intent and desire on the part of our patients.
Indeed, I would argue that the therapist sets the goals in
therapy more than does the patient.

Halleck put the matter thus:

At first glance, a model of psychiatric [or psychological]

practice based on the contention that people should just be
helped to learn to do the things they want to do seems
uncomplicated and desirable. But it is an unobtainable model.
Unlike a technician, a psychiatrist [or psychologist] cannot
avoid communicating and at times imposing his own values
upon his patients. The patient usually has considerable
difficulty in finding the way in which he would wish to
change his behavior, but as he talks to the psychiatrist [or
psychologist], his wants and needs become clearer. In the
very process of defining his needs in the presence of a figure
who is viewed as wise and authoritarian, the patient is
profoundly influenced. He ends up wanting some of the
things the psychiatrist [or psychologist] thinks he should want
(1971, p. 19).

664
Not Can but Ought
As mentioned below in my discussion of a critique by Sturgis
and Adams (1978), there is an important and oft-overlooked
distinction between being able to achieve a goal and whether
it is proper to try to do so. Empirical evidence as to whether
we can change sexual orientation is not relevant to whether
we ought to — except that we ought not to engage in a given
change effort when there is no evidence that we can actually
do so. This may well be the case with conversion therapies.
The ethical argument against an ineffective treatment is that
patients are bound to be disappointed and likely to feel even
worse and “sicker” if they have made an effort to alter
something that cannot be changed. The patient has not only
failed to achieve a goal that has been set forth by the therapist
as important but is likely to come away from the unsuccessful
therapy continuing to believe that their behavior is bad and
that they are really hopeless and unworthy.

But the two domains — empirical and ethical — are best kept
separate.

Psychotherapy, Politics, and Morality

And this takes us to the final aspect of my argument. I hadn’t
considered myself a community psychologist until the
formulation of my brief against conversion therapy, but I
think the characterization is apt. In Rappaport’s (1977) terms,
I am working at an institutional level, which is the domain of
community psychology. In contrast, most therapists operate at
the individual level. An institutional analysis of human

665
problems examines those values and ideologies that guide the
decision-making of a society. Individual therapy work, in
contrast, assumes that society is benign and that
psychological suffering can best be alleviated by helping the
patient adjust to prevailing values and conditions. My
underlying assumption is that issues surrounding therapy for
homosexuality should be addressed at an institutional level,
and that greater societal acceptance of homosexuality as a
normal variation of human sexuality rather than as a problem
that needs to be fixed will, in fact, redound to the benefit of
the individual by reducing the discrimination and oppression
described earlier that, I firmly believe, accounts for the
distress that can be associated with homosexuality and
ultimately the desire of some homosexual individuals to seek
sexual reorientation.

Do therapists have some kind of abstract responsibility to

satisfy a patient’s expressed desires and wishes, as asserted by
some (e.g., Sturgis & Adams, 1978)? No. Therapists constrain
themselves in many ways when patients ask for assistance,
and under some circumstances, therapists are even legally
required to break the confidentiality that is inherent in the
relationship. In any event, requests alone have never been a
sufficient justification for providing a particular service to a
patient.

Finally, am I arguing against trying to help homosexuals in

therapy? Not at all. It is one thing to argue that therapists
should not try to alter patients’ sexual orientation; it is quite
another to suggest that therapists should not work
therapeutically with people who are gay or lesbian. (This
seems straightforward enough, but over the years some critics
have alleged that I have urged people not to treat

666
homosexuals at all.) Indeed, the implication of my thesis is
that therapists consider seriously the problems in living
experienced by people who happen to prefer members of their
own sex as sexual partners. For example, while a gay person
may be depressed because his sexual orientation is mocked or
attacked and he feels insecure about standing up for himself,
gay people also get depressed because their professional
aspirations are thwarted by circumstances having nothing to
do with their sexual orientation. And it would be nice if
alcohol abusers who happen to be homosexual could be
helped to reduce their excessive drinking without having their
sexual orientation questioned. Freed of the inclination of
trying to alter a homosexual’s sexual preferences, therapists
will find many other ways that they might help that individual
lead a more fulfilling life.

Aftermath of The Paper

To return to the circumstances of my AABT presidential
address in 1974, the immediate aftermath was pretty
emotional. The audience had been very attentive, with the
silence deepening when I articulated the main point that we
should not be engaging in sexual conversion therapy even
when the patient asked for it. (Friends commented afterwards
that one gets that kind of silence when everyone in a room
full of 1,000 people stops breathing at the same time.) There
was a reception of sorts right after the talk, and I recall some
colleagues seeking me out to shake my hand and others
keeping their distance, with looks on their faces too complex
to interpret. But the most memorable reaction came from a
young woman who approached with glistening eyes and told
me that she could not believe what she’d just heard and that

667
she just wanted to thank me. I’ve been told that other people
who have been personally affected by conversion efforts and
their promulgation reacted similarly, albeit privately. I have
found these reactions very gratifying, especially as the years
have gone by and I have seen the argument become, if not
universally accepted, at least more mainstream and one that
can no longer be ignored.

It may or may not have been assumed by some that I was gay.
Besides some occasional heterosexist kidding from a
colleague or friend, I’m not aware of this consequence of
which I’d been forewarned (not that it mattered to me one
way or the other). And of course this admonition assumes that
only a gay person would hold the point of view against sexual
conversion therapies that I’d articulated — a position that I’ve
always seen as a strategy, perhaps unconsciously employed,
to denigrate the message by denigrating the messenger.

More important is what happened a month later when I

submitted for publication a manuscript based on my AABT
address. For reasons that I hope are obvious, I selected the
American Psychologist. Only a week or two after sending it
in, I received a letter from the editor handling the manuscript
(an APA staff person of no scholarly credentials that I was
aware of) that he had decided not to send it out for review
because it was not “of general enough interest” to warrant
consideration for a journal sent to all APA members as part of
their dues. Think about this. I was not surprised that he was
offended by the content of the paper — and, yes, I am
presuming that this was the reason he rejected it without
obtaining input from appropriate referees — but I was taken
aback at the peremptory judgment that a paper examining the
ethical bases of psychotherapy as applied to the case of sexual

668
conversion treatment was not of “general enough interest” to
an organization like APA.

Well, no one likes rejection letters, but I did my best to let go

of my pique and decided to submit the same manuscript to the
Journal of Consulting and Clinical Psychology, edited at the
time by Brendan Maher. Again I got a very speedy response
in a thin envelope, and I feared for the worst as I opened it.
But Maher’s decision could not have been more different or
gratifying. He told me that he wanted to publish it without
having it vetted by outside reviewers, provided that I would
agree to his inviting several accompanying critiques. I could
not have been more pleased. At his request, I made two
suggestions: Irving Bieber, who I was confident would
excoriate my paper (which he did); and Seymour Halleck,
whom I had relied on extensively in formulating my argument
on the politics of therapy and whose opinion of my effort I
was certain the readership would be interested in. The
commentaries followed my paper, which was published as the
lead article (Davison, 1976). Interestingly, Halleck’s
comments did not, as I read them, fully embrace the
conclusion I had come to, but his commentary was, I think,
the most supportive of my effort.

A year after its publication, Maher sent me a manuscript to

review for JCCP. It was a critique of my article by Ellen
Sturgis and Henry Adams (Sturgis was Adams’ graduate
student at the University of Georgia). I found the manuscript
to be an interesting and thoughtful paper on how better to
change people’s homosexual orientations. The only problem
was that I found it irrelevant to my earlier article, for the
question to me was not whether we can change sexual
orientation but whether therapists should help people do so.

669
Clearly my belief was and is that we should not. So I told
Maher that I would not be an appropriate reviewer because I
would have to reject the manuscript out of hand as not
relevant. His response was that he wanted to publish the paper
provided I write a rebuttal (instead of the review he had asked
me to write). This seemed a very sensible editorial decision,
and I agreed to do so. Basically “Not Can But Ought: The
Treatment of Homosexuality” (Davison, 1978) responded to
Sturgis and Adams (1978) in the aforementioned fashion, that
is, that their paper was irrelevant to my argument. I don’t
believe my rebuttal was convincing to the authors, but I found
it interesting some years later to be told by Sturgis that she
had changed her views on the matter and now agreed with my
position. Adams, on the other hand, continued to believe that
therapists have an obligation to change people’s sexual
orientations if they seek such treatment. Interestingly, he and
his students conducted some very interesting and ingenious
research on homophobia, a focus that I was delighted to see
for his considerable research skills.4

Importance of My AABT Presidential

Address
It is both a treat and an embarrassment to be asked to
comment on the importance of one’s work. The only thing
one can really do is suspend modesty and try to comment on
it as if it were the work of someone else. I will try to do that.

670
Empirical versus Ethical Questions
I think my paper, and the rebuttal to the Sturgis-Adams
critique, have contributed to a clearer understanding of the
difference between what we as psychotherapists can or think
we can do and what we ought to be doing. It is surprising to
me how difficult it is for some folks to see this essential and
simple difference. In my teaching I sometimes use an
intentionally bizarre example to make the point. I tell students
that I have a one-session cure for any mental/emotional/
behavioral problem. In fact, it works in much less than one
session. Indeed, it works in much less than one minute. It is a
bullet in the head of the patient. With death comes an end to
all the person’s psychological suffering and/or maladaptive
behavior. No more panic attacks, no more depression, no
more disordered thinking, no more shy withdrawal, no more
non-assertiveness, no more autistic aloneness, no more
psychopathic finagling, no more aggression. All gone in an
instant.

So, what’s the problem?

The concepts of values and biases are not as anathema in

professional circles now as I found them to be when I was in
graduate school in the mid-1960s. This is a good thing, and
perhaps my paper has contributed to the clarification of the
issue, whether or not people agree with the particulars of my
argument.

671
The Therapist as Secular Priest
Related to this point are Perry London’s writings on moral
issues in psychotherapy (e.g., London, 1964). This influence
from my graduate school days did not show up fully until I
became obsessed with the sexual conversion issue. As
indicated earlier, his concept of therapist as secular priest
defines our role as inherently moral, whether we like it or not.
Especially behavior therapists unabashedly try to shape the
patient in ways that they believe will benefit the patient and
not infringe on the rights and sensibilities of others. But we
also are good at engineering what the patient ends up
wanting, as Halleck said so eloquently in his 1971 book. I
believe that my article has helped sensitize people to the
issue, regardless of how they think about it. As a teacher it is
enough for me to know that I may have helped frame the
debate and made it legitimate, if not actually necessary, to
consider the influence that therapists have on their patients,
even when therapists think of themselves as hands-off when it
comes to therapeutic goals. I just don’t believe that patients
don’t get shaped in this way. At the very least, I think it is
better to assume this shaping rather than, as we have been
doing, assume its absence.

Liberalization on the part of the APA

and APA re Homosexuality
It’s possible that my 1974 address and the publications based
on it played some role in discussions that led stagewise to the
dropping of homosexuality entirely from the DSM as well as
to the recent position of APA against sexual conversion

672
therapies. I am not in a position to know this, but friends and
colleagues have suggested this to be the case. Certainly my
own “conversion” in 1973-1974 took place at a time that
changes in organizational viewpoints were occurring. I cannot
help but be pleased if the position I took was at all
instrumental.

Fewer Requests for Sexual

Reorientation and Fewer Articles in
the Professional Literature
Over the past 25 years there seems to have been a sharp
decline in people seeking conversion therapy and there
certainly has been a decline in articles published on the
subject in the professional mental health literature (Campos &
Hathaway, 1993). With respect to the latter, one can inspect
the tables of contents of journals such as Behaviour Research
and Therapy, Journal of Abnormal Psychology, and Journal
of Consulting and Clinical Psychology, as well as the titles of
psychotherapy books, and readily verify the decline. This
does not mean that some therapy efforts do not continue to
involve attempts at sexual reorientation — nearly all that
happens in therapy settings remains hidden from view, with
practically none of it seeing the light of publication. But I
suspect the incidence is down, consistent with the decrease in
our journals and professional books.

673
Psychosocial Interventions as Part of
Social Institutions
As I argued in my original paper, an institutional perspective
is important in understanding the conduct of psychotherapy.
As private and walled-off-from-the-world as outpatient and
some inpatient mental health intervention is, therapists and
patients do not work together in a social vacuum. As Halleck
argued in 1971, the decisions made in the consulting room
reflect and have effects on the politics and social fabric of the
place and time in which therapy is conducted. Therapists’
behavior is constrained by multiple factors — from
theoretical orientation, to personal taste, to religious values, to
legal requirements and strictures, and most recently to
reporting requirements and treatment decisions from
insurance companies. Patients’ behavior is also influenced by
multiple factors, and the emphasis in my writings on
homosexuality is on the manner in which societal prejudices
and biases shape the very way people come to understand
what is wrong and what is right about themselves, what they
might wish to change and what they might prefer to leave
alone. I continue to focus on the specific issue of
homosexuality because so many people have been and
continue to be hurt by prejudice and discrimination. But as I
hope is clear, the issues are much more general, going to the
heart of how researchers and clinicians set their professional
agendas, which in turn affect what they learn and the
decisions they make. I believe and hope that the position I
took in 1974 has contributed to the debate.

674
References
Bandura, A. (1969). Principles of behavior modification. New
York: Holt, Rinehart, & Winston.

Begelman, D. A. (1975). Ethical and legal issues of behavior

modification. In M. Hersen, R., Eisler, & P. M. Miller (Eds.),
Progress in behavior modification (pp. 159-189). New York:
Academic Press.

Bieber, I., Dain, H. J., Dince, P. R., Drellich, M. G., Grand,

H. G., Gundlach, R. H., Kremer, M. W., Rifkin, A. H.,
Wilbur, C. B., & Bieber, T. B. (1962). Homosexuality: A
psychoanalytic study. New York: Basic Books.

Campos, P. E., & Hathaway, B. E. (1993). Behavioral

research on gay issues: 20 years after Davison’s ethical
challenge. The Behavior Therapist, 16, 193-197.

Churchill, W. (1967). Homosexual behavior among males: A

cross-cultural and cross-species investigation. New York:
Hawthorn Books.

Davison, G. C. (1968). Elimination of a sadistic fantasy by a

client-controlled counterconditioning technique. Journal of
Abnormal Psychology, 73, 84-90

Davison, G. C. (1974, November). Homosexuality: The

ethical challenge. Presidential address to the annual
convention of the Association for Advancement of Behavior
Therapy, Chicago.

675
Davison, G. C. (1976). Homosexuality: The ethical challenge.
Journal of Consulting and Clinical Psychology, 44, 157-162.

Davison, G. C. (1978). Not can but ought: The treatment of

homosexuality. Journal of Consulting and Clinical
Psychology, 46, 170-172.

Davison, G. C. (1991). Constructionism and morality in

therapy for homosexuality. In J. C. Gonsiorek & J. Weinrich
(Eds.), Homosexuality: Research findings for public policy
(pp. 137-148). Beverly Hills, CA: Sage.

Davison, G. C., & Friedman, S. (1981). Sexual orientation

stereotypy in the distortion of clinical judgment. Journal of
Homosexuality, 6, 37-44.

Davison, G. C., & Lazarus, A. A. (1995). The dialectics of

science and practice. In S. C. Hayes, V. M. Follette, T. Risley,
R. D. Dawes, & K. Grady (Eds.), Scientific standards of
psychological practice: Issues and recommendations (pp.
95-120). Reno, NV: Context Press.

Davison, G. C., & Liebert, R. M. (1971). Behavior therapy

for homosexuality [16 mm. film]. Psychological Cinema
Register, Pennsylvania State University.

Davison, G. C., & Neale, J. M. (2001). Abnormal psychology

(8th ed.). New York: Wiley.

Fassinger, R. E. (1991). Counseling lesbian women and gay

men. The Counseling Psychologist, 19, 157-176.

676
Fassinger, R. E., & Richie, B. S. (1997). Sex matters: Gender
and sexual orientation in training for multicultural
competency. In D. B. Pope-Davis & H. L. K. Coleman (Eds.),
Multicultural counseling competencies: Assessment,
education and training, and supervision (pp. 83-110).
Thousand Oaks, CA: Sage Publications.

Garnets, L., Herek, G. M., & Levy, B. (1990). Violence and

victimization of lesbians and gay men: Mental health
consequences. Journal of Interpersonal Violence, 5, 366-383.

Goldfried, M. R., & Davison, G. C. (1976). Clinical behavior

therapy. New York: Holt, Rinehart, & Winston.

Goldfried, M. R., & Davison, G. C. (1994). Clinical behavior

therapy. [Exp. ed.]. New York: Wiley-Interscience.

Halleck, S. L. (1971). The politics of therapy. New York:

Science House.

Herek, G. M. (1989). Hate crimes against lesbians and gay

men: Issues for research and policy. American Psychologist,
44, 948-955.

Herek, G. M. (1994). Assessing heterosexuals’ attitudes

towards lesbians and gay men: A review of the empirical
research with the ATLG scale. In B. Greene & G. M. Herek
(Eds.), Contemporary perspectives on lesbian and gay issues
in psychology (pp. 206-228). Newbury Park, CA: Sage.

Herek, G. M., Gillis, R., Kogan, J. C., & Glunt, E. K. (1996).

Hate crime victimization among lesbian, gay, and bisexual
adults. Journal of Interpersonal Violence, 12, 195-215.

677
Herrell, R., Goldberg, J., True, W. R., Ramakrishnan, V.,
Lyons, M., Eisen, S., & Tsuang, M. T. (1999). Sexual
orientation and suicidality: A co-twin control study in adult
men. Archives of General Psychiatry, 56, 867-874.

London, P. (1964). The modes and morals of psychotherapy.

New York: Holt, Rinehart & Winston.

London, P. (1969). Behavior control. New York: Harper &

Row.

Martin, D., & Lyons, P. (1972). Lesbian/woman. San

Francisco: Glide Publications.

Meyer, I. (1995). Minority stress and mental health in gay

men. Journal of Health Sciences and Social Behavior, 36,
38-56.

Rappaport, J. (1977). Community psychology: Values,

research, and action. New York: Holt, Rinehart, & Winston.

Silverstein, C. (1972). Behavior modification and the gay

community. Paper presented at the annual convention of the
Association for Advancement of Behavior Therapy, New
York City.

Silverstein, C. (1977). Homosexuality and the ethics of

behavioral intervention: Paper 2. Journal of Homosexuality,
2, 205-211.

Sturgis, E. T., & Adams, H. E. (1978). The right to treatment:

Issues in the treatment of homosexuality. Journal of
Consulting and Clinical Psychology, 46, 165-169.

678
Wilson, G. T., & Davison, G. C. (1974). Behavior therapy
and homosexuality: A critical perspective. Behavior Therapy,
5, 16-28.

Wolpe, J. (1958). Psychotherapy by reciprocal inhibition.

Stanford: Stanford University Press.

Footnotes
1
For helpful comments on an early draft of this paper, I thank
Asher Davison. This article is dedicated to the memory of one
of my mentors and a best friend, Perry London.
2
This brings to mind something that Jerome Bruner said in a
lecture in his cognitive psychology course back in 1959. He
was discussing concept formation and how, once we have
attained a concept of something, it is hard to recall what life
was like before that understanding. I think his example was
that we look at a chair, consider what it is, and try to
remember what it looked like before we knew it was a chair.
In an analogous fashion, I came away from my yearlong
clinical apprenticeship with Lazarus with a new
understanding of behavior therapy, different from what I had
had before seeing him in action with patients.
3
In those days behavior modification encompassed — in the
view of many laypersons like Senator Ervin’s committee —
psychosurgery and electroconvulsive shock therapy. The
reason was that these and other techniques modified behavior.
This was the kind of misconception we were dealing with at
the time.

679
4
Sadly, Hank Adams died a few months before the present
paper went to press.

680
681