Investigating Psychology

key concepts | key studies | key approaches

This publication forms part of the Open University Psychology teaching materials. Details of Open University
modules can be obtained from the Student Registration and Enquiry Service, The Open University, PO Box 197,
Milton Keynes MK7 6BJ, United Kingdom (tel. +44 (0)845 300 60 90; email [email protected]).
www.open.ac.uk
Investigating Psychology
key concepts | key studies | key approaches
Edited by Nicola Brace and Jovan Byford
 Published by Oxford University Press, Great Clarendon Street, Oxford, OX2 6DP in association
with The Open University, Walton Hall, Milton Keynes, MK7 6AA, United Kingdom.

Oxford University Press is a department of the University of Oxford. It furthers the University’s
objective of excellence in research, scholarship, and education by publishing worldwide in
Oxford New York
Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne
Mexico City Nairobi New Delhi Shanghai Taipei Toronto
with offices in
Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan
Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam
Oxford is a registered trade mark of Oxford University Press in the UK and in certain other
countries.
Published in the United States by Oxford University Press Inc., New York

First published in 2010 by The Open University

This edition published in 2012 by Oxford University Press in association with The Open University.
Copyright © 2010, 2012 The Open University
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system,
transmitted or utilised in any form or by any means, electronic, mechanical, photocopying,
recording or otherwise, without written permission from the publisher or a licence from the
Copyright Licensing Agency Ltd. Details of such licences (for reprographic reproduction) may be
obtained from the Copyright Licensing Agency Ltd, Saffron House, 6–10 Kirby Street, London
EC1N 8TS
(website www.cla.co.uk).
Edited, designed and typeset by The Open University.
Printed and bound in the United Kingdom by Bell & Bain Ltd, Glasgow.
British Library Cataloguing in Publication Data available on request.
Library of Congress Cataloguing in Publication Data available on request.
ISBN 978–0–19–965176–4
2.1
Contents
General introduction 1

Part 1
Introduction 11
Chapter 1 Exposing the authoritarian personality
Jean McAvoy 15
Chapter 2 Just following orders?
Philip Banyard 57
Chapter 3 Learning from watching
John Oates 99
Conclusion 139

Part 2
Introduction 145
Chapter 4 Changing behaviour
Frederick Toates 149
Chapter 5 Determined to love?
Deborah Custance 189
Chapter 6 Making friends
Charlotte Brownlow 233
Conclusion 271

Part 3
Introduction 277
Chapter 7 Language and the brain
Frederick Toates 281
Chapter 8 Paying attention
Helen Edgar and Graham Edgar 321
Chapter 9 Witnessing and remembering
Graham Pike and Nicola Brace 361
Conclusion 403

Final note 407

Acknowledgements 415
Index 417
General introduction
 Defining psychology

Defining psychology
Here you are at the start of a book entitled Investigating Psychology. As you
have chosen to study psychology, it is likely that you have an idea about
what the subject is about. It could well be that some of what you know
about psychology, and why you are interested in it, has come from the
newspapers or television. Psychology and psychologists have become a
regular feature in the media in relation to a broad range of issues.
Consider the following examples:
. ‘Curling up for an afternoon nap can improve the brain’s ability to
learn by clearing out cluttered memory space, psychologists say.’
(The Guardian, 22 February 2010)
. ‘A “dark side” to the internet suggests a strong link between time
spent surfing the web and depression, psychologists have claimed.’
(Daily Telegraph, 3 February 2010)
. ‘Primary school pupils are better behaved within the classroom than
in the 1970s, says a long-term study by educational psychologists.’
(BBC News online, 24 November 2008)

This is just a small glimpse of what has been reported in the press.
Psychologists regularly appear on television, either on news programmes
or in documentaries, to comment on a host of issues, such as the
influence of the internet, child development, attraction, animal
behaviour, ageing, stress, mental health, personality, differences between
men and women, reasons for committing crimes, offender rehabilitation,
and the belief in the paranormal.
So psychologists seem to comment on just about anything to do with
humans – their behaviour, personality characteristics, mental state and
abilities. This gives us a hint as to what psychology is and what
psychologists do. Psychology could be defined as the scientific study of
behaviour and mind.

Psychology as a scientific endeavour

When you listen to a psychologist or read about what they do, the
claims they tend to make are often associated with the terms ‘evidence’
or ‘research’. Even if psychologists sometimes say things that sound
common-sensical, their arguments are almost always made on the basis
of the results of studies. In other words, psychologists will usually
present arguments based on evidence gathered through research.

3
General introduction

Let us look at the three news reports again and explore how the specific
conclusions about the benefits of napping, the link between internet use
and happiness, and behaviour in the classroom were arrived at.
In relation to the benefits of an afternoon nap, this newspaper article
reported a study by Matthew Walker and colleagues from the University
of California, Berkeley. In this study, thirty-nine healthy young adults
who agreed to take part were split into two groups of approximately
equal size. One group became the ‘nap’ group and the other the ‘no­
nap’ group. In the first phase of the study, all thirty-nine volunteers
completed a learning task designed to test their memory. They
completed the task at noon, and both the ‘nap’ and ‘no-nap’ groups
performed equally well. Two hours later, at 2 p.m., the ‘nap’ group had
a 90-minute sleep, whilst the ‘no-nap’ group remained awake. Then, at 6
p.m., both groups participated in more learning tasks. The researchers
found that those who were in the ‘nap’ group performed better than
those in the ‘no-nap’ group. The explanation offered is that the 90
minutes of sleep allowed the brain of those in the ‘nap’ group to ‘store’
the learning that took place at noon, thus making room for the new
learning that occurred at 6 p.m.
This type of investigation is known as an experiment. You may remember
carrying out experiments at school for scientific subjects such as
chemistry, biology and physics. But experiments are not limited to the
natural sciences; psychologists make extensive use of them too. In fact,
an experiment doesn’t require test tubes and other laboratory apparatus.
It involves, instead, carefully manipulating whatever it is you are
interested in studying and then measuring the effect that this
manipulation has had on something else. In the above example, the
psychologists manipulated the volunteers’ experience so that half had a
nap and the other half did not, and then they measured the effect this
had on learning.
Throughout this book, you will be introduced to this method of
investigation in more detail, and learn about some very well-known,
‘classic’ experiments that have been conducted to investigate a number
of different issues, including obedience, learning, attention and memory.
However, experiments are just one of many types of method that
psychologists use when carrying out their investigations.
The second newspaper article used a different method to explore the
link between the amount of time spent online and happiness. This claim
was based on a study conducted by psychologists Catriona Morrison

4
 Defining psychology

and Helen Gore from the University of Leeds. Their study involved the
use of a questionnaire which was sent out to 1319 young people and
adults. Of this number, eighteen individuals were deemed to be addicted
to the internet. A high proportion of these internet addicts were found
to be depressed.
This research finding was therefore based on a type of investigation that
involved asking people to respond to a list of questions. In fact, they
were asked several sets of questions, including one set about what they
used the internet for, another set about how much they used the
internet, and a third that had been designed specifically to measure
whether or not they suffered from depression. By comparing the
answers to the different sets of questions, researchers were able to draw
conclusions about possible links between internet use and depression.
So, questionnaires are another type of method that psychologists can use.
In the first chapter of this book, you will read more about the
construction and use of questionnaires. An important point to consider
is what type of conclusions can be drawn from questionnaire-based
studies. In the above case, for example, it is difficult to tell whether
internet addiction causes depression, or the other way around. As
Catriona Morrison explains (Daily Telegraph, 3 February 2010): ‘Our
research indicates that excessive internet use is associated with
depression, but what we don’t know is which comes first – are
depressed people drawn to the internet or does the internet cause
depression?’ This issue of causality is an important one in psychological
research and is discussed in several chapters of Investigating Psychology.
What about the claim in the third article, concerning pupils’ behaviour
in the classroom? The investigation was led by Brian Apter, an
educational psychologist employed by the Wolverhampton City Council.
With his collaborators he studied the behaviour of primary school
pupils in 141 classrooms. Their findings were then compared with those
obtained in earlier studies of classroom behaviour that have been
routinely conducted since the 1970s. From these comparisons, it was
possible to infer an improvement in classroom behaviour that began in
the mid 1980s and that then continued for the following two decades.
The reason for this improvement was thought to be better teaching and
more verbal engagement with pupils.
This study used a method of investigation known as observation.
Information regarding the pupils’ behaviour was obtained by observing
and recording the number of pupils who were ‘on-task’ and following

5
General introduction

the teacher’s instructions, and the number of pupils who were deemed
to be ‘off-task’ and not following the teacher’s directions. In several
chapters you will read about studies that used observations, and learn
more about how this method can be used to arrive at reliable
conclusions about the behaviour that is being investigated.

Diversity within psychology

The previous section described three different methods that are used in
psychological investigations: experiments, questionnaires and
observation. As you progress through the chapters of this book you will
come across other methods. For example, in Chapter 1 you will read
about the use of interviews to gather information about people’s views on
a range of topics, including their childhood. This particular approach
can be used to obtain a much more in-depth understanding of people’s
attitudes, beliefs and dispositions, and is one that is increasingly adopted
in psychology.
Therefore, psychologists vary in the approach they take to their
research, with some adopting the experimental method typical of the
natural sciences, and others engaging in the sorts of investigation (such
as questionnaires and interviews) that are more typical of the social
sciences. Crucially, however, this is not the only kind of variability that
exists within psychology. There are also many sub-disciplines, which
differ in respect of their main focus of study. As you progress through
the chapters, you will be introduced to some of the main sub­
disciplines, including social psychology, developmental psychology,
biological psychology, cognitive psychology and the psychology of
individual differences. Although there is some overlap between them,
each sub-discipline focuses on a different aspect of human behaviour,
characteristics, mental states or abilities, and draws on methods of
investigation that are appropriate to its particular focus.

The structure of this book

The primary aim of Investigating Psychology is to illustrate how
psychologists go about their investigations and to give you a taste of the
range of topics they address. The book is divided into three parts, each
consisting of three chapters and a short introduction and conclusion.
Each part has a common theme running through the chapters. Part 1
looks at what makes people do harm to others. Part 2 considers how
other people influence or shape what we do, while Part 3 explores

6
 Defining psychology

internal mental processes, namely language, attention and memory, and

looks at different ways in which they can be investigated.
Each chapter follows a common structure and is divided into five parts:
. Section 1 introduces the specific topic explored in the chapter, and
how this relates to a particular problem relevant to everyday life.
. Section 2 describes a pioneering study, one that is regarded today as
being a classic piece of psychological research.
. Section 3 explores in more detail a particular feature or underlying
assumption of the classic study that is of broader relevance to
psychology.
. Section 4 considers developments in psychology since the classic
study was completed.
. Section 5 reflects on the broader relevance of the classic study and
the more recent findings to the particular problem set out in
Section 1.

The studies explored in the book are a very small subset of the research
conducted by psychologists to date. However, the material has been
carefully chosen to provide a taster of the main sub-disciplines and the
range of methods used by psychologists. By exploring in detail nine
classic studies in distinct areas of psychology and by describing the
work that followed, we hope to give you an insight into the diverse
nature of the discipline. Also, you will notice that included in the book
are studies conducted as early as the nineteenth century, across the
twentieth century and into the twenty-first century. So, in reading the
chapters you will get a sense of how psychology has evolved over the
years, and how our understanding of the human mind and behaviour
has become more sophisticated, although many questions remain
unanswered.
In addition to the five sections, each chapter contains activities that you
are invited to undertake. These should not take very long to complete
and they will increase your engagement with, as well your understanding
of, the material covered in the chapter. There are also two boxes in
each chapter, focusing on specific points relevant to the topic under
discussion. One of these, entitled ‘Why do it this way?’, explores the
reasons or assumptions underpinning the method used in the classic
study described in that chapter, and highlights some broader issues
relevant to psychological research more generally.

7
General introduction

Finally, the classic study covered in each chapter addresses a particular

everyday or real-world issue. For instance, Chapters 1 and 2 explore
whether the conduct of Nazi war criminals is best explained in terms of
personality or situational influences, Chapter 5 examines a number of
important questions related to the nature of an infant’s attachment to
their caregiver, while Chapter 8 considers if it is safe to drive while
using a mobile phone. This link to everyday concerns will help you to
get a sense of the broader relevance of psychological research, and the
fact that psychology has always been responsive to practical problems
faced by society as a whole.
We hope that you will enjoy finding out about what psychologists do,
and how their research sheds light on everyday life.
Jovan Byford and Nicola Brace

8
Part 1
 Introduction

Introduction
‘What makes people do harm to others?’ This might sound like a
strange and somewhat negative question with which to begin the
journey of discovering psychology. Why not look at more positive
themes – love, happiness, friendship or altruism? Psychology must have
something to say about these, inherently psychological, yet more
affirmative and upbeat topics.
While this might be the case, it is also true that within psychology there
has traditionally been a greater emphasis on trying to explain, account
for and suggest ways of changing the more negative aspects of human
behaviour. This is in part because society as a whole tends to call upon
experts, psychologists included, when things go wrong. At the same
time, psychologists and their research interests are, as you will see,
deeply rooted in the historical and social context. Scholarly attention is
often drawn to the problems and challenges facing society, so the
questions that psychologists choose to address, as well as the way they
go about seeking answers, are determined by the world that they
inhabit.
It is therefore not surprising that after 1945, when the world became
aware of the atrocities committed by the Nazis and their allies during
the Second World War, psychologists became particularly aware of the
need to investigate and explain the potential psychological roots of the
excesses of human violence and aggression, including mass murder and
genocide. Subsequent events, such as the massacre of civilians in the
village of My Lai in Vietnam committed by a US military unit in 1968,
or the genocides in Cambodia (1975–79), Rwanda (1994) and Bosnia
(1995), suggested that the events of the Second World War were in fact
not an isolated case, but that the tendency to cause harm to other
human beings on a massive scale has become a stable aspect of the
human condition. What is more, people do harm to others also on a
more ‘mundane’, everyday level, and not just in war. Horseplay in the
schoolyard can deteriorate into bullying, aggressive intimidation and
violence. Violence of one kind or another takes place regularly in city
centres, schools, playgrounds, and homes. Although the media often
exaggerate the frequency of anti-social behaviour and violent crime,
these phenomena are nevertheless around us and represent a cause for
concern.

11
Introduction to Part 1

One thing that people often ponder over when hearing of stories of
aggression and violence are the origins of the perpetrators’ actions, and
the question from the beginning of this introduction gets posed: ‘What
makes people do harm to others?’ Significantly, the explanations that
spring to mind are mainly psychological. Is it personality? Are there
certain kinds of people who are more likely to harm others because of
their character and inbuilt dispositions? Or is the situation to blame? Is
the context in which an individual might find themselves the main
determinant of their behaviour? If so, how does the environment
influence an individual? Also, how do people become violent? Are the
roots of violence to be sought in early childhood, in the relationship
with parents and caregivers? Or is violence simply learned, from
watching others and from living in a society in which it is increasingly
difficult to avoid violent films and computer games?
These are some of the issues that are considered in Part 1 of Investigating
Psychology. In Chapter 1, Jean McAvoy examines the research on the
authoritarian personality which was carried out in the 1940s by a group
of researchers in the USA. Theodor Adorno, Else Frenkel-Brunswik,
Daniel Levinson and Nevitt Sanford were keen to uncover aspects of
personality that make some people predisposed to embrace the kind of
extremist, violent and racist ideology that was espoused in Germany
under Hitler. In their quest for this authoritarian personality they used a
complex method involving thousands of questionnaires and more than a
hundred in-depth interviews. The study will be used to illustrate two
issues relevant to research on personality: first, how personality can be
measured, and, second, how these differences in personality develop and
where they come from. As you will see, to address the latter question,
Adorno and colleagues drew on the psychoanalytic theory of Sigmund
Freud. Jean McAvoy also explores some contemporary developments in
the study of authoritarianism which account for this phenomenon in
very different ways.
Chapter 2, by Philip Banyard, examines one of the best-known studies
in the history of psychology, the obedience studies carried out by
Stanley Milgram in the 1960s. In his research Milgram demonstrated the
lengths to which people are willing to go just because someone tells
them to do something. The focus of Milgram’s work was on the power
of the situation and the extent to which doing harm is a function not so
much of who we are, but of the circumstances we find ourselves in.
Philip Banyard also discusses the ethics of Milgram’s research,
specifically whether the author of the study took sufficient precautions

12
 Introduction

to protect the integrity and well-being of the participants in the

experiment. The chapter also considers developments in obedience
research since Milgram, and in particular how psychologists have tried
to overcome the ethical issues posed by Milgram’s obedience studies.
How does one investigate obedience without compromising the well­
being of participants?
In Chapter 3 John Oates explores the work of Albert Bandura on the
effects on children of the media’s portrayal of violence. Through a
discussion of a classic study, John Oates introduces the broader notion
of social learning – learning through observation and imitation. Chapter 3
also discusses the role that psychologists can, should and do play in
informing government policy on regulating the portrayal of violence in
the media, especially given the attention that the media itself places on
the apparent causal link between youth crime and aggression and
exposure to violent games and films.
We hope that you will enjoy the chapters and that you will gain not just
a greater understanding of what makes people do harm to others, but
also an appreciation of what psychologists do and of the challenges they
face as they seek to explain and understand human behaviour.

13
Chapter 1
Exposing the
authoritarian personality
Jean McAvoy
Contents
1 Introduction 19
1.1 Approaches to personality 20
1.2 Personality and ‘authoritarianism’ 23
2 The authoritarian personality 26
2.1 Measuring personalities 27
2.2 Interpreting the scales: composing an authoritarian
personality 32
2.3 Explaining the causes of ‘authoritarianism’ 33
3 Exercising caution: evaluating the ‘authoritarian
personality’ study 39
3.1 Focusing on method: the questionnaires 39
3.2 Focusing on method: the interviews 42
3.3 Authoritarianism as personality? 42
4 Revivals and revisions of the authoritarian
personality 47
4.1 Altemeyer and right-wing authoritarianism 48
5 Final thoughts 52
References 55
Chapter 1 Exposing the authoritarian personality

Aims and objectives

By the end of this chapter you should be able to:
. outline what psychologists mean when they refer to the term
‘personality’
. describe the notion of the ‘authoritarian personality’ and how it has
been measured
. discuss the methodological problems of using scales to measure
personality
. outline alternative approaches which have sought to explain
‘authoritarian’ behaviour.

18
 1. Introduction

1. Introduction
A central task for psychology is to try to explain people’s behaviour. A
lot of what psychologists do involves addressing the question: ‘Why do
people do the things they do?’ An important factor in human behaviour
which psychologists have traditionally been interested in is personality. Personality
There are many definitions of personality but, in general terms, it refers A person’s stable and
to a set of stable and enduring individual characteristics or inner enduring traits and
characteristics, which
dispositions that lead people to behave in a steady way over time and
lead them to behave in
maintain a consistent orientation to other people and the world around a steady way over time.
them.

Activity 1.1
Spend a few minutes thinking about someone you know. How would you
describe their personality? Do you think their personality influences the
way they behave? Think about some of the things they do that you
consider are down to their personality.

What kinds of things did you come up with? What personality

characteristics does the person you have chosen have? What terms did
you use to describe him or her? Did you use terms like ‘extrovert’,
‘serious’, ‘competitive’, ‘bossy’, ‘liberal’, ‘realistic’ or ‘a good listener’?
You probably came up with a completely different set of terms. This is
unsurprising given that there are thousands of words available to us that
relate to aspects of personality, and people use them in everyday life to
describe themselves and others, attribute motives for behaviour, and
make sense of the world. The fact that there are so many different
terms to describe personality is in itself illustrative of its importance. It
has been found, for instance, that people, especially in Western societies,
have a tendency to attribute causes of someone’s behaviour to their
personality, rather than to the situation, even in instances where there is Fundamental
attribution error
overwhelming evidence that a situation is more important (Ross, 1977).
The tendency to
For example, if people see a stranger slip and fall in the street, they are explain the causes of
more likely to attribute the accident to the person’s ‘clumsiness’ or other people’s
‘carelessness’ (a feature of their personality) than to the fact that the behaviour as a product
ground is slippery (an aspect of the situation). They will tend to do so of their internal
even when a situational explanation would have been more appropriate characteristics and
dispositions rather than
(such as if the pavement is visibly wet). Lee Ross (1977) has called this
external situational
preference towards a personality-based explanation for behaviour the factors.
fundamental attribution error. Personality therefore appears to be

19
Chapter 1 Exposing the authoritarian personality

central to the way in which, in everyday life, many people view and
interpret the causes of other people’s actions, attitudes, choices, etc.

Figure 1.1 Prior to the 1983 general election, Neil Kinnock, then leader of the
Labour Party, was walking on the beach when he tripped and fell during a
photo opportunity. Political commentators suggested the incident contributed
to Kinnock’s party losing the general election because people saw his fall as
a reflection of his character

Psychologists tend to be more cautious about the role of personality in

human behaviour. Very few psychologists today would claim that
behaviour can be explained exclusively in terms of personality
characteristics, and, instead, they have warned against the dangers of
focusing too much on personality. Nevertheless, personality is widely
acknowledged as playing an important role in shaping the way people
are and how they interact with the world around them.

1.1 Approaches to personality

For psychologists interested in personality, the first task has been to
define, categorise and measure personality. Because the term
Individual differences ‘personality’ encompasses characteristics and dispositions that are
Any characteristics that susceptible to individual variation (no two people have identical
are susceptible to
personalities), the efforts to define and measure personality are part of a
variation between
individuals; for sub-discipline of psychology concerned with the study of individual
example, personality or differences. As mentioned previously, there are thousands of words
intelligence. describing personality characteristics. Can these be reduced to a smaller

20
 1. Introduction

and therefore more manageable cluster of personality traits? For

instance, if it could be established that people who are ‘affectionate’
tend also to be ‘generous’, ‘gentle’ and ‘modest’, then maybe all these
different terms to describe that aspect of a person’s personality are not
necessary. Instead, ‘agreeableness’ could be used as a broader category
that encompasses a range of characteristics, including how ‘gentle’ or
‘generous’ or ‘affectionate’ someone is. If this rule is followed, it is
possible to reduce the vast number of terms to a small number of
categories that can be used to describe and measure human personality.
This is precisely the approach taken by Lewis R. Goldberg, for example,
who, in the 1980s, devised what are known as the ‘Big Five’ personality
factors (Goldberg, 1981). According to Goldberg, personality can be
divided into five dimensions, namely Extraversion (which covers things
like sociability, assertiveness, adventurousness), Agreeableness
(mentioned above), Conscientiousness (things like orderliness and
reliability), Emotional stability (things like security, irritability,
emotionality) and Intellect (imagination, perceptiveness, etc.). The
assumption here is that each of the five factors is a continuum (a
seamless progression ranging from extreme extraversion to extreme
introversion, for example, or emotional stability to emotional instability)
with each person’s personality being definable in terms of where they
are along each of the five dimensions. As well as categorising and Personality Theory
defining personality in terms of a smaller number of factors, Goldberg A set of propositions
also devised a way of measuring personality. He developed a about the structure
questionnaire which is used to establish how extrovert, agreeable, and/or development of
emotionally stable, etc. a person is. personality that forms
the basis of a coherent,
Other psychologists researching personality have also tried to reduce evidence-based
personality to a smaller number of core characteristics, but have come explanation.
up with different dimensions. The personality theory put forward by
Paul Costa and Robert McCrae (1992), for instance, also has five
Extraversion
factors, but theirs are slightly different from those of Goldberg. Hans A personality type
Eysenck (1967), on the other hand, has put forward a theory which characterised by
states that personality consists of just three dimensions. There is some outgoing and
overlap between the different theories – Costa and McCrae, Eysenck gregarious behaviour. It
and Goldberg all have Extraversion as a factor – but there are also forms an extraversion–
introversion continuum
differences. How many dimensions a theory proposes and how these
with introversion, at the
are defined depends very much on the method that the researcher uses opposite pole, typified
to derive them and the various assumptions that guide their research. by reserved and
inward-looking
Importantly, however, psychologists have not just been interested in
behaviour.
describing people’s personalities or measuring them. They also looked at

21
Chapter 1 Exposing the authoritarian personality

how personality is formed. Why do people have different personalities?

What causes particular personality characteristics to develop in certain
people but not others?
A classic example of a personality theory that focused on development
rather than on measurement is the work of Sigmund Freud. Freud was
a psychoanalyst working in Vienna in the first half of the twentieth
century and is credited as the father of the psychoanalytic movement.
Psychoanalysis Psychoanalysis is an approach within psychology that explores the
A set of theories and ways in which human behaviour and thinking are influenced by
therapeutic methods unconscious processes. Although Freud’s immediate concern was the
exploring the
mental health of his patients, he also used his therapeutic practice to
unconscious processes
influencing human think about the human mind more generally and about what makes us
behaviour. who we are. Notably, Freud lived before psychologists became
interested in the details of categorising and measuring personality. He
was interested instead in where personality comes from, and used his
extensive experience as a psychoanalyst to theorise about the
development of personality.

Figure 1.2 Sigmund Freud (1856–1939)

The main tenet of Freud’s theory is that personality consists of two

broad elements: the conscious and the unconscious. He argued that much of
what a person is like (i.e. their outward personality of the kind

22
 1. Introduction

measured by personality tests) is determined by drives, impulses and

motivations that are not accessible to conscious awareness. Outward
manifestations of personality are therefore just the tip of the iceberg,
one that lies at the top of a much larger structure – the unconscious.
Also, Freud stipulated that personality is formed in childhood through
the way an individual experiences and reacts to people around them.
According to psychoanalytic theory, the growing child goes through
several stages of development. At each stage he or she experiences
gratifications and frustrations. For example, according to Freudian
theory the first stage of development is the ‘oral stage’ when a baby’s
interaction with the outside world is limited to the satisfaction of its
basic need: the need to eat. At this stage, the baby concentrates on
learning about its mouth, and all the pleasures and discomforts
associated with it – such as sucking, biting or hunger. Normally the
child will progress from this stage after weaning. However, if a child
encounters some difficulty at the oral stage, and its progress through it
is inhibited in some way, the child will become ‘fixated’ at this stage and
this will shape its adult personality. Thus, if an infant is weaned too
early, for example, she or he might grow into an adult who is
unconsciously motivated to seek extra oral gratification, such as to
overeat, for example. Now, I have presented this in quite a simplistic
way. You might not be surprised to hear that Freud’s theories are a lot
more complicated and sophisticated than this, but this just introduces
two assumptions underpinning the psychoanalytic theory of personality.
First, infant and early childhood experiences (including the relationship
with parents) have a strong influence on personality formation. Second,
besides the conscious aspect of personality, and the visible traits,
behavioural tendencies and dispositions, there is also the world of the
unconscious, which is inaccessible to conscious awareness.

1.2 Personality and ‘authoritarianism’

This chapter examines a specific theory of personality, one that, rather
Authoritarian
than looking at personality generally, sought to describe, measure and personality
explain a specific kind of personality: the authoritarian personality. A kind of personality
During the Second World War, the world witnessed atrocities on a typified by obedience to
horrific scale. Millions of people were murdered by the Nazis in the authority, strict
most brutal ways. Among them were six million European Jews, most adherence to rules, and
hostility towards
of whom were systematically murdered in Nazi concentration camps.
anyone different from
After the war, psychologists sought to explain what it is that leads some oneself.
people to take part in such extreme violence. One question that they

23
Chapter 1 Exposing the authoritarian personality

asked is whether there is a particular kind of person who is more likely

to commit this kind of aggression.

Figure 1.3 Do violent and destructive movements and ideologies such as Nazism attract only certain
kinds of people?

An answer to this question was proposed by a group of researchers

working at the University of California, Berkeley, in the 1940s. Theodor
Adorno, Else Frenkel-Brunswik, Daniel Levinson and Nevitt Sanford
suggested that some people have a specific kind of personality that
makes them susceptible to extremist ideologies and predisposes them to
commit acts of aggression and murder in the name of political ideology.
They called this the authoritarian personality. In this chapter you will read
about the idea of authoritarian personality in more detail. How did
Adorno and colleagues go about looking for the authoritarian
personality? How did they measure it, and most importantly how did
they explain its roots? What is it that makes people become
authoritarian? Finally, how valid is the claim that participation in mass
violence is attributable to a particular kind of personality?

24
 1. Introduction

Summary
. Psychologists largely agree that personality plays an important part in
how people behave.
. Many theories have been developed that try to define and measure
personality.
. Freud argued that personality consists of two main elements: the
conscious and the unconscious.
. After the Second World War, Adorno and colleagues examined what
makes people susceptible to extreme ideologies such as Nazism.
Their research led to the concept of the authoritarian personality.

25
Chapter 1 Exposing the authoritarian personality

2 The authoritarian personality

In the 1940s, Nevitt Sanford was a psychology professor working at the
University of California at Berkeley. He recruited a PhD student, Daniel
Levinson, to work on constructing a method of measuring anti-
Anti-Semitism Semitism (Smith, 1997). Later, Sanford invited Else Frenkel-Brunswik
Prejudice and hostility and Theodor Adorno to join his project and collaborate with him as
towards Jews. part of a research team. Both Adorno and Frenkel-Brunswik had
witnessed the rise of fascism and anti-Semitism first hand. Adorno
Fascism
was an eminent sociologist in Germany, and Frenkel-Brunswik was a
A political ideology or child psychologist in Austria. Both of them fled Germany in the 1930s
regime marked by and made their way to America. Like many scholars at this time,
extreme nationalism especially those who had been affected personally by the events in
and racism, Europe leading up to and during the Second World War, the research
centralisation of team were trying to explain how it was that the Second World War saw
authority and the
suppression of political
so many people join in such extraordinary inhumanities and on such a
opposition. large scale.
But, while the research team were motivated by the desire to understand
what had happened in Europe between 1933 and 1945, they conducted
their research in the United States, in the years after Nazism had
already been defeated. Their starting point was that anti-Semitism and
fascism were not solely a ‘German’ problem, as some had speculated.
Instead, as Adorno and his colleagues soon realised, they had no
difficulty in finding Americans whose outlook was such as to indicate
that they would readily accept fascism if it should become a strong or
respectable social movement. In other words, people with a
predisposition to accept violence towards anyone different from
themselves or embrace an extremist political ideology were to be found
anywhere, not just in Germany. It is this predisposition for fascism that
Adorno and colleagues called the authoritarian personality.
This study by Adorno, Frenkel-Brunswik, Levinson and Sanford,
published in 1950, quickly had an enormous impact. A review by
Brewster Smith in 1950 described it as ‘certainly the most extensive and
sophisticated research on [anti-Semitism] yet contributed by
psychologists’ (quoted in Smith, 1997, p. 160). Today, the description of
someone as ‘authoritarian’ has entered the colloquial language, and is
used to describe the kind of attitudes and dispositions that Adorno
et al. (1950) sought to explain. In the following sections I will look at
how Adorno, Frenkel-Brunswik, Levinson and Sanford conducted their
investigation, and why they did it the way they did.

26
 2 The authoritarian personality

Before doing so, it is worth drawing attention to the issue of authorship

in the authoritarian personality study. Theodor Adorno has become
almost synonymous with the research on the authoritarian personality.
And yet it was Nevitt Sanford who instigated this research, with
Adorno joining the project last, after Levinson and Frenkel-Brunswik.
The reason why the study is known under Adorno’s name is purely and
simply because ‘A’ comes first in the alphabet! When they were
preparing the almost 1000-page book in which they published the
results of their work, the authors took the decision to list the
researchers’ names alphabetically. So the study became known as
Adorno et al. (1950). There is a particular irony in this, though. A little
while earlier Adorno had changed his name, shortening it from
Wiesengrund-Adorno to simply Adorno. But for this change, Adorno
would have been the fourth named author on the study and we might
now be referring to authoritarian personality as the work of ‘Frenkel-
Brunswik and colleagues’ (Altemeyer, 1996).

Figure 1.4 Theodor Adorno’s name became principally associated with the
authoritarian personality study, although he was not the main contributor

2.1 Measuring personalities

The idea of ‘measuring’ personality is quite a strange one in some ways.
Think back to the description of someone you know which you

27
Chapter 1 Exposing the authoritarian personality

provided for Activity 1.1. Perhaps describing the personality of the

person you chose felt reasonably straightforward. But how would you
go about ‘measuring’ their personality? In your description, did you use
words like quite bold, very conscientious, always enthusiastic, or sometimes
relaxed? These kinds of words help us get some rough sense of how
much someone might exhibit a particular characteristic, but they are
quite broad terms, and not very precise. More importantly, they are
based on impression rather than objective measurement. In order to
make meaningful comparisons between two people’s personalities (for
instance, John is more ‘extrovert’ than Judy) or assess an individual’s
personality in relation to the population as a whole (William is more
‘agreeable’ than the average person), psychologists need quite precise
methods for actually measuring different aspects of personality.
Measurement was the first issue that Adorno and colleagues set out to
resolve. They began with the assumption, common in research on
Attitudes personality, that people’s attitudes are a reflection of their underlying
A person’s beliefs and personality characteristics. Adorno and his colleagues argued that a way
feelings about issues, of tapping into an aspect of human personality is by measuring
events, objects or
corresponding attitudes. So they began by trying to measure attitudes, in
people, which are
thought to influence order to draw some conclusions about underlying personality.
behaviour. Adorno and colleagues were interested in characteristics that they
believed accounted for much that happened in Europe during the
Second World War. Jewish people had been a particular target under the
violent Nazi regime so Adorno et al. developed ways of measuring
attitudes towards Jews in particular, but also towards anyone ethnically
different, as well as attitudes towards traditional, conservative political
beliefs.
Adorno et al. (1950) argued that these three attitude sets were all
intrinsically related to one another so that people who endorse
conservative political beliefs tend also to exhibit intolerance towards
those who belong to other ethnic groups, as well as towards Jews.
Scale
A set of items such as To measure these attitude sets, Adorno et al. used scales. In
questions on a psychological research a scale refers to a set of items – statements – to
questionnaire which which participants are asked to respond. The individual items on the
combine to measure a
scale are all related in meaning and together they combine to add up to
bigger construct (e.g. a
personality a bigger construct. So, for example, in the ethnocentrism scale devised
characteristic) that by Adorno et al., there are a set of statements about people from
cannot be measured minority ethnic groups and about people from the dominant American,
directly. white, Christian group in the 1940s and 1950s (see Table 1.1 for

28
 2 The authoritarian personality

examples). Participants in the study were asked to rate how much they
agree with each statement. These individual statements are said to
combine to represent a bigger characteristic – ethnocentrism – Ethnocentrism
meaning favouring one’s own group at the expense of another. The Belief in the superiority
combined ratings on individual items are then used to derive each of one’s own ethnic
group or culture.
participant’s ethnocentrism score.

Table 1.1 Examples from the three attitude scales devised for the
authoritarian personality study

From the Anti-Semitism scale (AS-scale):

It is wrong for Jews and Gentiles to intermarry.

It is sometimes alright to ban Jews from certain apartment houses.
The best way to eliminate the communist menace in this country is to
control the Jewish element which guides it.

From the Ethnocentrism scale (E-scale):

In view of the present national emergency, it is highly important to limit
responsible government jobs to native, white, Christian Americans.
There will always be superior and inferior nations in the world and, in the
interests of all concerned, it is best that the superior ones be in control of
world affairs.
There will always be wars because, for one thing, there will always be
races who ruthlessly try to grab more than their share.

From the Politico-Economic Conservatism scale (PEC-scale):

Young people sometimes get rebellious ideas, but as they grow up they
ought to get over them and settle down.
In general, full economic security is harmful; most men wouldn’t work if
they didn’t need the money for eating and living.
Character, honesty, and ability will tell in the long run; most people get
pretty much what they deserve.

Source: Adorno et al., 1950, pp. 69–70, 110–11, 158

Table 1.1 shows some examples of the kinds of statement included in

the attitude scales. Remember, Adorno and his colleagues were trying to
find out the extent to which ordinary American citizens would agree
with overtly anti-Semitic, ethnocentric and conservative ideas.
Participants were presented with the scales in the form of
questionnaires and were asked to indicate their own levels of agreement

29
Chapter 1 Exposing the authoritarian personality

or disagreement with each item, saying whether they felt ‘slight support’
for the statement, ‘moderate support’, ‘strong support’, ‘slight
opposition’, ‘moderate opposition’ or ‘strong opposition’.
You may have noticed that these options do not invite participants to
offer any kind of ‘neutral’ response. The researchers argued that when
surveys such as this allow space for neutral responses, such as ‘neither
support nor oppose’ or ‘don’t know’, this middle-ground option tends
to attract a lot of responses from participants. The researchers here
wanted to ‘force’ participants to take a position either in favour of or
against the statements and therefore omitted this ‘neutral’ category.
Once participants had completed the scales indicating their levels of
agreement or disagreement, the researchers scored participants’
responses on a range from 1 to 7:

Strong opposition 1 point

Moderate opposition 2 points
Slight opposition 3 points
Slight support 5 points
Moderate support 6 points
Strong support 7 points

You can see that points run from 1 to 7, with the middle point – 4 –
being omitted. Four points were assigned only when participants did not
register any response to an item. Thus, although participants were not
given the option of a neutral response, their failure (or refusal) to rate a
statement was interpreted as neutrality and was therefore awarded the
score of 4.
The ratings of individual items within a scale were then added up to
give an overall score for each participant. The higher the score, the
more the person was said to exhibit the characteristic in question. So, a
person who tends to agree with (indicate support for) the statements in
the ethnocentrism scale would receive a high score, indicating a high
level of prejudice towards people from ethnic groups other than that to
which that person belonged.

30
 2 The authoritarian personality

Figure 1.5 Absence of ethnocentrism: choosing to mix together with other

ethnic groups

Having established scales to measure attitudes towards Jews, other

ethnic groups and conservative political ideas, Adorno et al. developed
an additional scale, the potential for fascism scale. This scale, known as
the F-scale, was central to Adorno et al.’s study. It included items that F-scale
made no explicit reference to racial or ethnic minorities, or conservative A measure of
politics. It was a measure not of attitudes but of personality personality
characteristics
characteristics underpinning potential for fascism.
underpinning potential
How did Adorno et al. go about developing the F-scale? Their starting for fascism.
point was that this personality measure, while containing items that do
not mention explicitly either Jews or other ethnic groups, would
nevertheless yield scores that correspond to those obtained on the
various attitude measures. This is because, they argued, there is a strong
relationship between the potential for fascism and prejudiced attitudes: a
person who scores highly on the anti-Semitism and ethnocentrism scales
should also obtain a high score on the F-scale, and vice versa. (They
were less concerned with the political and economic attitudes, as these
were believed to be less directly linked to potential for fascism, as a
personality measure.) So, Adorno et al. came up with a large number of
more general statements about obedience, respect for authority and
traditions, resistance to novel or unusual lifestyles, disapproval of sexual
freedoms and gender equality, and so on (see Table 1.2 for examples).
These statements were administered in a questionnaire to over 2000

31
Chapter 1 Exposing the authoritarian personality

respondents, along with the other three scales, the ones that measured
anti-Semitism, ethnocentrism and political and economic attitudes.
Based on the responses that they collected, Adorno et al. identified
those items on the potential for fascism measure that were most
strongly related to the anti-Semitism and ethnocentrism scores. This
enabled them to draw together a final version of the F-scale containing
those items that best predicted a person’s anti-Semitism and
ethnocentrism score. The F-scale, which measured the potential for
fascism, lay at the core of the authoritarian personality.

Table 1.2 Examples from the Potential for Fascism scale (or F-scale)
devised for the authoritarian personality study

No insult to our honor should ever go unpunished.

Obedience and respect for authority are the most important virtues
children should learn.
Books and movies ought not to deal so much with the sordid and seamy
side of life; they ought to concentrate on themes that are entertaining or
uplifting.

Source: Adorno et al., 1950, pp. 226–7

The authors of the study were eager to make it clear, however, that
what the F-scale was measuring was the potential for fascism, not
fascism itself. After all, they were not conducting research on active
members of a fascist movement or German Nazis, but white, middle­
class Americans from around San Francisco. What they wanted to tap
into, therefore, are the kinds of personality characteristics that
predispose individuals to become seduced by fascist ideology. The
investigators argued that the score on the F-scale can determine the
extent to which a participant exhibits personality characteristics –
namely authoritarianism – that would make them likely to support
oppression and violence towards minority groups.

2.2 Interpreting the scales: composing an

authoritarian personality
Adorno et al. argued that participants scoring highly on the F-scale were
revealing an authoritarian personality; and, in contrast, those with low
scores were revealing a non-authoritarian, more liberal, personality. The
authoritarian personality was described as consisting of nine

32
 2 The authoritarian personality

characteristics. These included things such as deference to authority,

obedience to rules, rigid adherence to ‘traditional’ values, the rejection
of anything unconventional, rigid thinking and hostility to others,
especially those appearing weaker in some way. Also linked to
authoritarianism were some less obvious characteristics. For instance,
people with an authoritarian personality tend to perceive the outside
world as threatening, and they are said to be more superstitious, and
overly concerned with sexual propriety.
According to Adorno et al. (1950, p. 228), these characteristics were
interconnected and formed a ‘single … more or less enduring structure
in the person’. This ‘structure’ – the authoritarian personality – provides
the framework through which some people make sense of the world
and through which their actions are shaped. More importantly, it
predisposes individuals to prejudice, and makes them receptive to anti­
democratic, fascist and right-wing ideologies.
Crucially, Adorno et al. were not saying that everyone who scored
highly on the authoritarian personality scale was a fascist; or that
everyone with an authoritarian personality would commit violent acts
against others in the name of political ideals. Rather, they were saying
that, where people high in authoritarian characteristics found themselves
in a society that tolerated prejudice towards weaker others, they would
be the ones most likely to endorse those prejudiced ideas, be swayed by
arguments that suggested the ideas were reasonable, and then act on
them. So authoritarianism was a predisposition for fascism which would
become realised if the political climate of the day made right-wing
ideologies socially acceptable.

2.3 Explaining the causes of ‘authoritarianism’

So far, you have read about the way Adorno et al. set out to discover
people’s attitudes, and used these to say something about their
personalities. However, you may also have noticed that while the scales
were able to measure authoritarianism and locate people on a range from
‘low in authoritarian tendencies’ to ‘high in authoritarian tendencies’,
these measures do not explain where the differences come from. They do
not tell us why some people present these characteristics suggesting an
authoritarian personality, and others do not. So, where does an
authoritarian personality come from?
To answer this question, Adorno et al. decided they needed a more in­
depth, individual approach. In the first phase of the study, in which they

33
Chapter 1 Exposing the authoritarian personality

developed the F-scale and looked at how it is related to prejudice,

Adorno et al. administered questionnaires. In the second phase, where
they looked for the underlying causes of authoritarianism, they decided
Interview to conduct interviews. Adorno et al. (1950, p. 291) argued that an
A ‘conversation with a interview allows the researcher to ask the participant open questions
purpose’, designed to about ‘what he thinks about himself, about his hopes, fears and goals,
gather in-depth
about his childhood and his parents, about members of the other sex,
information from
research participants. about people in general’. The interviews therefore offered the
Interviews may be participant in the study the necessary ‘scope and freedom of expression’
structured, asking the (p. 291).
same questions of all
participants, or semi­ As you can imagine, interviews are much more time-consuming than
structured or questionnaires. In the first phase of the study, Adorno et al.
unstructured, allowing administered the questionnaires to more than 2000 respondents. It
interviewers to adapt would have been impossible to interview all of them. Instead, they
questions according to picked a sample of 150. Roughly half were from those who scored
participants’ responses.
highly on the authoritarianism scale, and half from those who obtained
low scores. This gave the researchers two distinct groups to work with
so that they could look for patterns of similarities and differences
between them. These 150 participants were interviewed in depth to find
out why they had or had not developed an authoritarian personality. As
Adorno and colleagues put it, they wanted to find out the following:

Does the family constellation differ in the typical prejudiced home

as compared with the typical unprejudiced home? Do prejudiced
individuals tend to have different images of their parents than do
unprejudiced ones? How does the handling of discipline vary in
this respect? Do prejudiced and unprejudiced individuals differ in
their sex life, their way of choosing friends, their values, their
general cognitive and emotional approach to life?
(Adorno et al., 1950, pp. 291–2)

The interviews lasted an average of two hours each. They covered wide­
ranging questions. Participants were asked about their work, income and
religion, and questioned about politics and attitudes to minority groups.
But the questions that interested the researchers most were those about
the participants’ families, their early childhoods, and their sexual and
social relationships. This was because the researchers were heavily
influenced by psychoanalysis. Psychoanalytic theory argues that the
importance of childhood and parenting practices is crucial to the way
personality develops. Moreover, the relationship styles developed in early

34
 2 The authoritarian personality

childhood become the template for adult relationships too. Therefore,

the research team set out to explore connections between the adult
scores on authoritarian personality scales and information gleaned from
interviews about early childhood relationships.
By comparing the interview data from the two groups, Adorno et al.
observed that respondents who scored highly on authoritarianism
tended to report growing up in stricter family environments, compared
with those who exhibited low authoritarianism. They were also more
likely to report having parents (or other primary caregivers) who had
been harsher disciplinarians, who demanded obedience, and who were
quick to punish infractions. Adorno and colleagues argued that, in
response to this harsher upbringing, children developed feelings of both
love and hate towards their strict parents. How and why this happened
was explained in terms of the psychoanalytic theory. The researchers
argued that children experienced a deep desire to please much-loved
parents; but the strict and quick discipline they received also induced a
hatred and resentment towards the disciplining parent. This tension
between simultaneous love and hate produced strong and conflicting
emotions in children, which were both psychically uncomfortable to
bear and difficult to reconcile. As a consequence, the feelings of hate
for the parents were repressed; that is, buried in the unconscious. The
outcome of this process is that the child was left idealising and revering
the harsh parent. Crucially, however, Adorno et al. argued that the
feelings of anger and hostility, although hidden away in the unconscious,
continued to have influence. They were displaced on to others –
especially those whom it was ‘safer’ or easier to hate and despise – in
other words, anyone who was seen as weaker and more vulnerable than
the self.

35
Chapter 1 Exposing the authoritarian personality

Figure 1.6 Roots of authoritarianism?

Perhaps you can see the way this argument is going? Children of strict
authoritarian parents, in order to deal with their environment,
underwent a number of unconscious processes which resulted in their
idealising the authority figures that were their parents, and redirecting
hatred, fear and aggression to weaker others. This, according to the way
Adorno et al. interpreted psychoanalytic theory, became the enduring,
albeit unconscious, pattern for organising their future adult relationships
too. Authority is to be revered and anyone perceived as weaker, or
unconventional, could become the target for hate and aggression.

36
 2 The authoritarian personality

Box 1.1 Why do it this way?

Adorno et al. (1950) used a mixture of methods, using both scale
measures and in-depth interviews. The scales allowed the
researchers to test large numbers of people quickly and
systematically and the questions could all be presented in the same
way to all participants. Then, scoring responses along a numerical
scale allowed the researchers to transform a complex series of
questions about beliefs and attitudes into a single numerical score.
The fact that this method involved some kind of measurement
means that the data collected was quantitative data. Transforming Quantitative data
people’s responses into scores (i.e. numbers) made it possible to Data that can be
make direct comparisons with others’ scores. measured, counted, or
expressed in numerical
The interviews, on the other hand, allowed the researchers to terms, for example,
collect much ‘richer’ data. The scales had constrained participants scores, ratings or
to responding only to those items presented by the researchers, percentages.
and limited them to simple indications of levels of agreement or
disagreement. In contrast, the interviews allowed participants to
present, in their own words, their life histories, their experiences,
their own ideas and thoughts. Data collected in this way is referred
to as qualitative data. Qualitative data
Data that is not in
The two different methods used by Adorno et al. reflected the two numerical form. There
traditions in psychology that influenced the researchers. One was are different kinds of
psychoanalysis, which uses clinical interviews and free-flowing qualitative data, for
conversation as a way of investigating the human mind, and the instance interview
other was the approach to studying personality, which focuses on material, written text
measurement and quantification. such as newspaper
articles, or diaries.

37
Chapter 1 Exposing the authoritarian personality

Summary
. Adorno et al. used a series of questionnaires to measure attitudes to
Jews, minority groups and conservative beliefs, and the potential for
fascism.
. The researchers also used interviews informed by psychoanalytic
theory to identify the causes of these authoritarian personalities.
. Adorno et al. argued that harsh discipline in early childhood caused
children to unconsciously project feelings of hate onto people
different from themselves, and this became the template for their
adult relationships.

38
 3 Exercising caution: evaluating the ‘authoritarian personality’ study

3 Exercising caution: evaluating the

‘authoritarian personality’ study
The authoritarian personality study has had tremendous impact since its
publication in 1950. Over 2000 articles were published in response in
the first forty years alone (Meloen, 1991). On the other hand, the idea
of the authoritarian personality has been subject to sustained criticism.
This criticism takes two main forms: a critique of the methods Adorno
et al. used, and a critique of their interpretations of the data. I will now
examine these in turn.

3.1 Focusing on method: the questionnaires

The first of the main criticisms of Adorno et al.’s method centres on
the construction of the F-scale. Psychologists who have used
questionnaires and scales to measure attitudes and personality have
uncovered something called the acquiescence response bias. They Acquiescence
found that participants can show a tendency to agree with statements response bias
presented in attitude scales, regardless of the topic. There is, in other A tendency to agree
with statements
words, an inclination, among some people, to simply ‘go along’ with the
presented in scales,
ideas expressed in the questionnaire items. The possibility that people regardless of content.
may exhibit an acquiescence response bias is sufficiently well established
that it is now a major concern for the way questionnaires are
constructed.
Why is acquiescence a problem for scale construction? How is it
relevant to Adorno et al.’s study? Recall that the F-scale was designed
by the authors so that all the statements run in the same direction. If
you look back at examples in Table 1.2, you will see that each statement
is worded in such a way that agreement always points towards
authoritarianism. Because of the acquiescence response bias, it is
impossible to determine, from looking at the scores, if participants who
scored highly on the scale really are particularly authoritarian, or
whether they are simply the sort of people who agree to any statement
encountered in questionnaires. Therefore, it cannot be ascertained for
sure whether the scale in fact measures what Adorno et al. assumed
that it measures, namely authoritarianism.

39
Chapter 1 Exposing the authoritarian personality

Box 1.2 Acquiescence response and the problem

of bias
Questionnaires have been popular in psychology and the social
sciences more generally partly because they are easy to administer
to a large number of people, they can generate a great amount of
data relatively quickly, and they are relatively inexpensive. However,
none of this is much use if the questionnaire is not well constructed
in the first place. Any bias resulting from the way a questionnaire is
constructed undermines the conclusions that can be drawn from it.
The F-scale underpinning the authoritarian personality study has
been criticised for eliciting just such a bias: the ‘acquiescence
response’. This refers to a tendency people have to agree with
statements put to them by researchers. Consequently, the extent to
which people might really agree with the items in the questionnaire
may be artificially inflated by the acquiescence response.
This has serious implications for the F-scale. Participants were
asked to complete the scale by saying to what extent they agreed
or disagreed with a series of statements. But, on the F-scale, items
were all worded in such a way that ‘authoritarian’ views were
always represented by agreement with the content. Are participants
indicating agreement with an authoritarian statement because this is
actually their opinion? Or are they just going along with the ideas
expressed? If there was any ‘acquiescence response’ in the original
F-scale responses it would have biased the results and
exaggerated the appearance of commitment to authoritarian ideas.

Activity 1.2
Think for a moment how you might solve the problem outlined in Box 1.2.
What could the researchers do to control the possibility that
acquiescence responses might bias the scores on the scale? Remember,
the problem is that all the items on the scale run in one direction and
‘acquiescing’ would score the same as ‘agreeing’ with authoritarian items.

The solution to this problem of acquiescence bias is to include a mix of

‘positive’ and ‘reverse’ statements. This means rewording some scale
items in such a way that a participant would have to disagree with a
statement in order to exhibit authoritarian ideas. Several studies have
done just this. In one example, Bernard Bass (1955) reversed half of the

40
 3 Exercising caution: evaluating the ‘authoritarian personality’ study

items on the F-scale. The column on the left shows some of the
original F-scale statements. The column on the right shows the same
statements, but this time in the reversed form:

Table 1.3

Examples from the F-scale Examples from Bass’s (1955)

devised by Adorno et al. (1950) reversed F-scale
Familiarity breeds contempt. Familiarity does not breed contempt.
No weakness or difficulty can hold Weaknesses and difficulties can
us back if we have enough will hold us back; will power is not
power. enough.
If people would talk less and work If people would discuss matters
more, everybody would be better more before acting, everybody
off. would be better off.

On Bass’s revised scale, if participants are being consistent in their

support for authoritarian ideas, they would agree with the statements on
the left, which run in favour of authoritarian ideas; but they would
disagree with the statements on the right, which run counter to
authoritarian ideas. Any tendency towards an acquiescence response bias
has been balanced out across both negative and positive statements.
So, if the responses overall continue to show support for authoritarian
ideas, investigators can be more confident that they have identified a
pattern in participants’ thinking rather than simply a tendency to ‘go
along’.
As Table 1.3 shows, the acquiescence response bias can be avoided if
statements that run in different directions are included, within the same
scale. This has since become established practice in psychology. In the
case of authoritarian personality, this means that the F-scale would
include some statements that are worded in such a way that disagreement
points to authoritarianism. Several subsequent studies did precisely this
and found that once the effect of the acquiescence bias is removed,
participants tend to score lower on the F-scale, and therefore exhibit
fewer specifically authoritarian tendencies (Christie, 1991). Nevertheless,
many of these later studies continued to find substantial support for the
idea of authoritarianism, and authoritarian personality, although these
tendencies were less pronounced than Adorno et al. originally thought.
So, while the acquiescence response bias clearly influenced Adorno
et al.’s findings, it did not completely invalidate the idea of authoritarian
personality.

41
Chapter 1 Exposing the authoritarian personality

3.2 Focusing on method: the interviews

A second criticism of the authoritarian personality study refers to the
interviews conducted by Adorno et al. in the second stage of the study.
Robert Rosenthal (1966) raised the concern that the researchers’
Confirmatory bias findings were affected by confirmatory bias. Confirmatory bias refers
A tendency to pay most to the way in which researchers can unintentionally (or otherwise) shape
attention to those the outcomes of their study to match the expectations they had at the
features of a
start. Remember that the interviews were conducted with two sets of
phenomenon that
appear to confirm prior respondents: one group was selected because they scored highly on the
expectations. authoritarianism scale, with the second consisting of participants who
obtained low scores. The methodological concern raised by Rosenthal
(1966) and indeed by others is that researchers who conducted the
interviews knew, for each and every interviewee, whether they belonged
to the high-scoring or low-scoring group. What is more, they knew the
basic hypothesis of the research; that is, the types of difference that
Adorno et al. expected to find. This knowledge may have affected, even if
only inadvertently, the content of the interview and the interpretation of
what the participants said. Note that there is no suggestion here that
Adorno et al. sought to fix the results of their findings. The
confirmatory bias often takes a very subtle form, and occurs without
the researcher’s conscious intent. This is why it is common practice
Double blind today to introduce what are known as double blind procedures. This
A research design means that both the interviewer and the interviewee would be unaware
where neither the of (or be ‘blind’ to) both the hypothesis of the study and which group
participants nor the
the interviewee belongs to. That way, the results could not be affected
investigator know
which group the by the interviewer’s (or the interviewee’s) pre-existing knowledge of, or
participants belong to, assumptions about, the study.
thus reducing the risk
of bias in measures and
interpretations.
3.3 Authoritarianism as personality?
Criticism of Adorno et al.’s study was not limited to methodological
issues. There were those who questioned whether the phenomenon that
the authors of the study sought to identify, measure and explain is an
enduring feature of personality at all. Others challenged the assumption
that the authoritarian personality is useful in shedding light on the type
of behaviour that led Adorno et al. to embark on their research in the
first place.
Let’s look at the second criticism first, before returning to the first one
in Section 4. If authoritarianism was the main cause of the events in
Nazi Germany between 1933 and 1945, then it means that there was a

42
 3 Exercising caution: evaluating the ‘authoritarian personality’ study

higher prevalence of authoritarian personalities in Germany compared

with other countries where Nazism did not develop. Given that the
authoritarian personality was researched after the end of the war, this
proposition could not be tested directly. However, Thomas Pettigrew
(1958) conducted a study in which he explored the levels of
authoritarian personality in South Africa and in different parts of the
USA. He wanted to see if there was a difference in levels of
authoritarianism between: apartheid South Africa, where racist policies
permeated every level of society; the southern USA where the policy of
segregation was still in place; and the northern USA where racial
discrimination was less prominent. If Adorno and colleagues were right,
and prejudice within a society can be accounted for by authoritarianism
among the population, then there should be differences between the
three contexts. What did the study reveal? Pettigrew found that, while
participants in the southern USA and South Africa were more racist
than their counterparts in the northern parts of the USA, authoritarian
personality levels, as measured by the F-scale, were similar across the
three groups. In each context, only around 10 to 15 per cent of the
population could be described as having an ‘authoritarian personality’.
So, there was little evidence of a link between the prevalence of
authoritarian personality and actual levels of prejudice within a society
(Duckitt, 1989). Authoritarianism alone was not sufficient to account
for the differences in prejudice and the levels of support for racial
discrimination.

Figure 1.7 Racial segregation in South Africa and the American South in the
1950s

43
Chapter 1 Exposing the authoritarian personality

This finding is important because it suggested that the prevalence of

high authoritarianism or potential for fascism as measured by the F­
scale is not necessary to produce fascist movements or sustain racist
regimes or political cultures. Pettigrew’s research appears to suggest that
the potential for fascism exists as a personality characteristic in every
society, and only in a minority of cases. But how then does fascism
become a mass movement? When Hitler came to power in 1933, he
attracted nearly 44 per cent of the popular vote. Even if it could be
assumed that all 10–15 per cent of the German population who were
authoritarian voted for Hitler, this only accounts for one-third of his
overall support, and means that two-thirds of his supporters were not
authoritarian (Billig, 1978). There was clearly something above and
beyond authoritarianism that accounts for the support which existed in
Germany in the 1930s for a fascist political party, its programme and its
leader.
A related criticism has been directed at the psychoanalytic interpretation
of the causes of authoritarianism. According to the original study, the
reasons why people act as they do are said to be hidden in the
unconscious. Crucially, because the processes underpinning
authoritarianism are unconscious, people are prevented from being able
to direct them at will, and the ability to exercise choice and agency is
limited. Unconscious processes keep individuals fixed in particular kinds
of personality, and lead them to repeat the same kinds of behaviour. If
this is so, then once formed, authoritarian personalities would tend to
remain that way.
There is, however, substantial evidence suggesting that deference to
authority, prejudiced thinking and other features of the authoritarian
personality fluctuate in any given population. Gerda Lederer (1993), for
instance, noted that authoritarian tendencies diminished in Western
Germany after the Second World War, when that country began its
transition towards democracy. If authoritarian personality is a stable
quality, as Adorno et al. suggested, then sudden reductions in
authoritarianism would be difficult to account for. Similarly,
authoritarianism can vary according to experience: it can increase or
diminish in response to specific social events or situations, sometimes
even within a short period of time (Minard, 1952; Pettigrew, 1958). An
explanation based on the idea of personality – which implies stable
predispositions and patterns of behaviour – does not seem to be able to
account adequately for such fluctuations.

44
 3 Exercising caution: evaluating the ‘authoritarian personality’ study

These criticisms of the authoritarian personality are important for a

number of reasons. They all seem to point to the fact that, although
Adorno et al. were clearly onto something with the idea of
authoritarianism, an authoritarian personality appeared to be neither
necessary nor sufficient for explaining prejudice, discrimination or
violence towards minorities. Evidently, social environment, cultural
norms and political context all have a key role to play. Remember,
however, that Adorno et al. argued from the outset that authoritarian
personality was a predisposition, something that is likely to manifest
itself only in certain socio-economic and political conditions. So they
were aware of the dynamic interplay between personality and context,
but prioritised personality. The balance of evidence appears to suggest,
however, that it is other factors, above and beyond personality, which
are the more important determinants of prejudice.
The criticisms of the authoritarian personality study should not detract
from its importance in the history of psychology. While echoing many
of these criticisms, Michael Billig (1978, p. 36) referred to the work of
Adorno et al. as ‘the single most important contribution to the
psychology of fascism’ and, more importantly, as ‘a major landmark in
the history of psychology’. This is because of the study’s sheer scope,
ambition and scale, and the fact that it sought to bridge the gulf that
separated the psychoanalytic tradition and the quantitative approaches to
personality (see Box 1.1). In that sense, Adorno et al. were pioneers,
keen to explain what they considered to be the most destructive
phenomenon in the history of the world. Therefore, they should be
evaluated on the basis not only of how much they got right, but of how
much they inspired others to think about prejudice and develop new
theories about the phenomenon they wanted to understand.

45
Chapter 1 Exposing the authoritarian personality

Summary
. The authoritarian personality study has been criticised both for the
methods used and for the interpretations put on the results.
. Subsequent research that addressed the issue of the acquiescence
response bias in the design of the F-scale found that
authoritarianism is a valid phenomenon and that some individuals
display characteristics associated with authoritarian personality.
. Confirmatory bias was identified as a potential problem during the
interviews because the interviewers knew participants’ scores from
the surveys.
. Later research questioned whether authoritarian personality is either
necessary or sufficient to explain events such as those in Nazi
Germany.

46
 4 Revivals and revisions of the authoritarian personality

4 Revivals and revisions of the

authoritarian personality
Soon after Adorno et al. completed their study, American society, and
the Western world more generally, found itself fighting another war, this
time a cold war with the communist Soviet Union. While Adorno et al.
had believed that fascism posed the greatest threat to society, others
argued that the threat of a possible nuclear war involving the USA and
the Soviet Union meant that communism posed the greater threat. So,
from a psychological perspective, authoritarianism needed to be
explained when it appeared both on the right of the political spectrum
(linked with racism, anti-Semitism and conservative attitudes), and also
on the left, which was seen as equally ‘authoritarian’, inflexible and
undemocratic, but not ‘fascist’ in the conventional sense.

Figure 1.8 Communist Soviet Union: a new threat to the Western world

To address this issue, Milton Rokeach (1960) reconceptualised the idea

of ‘potential for fascism’ (as measured by the F-scale) as a more general
authoritarianism termed ‘dogmatism’, a concept which could encompass
both right-wing conservatism and left-wing communism. More
importantly, Rokeach rejected the psychoanalytic explanation, originally

47
Chapter 1 Exposing the authoritarian personality

proposed by Adorno et al. He argued that the various features of

authoritarianism were the product not of unconscious processes but of
Cognitive style a distinct cognitive style, a particular way of structuring and processing
The habitual way a information. Those high in authoritarianism, or dogmatism, were said to
person processes be closed-minded. This means that they were less able to evaluate new
information.
information and more likely to maintain contradictory beliefs depending
on the authority behind the message. In other words, they tended to
look at the source of the message rather than its logic or plausibility. So,
if an authority figure presented inconsistent information it would be
endorsed purely because of the status of that figure. In contrast, those
low in authoritarianism/dogmatism were more open-minded, meaning
they would be more able to evaluate information independently,
recognise inconsistencies, and be more swayed by the content of the
message rather than the authority of the source. In essence, open and
closed cognitive styles were characterised by ‘the ability (or inability) to
discriminate substantive information from information about the source,
and to assess the two separately’ (Rokeach, 1960, p. 60). Crucially,
dogmatism was seen as a process of cognitive function – of information
processing and reasoning – rather than as anything to do with
personality, unconscious processes or political ideologies.
Rokeach’s approach had the advantage of extending exploration of
extremes of prejudice and hatred to other dogmas. However, it was still
essentially a personality theory, in that it located the roots of dogmatism
within the individual and their relatively stable cognitive style. For many
critics, though, this emphasis on personality did not address the main
shortcoming of the Adorno et al. study, namely the issue of the
interplay between the person and the context.

4.1 Altemeyer and right-wing authoritarianism

By the mid 1960s research into authoritarianism had begun to wane. By
that time, researchers were looking beyond the concepts of an
authoritarian personality for explanations of prejudice. Also, research by
Stanley Milgram, which you will read about in Chapter 2, showed that
tendencies to engage in acts of mass violence of the kind that Adorno
et al. sought to explain might not be attributable to personality at all.
Instead, it might be more to do with the situation in which people find
themselves.
However, the idea of authoritarianism has not been completely
superseded by other explanations and approaches. Although there were

48
 4 Revivals and revisions of the authoritarian personality

clear problems with the original study, there were those who were
reluctant to abandon the concept altogether. There were aspects of the
idea of ‘authoritarianism’ that were worth retaining.
The person most responsible for rekindling the interest in
authoritarianism in recent decades is Bob Altemeyer, who revised and
updated Adorno et al.’s original work. Like Adorno et al., Altemeyer
was interested only in authoritarianism on the right. Altemeyer (1981)
argued, however, that the definition of authoritarianism in the work of
Adorno and colleagues was too broad. As you will remember, the
original study proposed that authoritarianism consisted of a wide range
of characteristics, including things like superstition, interest in sexual
propriety, etc. Altemeyer argued that it is possible for someone to be
superstitious without being authoritarian, just as someone could be
authoritarian without being overly concerned about sexual behaviour.
Altemeyer went on to revise the original F-scale and develop his own
measure – the right-wing authoritarianism scale. This new and updated
measure addressed the methodological flaws of the original scale,
balancing positive and reversed items. In addition, Altemeyer’s definition
of authoritarianism was much simpler, consisting of just three
characteristics:
. authoritarian submission, which means deference to authority, such as
government officials and religious leaders
. authoritarian aggression, which means supporting extreme action
against anyone who deviates from what is considered ‘normal’
. conventionalism, which means a preference for established ideas and
practices and a resistance to anything that might appear new or
different.

Altemeyer found that people who scored highly on the right-wing

authoritarianism scale tended to be conservative, favour conventional
gender roles, support harsh punishment for criminals, and express
hostility to homosexuality and prejudice against people from groups
other than their own. In that sense, the similarities between right-wing
authoritarianism and the original idea about authoritarian personality are
clear. However, there is a key difference in the way the two studies
explain the causes of authoritarianism. Like Rokeach (1960), Altemeyer
rejected the psychoanalytic explanation of authoritarianism which
emphasised the role of the unconscious. Instead, Altemeyer proposed
an explanation based on the idea of ‘social learning’. You will read more
about social learning in Chapter 3. For the time being, it is sufficient to

49
Chapter 1 Exposing the authoritarian personality

know that for Altemeyer the attitudes revealed on the right-wing

authoritarianism scale were reflections of values observed and learned in
childhood from strongly disciplinarian parents. No unconscious
processes were involved. Instead, children adopted certain patterns of
behaviour from their parents, including placing a high value on
obedience to authority and displaying intolerance and aggression
towards anyone and anything unconventional. More importantly,
Altemeyer does not see right-wing authoritarianism as fixed in a
person’s personality, but more as a cluster of attitudes that are learned
in childhood. Also, he argued that how, where and when these attitudes
– that is, right-wing authoritarianism – will be demonstrated depends on
the context and situational factors that bring them to the surface.

Activity 1.3
Pause for a moment and think again about the work of Rokeach and
Altemeyer. How do their ideas differ from those of Adorno et al.? Are
there any similarities in the way in which they approached their research
question?

An important similarity is that, just as in the original study by Adorno

et al. (1950), subsequent research also took a two-pronged approach.
Both Rokeach and Altemeyer sought to measure the specific aspect of
personality underpinning authoritarianism and then to explain why some
people are more or less authoritarian or, in the case of Rokeach,
‘dogmatic’. You probably noticed also that a major difference between
the three is in the domain of the explanation. Adorno favoured the
psychoanalytic explanation, focusing on what happens in the
unconscious. Rokeach interpreted ‘dogmatism’ as a reflection of a
specific way of processing information. Altemeyer adopted a social
learning approach. What this demonstrates is that these three
researchers have been influenced by different theories, and that, as is
often the case in psychology, the same problem can be approached
from different perspectives.
However, all three studies also share common shortcomings. They are
all, ultimately, more descriptive than explanatory (Jones, 2002). They
describe what authoritarianism looks like, but have all struggled to
explain clearly and convincingly where it comes from. In fact, evidence
to support the social learning, cognitive style or psychoanalytic
explanations has been at best circumstantial.

50
 4 Revivals and revisions of the authoritarian personality

Also, none of the three explanations has been able to adequately explain
the key question that underlies much of the discussion in Part 1 of the
book. What is the relative importance of personality in determining
human behaviour, and how important are other factors, such as the
situation or a person’s social environment? While the importance of
these other factors is acknowledged in the accounts of authoritarianism
(although often indirectly), how they interact with personality is seldom
addressed in any detailed way.

Summary
. Rokeach suggested that both right- and left-wing authoritarianism
were based on a closed-minded cognitive style, rather than the
unconscious.
. In the 1980s, Altemeyer devised a ‘right-wing authoritarian scale’
and offered a revised description of the authoritarian personality that
he attributed to social learning.
. Critics argue that these studies overemphasise personality
explanations for behaviour and that more attention should be given
to the context in which behaviour occurs.

51
Chapter 1 Exposing the authoritarian personality

5 Final thoughts
The chapter has posed the question: to what extent could the extreme
Nazi violence towards European Jews in the Second World War be
explained in terms of a particular personality type? You read about the
pioneering work of Adorno et al., which sought to measure and explain
personality characteristics that predispose individuals to become part of
far-right, fascist movements. You also read about subsequent research
that moved away from personality-based explanations and also
broadened Adorno et al.’s work to cover left-wing as well as right-wing
authoritarianism.
So, can the causes of mass violence be understood in terms of personal
characteristics of the perpetrators? While the work on authoritarianism
goes some way to explaining the tendency in some people to support
totalitarian political ideologies, the authoritarian personality has been
shown to be neither a sufficient nor a necessary factor in determining
whether someone will take part in violence. Although Adorno et al.’s
study has been criticised both on methodological grounds and for its
emphasis on personality, it has nevertheless been influential in
highlighting the need for psychology to investigate phenomena such as
authoritarianism, fascism and genocide.
Although Adorno et al. were interested in explaining the root causes of
Nazi atrocities, their work has a distinct contemporary relevance. The
defeat of the Nazis in 1945 did not bring an end to fascist ideology.
Today, almost seventy years later, there are still extremist neo-Nazi
groups – the skinheads, Blood & Honour, Combat 18 and others –
who promote the ideas of racial superiority and violence towards
minority groups. At the time of writing, a number of European
countries have witnessed a rise in the popularity of right-wing political
parties who espouse extremist views and maintain links with violent
neo-Nazi groups. In the UK, the British National Party, a party whose
leaders promote racist, anti-Semitic and homophobic views, won almost
a million votes at the 2009 European elections and won two seats in
the European Parliament. Explaining fascism, or what Adorno et al.
called authoritarianism, is therefore something that remains of relevance
today.

52
 5 Final thoughts

Figure 1.9 Fascism today

The key issue, however, is whether the explanation for fascism should
be sought, as Adorno et al. suggested, in personality. In many ways
their attempt at explaining fascism was a product of the historical time
that they inhabited. Evidence of the Holocaust that emerged after the
war was so shocking that people wanted to believe that the Nazi
perpetrators were quite different from anyone else; that they were
particular kinds of people with profoundly flawed personalities. But a
personality explanation has been shown to be problematic. Personality is
understood to be quite a stable property, in which case personality
theories of prejudice struggle to account for the rapid rise in anti-
Semitism after Hitler came to power in Germany, and the reduced levels
after the war.
All of the studies referred to in this chapter tried to find ways to
measure attitudes to authority. The results were then interpreted in
psychoanalytic terms, in terms of how information is processed, or by
reference to socially learned attitudes and patterns of behaviour.
However, scoring highly on a particular scale does not mean that
individuals will go on to act in particular ways. Pettigrew showed that
individuals with the same apparent traits might behave differently
depending on the culture in which they live, so responses are about
opportunity and cultural values as well as about the individual.

53
Chapter 1 Exposing the authoritarian personality

Also, authoritarianism poses a problem in society as part of a group

phenomenon. It manifests itself in far-right movements, religious cults,
neo-Nazi groups, radical political parties, etc., many of which continue
to exercise influence in contemporary society. Authoritarian individuals
are therefore much less of a concern than authoritarian movements or
groups. What this suggests is that a perspective moving away from the
individual, towards exploring the context in which authoritarianism
appears, and how it is fostered within groups and institutions, might be
more fruitful. As Richard Schermerhorn (1970, p. 6), one of the critics
of the individualist approach, argues, ‘prejudice is a product of
situations, historical situations, economic situations, political situations
… and not a little demon that emerges in people simply because they
are depraved’.
As you have been reading about the authoritarian personality in this
chapter you may have thought that it is characterised by a collection of
extremely negative traits. Many commentators would agree with you
(Oesterreich, 2005). Indeed, the term ‘authoritarianism’ is used today
not just to describe various right-wing extremist groups such as the
neo-Nazis or the skinheads but also as a term of insult. The traits
associated with authoritarianism appear to be so negative that the
temptation is to imagine that ‘authoritarians’ are always other people,
never ‘us’. It would be comforting to suppose that ordinary people were
not like these extraordinary authoritarians, the ones who are susceptible
to obeying the extreme demands of authority figures. However,
psychological research has shown us that, in certain circumstances,
obedience to the punitive demands of authority figures is a feature of
quite ordinary people. The next chapter will take up this story.

Summary
. Fascism and right-wing extremism are still present in most Western
societies, which means that the study of authoritarianism has a
contemporary relevance.
. Any explanation of authoritarianism needs to take into account
historical, political and ideological factors and not just those located
within the individual.

54
 References

References
Adorno, T.W., Frenkel-Brunswik, E., Levinson, D.J. and Sanford, R.N. (1950)
The Authoritarian Personality, New York, NY, Harper.
Altemeyer, B. (1996) The Authoritarian Specter, Cambridge, MA, Harvard
University Press.
Altemeyer, B. (1981) Right-wing Authoritarianism, Winnipeg, University of
Manitoba Press.
Bass, B.M. (1955) ‘Authoritarianism or acquiescence’, Journal of Abnormal and
Social Psychology, vol. 51, no. 3, pp. 616–23.
Billig, M. (1978) Fascists: A Social Psychological View of the National Front, London,
Harcourt Brace Jovanovich.
Christie, R. (1991) ‘Authoritarianism and related constructs’ in Robinson, J.P.,
Shaver P.R. and Wrightsman L.S. (eds) Measures of Personality and Social
Psychological Attitudes, San Diego, CA, Academic Press.
Costa, P.T. and McCrae, R.R. (1992) NEO PI-R Professional Manual, Odessa,
FL, Psychological Assessment Resources.
Duckitt, J. (1989) ‘Authoritarianism and group identification: a new view of an
old construct.’ Political Psychology, vol. 10, no. 1, pp. 63–84.
Eysenck, H.J. (1967) The Biological Basis of Personality, Springfield, IL, Thomas.
Goldberg, L.R. (1981) ‘Language and individual differences: the search for
universals in personality lexicons’ in Wheeler, L. (ed.) Review of Personality and
Social Psychology, vol. 2, London, Sage.
Jones, M. (2002) Social Psychology of Prejudice, Upper Saddle River, NJ, Pearson
Education.
Lederer, G. (1993) ‘Authoritarianism in German adolescents’ in Stone, W.F.,
Lederer, G. and Christie, R. (eds) Strength and Weakness, New York, NY,
Springer.
Meloen, J.D. (1991) ‘The fortieth anniversary of “The Authoritarian
Personality”’, Politics and the Individual, vol. 1, no. 1, pp. 119–27.
Minard, R.D. (1952) ‘Race relations in the Pocahontas coal field’, Journal of
Social Issues, vol. 8, no. 1, pp. 29–44.
Oesterreich, D. (2005) ‘Flight into security: a new approach and measure of the
authoritarian personality’, Political Psychology, vol. 26, no. 2, pp. 275–97.
Pettigrew, T.F. (1958) ‘Personality and socio-cultural factors in intergroup
attitudes: a cross-national comparison’, Journal of Conflict Resolution, vol. 2, no. 1,
pp. 29–42.
Rokeach, M. (1960) The Open and Closed Mind, New York, NY, Basic Books.

55
Chapter 1 Exposing the authoritarian personality

Rosenthal, R. (1966) Experimenter Effects in Behavioural Research, New York, NY,

Appleton-Century-Crofts.
Ross, L. (1977) ‘The intuitive psychologist and his shortcomings: distortions in
the attribution process’ in Berkowitz, L. (ed.) Advances in Experimental Social
Psychology, vol. 10, New York, NY, Academic Press.
Schermerhorn, R.A. (1970) Comparative Ethnic Relations, New York, NY,
Random House.
Smith, M.B. (1997) ‘“The authoritarian personality”: A re-review 46 years later’,
Political Psychology, vol. 18, no. 1, pp. 159–63.

56
Chapter 2
Just following orders?
Philip Banyard
Contents
1 Introduction 61
1.1 Early influences 62
1.2 Adolf Eichmann and the banality of evil 64
2 Milgram’s obedience study 67
2.1 The set-up 67
2.2 The results 72
2.3 The variations 74
3 Milgram’s study and ethics 77
3.1 Ethics 77
3.2 The case against Milgram 79
3.3 The case for the defence 80
3.4 The judgement 82
4 Obedience research after Milgram 84
4.1 Replications 84
4.2 Recent work 86
5 The implications of Milgram’s work on obedience 89
5.1 Situation vs personality 89
5.2 Just ordinary men? 90
5.3 Conclusion 92
References 96
Chapter 2 Just following orders?

Aims and objectives

By the end of this chapter you should be able to:
. describe the research of Stanley Milgram on obedience
. outline some of the ethical challenges psychologists encounter when
devising studies on human behaviour
. discuss the relative importance of situational influences on human
behaviour
. appreciate how Milgram’s studies have informed the debate on the
phenomenon of obedience, and social behaviour more generally.

60
 1 Introduction

1 Introduction

Figure 2.1 Doing what is expected: the power of the situation

Activity 2.1
Take a closer look at the three photographs in Figure 2.1 and describe
what you see. All three photographs show people engaged in some sort
of action. The first shows soldiers in battle, who are charging having
been instructed to do so by an officer. The second image is from a
football match and shows Nottingham Forest supporters chanting and
waving their arms in a uniform fashion. The third, taken in China, shows

61
Chapter 2 Just following orders?

a group of cyclists waiting for the signal from a traffic warden before
crossing a junction. What do you think accounts for the behaviour of
these individuals? What is it about being in the military that makes
people willing to put themselves in harm’s way, or take the lives of
others, just because someone else issues an order? Do you think that
football fans behave in the way shown here when watching the game at
home on the TV? What is it about the traffic wardens that makes people
listen to their commands and obey?

In all three situations depicted in the photographs, people appear to

know what is expected of them in the situation in which they find
themselves, and are behaving accordingly. This is not in itself surprising.
Even in the most mundane situations in everyday life, people seem to
realise that situations, and the presence of others, impose different
demands on their conduct. For example, if they think that someone can
see them, most people behave differently from how they behave if they
are alone. Schoolchildren’s behaviour will change, often instantly and in
a dramatic way, as soon as the teacher leaves the classroom. There are
also places, such as airports, where people, regardless of their
personality characteristics or political views, appear to be especially
ready to accept unquestioningly the authority of uniformed personnel
and follow instructions to take off their shoes, open their bags, be
searched, etc.
In Chapter 1 you learned about the role that personality plays in
determining behaviour. What the examples in the previous paragraph
appear to suggest is that people’s actions can be accounted for by other
factors, including the situation in which they find themselves, and who
else is present. In this chapter I explore the issue of situational influence
Obedience in more detail, using as an example the work on obedience to
Complying with the authority by Stanley Milgram. There are very few studies in psychology
demands of others, that demonstrate, in such a dramatic and disturbing way, the power of
usually those in
the situation.
positions of authority.

1.1 Early influences

Stanley Milgram was born in New York City in 1933 to working-class
Jewish parents who had emigrated to the USA from Europe. His first
degree was in political science but after a crash course in psychology he
started a doctorate in psychology at Harvard University in 1954. In
1961 Milgram began work on a study that would define social
psychology for the next fifty years. The key question that Milgram

62
 1 Introduction

wanted to explore is what makes people do evil things. He devised a

unique task, an experiment, that placed people in a moral dilemma Moral dilemma
where they had to choose between doing what they were told and doing A situation in which
the ‘right thing’. The reason his research became so well known is that each possible course of
action breaches some
it held up a mirror for people to see themselves in, and the image was
otherwise binding
not as nice as they expected. moral principle.

Figure 2.2 Stanley Milgram (1933–1984)

Just like Adorno et al.’s (1950) study, Milgram’s work was inspired by
the big moral question in the middle of the twentieth century, namely
how the horrors of the Second World War could have happened and
how they could be prevented in the future. The question concerned the
behaviour of the Nazis and their allies who carried out the mass
slaughter throughout Europe. An important aspect of the crimes
committed by the Nazis was their systematic nature. Millions of victims
of Nazism perished in large concentration camps which were essentially
factories of death. How could people go about carrying out the
systematic murder of thousands of other human beings? Given that
among the victims of Nazism were over six million European Jews, this
was a personal question as well as a scientific one for Milgram, who

63
Chapter 2 Just following orders?

was very conscious of his Jewish heritage. In a letter from France to a

school friend he wrote:

My true spiritual home is Central Europe. … I should have been

born into the German-speaking Jewish community of Prague in
1922 and died in a gas chamber some twenty years later. How I
came to be born in the Bronx Hospital I’ll never quite understand.
(Quoted in Blass, 2009, p. 39)

Who were the people who carried out the atrocities against Jews in
concentration camps? Were they ‘monsters’? A widespread assumption
at the time was that those behind the Holocaust must have been
different from most of humanity, otherwise the world would be full of
evil. The work of Adorno and others, which you read about in
Chapter 1, reflected this view: the Nazis, it was assumed, had a distinct,
identifiable and measurable personality, one that was rooted in
childhood development. The implication of this assumption was that
people who carried out those atrocities, who have an authoritarian
personality, were somehow ‘different’. This was in many ways a
comforting explanation as it implied that the inclination towards such
evil acts was limited to a small majority of the population: those who
would score high on the ‘F-scale’.

1.2 Adolf Eichmann and the banality of evil

At the time of Milgram’s studies a trial was taking place in Jerusalem of
one of the Nazi leaders involved in the Holocaust. Adolf Eichmann had
been part of this project from the beginning, and in 1942 he was given
the job of transportation administrator, which put him in charge of all
the trains that carried Jews from around occupied Europe to a small
number of death camps in Poland. After the war, Eichmann fled to
Argentina where he was captured in 1960, and a year later he was put
on trial in Israel. He was found guilty and hanged.
The question that Milgram asked, and that also became the centre of
Eichmann’s trial, concerned what sort of person could carry out such
evil actions. The analysis of the trial that interested Milgram was the
writing by Hannah Arendt, a well-known philosopher and writer who
reported on the trial for The New Yorker. In her reports, Arendt focused
on how ordinary Eichmann appeared to be. She had expected the
person who carried out monstrous acts to look and behave like a

64
 1 Introduction

monster, but the reality, as observed by her, was very far from this.
Eichmann came across as a bland, simple and passionless man – not a
monster at all but an ordinary, petty bureaucrat who claimed to have
been ‘doing his job’, which in his case involved killing hundreds of
thousands of people. This observation led Arendt (1963) to refer to the
‘banality of evil’. Evil acts, she argued, do not require people to be
intrinsically ‘evil’. All that is needed is for people to be prepared to
carry out orders, and obey authority. Interestingly, this was the essence
of Eichmann’s defence: he claimed that he was just following orders.

Figure 2.3 Adolf Eichmann: an evil man or an efficient bureaucrat who was
merely ‘following orders’?

Since Arendt’s reports were published as a book, the notion of the

‘banality of evil’ has received some criticism, especially in relation to
Eichmann. It turned out that Arendt attended only the start of the trial,
and heard Eichmann and his defence team as they tried to convince the
court that he was merely following orders. If she had stayed to witness
more of the court case, she would have come to a very different
conclusion. As it turned out, Eichmann was not just following orders
but was an innovator in ways to transport people to their deaths more
efficiently. He had, at times, gone beyond his remit and continued with
the slaughter of Jews even after he was ordered by his superiors to stop.
He was also aware that other people considered his actions to be

65
Chapter 2 Just following orders?

wrong, but even when the war was over he failed to show any remorse
(Haslam and Reicher, 2008). Eichmann was evidently not an ordinary
man. Nevertheless, Milgram was intrigued by Arendt’s writing and saw a
similarity between the idea of the ‘banality of evil’ and a phenomenon
that he observed in his research in the psychology laboratory at Yale
University in Connecticut, USA. The topic of his research was
obedience to authority.

Summary
. Behaviour may be influenced by the situation a person finds
themselves in, as well as by personality.
. The ‘banality of evil’ hypothesis suggests that in certain
circumstances ordinary people can carry out extraordinary crimes,
and that this may have been the case with some Nazi officials.
. Stanley Milgram’s research explored obedience to authority.

66
 2 Milgram’s obedience study

2 Milgram’s obedience study

Milgram was one of the most innovative and productive social
psychologists of his generation, who undertook a variety of studies that
explored social psychological aspects of everyday life. However, he is
largely remembered for one dramatic piece of work – the obedience
studies. The best way to get inside this study is to imagine that you are
one of the participants taking part in Milgram’s experiment. So read on
with that in mind.

2.1 The set-up

It’s 1961 and you are arriving at the doors of the Psychology
Department of the prestigious Yale University in the USA. The reason
you are here is that you replied to an advert in the local paper asking
for volunteers to take part in a study on memory. The advert (see
Figure 2.4) offered a fee plus expenses and said that you would be paid
on arrival at the laboratory.
As you walk through the doors you are met by a serious-looking man in
a laboratory coat who turns out to be the experimenter. He introduces
you to a genial middle-aged man who is described as a fellow volunteer.
The experimenter explains that the study will involve one of the
volunteers taking on the role of a ‘teacher’ and the other taking on the
role of a ‘learner’. As part of the experiment, the ‘teacher’ will engage
the ‘learner’ in a simple memory task. The ‘learner’ and the ‘teacher’
will be in different rooms and will communicate through microphones
(see Figure 2.5). The experimenter reveals that the study is designed to
investigate the effect of punishment on learning. The ‘teacher’ will be
asked to administer an electric shock to the ‘learner’ every time the
latter makes an incorrect response on the memory task.
To select who will be the ‘teacher’ and who will be the ‘learner’, you
draw slips of paper. You pick out the ‘teacher’ slip. You then watch as
the ‘learner’ is strapped into a chair, and you hear the experimenter tell
him that ‘although the shocks can be extremely painful, they cause no
permanent tissue damage’. The experimenter now gives you a sample
shock of 45 volts to show you what the ‘learner’ will experience during
the study. The shock is unpleasant, but short of being painful.

67
Chapter 2 Just following orders?

Figure 2.4 The advert used to recruit volunteers for Milgram’s study

68
 2 Milgram’s obedience study

The experimenter then takes you into the adjacent room and sits you
down in front of an impressive-looking apparatus that will be used to
administer the shocks (see Figure 2.6).

Figure 2.5 The layout of the experiment at the Yale laboratory

The shock generator consists of a row of switches that run in 15 volt

increments from 15 volts through to 450 volts. Under the label for each
switch are some descriptive words, such as ‘slight shock’ (15 volts),
‘moderate shock’ (75 volts), ‘strong shock’ (135 volts), ‘very strong
shock’ (195 volts), ‘intense shock’ (225 volts), ‘extremely intense shock’
(315 volts), ‘danger: severe shock’ (375 volts) and finally ‘XXX’ (435
volts). Suddenly, this looks quite serious and you probably hope that
you don’t have to go very far up the scale. This is especially so given
that you received the 45-volt shock, and you know that this was
unpleasant enough. The last switch on the shock generator administers
an electric impulse ten times as strong!

69
Chapter 2 Just following orders?

Figure 2.6 The ‘shock generator’ used in Milgram’s experiment

In the first phase of the experiment, the experimenter asks you, the
‘teacher’, to read a series of word pairs to the ‘learner’ who is expected
to memorise them (for instance, ‘green-grass’ , ‘blue-sky’, ‘nice-day’). In
the second phase, the test phase, you are asked to read out the first
word of the pairs (e.g. ‘green’), followed by four possible responses
(‘grass, hat, ink, apple’). If the ‘learner’ identifies the paired word
correctly, you are to move on to the next word pair on the list. If the
answer is wrong you have to tell the ‘learner’ the correct answer,
indicate the level of punishment you are going to give them (starting
with 15 volts), and flick the appropriate switch on the shock generator.
For every subsequent incorrect answer, you are told to move one switch
up the scale of shocks.
The experiment starts. To begin with everything is fine and the ‘learner’
gets most of the answers right. You have only used the shock generator
a couple of times, and at this stage the shocks are mild. Then the
‘learner’ starts to get the answers wrong and you are moving up the
shock scale into the ‘strong shock’ range. Although you cannot see the
‘learner’ you can hear him and as the shocks increase he starts to shout
out. You have heard him grunt at the low voltage but now he is starting
to ask to be let out. At 120 volts you hear him shout out in an agitated
tone, complaining that he is in pain, and at 150 volts he asks to be
released.
Suddenly, you feel uncomfortable and you decide to stop. The
experimenter, the man in the grey coat, objects and asks you to carry
on, in spite of the ‘learner’s’ protestations.

Activity 2.2
What do you think you would do in this situation? At what point would
you stop? 200 volts? 150 volts? Would you respond to the cries of your

70
 2 Milgram’s obedience study

fellow volunteer or would you complete the job you agreed to do and
carry out the instructions of the experimenter?
How many people do you think would continue to follow the orders? At
what point do you think people would stop?

Before Milgram carried out the study, he posed the same questions as
in Activity 2.2 to different groups of people, including ordinary
members of the public, college students, psychologists and psychiatrists.
He asked them to speculate on how far they thought most people
would go if asked to administer shocks. Most ordinary people said that
participants would generally refuse to administer shock, or at least not
go very far beyond the point where the ‘learner’ experienced pain. Also,
most said that participants should rebel, and that they should not
continue beyond around 150 volts. Among the professional groups,
there was widespread agreement that nobody taking part in the study
would go all the way.
You will be relieved to know that in the actual study carried out by
Milgram, no person was hurt during the procedure, and the only actual
shock administered was the 45-volt ‘tester’ given to the ‘teacher’. In
fact, the whole situation was staged. The role of ‘experimenter’ was
played by a 31-year-old biology teacher. The ‘learner’, presented as a
‘fellow volunteer’, was in on the deception and was merely playing the
part. In reality, he was a 47-year-old accountant, who was chosen for
the role because he appeared mild-mannered and likeable. He was not
the sort of person one would want to see hurt. The drawing of slips of
paper was fixed to ensure that the ‘naive participant’ was always cast in
the role of the ‘teacher’, and the ‘shock generator’ was simply a
simulator. The sounds (the moans and cries) that the participants heard
were a recording played from the adjacent room. Importantly, however,
the deception was so good that participants believed that they were
actually administering shocks. So the study presented an ingenious way
of discovering how far people would be willing to go, just because a
psychological experiment on ‘the effects of punishment on learning’
demanded it. Most people like to think that they (and people around
them) would not go very far. But what happened when Milgram actually
placed people in that position?

71
Chapter 2 Just following orders?

2.2 The results

In the first instance Milgram conducted the study on a sample of forty
participants, all of them male. Each played the role of ‘teacher’ in the
situation described in Section 2.1. Each participant went through the
identical experimental procedure: all forty heard the same instructions,
encountered the same ‘experimenter’ and ‘learner’, heard identical (pre­
recorded) cries from the next room. The ‘experimenter’ in the grey lab
coat offered the same words of encouragement. The sessions were
filmed and notes were taken by observers looking through an
observation mirror.

Figure 2.7 Scenes from the Milgram study

Milgram found that, of the forty participants who took part in the
study, all obeyed up to 300 volts, the twentieth switch on the shock
generator. This is the point at which the ‘learner’ was heard screaming:
‘I absolutely refuse to answer any more. Get me out of here. You can’t
hold me here. Get me out. Get me out of here.’ However, only five of
the forty participants refused to continue beyond this point. Four gave
only one more shock before breaking off, with an additional five
stopping between 315 volts and 435 volts. But as many as twenty-six
continued to the end of the scale and administered the maximum 450
volts. This is despite the fact that, at 330 volts, they had already heard
intense and prolonged screaming: ‘Let me out of here. Let me out of
here. … Let me out of here. You have no right to hold me here. Let
me out! Let me out!’ Shocks beyond 330 volts were accompanied by
eerie silence. Nevertheless, twenty-six ordinary members of the public
from Connecticut administered the maximum shock and continued to
do so until the experimenter called a halt to the proceedings.
As well as counting the number of participants who went all the way on
the shock generator, Milgram also observed their reactions. Participants

72
 2 Milgram’s obedience study

who took part in the study generally displayed signs of nervousness and
tension. Many were visibly uncomfortable and probably would not have
continued had they not heard the experimenter say things like ‘Please
continue’, ‘Please carry on’, ‘It is absolutely essential that you continue’
or ‘You have no choice; you must go on’. At the end of the study,
many of the obedient participants heaved sighs of relief or shook their
heads in apparent regret. Some even had laughing fits during the
experiment, probably brought on by anxiety. Milgram (1963, p. 375)
wrote that ‘full-blown, uncontrollable seizures were observed for 3
subjects. On one occasion we observed a seizure so violently convulsive
that it was necessary to call a halt to the experiment’. (You may have
noticed that in this quote Milgram refers to people who took part in his
study as ‘subjects’. This was common practice in psychology in
the 1960s. Today the word ‘participant’ is used instead as the word
‘subject’ is considered demeaning, and lacking in respect towards
volunteers on whose participation much of psychological research
ultimately depends.)
Do Milgram’s findings seem plausible to you? Ordinary members of the
public were prepared to administer electric shocks to another person on
the mere (albeit persistent) request of a man in a laboratory coat. They
did so despite the protests from the ‘victim’ and continued even after
the supposed recipient of the shocks went quiet. Before the study, when
Milgram asked his fellow professionals to predict how many participants
would refuse to go all the way, they said that all of them would do so.
In reality only 35 per cent did. In Milgram’s study, the average voltage
at which participants stopped shocking the ‘learner’ was 368 volts.
Members of the public predicted that people would stop at around 140
volts. This is a remarkable discrepancy. It is therefore not surprising
that Milgram’s research went on to provoke considerable debate.

Box 2.1 Why do it this way?

Milgram’s obedience work is remarkable, not only because of the
important questions it sought to explore, but also because it is a
fine example of good experimental procedure in social psychology.
The most important feature of any laboratory experiment is its
controlled nature. Note that every person who took part in Milgram’s
research underwent an identical experience. All participants
received the same instructions, encountered the same individuals
(the ‘experimenter’ and the fellow ‘volunteer’) and heard identical

73
Chapter 2 Just following orders?

cries and protestations from the ‘learner’. To ensure consistency in

the experimental procedure, Milgram even recorded the anguished
cries in advance, and played them to participants from a tape.
This equivalence of experience across the forty participants was
essential if meaningful comparisons were to be made. It ensured
that any difference in behaviour observed in the study could not be
attributed, for instance, to the fact that some participants heard
louder or more desperate cries than others. For similar reasons,
Milgram used the same ‘learner’ and ‘experimenter’ with each
participant. He wanted to ensure that none of the results could be
accounted for by differences in the personality or the demeanour of
the confederates.
Another interesting aspect of Milgram’s research is that he recruited
participants from the general public, using a newspaper advert. At
the time (and still now in many psychology departments)
participants tended to be recruited mainly from among the student
population. However, Milgram was interested in exploring the level
of obedience to scientific authority among ordinary members of the
public – that is, people with no direct link to the university or
research environment – and so he recruited from the general public.
Finally, in Milgram’s original study, all forty participants were male.
Why do you think this was the case? This was not because Milgram
wanted to exclude women from his research. He later conducted
further studies in which he explored gender differences in
obedience. In the initial study, however, he decided to control for the
potential effects of gender on the findings by limiting the sample to
men.

2.3 The variations

The findings of Milgram’s original study highlighted the phenomenon of
obedience, but it could not reveal what it is about the situation that
made participants administer potentially lethal shocks to a fellow human
being. To address this question Milgram carried out further research in
which he introduced subtle variations to the original procedure. By
examining the effects of these variations on levels of obedience, he was
able to isolate specific aspects of the situation that might influence
whether participants obey or not.

74
 2 Milgram’s obedience study

By the time Milgram completed his research in 1962 he had processed

800 people through nineteen variations of the original design. For
instance, in one variation, Milgram introduced into the proceedings a
dialogue about a heart attack. He wanted to see whether alerting the
participants to the impact of the shocks on the ‘learner’s’ health might
reduce obedience. Note that all other aspects of the original study were
preserved. Interestingly, the conversation about the heart attack made
no real difference. Twenty six out of the forty participants still
continued to 450 volts, although those who stopped did so at a lower
voltage with five stopping as soon as the ‘learner’ asked to be let out.
So, the reference to the heart attack made those who disobeyed do so
earlier, but it did not prevent the more obedient participants from going
all the way.
Milgram also varied the proximity of the ‘learner’ and ‘teacher’. In one
variation he put them in the same room, while in another he required
the ‘teacher’ to hold the ‘learner’s’ arm down on a plate to receive the
electric shock. This manipulation had a clear effect. Milgram found that
the closer you place the ‘teacher’ to the ‘learner’, the fewer shocks the
‘teacher’ is likely to administer. Equally, the further you place the
‘learner’ away from the ‘teacher’, the less the impact their pleas are
likely to have.
Equally crucial was the presence of the authority figure. In one
variation, the ‘experimenter’ in the grey coat pretended to have to leave
the experiment owing to some emergency and was replaced by a person
in plain clothes, who was not a scientist. Only 20 per cent of
participants went all the way and gave the ‘learner’ 450-volt shocks.
Similar results were obtained when orders were given by phone. The
physical presence of an authority figure was therefore crucial.
In another variation Milgram placed two ‘experimenters’ in the room.
One told the participants to continue (as in the original study), while the
other told them to stop. In this variation, all the participants stopped
giving the shocks very early on. This showed that an absence of a clear
authority figure reduces obedience.
Milgram also conducted a version of the experiment in which he placed
a second ‘teacher’ in the room, although this one was a stooge
instructed to obey until the end. In this variation all the participants
went along with the confederate and shocked up to 450 volts! So the
mere presence of another obedient ‘volunteer’ made all the participants
go all the way.

75
Chapter 2 Just following orders?

One of the main conclusions of Milgram’s work was that under certain
conditions involving the presence of authority, people suspend their
capacity to make informed moral judgments and defer responsibility for
their actions to those in authority. When people are in this particular
frame of mind, the nature of the task that they are asked to perform
becomes largely irrelevant, and the main determinant of their actions is
the commands of the authority figure.

Summary
. Milgram found that most people would administer potentially lethal
levels of shock to another human being, just because they were told
to do so by an authority figure.
. The use of a controlled experimental procedure enabled Milgram to
explore different aspects of the situation that influence the extent to
which people will obey authority.
. Two key factors in obedience are the presence of a clear authority
figure, and the distance between the person administering the shock
and the ‘victim’.

76
 3 Milgram’s study and ethics

3 Milgram’s study and ethics

At the beginning of Section 2 you were asked to put yourself in the
shoes of one of the participants in Milgram’s research. How do you
think being a participant in the study felt? As you already read, many of
the participants were visibly uncomfortable during the procedure. This
is one of the reasons why the study created a storm, starting with a
hostile review of the research in a newspaper, the St. Louis Post-Dispatch.
The newspaper criticised Milgram and Yale University for putting the
participants in such a stressful situation. It claimed that Milgram
violated the rules of ethics which guide psychological research. The
charge was repeated in academic circles, and led to Milgram’s
application to join the American Psychological Association being put on
hold for a year. Milgram made a robust rebuttal of the charges and the
debate about the issues led to the introduction of new codes of good
practice for psychologists.

3.1 Ethics
Before we look at the arguments that swirled around the obedience
study we need to consider what we mean by ethics. It all starts with Ethics
morals, which are rules to guide our behaviour. These rules are based on Principles that determine
a number of socially agreed principles which are used to develop clear right and wrong
conduct. In
and logical guidelines to direct behaviour. They also contain ideas about
psychological research,
what is good and desirable in human behaviour. Ethics, in the context of ethics refers to the
psychological research, refers to a moral framework that governs what codes and principles
psychologists can and cannot do. that researchers should
adhere to.
The first generally accepted code of ethics for research on humans was
devised in 1947 as a response to the very events that provoked
Milgram’s research. During the Second World War (1939–45), under the
Nazi regime, research was carried out on human beings that led to
many deaths, deformities and long-term injuries. Revelations about this
research were as great a shock for the post-war world as the death
camps, because these acts of brutality and murder were conducted by
doctors and scientists.
After the war the victors held a series of trials, in the German city of
Nuremberg, of people who had taken part in the worst excesses of the
horrors that had swept across Europe. Among them were twenty-three
doctors involved in the brutal experiments. Sixteen of them were found
guilty, of whom seven were sentenced to death. Significantly, the

77
Chapter 2 Just following orders?

judgement included a statement about how scientists should behave

when experimenting on other humans. This is referred to as the
Nuremberg Code (see Table 2.1) and it became the basis for future
ethical codes in medicine and psychology.

Table 2.1 The Nuremberg Code (1946)

1 The voluntary consent of the human subject is absolutely

essential

2 The experiment should yield fruitful results for the good of

society, that cannot be obtained by other means

3 The experiment should be based on previous research so that

the anticipated results can justify the research

4 All unnecessary physical and mental suffering should be

avoided

5 No experiment should be conducted where there is reason to

believe that death or disabling injury may be the result

6 The degree of risk should also be less than the potential

humanitarian importance of the research

7 Adequate precautions should be in place to protect the

subjects against any possible injury

8 Experiments should only be conducted by qualified persons

9 The human subject should always be at liberty to end the

experiment

10 The scientist in charge should be prepared to terminate any

experiment if there is probable cause to believe that
continuation is likely to result in injury or death
Informed consent
The principle in
psychological research Source: adapted from Katz, 1972, pp. 305–6
whereby participants
must be given Four key principles emerged from the Nuremberg Code. First,
comprehensive
participants must be able to give informed consent to the procedure.
information concerning
the nature and purpose Second, they must retain the right to withdraw from the study whenever
of the research and they want. Third, the welfare of the participant must be protected wherever
their role in it, in order possible. The fourth principle is the most difficult to interpret because it
that they can make an concerns the costs and benefits of the study. It says that any risks to the
informed decision participants must be greatly outweighed by the possible benefits for the
about whether to
greater good.
participate.

78
 3 Milgram’s study and ethics

3.2 The case against Milgram

Activity 2.3
Before you go on to read about the criticism of Milgram’s obedience
studies, try to think through all the issues relating to ethics that are
raised by this work. In what way were the participants deceived, or
harmed? Did they have the right to withdraw? Do you think that in
Milgram’s case the ends justify the means? Do the benefits of the study
justify the costs? Do you think that the results of the study are worth the
pain and discomfort caused to the participants?

Among those who were highly critical of Milgram’s study was fellow
psychologist Diana Baumrind. She started her critique by noting the
dilemma that all research psychologists face: ‘Certain problems in
psychological research require the experimenter to balance his career
and scientific interests against the interests of his prospective subjects’
(Baumrind, 1964, p. 421).
Baumrind challenged Milgram on whether he had properly protected
the welfare of the participants. She used direct quotes from Milgram’s
original report to illustrate the lack of regard she said was shown to the
participants. In particular, she noted the detached manner in which
Milgram described the emotional turmoil experienced by the volunteers.
For example:

In a large number of cases the degree of tension [in the

participants] reached extremes that are rarely seen in
sociopsychological laboratory studies. Subjects were observed to
sweat, tremble, stutter, bite their lips, groan, and dig their
fingernails into their flesh. These were characteristic rather than
exceptional responses to the experiment.
(Milgram, 1963, p. 375)

In Baumrind’s view, and in the view of numerous others, the levels of

anxiety experienced by participants were enough to warrant halting the
experiment. What is more, just because someone volunteers to take part
in the study (i.e. gives informed consent at the start of the study), it
does not mean that the researcher no longer has responsibilities towards
them and their well-being. On the principle of cost–benefit, Baumrind
challenged the view that the scientific worth of the study balanced out

79
Chapter 2 Just following orders?

the distress caused to the participants. She acknowledged that some

harm to participants might be a necessary part of some research – for
example, when testing out new medical procedures – as in those cases
results cannot be achieved in any other way. Social psychology, however,
is not in the same game as medicine and is unlikely to produce life­
saving results. The strength of the conclusions does not, therefore,
justify harming participants. Milgram related his study to the behaviour
of people who worked in the Nazi death camps and suggested that his
study illuminated the way that ordinary people living ordinary lives are
capable of playing a part in destructive and cruel acts. Baumrind
dismissed this justification for the study and suggested there are few, if
any, parallels between the behaviour in the study and the behaviour in
the death camps.
Baumrind went on to make a further criticism by considering the effect
of this work on the public image of psychology, and suggested that it
would be damaged because the general public would judge that the
participants were not protected or respected.
A further potential problem with Milgram’s experiment concerns the
participants’ right to withdraw. Do you think that this principle, embedded
in the Nuremberg Code, was sufficiently observed in Milgram’s
research? Recall that one of the key aspects of the experimental
procedure was that whenever a participant demonstrated a reluctance to
carry on with administering the shocks, they were told by the
‘experimenter’ in the grey coat ‘you must go on’, or ‘you have no
choice; you must go on’. It might be argued that telling a participant
that they ‘have no choice’ but to continue with the experiment
contravenes the right to withdraw, which is enshrined in the ethics code.
To be fair, fourteen of the forty participants in the original study did
withdraw, in spite of being told that they had no choice, so it could be
argued that, ultimately, the participants did have a choice. It is just that
making that choice was made more difficult by the presence of the
‘experimenter’ and by his prods. After all, the study was about
obedience, and the instructions from the ‘experimenter’ were essential
to the investigation. Exercising or not exercising the right to withdraw is
what the study was about.

3.3 The case for the defence

Milgram made a series of robust defences for the study, starting with a
response to the newspaper article that first raised concerns. He

80
 3 Milgram’s study and ethics

dismissed the accusation that participants were severely traumatised by

the experience. He argued that ‘relatively few subjects experienced
greater tension than a nail-biting patron at a good Hitchcock thriller’
(quoted in Blass, 2007). This was rather disingenuous, given his other
descriptions of their reactions (see above). However, Milgram made a
more measured response to the academic arguments. He pointed out,
for instance, that he could not have known the outcome of the research
before he started. As you already read, before embarking on the study
he asked fellow professionals how they expected people to behave, and
they predicted that participants would not continue to obey and
administer severe shocks to the ‘learner’.
More importantly, Milgram was not oblivious to the psychological needs
of his participants and was aware of the potential harm caused by the
study. Immediately after the study, its true purpose was revealed to the
participants. They were interviewed and given questionnaires to check
they were all right. A friendly reconciliation was also arranged with the
‘victim’ whom they thought they had shocked. This procedure, known
as debriefing, is commonplace today, but this was not the case in Debrief
the 1960s. So, in this respect at least, Milgram was ahead of the game in A post-research
terms of ethics procedures (Blass, 2004). interview designed to
inform the participant
Milgram also conducted a follow-up survey of the participants one year of the true nature of
after the study, to ensure that there was no long-term harm the study. It may also
(Colman, 1987). The results showed that 84 per cent said they were be used to gain useful
feedback about the
‘glad to have been in the experiment’, and only 1.3 per cent said they procedures in the study.
were very sorry to have taken part. Milgram also described how the
participants had been examined by a psychiatrist who was unable to find
a single participant who showed signs of long-term harm. Morris
Braverman, a 39-year-old social worker, was one of the participants in
Milgram’s experiment who continued to give shocks until the maximum
was reached. He claimed, when interviewed a year after the experiment,
that he had learned something of personal importance as a result of
being in the experiment. His wife said, with reference to his willingness
to obey orders, ‘You can call yourself an Eichmann’ (Milgram, 1974,
p. 54).
Milgram’s basic defence was that the harm to the participants was not
as great as it might appear, and for some of them the change in their
understanding of their own behaviour and the behaviour of others was
a positive event. He makes a further defence that we have to treat all
people with respect and that this involves allowing them to make

81
Chapter 2 Just following orders?

choices even if those choices are not always for the best. In direct
response to Baumrind’s criticisms he wrote:

I started with the belief that every person who came to the
laboratory was free to accept or to reject the dictates of authority.
This view sustains a conception of human dignity insofar as it sees
in each man a capacity for choosing his own behavior.
(Milgram, 1964, p. 851)

3.4 The judgement

So what do you think should be the final judgement on the ethics of
Milgram’s study? As you can see from the debate between Milgram and
Baumrind, ethics is something that psychologists debate and often
disagree on. Ethics principles, like all rules, are subject to interpretation
and disagreement.
And yet, while individuals might have their personal view about whether
a piece of research is ethical or not, what really matters is the judgment
of institutions that regulate the profession. In the USA the regulatory
body is the American Psychological Association. Its equivalent in the
UK is the British Psychological Society. These institutions have ethics
committees which issue guidelines and codes of conduct related to
ethics in research and can reprimand researchers who can be shown to
have violated the rules. At the time of Milgram’s study, his research was
investigated by the ethics committee of the American Psychological
Association, who eventually came to the conclusion that it was ethically
acceptable. Notably, however, Milgram’s studies could not be carried out
today, as the ethics guidelines have become more restrictive since
the 1960s.
Finally, one further issue regarding Milgram’s study is worth pointing
out. Although the ethics of Milgram’s research have been questioned, it
could be argued that the obedience study, more than any other study in
psychology, demonstrated why ethics are important. Recall that what
Milgram’s study showed was that ordinary people were willing to harm
another human being just because they were told to do so by a person
they believed was a psychologist, and because doing so was supposedly
‘required by the experiment’. This shows that people generally are ready
to give scientists the benefit of the doubt and go along with what they
are doing, even when it involves harming individuals. This in itself

82
 3 Milgram’s study and ethics

illustrates how important it is to have some moderation of scientific

activity, and have limits imposed on what scientists can and cannot do.

Summary
. Psychologists have a duty of care towards participants and must
ensure that their well-being is preserved throughout a study.
. Participants must be asked to give informed consent before taking
part in research and have a right to withdraw at any point.
. Milgram’s obedience studies kick-started an ethics debate in
psychology and highlighted the need for the development of more
stringent guidelines for the conduct of research psychologists.
. Although Milgram’s obedience study was judged to be ethical at the
time of publication, it would be in violation of the strict ethics
guidelines in place today.

83
Chapter 2 Just following orders?

4 Obedience research after Milgram

4.1 Replications
Milgram’s research was carried out in Connecticut in the 1960s, on a
sample that consisted exclusively of men. Can we say, therefore, that
Milgram’s findings about obedience are limited to middle-aged American
men, or were they, as he seemed to assume, universal? To address
Replication questions such as these, psychologists use replication: they carry out
When a result from a additional research, on different populations, in different historical and
research study is found social contexts, and look at whether similar results can be obtained. For
again in a subsequent
instance, Milgram himself addressed the gender issue by carrying out a
study. Replication is
important to study identical to the original one, but with female participants. He
establishing the veracity found that women were as likely to administer electric shocks as men.
of a finding.
Further replications of Milgram’s study have been conducted in many
countries, obtaining a range of results. Peter Smith and Michael Bond
(1993) produced a review of twelve of these studies and found a fair
degree of cultural variation in obedience, ranging from 92 per cent
obedience to only 12 per cent. Smith and Bond offered two conclusions
from their review. First, substantial numbers of people from a variety of
countries will harm other people on the instructions of an authority
figure. Second, the variability in the levels of obedience can be explained
by the social context of the study and by the often subtle differences in
the wording of the orders given by the authority figure. This second
factor confirms the basic finding of Milgram’s study: people are not
blindly obedient to authority, but respond to the specific social and
physical context in which they receive the orders.
Other researchers sought to identify specific processes that encourage
obedient behaviour. Don Mixon (1972) carried out a study in which he
used a role-playing method. People taking part in the study acted out
the parts of ‘teacher’ and ‘learner’. There was obviously far less
deception here (participants knew that they were taking part in role
play), but it enabled those taking part to consider what might have gone
through a participant’s mind in the original study. One of the findings
concerned the role in obedience of the laboratory-coated ‘experimenter’.
It emerged that, in evaluating the appropriateness of their own
behaviour, participants often looked to the authority figure – the
‘experimenter’ – for clues. The calmness audible in the ‘experimenter’s’
responses and verbal prods to the ‘teacher’ suggested that there was less
of an emergency situation than appeared. The ‘experimenter’ showed no

84
 4 Obedience research after Milgram

concern as the ‘learner’ started to scream. This inaction might have

indicated to the ‘teacher’ that the screams were not unusual, or
dangerous, especially as the ‘teacher’ is aware that the ‘experimenter’
had already carried out the study with other participants. Perhaps this is
one of the features of the situation that encourages obedience in the
‘teacher’.
Another challenge to Milgram is whether the study would work outside
the laboratory. Would comparable levels of obedience be observed in
real-life situations?

Activity 2.4
It is worth stopping at this point to reflect on the situations in everyday
life where people are required to do as they are told. Think of two or
three occasions in the last week where you complied with a request, or
an order, to do, or stop doing, something. For example, perhaps you
stopped at red traffic lights or agreed to work late to meet a deadline?
Are there any occasions when you obeyed where you later wondered
about whether that was the right course of action?
Also, can you think of particular professions where obedience is relevant
and which involve people taking orders from others?

A particularly interesting study using a real-life work environment was

carried out by Charles Hofling et al. (1966). They investigated how far
nurses would go in administering a fatal dose of a medication on
demand from a doctor. A bottle of dummy pills was labelled ‘Astroten’
(a non-existent drug) and placed in the ward medicine cabinet. The
bottle had a label which said the usual daily dose was 5 mg and the
maximum daily dose was 10 mg. The nurse on duty received a phone
call from a Dr Smith asking her to give his patient 20 mg of Astroten
straight away. He explained that he was in a hurry and wanted the drug
to have taken effect before he saw the patient, and that he would sign
the drug authorisation form when he came on the ward in about ten
minutes’ time.
This request broke hospital procedures because nurses were only to give
drugs with written authorisation, were not to exceed the maximum
dose, and were only to take instructions from people they knew. The
nurses in this situation had not come across the drug before, nor ‘Dr
Smith’. Remarkably, twenty-one out of the twenty-two nurses who were
given this request followed the instruction and prepared the drug for

85
Chapter 2 Just following orders?

the patient. When they were interviewed afterwards, some commented

that doctors frequently phoned instructions and became annoyed if the
nurse protested. Half the nurses said they had not noticed what the
maximum dosage was. Either way, it appears that the word of the
authority figure was enough to override the rules governing appropriate
conduct (and common-sense notions).

4.2 Recent work

The ethics issues raised by the original study have meant that complete
replications have been impossible in university psychology departments.
Recently, however, there have been two innovations that provide an
opportunity for the research to continue. The first involves virtual
worlds.
It is now common to have a virtual character. Many people in the UK
have played on game consoles where you can create your own
Avatar alternative, virtual identity known as an avatar, who is then the
A term used to character that plays your part in computer (and even online) games.
describe an alternative Online social networks and virtual worlds offer further opportunities to
virtual identity that is
create avatars. But how ‘real’ are they to us and could they be used as
used in three­
dimensional computer part of a psychology study?
games, or an image This is the question explored by Mel Slater and colleagues (2006) when
used to represent
oneself in online
they replicated the obedience study in a virtual environment. In their
communication. study, volunteers (who took the role of ‘teacher’) were required to
interact with a female virtual character (the ‘learner’) who took part in a
memory task (see Figure 2.8). As in the obedience study, when the
‘learner’ gave a wrong answer the ‘teacher’ had to administer a shock
and with each subsequent wrong answer to turn up the voltage.

86
 4 Obedience research after Milgram

Figure 2.8 Obedience in cyberspace

In one variation of the experiment, where the ‘teacher’ communicated

with the ‘learner’ by text and could not see her, all participants
completed the procedure and administered the twenty shocks. In
another variation, the ‘teacher’ spoke to the ‘learner’ and could see her
responses. In this variation some of the ‘teachers’ would not complete
the procedure. More importantly, on physiological measures such as
heart rate, they showed significant signs of stress during the procedure,
which was not the case when the communication was by text message.
In fact, the ‘teacher’s’ response in this interactive ‘virtual’ encounter was
not all that dissimilar to what it would have been if they were dealing
with a real person. When the ‘learner’ in the experiment asked them to
speak up, they spoke louder. Some of the ‘teachers’ showed frustration
when the ‘learner’ got the answers wrong, and when she objected to the
procedure some of the ‘teachers’ turned to the experimenter and asked
for guidance.
Designing this ‘virtual’ encounter allowed the researchers to model the
behaviour of people in real situations. The volunteers were always aware
that it was a simulation but described a range of emotional responses to
the situation. They felt genuine distress even though they knew they
were not hurting anyone. This work gives an opportunity for further
research into obedience and also raises a lot of questions about the
impact on us of virtual people in virtual worlds.
A second recent innovation has been to replicate just part of Milgram’s
procedure. Jerry M. Burger (2009) argued that the 150-volt level in the
original study was the key point at which most participants decided
whether to continue to obey. It is at this point that the ‘learner’ started

87
Chapter 2 Just following orders?

to scream and asked to be released. He argued that exploring how

people respond at this point allows reasonable extrapolations to be
made about their likely behaviour beyond this point. Burger put in a
number of ethical safeguards including screening the volunteers and
carrying out instant debriefing, and obtained approval for his study.
The measure of obedience in the Burger (2009) study was whether the
‘teachers’ would be prepared to go beyond 150 volts. The study
modelled Milgram’s original experiment and used his script, but it
terminated after participants made their choice at 150 volts. Burger
found that twelve ‘teachers’ stopped at that point and twenty-eight
continued. Although this shows a little less obedience than the original
study, it is a similar outcome.
Burger (2009), just like Slater et al., found a way to further explore the
issue of obedience using a procedure similar to that devised by
Milgram, but which addresses the main concerns about ethics. Both
approaches to the study of obedience have been well received by
scholars in the field.

Summary
. Replications of Milgram’s study have found high levels of obedience.
. Concerns about ethics have made it difficult to replicate Milgram’s
original study, which has necessitated the invention of creative ways
to study obedience.

88
 5 The implications of Milgram’s work on obedience

5 The implications of Milgram’s work

on obedience
5.1 Situation vs personality
The big lesson that is commonly drawn from the obedience studies is
that people are capable of acts of great inhumanity. Remember that one
of Milgram’s participants recalled his wife telling him that there was an
‘Eichmann’ inside him. The studies are often seen as suggesting that it
is not so much who you are that defines your behaviour as where you
are and who else is present (i.e. an authority figure). As Milgram writes
(1974, p. 205), ‘the social psychology of this century reveals a major
lesson: often, it is not so much the kind of person a man is as the kind
of situation in which he finds himself that determines how he will act’.
This was a shocking revelation at the time. When Milgram conducted
his survey before the study, he found that most respondents believed
that participants were not capable of inflicting lethal pain on another
person, and they thought that people should and indeed would disobey.
The popular press at the time also tended to label people who carried
out evil acts as monsters and somehow intrinsically different from the
ordinary man or woman. The obedience studies suggested otherwise.
However, is personality really irrelevant to obedience? Alan Elms,
Milgram’s research assistant, investigated this issue in more detail. He
invited forty of Milgram’s volunteers to return to the university to find
out more about them and their personalities, and to look for differences
between those who had been fully obedient and those who had defied
authority. On standard personality measures, Elms found no differences
between the two groups (Elms and Milgram, 1966; Elms, 1972). The
only measure that suggested a difference between obedient and defiant
participants was the authoritarian personality F-scale. This is not
surprising given that, in Chapter 1, you learned that one of the key
characteristics of the authoritarian personality is deference to authority
figures. One could expect that obedient volunteers would show more
authoritarian tendencies, and this is indeed what Elms found.
And yet, the difference in personality observed by Elms cannot explain
fully the pattern of results that emerged from the obedience studies.
People who obeyed in the original experiment did not do so solely
because of their personality, although their authoritarianism was, it
appears, a contributing factor. If personality was the sole cause of

89
Chapter 2 Just following orders?

obedience, then none of the variations introduced by Milgram in

subsequent studies would have made a difference. But they did. The
greater proximity of the victim made participants less obedient, just as
the presence of a uniformed ‘experimenter’ made them more so. This
goes to show that behaviour, and specifically how we respond to
authority, is not attributable to a single cause. Instead, it is the outcome
of a complex interplay of a range of factors, including personality and
the situation.

5.2 Just ordinary men?

If we go back to the Eichmann trial that was an influence on Milgram’s
research, the question that Milgram was asking, which also became the
centre of Eichmann’s trial, concerned what sort of person could carry
out such evil actions. Eichmann’s defence was that he was just following
orders. Milgram’s conclusion (1974, p. 189) was that ‘a substantial
proportion of people do what they are told to do, irrespective of the
content of the act and without limitations of conscience, so long as they
perceive that the command comes from legitimate authority’.
A similar argument was put forward by Christopher Browning (1992) to
explain historical evidence related to the activities of Reserve Police
Battalion 101. This was a German killing unit that is estimated to have
murdered around 40,000 Polish Jews during the Second World War.
However, unlike many other SS killing units, comprising hard core Nazi
sympathisers and anti-Semites, the Reserve Battalion consisted of
‘ordinary men’, German conscripts most of whom had no allegiances to
the Nazi Party or National Socialist ideology. And yet they took part in
most brutal murders. In his book, entitled Ordinary Men, Browning
draws on Milgram’s work to explore how the specific situation in which
members of the battalion were placed, together with the expectations
placed upon them by their superiors, were enough to make mass
murderers of them. So, not only was Milgram’s research motivated by
the events of the Second World War, but his work influenced the way in
which historians view the actions of the perpetrators.

90
 5 The implications of Milgram’s work on obedience

Figure 2.9 Just following orders? Deportation of Jews from the Warsaw
Ghetto

The applicability of Milgram’s work to the Holocaust has been

challenged by David Mandel (1998), who claims that, by suggesting that
obedient behaviour is inevitable in certain situations, Milgram provided
an ‘obedience alibi’ to perpetrators. Milgram’s emphasis on situational
factors, Mandel argues, takes responsibility away from the individual and
implies that they cannot be blamed for their actions. Mandel also argues
that there is more to obedience than just obeying orders, and points to
a specific incident described by Browning. On one occasion, Major
Wilhelm Trapp, the commander of the Reserve Police Battalion 101,
received orders for the battalion to carry out a mass killing of Jews,
including women and children. Trapp was distressed by this assignment
and told his men that those who did not feel up to it could be assigned
to other duties. Only two per cent of the unit took up Trapp’s offer.
Mandel argues that, according to Milgram’s research, more men should
have taken up the offer because there was no direct command from
authority to obey, because the task (shooting women and children) was
presented by the superior officer as gruesome, and because there were
other disobedient peers present.
Mandel’s critique of Browning’s application of Milgram’s research to the
actions of German military units is important because it points to the
fact that there is more to obedience than deference to an authority
figure issuing a direct order. An important additional factor is the

91
Chapter 2 Just following orders?

broader context and the institutional arrangements in which people, in

this case German soldiers, found themselves. In Milgram’s original
experiment, the fact that participants took part in a scientific experiment
at a highly regarded institution, Yale University, invariably contributed to
obedience levels. The obedience of nurses revealed in the study by
Hofling et al. (1966) can be attributed, in part, to the institutionalised
hierarchical organisation of hospitals and the fact that following
instructions from doctors is embedded in the job description of nurses.
In the specific episode discussed by Mandel, members of Reserve Police
Battalion 101 were not responding just to the immediate authority of
their superior officer, Major Trapp; their choice was embedded in a
broader culture of the military which has its own standards of loyalty
and solidarity. These included not just obeying orders, but also not
abandoning fellow soldiers when the unit is faced with a gruesome task.

5.3 Conclusion
However Milgram’s work is interpreted, it is undeniably one of the
greatest and most influential sets of experiments in social psychology. It
continues to challenge anyone who comes across it and still attracts
academic and political interest.

Box 2.2 Milgram’s other work

The impact of the obedience studies has been so great that most
people are not aware of the other studies that Milgram carried out.
They are remarkable for their originality and their breadth. Here are
some examples, described in more detail in Milgram (1992).

Everyday obedience
Milgram explored obedience in everyday life and in one study asked
strangers to give up their seat on the New York subway (see
Figure 2.10).

92
 5 The implications of Milgram’s work on obedience

Figure 2.10 Milgram on the New York subway

Familiar strangers
Milgram noticed that a strange phenomenon of city life is that we
regularly see people whom we recognise but never talk to. He
called these people ‘familiar strangers’. They might be the person
who always gets on the same bus as us, or whom we see in the
corner shop. Milgram got his students to carry out a novel study on
these familiar strangers. They chose a suburban railway platform
and photographed commuters waiting for a train. A few weeks later
the students gave the commuters the photograph and asked
questions about the people depicted. On average, the commuters
reported seeing four familiar strangers in the photograph, but the
average number that they had ever spoken to was only 1.5. Other
questions revealed that 47 per cent of the passengers had
wondered about the familiar strangers although less than a third
reported even feeling a slight inclination to start a conversation.

93
Chapter 2 Just following orders?

Dropped letter technique

Measuring attitudes and their effect on behaviour is very difficult but
Milgram devised a delightful technique to explore this. He left un­
posted letters in public places to see how many would be posted by
the people who found them. The letters addressed to individuals or
to socially admired organisations like medical research labs were
most commonly posted (over 70 per cent returned), but most of
those addressed to the Friends of the Nazi Party or the Friends of
the Communist Party were never seen again (only 25 per cent
returned).

Six degrees of separation

Milgram tested the ‘it’s a small world’ hypothesis by asking some
people from Kansas to send packages to a stranger in
Massachusetts, several thousand miles away. The senders were
told the stranger’s name, occupation, and roughly where they lived.
They were told to send the package to someone they knew on a
first-name basis who they thought was most likely, out of all their
friends, to know the target personally. That person would do the
same, and so on, until the package was personally delivered to its
target. Most of the packages didn’t make it but the ones that did
were able to get there in about six hops. This led to the famous
phrase ‘six degrees of separation’, which suggests that any two
people in the world can be connected by an average of six
acquaintances.

Milgram’s work represents social psychology at its boldest and most

engaged with the bigger questions about human behaviour. However,
concerns about ethics, and the controversy the experiments provoked,
mean that attempting similar research is difficult. Nevertheless, as you
read, this does not mean that obedience research has been abandoned.
It is just that more imaginative ways of conducting research have to be
created.
An important implication of Milgram’s work, and obedience research
conducted since, is that it is not reasonable to hope for simple answers
to the puzzles of human behaviour, especially its more destructive
aspects. One of the perplexing experiences of studying psychology is
that the more you find out, the more you realise that there are further
discoveries to be made. Don’t despair! If you look back, even across
only a hundred years, you will see that we now have a much better

94
 5 The implications of Milgram’s work on obedience

understanding of psychology. Just don’t expect simple answers to

questions concerning human mind and behaviour.

Summary
. Research on obedience suggests that human behaviour is the
outcome of a complex interplay of a range of factors, including
personality, the situation, and the broader social or institutional
context.
. Milgram’s work was not only motivated by a desire to understand
the actions of perpetrators of the Holocaust, but has also influenced
how some historians have interpreted their conduct.

95
Chapter 2 Just following orders?

References
Adorno, T.W., Frenkel-Brunswik, E., Levinson, D.J. and Sanford, R.N. (1950)
The Authoritarian Personality. New York, NY, Harper.
Arendt, H. (1963) Eichmann in Jerusalem, London, Penguin.
Baumrind, D. (1964) ‘Some thoughts on ethics of research: after reading
Milgram’s behavioural study of obedience’, American Psychologist, vol. 19, no. 6,
pp. 421–3.
Blass, T. (2004). The Man Who Shocked the World: The Life and Legacy of Stanley
Milgram, New York, NY, Basic Books.
Blass, T. (2007) ‘Memorable quotes’ [online], The Stanley Milgram Website,
www.stanleymilgram.com/quotes.php (Accessed 11 February 2010).
Blass, T. (2009) ‘From New Haven to Santa Clara: a historical perspective on
the Milgram obedience experiments’, American Psychologist, vol. 64, no. 1,
pp. 37–45.
Browning, C. (1992) Ordinary Men, London, Penguin.
Burger, J.M. (2009) ‘Replicating Milgram: would people still obey today?’,
American Psychologist, vol. 64, no. 1, pp. 1–11.
Colman, A.M. (1987) Facts, Fallacies and Frauds in Psychology, London, Unwin
Hyman.
Elms, A.C. (1972) Social Psychology and Social Relevance, Boston, MA, Little,
Brown.
Elms, A.C. and Milgram, S. (1966) ‘Personality characteristics associated with
obedience and defiance toward authoritative command’, Journal of Experimental
Research in Personality, vol. 1, no. 4, pp. 1282–9.
Haslam, S.A. and Reicher, S. (2008) ‘Questioning the banality of evil’, The
Psychologist, vol. 21, no. 1, pp. 16–19.
Hofling, K.C., Brotzman, E., Dalrymple, S., Graves, N. and Pierce, C.M.
(1966) ‘An experimental study in the nurse-physician relationship’, Journal of
Nervous and Mental Disorders, vol. 143, no. 2, pp. 171–80.
Katz, J. (1972) Experimentation with Human Beings, New York, NY, Russell Sage
Foundation.
Mandel, D.R. (1998) ‘The obedience alibi: Milgram’s account of the Holocaust
reconsidered’, Analyse und Kritik, vol. 20, no. 1, pp. 74–94.
Milgram, S. (1963) ‘Behavioral study of obedience’, Journal of Abnormal and
Social Psychology, vol. 67, no. 4, pp. 371–8.
Milgram, S. (1964) ‘Issues in the study of obedience: a reply to Baumrind’,
American Psychologist, vol. 19, no. 11, pp. 848–52.

96
 References

Milgram, S. (1974) Obedience to Authority: An Experimental View, London,

Tavistock.
Milgram, S. (1992) (eds J. Sabini and M. Silver) The Individual in a Social World:
Essays and Experiments (2nd edn), New York, NY, McGraw-Hill.
Mixon, D. (1972) ‘Instead of deception’, Journal for the Theory of Social Behavior,
vol. 2, no. 2, pp. 145–77.
Slater, M., Antley, A., Davison, A., Swapp, D., Guger, C., Barker, C., Pistrang,
N. and Sanchez-Vives, M.V. (2006) ‘A virtual reprise of the Stanley Milgram
obedience experiments, PLoS ONE, vol. 1, no. 1 [online],
www.plosone.org/article/info:doi%2F10.1371%2Fjournal.pone.0000039
(Accessed 11 February 2010)
Smith, P.B. and Bond, M.H. (1993) Social Psychology Across Cultures, Hemel
Hempstead, Harvester Wheatsheaf.

97
Chapter 3
Learning from watching
John Oates
Contents
1 Introduction 103
1.1 Multiple media and media violence 103
1.2 What’s the mechanism? 107
2 Social learning 109
2.1 The Bandura et al. (1963) experiment 109
2.2 The design of the experiment 112
2.3 The findings 114
3 Interpreting the results of the Bandura et al. (1963)
experiment 118
3.1 The behaviour that was observed 118
3.2 Ethics 121
4 The mechanisms behind social learning 123
4.1 Priming 123
4.2 Attitudes and beliefs 124
4.3 Scripts 125
4.4 Multiple influences 127
4.5 A note of caution 128
5 A call to action? 130
5.1 The Newson report 130
5.2 The Byron report 131
5.3 Conclusion 135
References 137
Chapter 3 Learning from watching

Aims and objectives

After studying this chapter you should be able to:
. discuss how portrayals of violence in different media may affect
human behaviour
. describe a key piece of research by Albert Bandura and colleagues
into children’s imitation of violent acts
. outline why findings of associations between events and behaviour
do not provide conclusive evidence of cause-and-effect relationships
. outline how and why experiments can identify causes of behaviour
. summarise the findings of psychological research into the topic of
media violence and behaviour
. outline the policies designed to protect children from negative
effects of screen violence.

102
 1 Introduction

1 Introduction
You have read in the previous two chapters that explanations of why
people do harm to others can be sought both in aspects of personality
and in situational factors. Also, you read about some of the research
that has given insights into how these different factors operate. There is,
however, another important influence on human behaviour that was
mentioned indirectly in earlier chapters. In Chapter 1, you encountered
the suggestion that early childhood experiences have a strong influence
on personality formation. Recognising that early experience can play a
role opens the possibility that some form of learning is involved. This is
precisely what Bob Altemeyer suggested when he proposed that the You read about Bob
attitudes measured by his right-wing authoritarianism scale reflected the Altemeyer’s work in
values observed and learned in childhood. You will read more about Chapter 1, Section 4.1.
learning in Chapter 4, but here I would like to focus on the notion that
children can learn new behaviours simply by observing other people’s
behaviour. This form of learning is known as social learning.
Researchers exploring social learning found that their work was Social learning
shedding new light on the way that different types of behaviour, A theory of learning
including aggressive behaviour, are learned by children. based on observing and
imitating the behaviours
of others.
1.1 Multiple media and media violence
The advent of digital media and the proliferation of technologies that
support their delivery, such as the internet, mean that children now have
easy access to lots of information. This ranges from educational
material, through various types of entertainment, to interactive online
experiences.
The development of visual animation techniques, more recently through
CGI (computer-generated imagery), has also led to the possibility of
including sequences in films and games that show emotionally intense
behaviour. For example, extreme violence can be portrayed in very
realistic ways. Coupled with developments in computing power and the
almost universal access to computers in developed nations, the growth
of interactive gaming has opened up new possibilities for social
interaction. Children and adults can now engage in activities that mimic
those performed in the ‘real world’. They can play in imaginary worlds
through the medium of the computer screen, and also engage with
actual others, either in simulation games or in virtual worlds.

103
Chapter 3 Learning from watching

Figure 3.1 Interactive gaming: a source of danger or an educational

opportunity?

Engaging in simulation games can be of great educational benefit, as

children are frequently asked to explore predictions and reflect on what
they do and how they relate to others on screen. One example is a
bridge-building game, which is designed to support children’s
understanding of physics. The aim is to build a bridge across a ravine
and the challenge is to get the bridge structure right. Similarly, virtual
worlds can provide a range of educational opportunities. The aim of
one particular cartoon-like world, directed at 8-to-12-year-olds, is to
teach them about road and traffic safety.

Activity 3.1
A national UK survey, looking at children’s use of the internet
(Livingstone and Bober, 2005), provided information on what children use
the internet for. Below is a list of activities for which the internet can be
used. Before reading on, scan the list. Which activity do you think would
be most popular with children and which least popular?
. obtain information on things other than school work
. help with school/college work
. send and receive emails
. play games online
. send and receive instant messages

104
 1 Introduction

. download music
. look for information on careers and further education
. look for products and shop online
. read the news
. chat rooms.

Sonia Livingstone and Magdalena Bober (2005) conducted a survey of

1511 children and young people aged from 9 to 19, and, among other
things, looked at their use of the internet. You might have judged the
last activity (chat rooms) to be popular, but in Livingstone and Bober’s
study, among those respondents who reported using the internet daily
or weekly, only 21 per cent said they accessed the internet to use chat
rooms. In fact this was the least popular activity of all. Indeed, the
activities above are listed according to their popularity, and so ‘help with
school work’ was the second most popular activity – 90 per cent
reported using the internet to do work for school or college, and 94 per
cent to get information on other things.
A serious concern, however, is the dangers that participating in the
digital world may present. One particular worry is the effects on
children and young people if they are able to access media content that
depicts violence. Since the early days of cinema, scenes of violence have
been commonplace in feature films. In the 1930s, a study showed that
crime was a major theme of 25 per cent of the 1500 films that were
analysed (Dale, 1935). There was also growing concern around this time
about violence shown in comics. Behind these concerns lay a notion
that there can be some sort of direct effect of viewing violent material
on screen, an effect that either predisposes the viewer to change their
attitude towards violence, for example by becoming more tolerant of
violence, or renders them more likely to behave in violent ways
themselves.
Exposure to media violence, either in films, on television or in
computer games, is regularly cited as a possible factor causing violent
behaviour in the real world. Several examples are explored in Box 3.1.

105
Chapter 3 Learning from watching

Box 3.1 Is media violence to blame?

Example 1
Following the murder of 2-year-old James Bulger by the 10-year­
olds Jon Venables and Robert Thompson in Merseyside, England,
in February 1993, there was widespread media speculation that the
murder was linked to the viewing, by the murderers, of a violent
video, Child’s Play 3. It was being suggested that there were
similarities between deaths portrayed in the film and James’s death,
although this was not the case. The Sun newspaper ran a front­
page story, with a large image of Chucky, the murderous doll in the
film, and a claim that the two 10-year-old boys had watched the film
not long before they committed the murder of James Bulger. While
there was no evidence that the boys had in fact watched this film, it
did emerge that the father of one of the boys had rented it a week
before the murder and that he had a collection of violent videos to
which the boys probably had access. This case, and the
widespread coverage of the suggestion that watching violent films
had in some way contributed to the boys’ aggression, led to a
national debate, and several video rental companies withdrew the
film.

Example 2
In 1999, two high-school students, Eric Harris and Dylan Klebold,
walked through the corridors of the Columbine High School in
Colorado, USA, carrying knives, guns and explosives. They
attacked and killed twelve students and a teacher. One of the killers
used a gun that he named Arlene, the same name as a character in
the video game Doom, which he had been playing with his
accomplice. Both teenagers spent a great deal of time playing and
developing new levels for Doom, and there were suggestions that
they had used a plan of the school for one of these. American
newspapers also drew links with a scene from the 1995 film The
Basketball Diaries, in which a trench-coated character shot several
students in a school hallway (as in the Columbine tragedy).

Example 3
The parents of Stefan Pakeerah, who was murdered by Warren
Leblanc in Leicester, England, in 2004, blamed a violent video
game called Manhunt, which they thought Warren was obsessed
with. The game, which is advertised as offering a ‘psychological
experience’, awards the player points for killing characters on the

106
 1 Introduction

screen. BBC News reported that a spokesman for the Entertainment

and Leisure Software Publishers’ Association said:

We sympathise enormously with the family and parents of

Stefan Pakeerah. However, we reject any suggestion of
association between the tragic events and the sale of the video
game Manhunt. The game in question is classified 18 by the
British Board of Film Classification and therefore should not be
in the possession of a juvenile. Simply being in someone’s
possession does not and should not lead to the conclusion that
a game is responsible for these tragic events.
(BBC News, 2004)

Figure 3.2 Exposure to violence in films and computer games: Child’s Play
and Grand Theft Auto

1.2 What’s the mechanism?

If it is assumed, for now, that there is a tendency for people to ‘copy’
behaviour that they watch on screen, the psychological questions that
can be asked are exactly how and why such a process operates, whether
it always operates or whether it works in some situations and not in
others. This is a series of questions about the ‘mechanism’ that lies
behind the process. It’s not difficult to agree that ‘copying’ other
people’s behaviour happens; after all, this is an essential part of most
forms of training. ‘Here, let me show you how to do it’ must have been
said billions of times in human history, quite apart from all the other

107
Chapter 3 Learning from watching

copying of others’ behaviour that takes place in everyday situations. But

this sort of learning, for it is a form of learning, has proved to be
something of a puzzle for psychologists to explain. It is one thing to
observe that a process happens, but altogether something else to be
able to explain it satisfactorily.

Summary
. An important form of learning involves observing other people.
. The increased availability of different media means that there are
many more opportunities to observe other people’s behaviour, both
positive and negative.
. There is general concern that observing violent behaviour may
encourage people who watch it to behave more violently.

108
 2 Social learning

2 Social learning
Classic research in the field of social learning theory, which has since
become widely known as the ‘Bobo doll studies’, was carried out in the
early 1960s by Albert Bandura and colleagues. Bandura was very
interested in different forms of learning, especially social learning, and
he predicted that in certain conditions children were likely to imitate
aggressive acts that they had observed. To explore this issue, Bandura
set up a series of laboratory experiments which involved looking at the
effects that exposure to a violent model had on children’s behaviour. As
several experiments involved seeing a model act aggressively towards a
‘Bobo doll’, an inflated five-foot-tall toy, these became known as the
‘Bobo doll studies’.
Albert Bandura was born in 1925, in Alberta, Canada. He studied
psychology as an undergraduate at the University of British Columbia
and as a postgraduate at the University of Iowa, in the USA. He took
his first academic post at Stanford University, USA, in 1953, where he
has stayed for his subsequent working life. Although he is perhaps best
known for his classic studies of observational learning, he has also
investigated other topics linked to social influence. He has recently been
applying his work in this area to practical matters. He works with
broadcasters in many countries, scripting public broadcasts on issues
such as teenage pregnancy, the transmission of AIDS, and literacy.

2.1 The Bandura et al. (1963) experiment

The research conducted by Bandura and colleagues sought to explore
the extent to which children would imitate aggressive behaviour that
they saw performed by another person, referred to as a ‘model’. They
were also interested in investigating what factors would affect any
imitation; for example, whether the child was male or female and
whether or not the model was of the same gender as the child. In one
study, which I will discuss in more detail next, Albert Bandura,
Dorothea Ross and Sheila Ross (1963) explored whether children would
be as likely to imitate the aggressive behaviour of a model depicted in a
film as they would the aggressive behaviour of a live model. This study
is therefore directly relevant to the debate about media and violence
which is the focus of this chapter.

109
Chapter 3 Learning from watching

Figure 3.3 Albert Bandura

A total of ninety-six children (an equal numbers of boys and girls)

participated in the experiment by Bandura et al. (1963), which was
carried out at Stanford University. The children, all of whom attended
the university nursery school, were aged between 35 and 69 months
(about 3–6 years). Their average age was 4 years and 4 months.
Each child took part in the experiment individually. However, before
the experiment began, Bandura and his colleagues allocated each child
to one of four different groups. There were twenty-four children in
each group. Although all ninety-six children underwent a similar
procedure, the researchers introduced important variations in the
procedure for each group, which enabled them to observe the effect of
certain variables on children’s behaviour. In psychological research,
Condition variations in the experimental situation are called conditions. So, the
A variation in the four groups could be referred to as the four conditions.
experimental procedure.
By comparing different Let’s look in more detail at what happened to the children who were in
conditions researchers Group 1. A female experimenter took each child into a room and sat
can make inferences them at a table with multi-coloured stickers, coloured paper and potato
about the effects of prints, which they were shown how to play with. Shortly afterwards, the
one variable on
another.
experimenter led another researcher, the model, into the room, seated
them at another table with a toy set, a mallet and an inflated Bobo doll,
and left the room. The model then played with the toys for a minute,
before acting out, over a period of about 10 minutes, a series of

110
 2 Social learning

aggressive behaviours towards the doll: sitting on the doll and punching
it on the nose; hitting it on the head with the mallet; throwing the doll
up in the air and kicking it. The model repeated these actions three
times, each time saying things like ‘Sock him in the nose’, ‘Hit him
down’, ‘Throw him in the air’, ‘Kick him’, ‘Pow’.

Figure 3.4 Images from the Bandura et al. (1963) study. The aggressive
behaviour of the adult model

The child was then taken to a second room, away from the first setting,
where he or she encountered many attractive toys. The child was told
that they could play with the toys, but, as soon as they started to play,
the experimenter told them that these were ‘the best toys’ and that she
was going to save them for some other children.
The child was told that they could play with toys in a third, adjoining
room, to which they went next. In this room there was a Bobo doll, a
mallet and a peg board, two dart guns and a hanging punchball. There
was also a variety of ‘non-aggressive’ toys: a tea set, crayons and paper,
dolls, toy bears, cars and trucks, and toy farm animals. All the toys were
arranged in the same positions before each child entered. Children were
allowed to play freely in this room for 20 minutes.
Have you guessed the purpose of this third room? Bandura and
colleagues recorded the play behaviour of the children in this room, and
this allowed them to see if the child was affected by the exposure to the
modelled aggressive behaviours. What do you think was the purpose of
the second room? What happened in this room was intended to subject
the child to a mildly frustrating event (showing the children attractive
toys that they were not allowed to play with) in order to, as the
researchers put it, ‘instigate aggression’.
The above procedure was repeated for each of the twenty-four children
in Group 1.
Children in Group 2 underwent a similar experience with one important
difference. Instead of being exposed to a live ‘model’ behaving in an
aggressive way with the Bobo doll, the aggressive behaviour was shown

111
Chapter 3 Learning from watching

in a film. The model and the behaviour were the same as those seen by
Group 1, but they were shown on a TV screen.
Again, the children in Group 3 had a slightly different experience. After
seating the child in the first room, the experimenter walked to a
television in the room, said ‘I guess I’ll turn on the colour TV’, and
pretended to tune the TV to a cartoon channel. What was then shown
was a film that started with music and some stage curtains being drawn
and the title Herman the Cat appearing on the screen. The film contained
footage of the same set of aggressive actions with the Bobo doll as in
Groups 1 and 2, but performed by a model dressed as a black cat. The
model moved in a ‘cat-like’ way, in a set with artificial grass, and a
‘fantasy-land’ background with brightly coloured trees, butterflies and
birds. The same aggressive words were spoken, but in a high-pitched,
You were introduced to
the controlled nature of
cartoon-like voice.
an experiment in The final group of twenty-four children (Group 4) were also introduced
Chapter 2, and you can
to the different rooms, but they were not exposed to any model
see this approach being
taken here too. behaving in an aggressive way with the Bobo doll.
To summarise:
Independent variable . Group 1 observed a live model behaving aggressively towards the
A variable that is Bobo doll.
manipulated by the . Group 2 observed a film of the live model behaving aggressively
experimenter to see
what effect this has on
towards the Bobo doll.
another variable. . Group 3 observed a film of a ‘fantasy’ model behaving aggressively
towards the Bobo doll.
. Group 4 did not observe any aggressive behaviour towards the
Experimental
condition
Bobo doll.
A condition in an
experiment where
participants are exposed
2.2 The design of the experiment
to a specific variation
This was a carefully planned experiment. Each of the four groups of
in the independent
variable. children experienced almost exactly the same set of events, in the same
rooms with the same people and toys. There was one key element that
varied across the groups: Bandura and colleagues varied the exposure to
Control condition violence. In an experiment, when something is purposely manipulated in
The ‘baseline’ this way, it is called an independent variable. Recall that the four
condition, against variations of the experimental situation are called conditions. The three
which experimental
conditions can be
conditions involving exposure to aggressive behaviour are defined as
compared. experimental conditions and the condition involving no exposure to
aggressive behaviour is defined as the control condition.

112
 2 Social learning

Why was there a control condition? This is needed as a comparison, to

see the extent to which children might behave aggressively towards the
Bobo doll when they had not seen a model acting aggressively towards
it. It is possible that providing children with a mallet means that they
might use this to hit something even when they’ve not seen a model use
it in this way. This is sometimes called the ‘baseline’ condition.
As well as manipulating exposure to violence, Bandura et al. made sure
that:
. There were equal numbers of boys and girls in the study as a whole,
and in each of the four groups.
. There was the same number of children in each group.
. The children across the groups were matched in terms of prior
ratings of aggressive behaviour (i.e. how aggressive they were more
generally).
. Half of the children saw a male model, the other half a female one.
This meant that researchers used two different confederates, a man
and a woman, who alternated throughout the study.

Activity 3.2
Take a couple of minutes to consider why it was necessary to take all
these issues into consideration.

Making good decisions on features such as these can make all the
difference between a poor research study and one that contributes
significantly to new knowledge. Taking the above points in turn, it is
important to note first that it may be that boys and girls differ in some
way in how they respond to modelled behaviour. If there were only a
few girls and lots of boys in the sample, it would not be possible to tell
with any confidence whether there were real gender differences. Also, if
boys and girls were not equally distributed across the conditions, it
would be difficult to disentangle the effects of gender from the effects
of the experimental manipulation. Similarly, if there were different
numbers of children in each group, and a difference was found for one
small group, that would make the finding difficult to interpret.
Examining prior levels of aggressive behaviour of the children is also
important. Prior to the study, the children in each group were matched
in respect of their aggressive behaviour with the children in the other
groups. This was done on the basis of ratings that had previously been

113
Chapter 3 Learning from watching

made, in their nursery school, of the levels of aggressive behaviour that

they displayed, and also of how much they tended to inhibit their own
aggressive impulses. This was another way in which the researchers tried
to ensure that the groups differed only in respect of their manipulation
of the independent variable. By matching the groups with regard to the
participants’ usual level of aggressive behaviour, the experimenters were
controlling for it so as to reduce or eliminate its possible effects.
Otherwise, it might confound the results. Variables that might influence
the result, if they are not controlled for, are called confounding
Confounding variable variables. It’s very important for researchers to recognise what
A variable that is not confounding variables there might be in a particular experiment and to
controlled by the control or eliminate them.
researcher but that can
affect the results. Finally, as well as manipulating the exposure to aggressive behaviour,
Bandura and colleagues purposely varied the gender of the ‘model’. In
the experimental conditions, the part of the ‘model’ was played by either
a male or a female researcher. The presentations were organised so that
half the children in each condition experienced the modelling by the
researcher of the same gender, and half experienced the modelling by
the opposite-gender researcher. This enabled the authors of the study to
explore whether children will be differently affected by a male model
compared with a female model.
An important part of designing a research study is thinking about
aspects like these. It is also important to bear these in mind when
critically examining research findings. Whether or not researchers have
taken this sort of care with designing a study can influence the validity
of their findings and conclusions.

2.3 The findings

Now that the design of the experiment has been explained in full, I
shall look at how the researchers obtained their data and at what they
found. Remember that they observed each child’s play in the third room
for 20 minutes. They did this through a one-way mirror and the
behaviour was video-recorded to permit thorough and careful analysis.
The 20 minutes were broken down into 5-second units, and each unit
Code/coding
A way of describing
was coded on the basis of the kind of behaviour the child was
observed behaviour displaying. In total, for each child there were 240 codings (20 minutes
using a set of equals 1200 seconds, and dividing 1200 by 5 gives you 240). These
predetermined codings were intended to accurately capture the amounts of aggressive
categories. behaviour shown by the children. The amount of aggressive behaviour

114
 2 Social learning

observed for each child was the dependent variable. It is referred to Dependent variable
as a dependent variable because its level is expected to depend on the A variable that is
experimental manipulation of the independent variable. expected to change as a
result of the
In research such as that conducted by Bandura et al. (1963), it is manipulation of the
necessary to define the behaviour being observed, before coding can independent variable.
start. Aggressive behaviour can take many forms, so the researchers
identified four different types, and focused on these:
. imitative aggression: acts that were clear and complete repetitions of
behaviour that had been performed by the model, such as punching
or kicking the doll, and accompanied by the same or very similar
utterances as in the first trials; for example, ‘Kick him’, ‘Pow’.
. partial imitative responses: aggressive behaviour involving the use of the
mallet on a toy other than the doll, or sitting on the doll but in a
non-aggressive way.
. nonimitative aggression: any aggressive acts not observed in the model’s
behaviour.
. aggressive gun play: shooting or aiming the toy gun.

The number of times these behaviours were observed for each child
was noted, and these were also added together to give a total aggression
score.
You might be wondering how easy and reliable it is to code behaviour
in this way. As a check, it is usual to have two people code some of the
recording, and then to calculate the degree of consistency between
them. This is what Bandura et al. (1963) did: 40 per cent of the trials
were coded independently by two observers to examine how reliable the
coding was. In more than nine out of ten observations the observers
agreed, so this was taken to indicate that the coding was sufficiently
reliable.

Activity 3.3
The researchers wanted to explore whether the exposure to a violent
model would increase the subsequent display of aggressive behaviour by
the children. Before describing their findings, I would like you to try to
predict what their observations revealed:
. Do you think that exposure to the aggressive model increased the
amount of aggressive behaviour displayed by the children, compared
with the control condition?

115
Chapter 3 Learning from watching

. Do you think there was a difference between the boys and the girls in
the amount of aggressive behaviour they exhibited?
. Do you think that the male models had more influence on the children
than the female ones?

The codings revealed that the children who participated in the three
experimental conditions (i.e. Groups 1–3 who were exposed to an
aggressive model) scored more highly in respect of the amount of
aggressive behaviour than children in the control condition
(i.e. Group 4). So, exposure to a violent model did increase the amount
of aggressive behaviour exhibited by the children. However, it made no
difference whether the model was seen live or on film. Also, in the
conditions where children saw the model on film it made no difference
whether the model was human or a fantasy, cat-like creature.

Figure 3.5. Images from the Bandura et al. (1963) study. Middle and bottom rows contain images of
child participants replicating the behaviour of the adult ‘model’ shown in the top row

Also, it was found that in all conditions, including the control condition,
the boys showed higher aggression scores than the girls. There were
also gender differences in respect of the different types of aggressive
behaviour (imitative aggression, partial imitative responses, nonimitative
aggression and aggressive gun play). For example, boys displayed more

116
 2 Social learning

gun play than girls in all conditions, while girls sat on the Bobo doll
much more than the boys in all conditions, but were much less likely to
punch it.
The researchers also found that the gender of the model made a
difference to the levels of aggression observed, although more so
among boys. The levels of aggression also varied depending on the
condition. For example, in the live model condition (but not in the film
model condition), a higher level of aggression was observed among
boys who saw the male model compared to those who saw the female
model. Among girls, the effect of the gender of the model was much
less pronounced.
Crucially, Bandura et al. were able to conclude that their findings
supported the main prediction of their study, namely that both the
live and the filmed displays of violence would increase the number of
aggressive acts subsequently displayed by children.

Summary
. Albert Bandura and colleagues have investigated children’s reactions
to portrayals of aggressive behaviour.
. In an experiment by Bandura et al. (1963), children observed
aggressive behaviour displayed by both live and filmed models.
. Exposure to the displays of aggression by both types of model led
to greater subsequent aggressive behaviour.

117
Chapter 3 Learning from watching

3 Interpreting the results of the

Bandura et al. (1963) experiment
From their results, the authors of this study concluded the following:
. Children’s behaviour can indeed be influenced by seeing violent acts
performed, and this influence manifests itself as increased levels of
violent behaviour shortly after exposure.
. This influence differs depending on various factors, such as the
gender of the child and of the person modelling the violent acts.

However, the question then arises of whether these findings are

generally applicable to all situations where aggression is observed, and
to all children. Another way of phrasing this is: to what extent do the
Generalise findings generalise? For example, this study was carried out with
To extend the findings children in quite a limited age range and in a specific situation. It would
of a single study to be of interest to know if there are differences, in the amount of
explain behaviour in
aggressive behaviour displayed, between children of different ages and
other situations or
settings. in different contexts.
This section discusses some of the questions that can be asked about
the generalisability of the findings of this study, and about the ethics of
this type of research.

3.1 The behaviour that was observed

The behaviours that were modelled, and those that the children then
later displayed in the third room, were all described as ‘violent’ or
‘aggressive’ acts. But were they really of the same nature or
consequence as actual violent, aggressive acts where these are directed
at other people? Perhaps children respond in one way to high-intensity,
acted behaviour towards an inflatable doll, but in quite different ways if
the violence is real and directed towards actual people. If children do
behave in aggressive ways with inanimate objects in play-like situations,
then does this mean they will also behave in similar ways with people?
There is little evidence that this is so. It is possible that playing out
violence with inanimate objects may actually reduce violence towards
people.
A second point to note about the observation of the children’s
behaviour is that it took place only a short while after exposure to the
aggressive behaviour of the model. Given that there was only a brief

118
 3 Interpreting the results of the Bandura et al. (1963) experiment

intervening period between the demonstration of violent behaviour and

the observation phase, it is valid to ask the question: how long would
the observed effects have lasted? The study by Bandura et al. (1963) did
not address this question. It may well be that the increases in violent
behaviour that were seen disappeared very quickly afterwards. They may
even have been confined simply to the research sessions; perhaps the
children thought that it was acceptable to behave in a violent way in this
particular setting and would not behave in this way in other settings.
Some subsequent research has suggested, however, that brief exposures
to violent acts can have lasting consequences in some circumstances
(Bandura, 1973).
It is also important to note that there was variation in the aggressive
behaviour among the children within the different conditions, from one
boy to another and from one girl to another. This is not so obvious
when averages are compared for groups assigned to different
conditions, but it does tell us that there are other factors at work,
probably involving differences between children, either in their
susceptibility to the influence, or in their willingness to show aggressive
behaviour. This study was important in part because it led to other
researchers asking these sorts of question. Bandura himself continued to
explore the effects that he had demonstrated in this study, focusing on
the differences in the characteristics of the models in particular. In later
studies he showed that there were two important variables that affected
the extent to which children would be influenced by the model. The
apparent status of the model had an effect, as well as whether or not
the model was seen to be rewarded for their aggressive behaviour
(Bandura, 1973).
Finally, one can question whether the rather unusual, strange setting of
the university laboratory, with adults doing strange and unfamiliar
things, with the female researcher apparently being unmoved by the
somewhat bizarre behaviour exhibited by the ‘model’, might have led to
some rather situation-specific reactions from the children. In the journal
article reporting this research, the authors say that ‘In order to minimize
any influence her presence might have on the subject’s behaviour, the
experimenter remained as inconspicuous as possible by busying herself
with paper work at a desk in the far corner of the room and avoiding
any interaction with the child’ (Bandura et al., 1963, p. 5). (As explained
in Chapter 2, ‘participants’ used to be referred to as ‘subjects’.)
However, this could have been seen by the children as suggesting that
the observed acts were acceptable in this situation and with this adult.

119
Chapter 3 Learning from watching

They might have taken this to mean that showing this sort of behaviour
was also acceptable and maybe even expected. It has even been argued
that the children may have sensed that being aggressive was what was
expected of them, and behaved accordingly (Borden, 1975).

Box 3.2 Why do it this way?

An important aspect of Bandura et al.’s (1963) study is that it was
one of the first experimental studies on media effects on behaviour.
There is a very dangerous fallacy in everyday debates about media
effects, and in how such effects are presented by the media. This is
that correlation is often taken to indicate cause. The term
Correlation correlation refers to an association between two events, meaning
An association between that they tend to occur together more often than one might expect
two events, meaning by chance; for example, if individuals who commit violent acts are
that they tend to occur also found to have watched a violent video shortly beforehand.
together more often There are indeed many studies, in different countries, that have
than one might expect shown correlations between the amount of exposure to screen
by chance.
violence and aggressive behaviour, and some of these studies have
been longitudinal, following children over several years
(e.g. Gunter, 2008). Given such an association, it would be easy,
Longitudinal studies
Studies that monitor but not necessarily correct, to jump to the conclusion that it was
and chart the viewing the violent video that caused an individual to become
development of violent (see Box 3.1).
psychological variables There are two important reasons why simply observing a correlation
over long periods of
between one thing and another is not adequate to show a cause–
time.
effect relationship. The first is that a correlation does not tell us
what the direction of effect might be. In other words, if A and B
Direction of effect usually occur together, does A cause B or B cause A? Does
Describes which of two viewing violent media content necessarily make aggressive
variables is the initial behaviour more likely? Could it be, instead, that children who are
cause and which is the already more aggressive (for whatever reason) are also more likely
resulting effect. to seek out and view violent media content? Second, a correlation
does not tell us whether or not the observed relationship could in
fact be the result of a third factor that affects both of the things that
are being looked at. For example, some children might be more
prone to violence than others because they are growing up in a
violent household, in which violent videos are readily available. In
this instance, the home environment could account both for the
increased aggression and for the exposure to violent media content.
Conducting experiments is a way of overcoming the difficulties in
interpreting observed correlations. The significance of an

120
 3 Interpreting the results of the Bandura et al. (1963) experiment

experimental study, if it is carefully designed, is that it can establish

causality. Differences between the conditions are very likely to have
been caused by the manipulation of the independent variable. This
is because of the care taken to make sure that participants in the
different conditions are as similar to each other as possible, either
by matching for important other factors, such as prior levels of
aggressive behaviour in the Bandura et al. (1963) experiment, or by
randomly assigning participants to the different conditions. The
other crucial characteristic of experimental studies is that the
experimental conditions are very carefully controlled to be the same
for each participant, and there is a control group. These essential
elements of experimental research seek to isolate the influence of
the variable of interest from the myriad of other variables that can
affect behaviour.
The fact that an experiment can help establish causality does not
mean that observing behaviour in the real world is unimportant, or
unscientific. Indeed, real-world observations are crucial for
developing explanations, but, unlike experiments, they cannot
isolate the cause–effect links underlying what is observed.

3.2 Ethics
Although ethics issues do not affect the generalisability of the findings,
it is nevertheless important to consider them when evaluating this study.

Activity 3.4
Think about the ethics of the Bandura et al. (1963) study in relation to
the following three principles of research ethics:
. participants must be able to give informed consent
. participants must retain the right to withdraw
. the welfare of the participants must be protected.

For each one, make notes on how well you feel that these principles
were followed by Bandura and colleagues in designing the experiment.

Bandura et al. (1963) do not mention any ethics review of their

research, while it is now a requirement that such a review would be
undertaken by a body such as an institutional research ethics committee.
This is especially so when research is conducted on children, as

121
Chapter 3 Learning from watching

additional issues have to be addressed. One of them concerns informed

consent. When conducting research on children, it is not enough to
obtain consent just from the child. The child’s legal guardian (parent or
caregiver) must also agree. In the case of the Bandura et al. (1963)
study, the children attended the university nursery school, so it is likely
that the relevant consent was obtained.
Perhaps the most significant ethics issue, however, concerns the effects
of participation in the study on the children. They were exposed to
events that they were very unlikely to experience in their everyday lives,
and it is possible that there were effects other than simply the increase
in aggressive acts towards the doll in the test phase. Also, it is easy to
find video material of the original studies on the internet, so the
identities of the children were not properly protected.
Finally, in the description of their study, the authors state that ‘It was
necessary for the experimenter to remain in the room during the
experimental session; otherwise, a number of the children would either
refuse to remain alone or would leave before the termination of the
session’ (Bandura et al., 1963, p. 5). This raises the issue of how well
the children’s right to withdraw was protected. Were the children freely
consenting to their participation? Were some children coerced?

Summary
. The extent to which the findings of Bandura et al. (1963) are
generalisable is questionable.
. Simply noticing that two elements are associated, for example
viewing video games and behaving violently, is not good evidence
that one causes the other.
. Experiments allow psychologists to examine cause–effect
relationships.
. Where potentially harmful or distressing topics such as media
violence are researched, ethics need to be especially carefully
considered.

122
 4 The mechanisms behind social learning

4 The mechanisms behind social

learning
The Bandura et al. (1963) experiment stands as a ‘classic’ demonstration
that children have a propensity to copy aggressive acts that they observe
in visual media, either as filmed actuality or in a cartoon-like form.
However, as I suggested in Section 3, the study also raised many
important questions. This is often seen as the hallmark of a good piece
of psychological research, indeed of scientific research more generally.
Bandura et al. offered a clear answer to a specific question, but they
also revealed new subtleties in human behaviour that were previously
not identified, and suggested new questions to be addressed by future
research.
Since 1963, there have been many studies and reviews of research into
the effects of media violence on children and young people (e.g. Calvert
and Wilson, 2008), and many psychologists have developed explanations
about how and why such effects come about. Bandura’s work, including
his work subsequent to the 1963 study, pointed clearly to a process in
which a model is more likely to be copied by children if the model is
seen as being of high status, gaining rewards for their aggression and
not receiving punishment. It also seems fairly well established that
children are more likely to copy if the model is more like themselves
(Bandura, 1973). Although these findings have not always been
replicated in other people’s research, in general other researchers have
confirmed Bandura’s findings (Ormrod, 1999). This gives us a starting
Mechanisms
point for thinking about how the process works, although it only tells The process or set of
us that the process is something to do with whether the individual processes that underpin
identifies with the violent model. To say that someone identifies with a particular
some other person is to say that the other person assumes a particular psychological
significance, and that they are seen as an important ‘reference point’ – phenomenon, such as
for example, as someone to look up to and to emulate. But it is social learning.
important to explore further how a media portrayal operates within the
person to raise the probability of subsequent copying. In psychology, Priming
these processes operating within the person are often called Where exposure to a
mechanisms. particular behaviour
renders more likely the
display of that
4.1 Priming behaviour in
subsequent similar
One way in which it has been suggested that social learning operates is situations.
through priming. Viewing an aggressive model creates memories of

123
Chapter 3 Learning from watching

this behaviour, and associated emotions and responses. This then means
that, when a similar setting or situation arises, a person is more likely to
respond in accordance with the stored memory (Berkowitz, 1993). In
other words, for a period after exposure these stored feelings and
behaviours are more likely to come to mind than other sorts of feelings
and behaviours.
The idea of priming suggests two things. First, the effect may be
relatively short-lived, as later experiences are likely to prime other,
different emotions and responses. Second, the expression of violence in
the everyday world depends on cues – that is to say, stimuli in the
environment – to trigger the primed behaviour. To have an effect, these
cues need to have some similarity to elements of the media depiction. It
is the association between the two that triggers the imitative behaviour.
For example, imagine a movie scene showing a person purposely
knocking a drink over someone, and that other person then punching
the first person. In real life, an accidental spillage of drink over
someone who had previously watched such a scene might prime that
person to at least think about responding aggressively (and possibly
punching the person who spilled the drink over them).
The idea of priming has been used primarily to explain the findings of
experimental studies that focus on short-term effects of exposure to
violence. However, some of Bandura’s and others’ later work showed
that children can still reproduce specific violent behaviours up to at
least eight months after they are shown them (Bandura, 1973). The
relevance of priming to these longer term effects of exposure is less
clear.

4.2 Attitudes and beliefs

Cognitive
To do with mental A mechanism by which longer-term effects might be explained is that
processes, such as exposure to violent content changes a person’s attitudes and beliefs.
perception, memory This is an alternative cognitive mechanism that involves thinking
and thinking. processes. Here, a person’s attitudes and beliefs towards specific
aggressive behaviours, that is to say whether they view these behaviours
in a positive or negative light, makes them more or less likely to
Desensitise
A process whereby
produce such behaviours themselves.
repeated exposure to a One way in which this might happen is by repeated exposure to
stimulus weakens an
portrayals of violence. Although initially viewed negatively, the repeated
initially strong
response. exposure may desensitise a person and lead them to develop less
negative attitudes towards aggressive behaviours. This in turn could

124
 4 The mechanisms behind social learning

make them less sensitive to viewing aggression and less inhibited about
acting in an aggressive way. In addition, exposure to depictions of other
people showing more positive attitudes towards violence might
encourage a person to shift their attitudes, through a process of social
influence.
There is some evidence for this mechanism, which has come mainly
from studies of the effects of repeated viewing of sexual violence.
These studies (e.g. Malamuth & Briere, 1986) have found changes in
attitudes and beliefs that are highly persistent, although, as with the
Bandura study, differences have been found between men and women
in how they respond. For example, in one study (Weisz and
Earls, 1995), men were found to become more accepting of
interpersonal violence and more attracted to sexual aggression. These
changes have also been found to depend very much on how the
violence is portrayed, whether there are positive or negative outcomes
for the perpetrator, and how the victim is shown to be affected by the
violence.

4.3 Scripts
Since Bandura and colleagues conducted the Bobo doll studies, there
has been increasing interest in psychology in the concept of script Script
learning as a way in which children and adults acquire complex An organised sequence
sequences of behaviour. Scripts are rich packages of sequences of of behaviours that is
associated with a
behaviours, intentions, meanings and emotions, which offer scenarios to
particular setting. A
deal with situations that arise in everyday life. This explanation has ‘restaurant script’, for
drawn on observations of children in pretend play, when they often example, would include
rehearse quite long and involved scenes, such as ‘going to the doctor’, a sequence of
where they are clearly building and rehearsing elaborate, routinised, behaviours appropriate
stories of everyday life. to eating in a
restaurant.

125
Chapter 3 Learning from watching

Figure 3.6 ‘Going to the doctor’: the role of scripts in pretend play

Clearly, having scripts available for everyday activities is an efficient

approach, as one does not have to invent anew how to behave in every
situation. How this plays out for the effects of violence in the media,
according to script theory, is that children who regularly and frequently
watch scenes of violence learn these sequences of behaviour as scripts.
They come to see aggression as a means of solving problems, and
hence these scripts become more likely to be used in real situations
(Huesmann, 1998).

126
 4 The mechanisms behind social learning

4.4 Multiple influences

One of the difficulties in the field of research into the effects of media
violence is that there are some studies that seem to show no influence
of watching violence on behaviour, while others show a large influence.
Also, it is difficult for a single study to control adequately the many
factors that affect how children and adults respond to screen violence.
So, is it possible to bring together, in some way, the data from the
different studies in this area and come up with an overall conclusion?
One way in which this can be done is by conducting a meta-analysis. Meta-analysis
A meta-analysis involves combining the data from different studies and A technique for
analysing it together, using complex statistical procedures. combining data from
different studies on the
There have been several such analyses in the area of research that you same topic, and
are reading about in this chapter. Some have focused on the effects of analysing it together, to
TV, some on those of film and others on those of computer games, but derive an overall
conclusion.
the largest and most comprehensive was carried out by Brad Bushman
and Rowell Huesmann (2006). Included in the meta-analysis was data
from studies of different media, including TV, film, computer games,
comics and music. Overall, the results confirmed those of previous
studies that looked at a smaller range of research and media types:
exposure to media violence was found to be associated with subsequent
violent or aggressive behaviour. However, this association was shown
not to be particularly strong. Exposure to violence on screen, while of
concern, is not the main cause of violent behaviour. It interacts with
other factors such as personality, aspects of the environment, and so
forth.
Anderson and Bushman (2002) attempted to draw together a diverse
range of factors to provide a complex explanation of how repeated
exposure to media violence might relate to aggressive behaviour. In
addition to the mechanisms outlined in this section, such as those
involving the influence of attitudes and beliefs, and of scripts, Anderson
and Bushman also highlighted other factors. These included the role
that the situation may play and the potential influence of the peer
group, as well as other variables including personality. So, various
processes might come together and affect the thoughts of the
individual, their mood, and the behaviour they display, which, together
with them selectively choosing environments in which aggression is
more likely to be experienced and evoked, would then lead to more
aggressive behaviour.

127
Chapter 3 Learning from watching

Figure 3.7 Violence between youths: the product of multiple influences

4.5 A note of caution

The meta-analysis by Bushman and Huesmann (2006) involved a range
of different types of study, including experiments, surveys and
longitudinal studies that followed people over varying lengths of time.
An important point is that some of these looked at correlations between
media use and aggressive behaviour. As I indicated in Box 3.2, finding
associations, even if they are supported across more than one study, is
not the same as clearly identifying that one factor causes another. In the
case of screen violence, there is no unequivocal proof that serious acts
of violence are directly caused by exposure to screen violence.
An example of a study that shows how cautious one must be when
attributing violent behaviour to media influence is a study by Ann
Hagell and Tim Newburn (1994). This study compared a group of
seventy-eight teenagers who had committed violent offences with a
group of teenagers who did not have a criminal record. The researchers
interviewed these young people about their media habits and, among
other things, asked them about their exposure to attitudes towards
violent content. They found that the offenders did not seem to have
any greater interest in screen violence. In fact, they appeared to have
less access to such material and as a result viewed less of it than the
comparison group. The offenders were less able to recall specific
characters and violent scenarios, and seemed generally to be less

128
 4 The mechanisms behind social learning

identified with this sort of material. In other words, it did not seem as
if media violence had much personal relevance to them. Evidently, it
was other factors that contributed to their violent behaviour.

Summary
. There are several processes through which media portrayals may
affect behaviour.
. Priming, desensitisation, attitude and belief change, and script
learning have all been considered as important processes through
which exposure to violent media may affect behaviour.
. Meta-analyses of a wide range of research findings have concluded
that observing media portrayals of violence has a weak association
with subsequent aggressive behaviour.

129
Chapter 3 Learning from watching

5 A call to action?
The question that all this research raises, even given the reservations
that I have expressed, is: what could be done to reduce the potential
impact on society of violent media? The authors of the major meta­
analysis described in Section 4.4 commented on their results as follows:

Although observing violence may increase aggression in the short

term for adults and children, long-term effects are most likely to
occur for children. Consequently, children need the most
protection from repeated exposures to violence. Infrequent
exposure is not likely to produce lasting consequences, but parents
particularly need to be urged to protect their children against the
kinds of repeated exposures that heavy play with violent video
games or immersion in violent TV programs is likely to produce.
(Bushman and Huesmann, 2006, p. 351)

This final section highlights that, as well furthering our understanding

of human behaviour, psychological research has also made a
contribution to government policies related to the media, and helped
formulate guidance for parents. I shall look at two reports that have
influenced policy on these matters. In fact, psychologists often find
themselves giving evidence, preparing reports and sitting on working
parties to develop advice based on reviews of research. The British
Psychological Society has a Parliamentary Officer whose role is to keep
up to date with government business and to help to ensure that the
Society provides experts who can contribute to consultations and
generally bring psychological knowledge to bear on social issues.

5.1 The Newson report

Following the murder of 2-year-old James Bulger in 1993, the Member
of Parliament David Alton commissioned Elizabeth Newson, a
developmental psychologist at Nottingham University, to produce a
report on the issue of negative media effects on children. The report,
‘Video violence and the protection of children’ (Newson, 1994), was
endorsed by a range of other professionals working in the area of
children’s health, and argued that there is evidence for a strong
connection between watching violence on screen and carrying out
violent acts.

130
 5 A call to action?

The report also acknowledged the difficulty faced by psychologists

seeking to establish a cause-and-effect relationship, highlighting that an
experiment seeking to look at the effects of images of extreme violence
on children would not be approved by an ethics committee. According
to Newson, exposure to such images would traumatise children, and
repeated exposure would lead to desensitisation, which means that the
children would be adversely affected by taking part in research.
This report was influential in that it encouraged the notion of
psychological harm to be included in government debates about video
censorship. It provided support to those seeking to amend the Criminal
Justice Bill to tighten video censorship. However, it was also very
controversial, sparking critical comment from others who felt that
evidence had been used selectively in the report. The debate raised by
the report continues to the present day. As you have read in Box 3.1,
it’s fairly common for press reports of a murder or murders to make a
link with the murderer having been found to have a collection of
violent videos, and for this to be brought into evidence in court.

5.2 The Byron report

In 2007, there was increasing public debate in the UK about children
and young people’s unrestricted access to media. The debate was
provoked by the rapidly growing access to the internet among children,
and the launch of ever more violent computer games. Grand Theft Auto,
in particular, was singled out as portraying extreme violence without
consequences for the screen characters and also as being widely used by
children younger than eighteen, the age rating for this game. The prime
minister at the time, Gordon Brown, commissioned another
psychologist, Tanya Byron, to carry out a review of the evidence on the
risks to children’s safety and well-being from exposure to material on
the internet and in video games that might be harmful or inappropriate.
The review was also to consider the effectiveness and adequacy of
existing measures that are supposed to protect children from such
exposure, and suggest any improvements or additional action. Another
objective was to help parents understand the risks of access to
inappropriate content and help them manage those risks.

131
Chapter 3 Learning from watching

Figure 3.8 Towards a greater protection of children in a digital world: the

Byron Review

The report of the Byron Review, Safer Children in a Digital World, was
published in March 2008 (Department for Children, Schools and
Families (DCSF), 2008a).
With regard to the potential negative effects of violent content in video
games, the author of the report reached the following conclusions,
which for the most part concur with points made in this chapter:
. There is some evidence from laboratory-based research connecting
playing violent video games with short-term aggression, but this may
not generalise to the real world. Also, it has not been established
that there are long-term effects.
. There is some evidence of a correlation between playing violent
games and aggressive behaviour or attitudes in the real world, such

132
 5 A call to action?

as hostility or the tendency to get into fights. However, it is not

possible to conclude from this evidence that one causes the other. So,
playing violent video games may cause aggression, or alternatively
more aggressive children may choose to play these games, or,
indeed, there may be another, third variable that explains both
choosing to play these games and behaving more aggressively.

As mentioned previously, the remit of the Byron Review was quite

broad and, in addition to the risks associated with playing violent video
games, it considered the risks associated with the internet.

Activity 3.5
Can you think of the different risks that exposure to the internet poses to
children’s safety? Start by thinking about the safety implications of
different websites that children might access.

As regards the internet, the review used a grid similar to that in

Table 3.1 to structure the wide range of factors that can form ‘risk’. In
your answer to the activity, you may have thought of some of these.

133
Chapter 3 Learning from watching

Table 3.1 Risk factors and internet use. The rows indicate three ways in which a child interacts with
the internet (as ‘recipient’, ‘participant’ and ‘actor’) while columns differentiate between four types of
content that might pose a risk to a child (‘commercial’, ‘aggression’, ‘sexuality’ and ‘values/ideology’)

Commercial Aggression Sexuality Values/Ideology

interests

Content(What is Advertisements Violent/hateful Pornographic or Biased

found on the Exploitation of web content unwelcome sexual information
web?)Child as personal content Racism
recipient information Misleading health
advice
Contact(Someone Harvesting of Being bullied, Being groomed Being supplied
making contact personal harassed or Arranging offline with misleading
with child)Child as information stalked meetings information or
participant Children being advice
tracked by
advertising
Conduct(Child Illegal Cyber-bullying or Creating and Providing
contacts downloading, harassing other uploading misleading
someone)Child as hacking internet users inappropriate or information or
actor illegal material advice

Source: adapted from Hasebrink et al., 2008

What becomes clear from looking at Table 3.1 is that discussion

regarding the potential negative influence of media has become
considerably more complex since the Bandura et al. study. The digital
world has increased the range of risks to which children might be
exposed, including increased exposure to sexually inappropriate content,
‘stranger danger’ and cyber-bullying. The report draws attention to the
fact that research cannot move at the pace of technological change, and
the study of long-term implications is particularly difficult when both
technology and its use are changing.
Importantly, the Byron Review highlights that children have the right to
participate in new interactive technologies and should be able to do so
in a safe and informed way. These technologies offer a range of
opportunities for learning and development, as well as frequently being
fun to engage with. The following point is made in the executive
summary to the report:

Having considered the evidence I believe we need to move from a

discussion about the media ‘causing’ harm to one which focuses

134
 5 A call to action?

on children and young people, what they bring to technology and

how we can use our understanding of how they develop to
empower them to manage risks and make the digital world safer.
(DCSF, 2008b, p. 2)

Figure 3.9 Introducing parental controls in software

The strategic objectives to emerge from the report therefore include

reducing the availability of harmful and inappropriate material in the
most popular areas of the internet, restricting children’s access to
harmful and inappropriate material and enhancing their resilience to the
material to which they may be exposed. This will involve better
regulation and better information and education.

5.3 Conclusion
This chapter has looked at the topic of aggressive behaviour and the
influence of media portrayals of violence on children. The evidence
related to this topic has led to much debate and discussion and
identified multiple influences that relate to aggressive behaviour. In this
final section you saw how the research conducted by psychologists has
helped to shape new policies designed to reduce potential negative
effects.
As you will see in subsequent chapters, social learning – that is, learning
through observing the behaviour of other people – is not the only way
in which children, and indeed adults, learn. But it is a potent form of
learning, and understanding its mechanisms can be very helpful, not just
in respect of our understanding of aggressive behaviour but also in
relation to how people, and most importantly children, acquire new and
useful skills and learn to engage in more positive aspects of behaviour.

135
Chapter 3 Learning from watching

Summary
. Psychological research on media and violence has made a
contribution to government policy on media regulation.
. Studying the effects of media on children is made difficult by its
changing nature and the ways in which children interact with it.

136
 References

References
Anderson, C.A. and Bushman, B.J. (2002) ‘Human aggression’, Annual Review of
Psychology, vol. 53, no. 1, pp. 27–51.
Bandura, A., Ross, D. and Ross, S.A. (1963) ‘Imitation of film-mediated
aggressive models’, Journal of Abnormal and Social Psychology, vol. 66, no. 1,
pp. 3–11.
Bandura, A. (1973) Aggression: a social learning analysis, Upper Saddle Place, NJ,
Prentice Hall.
Berkowitz, L. (1993) Aggression: Its Causes, Consequences, and Control, New York,
NY, McGraw Hill.
BBC News (2004) ‘Game blamed for hammer murder’, 29 July [online],
http://news.bbc.co.uk/1/hi/england/leicestershire/3934277.stm (Accessed 15
February 2010).
Borden, R.J. (1975) ‘Witnessed aggression: influence of an observer’s sex and
values on aggressive responding’, Journal of Personality and Social Psychology,
vol. 31, no. 3, pp. 567–73.
Bushman, B.J. and Huesmann, L.R. (2006) ‘Short-term and long-term effects
of violent media on aggression in children and adults’, Archives of Paediatrics and
Adolescent Medicine, vol. 160, no. 4, pp. 348–52.
Calvert, S.L. and Wilson, B.J. (eds) (2008) The Handbook of Children, Media and
Development, Chichester, Wiley-Blackwell.
Dale, E. (1935) The Content of Motion Pictures, New York, NY, Macmillan.
Department for Children, Schools and Families (DCSF) (2008a) Safer Children
in a Digital World: The Report of the Byron Review, Nottingham, DCSF
Publications; also available online at
http://publications.dcsf.gov.uk/default.aspx?
PageFunction=productdetails&PageMode=publications&ProductId=DCSF­
00334-2008& (Accessed 15 February 2010).
Department for Children, Schools and Families (DCSF) (2008b) Safer Children
in a Digital World: Executive Summary [online],
www.dcsf.gov.uk/byronreview/ (Accessed 15 February 2010).
Gunter, B. (2008) ‘Media violence; is there a case for causality?’, American
Behavioural Scientist, vol. 51, no. 8, pp. 1061–1122.
Hagell, A. and Newburn, T. (1994) Young Offenders and the Media: Viewing Habits
and Preferences, London, Policy Studies Institute.
Hasebrink, U., Livingstone, S. and Haddon, L. (2008) Comparing Children’s Online
Opportunities and Risks across Europe: Cross national Comparisons for EU Kids Online,
London, EU Kids Online [online],
www.lse.ac.uk/collections/EUKidsOnline/Reports/D3.2_ISBN.pdf (Accessed
15 February 2010).

137
Chapter 3 Learning from watching

Huesmann, L.R. (1998) ‘The role of social information processing and

cognitive schema in the acquisition and maintenance of habitual aggressive
behavior’ in Geen, R.G. and Donnerstein, E. (eds) Human Aggression: Theories,
Research, and Implications for Policy, New York, NY, Academic Press.
Livingstone, S. and Bober, M. (2005) UK Children Go Online: Final Report of Key
Project Findings, London, London School of Economics and Political Science.
Malamuth, N.M. and Briere, J. (1986) ‘Sexual Violence in the Media: Indirect
Effects on Aggression Against Women’, Journal of Social Issues, vol. 42, no. 3,
pp. 75–92.
Newson, E. (1994) ‘Video violence and the protection of children’, Journal of
Mental Health, vol. 3, no. 2, pp. 221–7.
Ormrod, J.E. (1999) Human Learning (3rd ed.), Upper Saddle River, NJ,
Prentice-Hall.
Weisz, M.G. and Earls, C.M. (1995) ‘The Effects of Exposure to Filmed
Sexual Violence on Attitudes Toward Rape’, Journal of Interpersonal Violence,
vol. 10, no. 1, pp. 71–84.

138
 Conclusion

Conclusion
In the first three chapters of Investigating Psychology you read about the
research of Theodor Adorno and colleagues on authoritarian
personality, Stanley Milgram’s obedience studies and the investigation by
Albert Bandura and colleagues into social learning and aggression. All
three chapters address a similar aim: understanding why people do harm
to others.
An important point that emerges from the chapters is that psychological
research is historically situated. The authors of the authoritarian
personality study, some of whom fled Nazi Germany in the 1930s, were
influenced by the events of the Second World War and the Holocaust.
Milgram too was influenced by his Jewish heritage and he saw parallels
between his findings and some of the issues that were emerging from
the trial of Adolf Eichmann. Bandura worked at a time when television
sets were introduced into American homes, and when access to violent
media was of increasing public concern. However, the research of
Adorno, Milgram and Bandura, although rooted in the period in which
they worked, raised many questions which continue to intrigue
psychologists today.
Another message that emerges from the chapters is that human
behaviour is inherently complex. Adorno and others sought to identify
aspects of personality that account for authoritarianism. Yet they were
also aware that the situation plays an important role in authoritarianism
and that, for the potential for fascism to be realised, the circumstances had
to be right. Equally, although Stanley Milgram sought to isolate aspects
of the situation that determine whether or not people will follow orders
and cause harm to others, the research by Elms showed that the
participants’ authoritarian dispositions also played a part. Personality
probably accounts, at least partly, for the fact that, while around two­
thirds of participants in Milgram’s study were prepared to administer
lethal shocks to a fellow human being, there was a third who were not.
Similarly, Bandura’s work on social learning provided evidence that
observing a violent model, even one depicted on film, makes a child
more likely to display aggressive behaviour. However, this observed link
between media and violence has been shown to be complex, involving
multiple mechanisms, which interact with a range of other factors
influencing aggressive behaviour. What all of these findings suggest is
that when we seek explanations for human behaviour – in this case for

139
Conclusion to Part 1 Conclusion to Part 1

why people do harm to others – the answer is unlikely to be related

only to personality, the situation or learning from others. These
different factors (and indeed others) act together to produce particular
and complex patterns of behaviour. It is down to psychologists to try to
disentangle the contribution of each of the different factors.
The first three chapters also offered you a taster of different sub­
disciplines of psychology. Chapter 1 introduced the psychology of individual
differences, a term used to denote research that looks at personality, and
at what make a person unique and different from others. Milgram’s
work is a classical study in experimental social psychology, in that it examines
the way in which situational variables and the interaction with others
affect human behaviour. The study by Bandura and colleagues is, on the
other hand, a study in developmental psychology, in that it focuses on
children’s behaviour and their psychological development. However, a
note of caution is necessary here. The boundaries between the sub­
disciplines are not fixed. For example, Adorno’s work was not just
about personality, but also about its roots in childhood development.
Also, authoritarianism research has had a far greater influence on social
psychologists interested in explaining prejudice and political extremism
than on researchers interested in personality. So, what Part 1 as a whole
illustrates is that, while there may be differences between different sub­
disciplines in psychology in respect of focus, emphasis or approach to a
topic, they nevertheless overlap and interact in their study of human
behaviour.
It is important to bear in mind that the chapters offered you a mere
snapshot of three strands of psychological research that have
investigated why people do harm to others. This was by no means a
comprehensive or exhaustive review of the research in this area. There
are many other ways of looking at issues of violence, aggression and
obedience. Biological psychologists, for example, have explored what
happens in the human brain when people engage in violent behaviour,
and whether specific processes in different parts of the brain might
differentiate people who have a tendency to engage in violence from
those who do not. Forensic psychologists also seek to understand
criminal and violent behaviour, with certain researchers focusing on
developing more sophisticated ways of profiling offenders. Clinical
psychologists too have been interested in the psychology of those who
harm others, exploring, for example, the relationship between mental
illness and criminal behaviour.

140
 Conclusion

In reading Part 1, you also encountered a number of different methods

used by psychologists to study the complexity of human behaviour. In
Chapter 1 on the work of Adorno and colleagues, you learned about
two different ways of studying personality: questionnaire-based
personality scales and in-depth interviews. In Chapters 2 and 3 you read
about the value of the controlled experiment as a method that allows
researchers to manipulate specific variables and measure their effect on
another. In the account of Bandura et al.’s work you also encountered
observation as a specific technique for gathering data: the Bobo doll
studies involved the systematic observation and coding of children’s
behaviour. In subsequent chapters you will learn more about these
methods and encounter new ones as a more complex picture of
psychology as a discipline begins to emerge.
When reflecting on method, it is important also to consider ethics.
Chapters 2 and 3 highlighted the need for psychologists to protect the
welfare of participants and to ensure that their integrity, and their status
as a willing participant, is never jeopardised. Observing the rules and
guidelines of ethics, however, protects the integrity not just of the
participants, but also of psychology as a scientific discipline. Crucially,
the fact that ethics has become embedded in the institutional framework
of psychological research is in large part the consequence of the earlier
debates about the limits of acceptable research. Milgram’s study in
particular pushed the boundaries of that debate. In Chapter 5 you will
read about another example of ethically questionable research,
concerning working with animals, which, while uncovering important
things about psychology, also made psychologists re-examine their own
practices.
Finally, the studies explored in Part 1 have highlighted that interacting
with others is an important influence on human behaviour. Research on
the authoritarian personality emphasised the role of interactions with
parents, while Milgram explored interactions with an authority figure.
Central to the notion of social learning are ‘models’, namely other
people, observed and imitated by an individual. Chapters in Part 2 will
examine an issue that is linked to this ever-present influence of others.
If others have the capacity to influence what people do, how does this
happen? Interestingly, as you will see in Chapters 4 and 5, some of the
answers to these questions have come from research on animals, not
humans.

141
Part 2
 Introduction

Introduction
Part 1 of Investigating Psychology highlighted the importance of interaction
with others – harsh parents in Chapter 1, an authority figure in
Chapter 2, and a violent ‘model’ in Chapter 3 – as a source of influence
on human behaviour. That people are influenced by others might seem
obvious. Many people cite family members or well-loved teachers as
major influences in their lives, as individuals who ‘made a difference’.
Others blame harsh or neglectful parents, or poor schooling, for missed
opportunities or adversity in life. Equally, few would challenge the idea
that friends are a major influence on people throughout life, and that
they affect, to some extent at least, what people do and how they do it.
What is less obvious in each of these cases, however, is how others
shape what people do.
This question is the key focus of Part 2 of Investigating Psychology. Let us
look at the example of teachers, parents and friends again. Teachers
exercise their influence because they are placed in charge of a child’s
education. They reward good behaviour and performance and
occasionally punish transgression. In doing so they teach a child to
differentiate appropriate from inappropriate behaviour and help them to
acquire knowledge about the world around them. Do parents influence
a child in the same way? While there is a clear overlap between the role
of the teacher and the parent in a child’s life, there is an obvious
difference. Most children appear to form a powerful bond with their
primary caregiver, usually a parent, very early on in life. This kind of
attachment is not characteristic of relationships with teachers or other
adults whom we encounter later in life. Similarly, unlike in the case of
the relationship with parents or teachers, people have at least some
choice of who their friends are. More importantly, they usually see them
as equals, that is as being on the same level as themselves in respect of
power relations. Not only does this make the pattern of influence
different, but questions inevitably arise about what a ‘friend’ is and how
friendships are formed and developed. Psychologists have explored
these and other issues to do with the influence of others, and as you
will see they have done so in quite different ways.
How behaviour is shaped is the specific focus of Chapter 4 by
Frederick Toates. In this chapter you will learn how reinforcement
makes a particular behaviour more likely to occur. Terms such as
‘reinforcement’ and ‘punishment’ have a common-sense meaning, but
you will be introduced to their use within a particular field of

145
Part 2 Introduction

psychology called behaviourism, which is concerned with how learning can

occur through these processes. You will be introduced to the work of B.
F. Skinner, who is one of the most influential psychologists of the
twentieth century. Skinner argued that learning through reinforcement is
common to all species, so much of his work involved studying animals,
mainly rats and pigeons. Skinner developed a special device, the Skinner
box, which provided a tightly controlled environment within which
animal behaviour could be studied in a systematic way. The principles
established by Skinner are used today in education, prison and
therapeutic settings, as well as by many pet owners!
Chapter 4 describes a number of different studies conducted on rats
and pigeons, and also cats and dogs, which were conducted to advance
our understanding of learning. Non-human animals have been used in
other psychological studies too. In Chapter 5, Deborah Custance
explores the attachment between the young and their primary caregivers
(researchers usually focus on the mother) and addresses the question of
how this bond arises. The chapter focuses on the work of Harry
Harlow, whose experiments on monkeys explored infant attachment. As
you will see, while Harlow’s findings led to profound changes to
Western attitudes to child rearing, his experiments also underlined the
need for ethics in psychological research with non-human animals. In
fact, Harlow is today considered to be one of the most controversial
figures in the history of psychological sciences, whose treatment of
animals was often cruel. Chapter 5 also describes more recent research
with human participants and looks at mothers’ interactions with their
infants and the different types of attachment that may be formed.
In the final chapter of Part 2, Chapter 6, Charlotte Brownlow explores
the important role of friendship in the lives of children. As you will see,
this is a relatively new area of research. One of the most influential
studies in this particular area, by Bigelow and La Gaipa, was conducted
in the 1980s. What makes this area especially interesting is that
researching friendship in children poses a number of methodological
questions. Building on what you already read in Part 1, this chapter will
revisit the distinction between qualitative and quantitative data and
explore the different ways in which qualitative data can be collected and
analysed to shed light on the issue of children’s friendship. The chapter
also considers cultural differences in friendship and how the role of
friends changes as children get older.
We hope that you will enjoy the chapters and gain not only a greater
understanding of the different ways in which others may influence us,

146
 Introduction

but also a sense of the different approaches and assumptions

psychologists bring to their research.

147
Chapter 4
Changing behaviour
Frederick Toates
Contents
1 Introduction 153
1.1 At the personal level 154
1.2 At the professional level 155
1.3 Identifying the role of learning 156
2 B.F Skinner and the foundations of behaviourist
psychology 158
2.1 Early life and scientific context 158
2.2 The emergence of a behaviourist psychology 161
2.3 The contribution of Skinner 162
2.4 The principle of reinforcement 167
2.5 Some phenomena associated with reinforcement 168
2.6 Punishment 168
2.7 Stimulus–response psychology 169
3 From the Skinner box to human behaviour 172
3.1 The basic issue 172
3.2 Extrapolation to humans: its validity and implications 173
3.3 The social, political and ethical issue 175
4 Behaviourism after Skinner 178
4.1 Different types of learning 178
4.2 The value of the Skinner box 179
4.3 Therapeutic procedures 180
5 The relevance of Skinner to today’s world 182
5.1 Addictions 182
5.2 Global survival 183
5.3 Concluding remark 186
References 187
Chapter 4 Changing behaviour

Aims and objectives

After reading this chapter you should be able to:
. recognise changes that are attributable to learning
. assess the influences on and contribution of B.F. Skinner
. explain what is meant by ‘behaviourist psychology’
. distinguish between classical and instrumental conditioning, and
positive reinforcement, negative reinforcement and punishment
. consider the validity of extrapolating from the behaviour of animals
in Skinner boxes to human society
. describe some contemporary developments that arise from Skinner’s
work.

152
 1 Introduction

1 Introduction
In Chapter 3, you met one form of influence on behaviour: imitation of
‘models’. It described how, by means of imitation, behaviour can
change. The chapter described this type of learning as ‘social learning’.
Children are said to learn to act violently, at least in part, by imitation.
The implication would appear to be that children who exhibit violence
might well have had a history of exposure to violent models. In this
way, violent behaviour might arise and be maintained by such exposure.
Conversely, it would seem to follow that non-violent children might owe
their peaceful behaviour, in part, to a history of exposure to non-violent
models.
The present chapter also concerns learning and tries to identify and
characterise some other influences on behaviour. In doing so, it seeks to
explain how certain behaviour arises and is maintained. Similarly, it is
also concerned with how, under certain conditions, behaviour changes.
This chapter will focus on an influence on behaviour that does not
involve the learner observing a model. Rather, it concerns the fact that
behaviour often appears to take its particular form because it has
consequences that gave rise to it and that maintain it. When these
consequences change, so also does the behaviour. This is best illustrated
by an example.
Suppose that you visit a friend regularly over the years. Why did this
behaviour start and why is it maintained? You might have started to
visit for various reasons, such as a wish to establish a new friendship or
out of sympathy for your friend. Whatever the reason for starting, it
would appear that visiting regularly had the consequence that you received
hospitality and good conversation. Suppose now that the friend stops
showing any hospitality or good conversation and you come to
terminate the visits. Common sense would suggest that this change in
behaviour (stopping the visits) is because of the changed consequences
of visiting. The purpose of this chapter is to put such common sense to
the test of scientific scrutiny and to investigate when and how
behaviour is initiated, maintained and changed as a result of its
consequences. This chapter will say more about the reasons that
behaviour continues once started, though it will not ignore the issue of
how behaviour gets started.

153
Chapter 4 Changing behaviour

The chapter also considers why sometimes behaviour fails to change in

spite of good intentions to do so. Could it be that the consequences of
the established behaviour are so strong as to resist efforts at change?

1.1 At the personal level

Everyone probably sooner or later encounters head-on the issue of
changing their own behaviour. Many people have types of behaviour
that they would rather lose and, conversely, good resolutions that prove
very difficult to put into practice. This may apply to you. For example,
you might wish to take more exercise and eat more healthily but good
intentions get lost along the way.
Many people try to alter the behaviour of those with whom they come
into contact: family, friends and work colleagues. People commonly
attempt to persuade their loved ones to quit smoking by a mixture of
emotional blackmail and the use of common-sense reasoning about the
price of cigarettes. However, some people try (I think wisely, but don’t
let me sway you so soon) to alter the consequences that follow the target
behaviour of the loved one: for example, putting money in a box for
each cigarette resisted.

Activity 4.1
Try listing a number of things that you would like to alter about your own
behaviour. Suppose that you have tried unsuccessfully to change
something. Can you suggest why you failed? What consequences might
have been maintaining the undesired behaviour?
Try to think of when you attempted to alter the behaviour of someone
you know. If it was successful, note why you think that it was. If not, why
do you think that you failed?

You will have arrived at a variety of examples. However, it is possible

that, whatever the example, you might have identified a conflict for the
control of behaviour. For example, if you have tried to quit smoking,
you might well have accepted all of the arguments about health and
expense. However, you have felt the urge to smoke that simply could
not be resisted. It seems that smoking has had consequences in the
past, for example, relaxation and a feel-good factor, which now stand in
the way of quitting. Changing these consequences in some way, possibly
by means of nicotine-free cigarettes, might help.

154
 1 Introduction

If you have tried to dedicate more time to study, you might have been
hijacked by some competing activity that could not be resisted.
Switching on a television has the consequence of producing instant
relaxation and distraction from other problems. When you have tried to
change the behaviour of someone else, they might well have reported
just such a conflict. In addition, of course, they might have been
resentful at what they saw as interference in their life.

1.2 At the professional level

It is the professional duty of some to try to change the behaviour of
others for whom they have responsibilities. The consequences of
behaviour can be manipulated to try to produce a desired change. For
example, teachers try to alter the behaviour of their pupils to get them
to learn and to avoid disturbances. Some schools make rewards the
consequence of good behaviour, as in earning vouchers for
conscientious attendance. The penal system tries to reform the
behaviour of prisoners, to get them to abandon crime. Rewards, such as
access to sports, television or days out, are offered for good behaviour.
Some people experience distress at their behaviour and request
therapeutic interventions to change it. Such intervention sometimes
takes the form of a therapist altering the consequences of the
individual’s behaviour.
Consider the following examples. A woman spends an excessive amount
of time washing her hands in a fruitless effort to eliminate what she
feels is their contamination. She is characterised as having an
‘obsessional disorder’. It seems that the behaviour has the consequence
of temporarily lowering the level of anxiety from where it was before
washing. It would also appear that this consequence can have such a
powerful effect that attempts to quit the behaviour are fruitless, even
though she might acknowledge that her behaviour is irrational.
Recognition that consequences could be maintaining certain behaviour
suggests possible interventions. For example, the therapist might sit
with the patient while urging her to refrain from washing, so as to
prevent the normal consequences. In this way, the strength of the
excessive behaviour may be weakened.
Suppose that a man cannot leave his home for fear of, say, coming into
contact with the feathers of a pigeon. He would be characterised as
having a ‘phobia’. One might speculate that the relaxation of being at
home is a consequence that helps to maintain this behaviour. Being at

155
Chapter 4 Changing behaviour

home might also have the consequence that, out of sympathy, friends
start to visit him. The therapist might try accompanying him as he
makes very brief excursions out of the house. The patient could earn
the consequence of praise for each small increase in length of time
outside the house. It might be arranged that visitors come only after he
successfully leaves the house for a short period. Hence, positive
consequences of going out are introduced.
Professionals have an idea concerning a desired form of behaviour and
seek to bring actual behaviour nearer to that which is desired by altering
the consequences of behaviour.

1.3 Identifying the role of learning

Learning usually occurs in parallel with other processes. Consider sports
training. Suppose that you are increasing your endurance over days and
weeks. It might be arranged that each increase in performance earns a
reward of praise from your coach. This increases your motivation to
strive harder. Your behaviour is observed to change. However, not all
of the change in behaviour can necessarily be attributed to learning. Part
of the improvement would most likely reflect adjustments by the body:
for example, the increased strength of muscles and the increased
efficiency with which the heart pumps blood around the body.
Alternatively, consider the example of the lowering in performance, such
as walking distance and speed, which often accompanies increasing age.
Again, this usually reflects such things as a weakening of muscles and a
lowered efficiency of the heart. The term ‘learning’ would seem
inappropriate in these cases and the changes would not form the object
of study by psychologists. However, of course, psychological
adjustments to such bodily changes are of concern to psychologists. For
example, people find ways of adjusting to the changes of ageing, such as
using a walking stick and taking on less strenuous challenges.
This section has identified an important psychological issue: it appears
that a factor that causes behaviour to be maintained is the consequences
that follow from it. Similarly, it appears that when behaviour changes, it
can be because of a change in these consequences. But how can these
assumptions be put on a firm scientific foundation? What characterises
those consequences of behaviour that have such effects? What sort of
experimentation would reveal the exact circumstances in which
consequences have these effects? The focus of this chapter is the
answers provided by the American psychologist B.F. Skinner.

156
 1 Introduction

Figure 4.1 B.F. Skinner (right) with the author of the present chapter (left)

Summary
. In addition to learning by imitation, it appears that behaviour arises,
is maintained and changes because of the consequences that follow
it.
. It appears that the consequences of behaviour can sometimes be
such as to make it difficult to change behaviour, in spite of good
intentions to do so.
. Not all changes in behaviour can be attributed to learning.

157
Chapter 4 Changing behaviour

2 B.F. Skinner and the foundations of

behaviourist psychology
2.1 Early life and scientific context
Burrhus Frederic Skinner was born in 1904, in the small Pennsylvania
railroad town of Susquehanna. As a child, he showed great interest in
exploring the natural environment and in devising mechanical gadgets.
Family life was basic and puritanical. The possible significance of these
events and interests for the adult Skinner are described later. On leaving
school, he attended Hamilton College, where he mainly studied
languages and literature, but with some psychology. Skinner later
obtained a place at Harvard University to do research in psychology.
There were several influences on the academic career of Skinner.
The first two influences deviate from the issue of the consequences of
behaviour. However, they serve to illustrate again that there are several
different forms of learning and they provide a historical context for
Skinner’s work.
Ivan Pavlov was a Russian scientist who researched what is known as a
Reflex reflex, an automatic reaction to a particular stimulus. The research
An automatic reaction concerned the reflex that underlies the production of saliva and
to a stimulus. digestive juices (in the stomach) by dogs when meat juice is dropped
into their mouths. The meat juice is the stimulus to the reflex. This
Stimulus
might not sound much like behaviour, but the study illustrated a very
A trigger to behaviour. general principle of how change could be produced automatically.
All hungry dogs react to the stimulus of meat juice by the reflexes of
producing salivation and digestive juices. Pavlov noticed that, after such
experience of getting food, the dog reacted by salivating even before the
food arrived in its mouth. For example, the sight of a colleague who
had regularly brought the meat would be sufficient to trigger salivation.
At first, this so-called ‘psychic effect’ was regarded as a nuisance, but
later Pavlov studied it in its own right.
In his famous experiment, the stimulus of a tone was sounded just
prior to the presentation of food. This combination of tone followed
immediately by food was presented on a number of occasions. Then the
effect of the tone, presented in the absence of food, was tested. The
tone alone triggered salivation and the release of digestive juices
(Pavlov, 1927). Here was a clear change in behaviour obtained under

158
 2 B.F. Skinner and the foundations of behaviourist psychology

rigorously controlled conditions. It is as if the dog had learned that the

tone meant food might follow.
John Watson, an American psychologist, brought Pavlov’s research to
the attention of American psychologists. He also conducted a famous
experiment himself (Watson and Rayner, 1920), the ethics of which you
might wish to question. An infant, called Little Albert, was presented
with a tame rat. Watson observed Albert’s reaction to the rat, this being
one of warm acceptance and no fear. He then observed the reaction of
Albert to a loud sound, banging a piece of metal behind Albert’s head.
Albert was observed to cry and to move away from the source of the
sound. As the next stage, Watson presented the rat to Albert just prior
to making the loud sound. Finally, Watson observed the reaction to the
rat on its own. Albert now showed a defensive reaction of moving away
from the rat, which exemplifies a change in behaviour as a result of
experience. Albert might be said to have learned to show a defensive
reaction to the stimulus of the rat.
The third influence on Skinner takes us back to the consequences of
behaviour. Edward Thorndike was an American psychologist who
pioneered the study of animal behaviour, specifically the behaviour of
cats in puzzle boxes (see Figure 4.2).

Figure 4.2 Thorndike’s puzzle box

Thorndike (1898) was interested in the nature of learning in animals

and whether this was fundamentally different from human learning. He
placed a hungry cat in a cage (the puzzle box) and put some food
outside it. By manipulating a latch, the cat could open the door and get

159
Chapter 4 Changing behaviour

to the food. Thorndike measured the length of time it took the cat to
escape from the box over a number of attempts. Figure 4.3 shows how
long it took an individual cat to escape across nine trials.

60

50
Escape time (seconds)

40

30

20

10

0
1 2 3 4 5 6 7 8 9
trials

Figure 4.3 Measure of learning for one cat, showing the time needed to
escape for each trial over nine trials

Activity 4.2
Have a look at Figure 4.3. The horizontal axis is numbered 1–9
representing each individual trial. The vertical axis shows escape time as
measured in seconds (0–60 seconds). The dots represent the escape
time for each trial. How would you describe the pattern of dots?
Which trial shows the longest escape time and which the shortest?
Which two trials differ most in respect of escape time?

As you can see, there is a lot of variation over the nine trials. The cat
took longer to escape on the first trial than on the second or
subsequent trials, and indeed the biggest difference is between the first
trial, where the cat took about 58 seconds to escape, and the second
trial, where the cat took just over 30 seconds. At the end of nine trials,
the cat is rapidly escaping, taking less than 10 seconds. The pattern of
dots in Figure 4.3 suggests that the cat gradually learned how to get
out. There was a steep drop from first to second trial, and then a
slower decline over the remaining trials, with a slight blip from fourth
to fifth trial where the time to escape increased slightly.

160
 2 B.F. Skinner and the foundations of behaviourist psychology

Based on findings like this, Thorndike argued that learning in this

situation consists of the formation of a link between the stimulus of the
latch and the response of manipulation of the latch. The strength of this
stimulus–response link is increased over trials because of the
consequence of escape from the box: the satisfaction of eating. As the
strength of this link increases, so the time needed to escape decreases
gradually.

2.2 The emergence of a behaviourist psychology

Up to the time of Watson, psychology had been based largely upon the
study of what happens in the mind, the so-called mental events. People
reported what was ‘on their minds’ and these accounts were
documented by psychologists (as in the case of Freud and his
followers). What is on your mind is private to you and therefore
described as subjective. I cannot observe it any more than you can
observe what is on my mind. So, such psychology was based on
subjective events. By contrast, behaviour is objective in that it can be
observed by psychologists and measured.
Watson revolted against the way of doing psychology that involved
subjective events (Watson, 1913). He argued that psychology should be
based upon only observable data, with a focus on behaviour. It should
be an objective science, rather as biology and chemistry are objective
sciences dealing only with observable phenomena. The experiments of
Pavlov and Watson fitted this way of doing science. Everything was
‘public data’ and measureable – in Pavlov’s case it was the food, the
tone and the salivation. In the study of Little Albert, the sound, the
presentation of the rat and Albert’s reaction were equally measurable.
Watson argued that a science of psychology should concern itself only
with such observable behaviour, and thereby this approach in
psychology acquired the name behaviourism. Behaviourism
An approach in
With Pavlov, new terminology entered the scientific vocabulary. The psychology, which
tone exemplified what became known as a ‘conditional stimulus’. The argued that only
meaning of ‘conditional’ is that there is a condition attached to the tone’s observable behaviour
power to trigger salivation; it has no intrinsic capacity to trigger such a should form the object
response. The condition is its earlier pairing with food on a number of of study.
occasions. By contrast, meat is not a conditional stimulus since no such
conditions pertain and it is therefore termed an ‘unconditional stimulus’.
The process of changing behaviour in this way became known as
‘conditioning’. Since Pavlov’s was the first type of conditioning to be

161
Chapter 4 Changing behaviour

studied (and others were to follow), it became known as classical

Classical conditioning or Pavlovian conditioning. Note that the dog does not
conditioning play a part in the sequence of when the external events occur. Both
Conditioning in which a tone and food were under the control of Pavlov.
stimulus acquires the
capacity to trigger a It was reasoned that links are formed between stimuli and responses
response by virtue of and hence this approach to understanding behaviour was described as
its pairing with an stimulus–response psychology.
unconditional stimulus.

2.3 The contribution of Skinner

Stimulus–response
psychology Skinner subscribed to the principles of behaviourism and wanted a
An approach to psychology based upon observable data. Though Skinner was aware of
understanding Pavlov’s research, his own interest was more in the tradition of
behaviour that
suggested that learning
Thorndike. In the experimental arrangement of Thorndike, things were
consists of the different from those of Pavlov and Watson, though the term
acquisition of new links ‘conditional’ could still be applied. Unlike Pavlov’s dogs, the cat in
between stimuli and Thorndike’s study had control over the sequence of events: getting to
responses. food was conditional upon the cat’s appropriate reaction, that is, turning
the latch. The consequence of the cat’s behaviour – access to food –
altered the cat, making turning the latch more likely in the future. The
cat was ‘instrumental’ in the sequence of events and hence this process
Instrumental of learning became known as instrumental conditioning.
conditioning
A form of conditioning
in which the outcome
depends upon the
action of the animal, as
exemplified by
obtaining food on
turning a latch or
start box
negotiating a maze.

goal box

Figure 4.4 Maze

Skinner (1979) performed experiments on rats running in mazes (see

Figure 4.4). The hungry rat is placed at the start box. On successfully

162
 2 B.F. Skinner and the foundations of behaviourist psychology

negotiating the maze, it reaches the goal box, where it finds a small
morsel of food. Finding food is the consequence of its behaviour.
Behaviour (taking the correct turns) is said to be instrumental in the
outcome (obtaining food). At first, the rat enters blind alleys. However,
with experience, it gradually becomes more proficient. The rat is said to
learn its way through the maze, learning being measured in two ways
(see Figure 4.5): (a) the decrease in the length of time taken to get from
start box to goal box or (b) the decline in the number of errors (blind
alleys entered).

180

160

140

120 6
time (seconds)

100 5

80 4
errors

60 3

40 2

20 1

0 0
1 3 5 7 9 11 13 15 1 3 5 7 9 11 13 15
trials trials
(a) (b)

Figure 4.5 Learning a maze as measured by (a) time and (b) errors

Experimentation with mazes takes a lot of time and so Skinner invented

a piece of apparatus that was later to earn the name ‘the Skinner box’.
Figure 4.6 shows the forms designed for the two favourite animal
species used by Skinner and those working in his tradition (known as
Skinnerians), namely rats and pigeons. When the rat presses the lever or Behaviour shaping
the pigeon pecks the key, a small pellet of food drops into the food tray The procedure of
and can be eaten. If a hungry animal were simply to be left alone in the giving reinforcement
apparatus, it would probably learn to perform the response that earns for successive
food, but this would take a long time. Things can be speeded up approximations to the
desired response.
enormously by a process known as behaviour shaping.

163
Chapter 4 Changing behaviour

(a)

(b)

Figure 4.6 (a) The Skinner box designed for rats; (b) The Skinner box designed for pigeons

I will describe shaping rats, though an identical logic applies to pigeons.

The desired response is lever pressing, and shaping consists of giving
food for successive approximations to this response. When the
experimenter presses a button, a pellet is given. On the first few
occasions, the rat is given a free pellet to familiarise it with the
apparatus and food delivery. Then the experimenter makes a demand by
giving a pellet only when its front legs or nose are anywhere near the
lever. This will normally persuade the rat to increase the amount of
time that it spends with its front end near the lever. Then the
experimenter makes the criterion for getting food harder: the rat must

164
 2 B.F. Skinner and the foundations of behaviourist psychology

make physical contact with the lever. (As rats are naturally curious
animals, they may well accidentally touch the lever.) After the rat makes
contact, the criterion is made still more stringent: the rat must lower the
lever. When it is reliably lowering the lever, the final stage of shaping
occurs. The apparatus is now on automatic control. Only a full lowering
of the lever such as to trigger the switch will deliver a pellet. If at any
stage the rat regresses, the experimenter can take the remedial step of
lowering the criterion and trying to ‘lift the rat up’ again.

Box 4.1 Why do it this way?

The Skinner box provides the extreme example of a controlled
environment. The rat has very little to do and this increases the
chances of its pressing the lever. The times of pressing are
recorded. This was originally on paper (as you can see in
Figure 4.6(a) and 4.7 below) but these days it would be on a
computer. It is an efficient way of working as regards the
psychologist’s time, since, after the initial phase of shaping, the
experiment runs itself. A number of rats can be run simultaneously.
The Skinner box minimises distractions from outside, so the effect
of the consequence of behaviour (getting the pellet) can be studied
in isolation from other influences such as imitation of another
animal.
In the Skinner box, an experiment can be performed in that an
independent variable can be manipulated and the effect on a
dependent variable observed. For example, by depriving the rat of
food for a period of time, its degree of hunger can be manipulated
(independent variable) and the effect on lever pressing observed
(dependent variable). Another possibility is to keep the level of
hunger constant but to alter the amount of force that it takes to
press the lever (independent variable) and to see what effect this
has on lever pressing (dependent variable). Normally, the
dependent variable is the frequency with which the rat presses the
lever.

165
Chapter 4 Changing behaviour

You might wonder what was so important about Skinner’s findings.

After all, circus trainers have known for centuries that behaviour can
be changed with the help of food. With the invention of the Skinner
box, however, behaviour was brought under rigorous scientific
control and the rules governing reinforcement and behaviour were
systematically studied. For behaviourists, the principles observed in
the Skinner box are not just about animal training: they are
applicable to real life situations, and underpin complex human
behaviour.

Figure 4.7 Skinner performing an experiment with a rat in the Skinner box

Since the food arrives as a result of the rat’s response, this is a form of
instrumental conditioning. However, it is a special form, since the
timing of pellet delivery is in the hands of the rat. The rat sets its own
pace. This is unlike the maze, where the experimenter picks the rat out
of the goal box and places it back in the start box each time. The form

166
 2 B.F. Skinner and the foundations of behaviourist psychology

of instrumental conditioning that is studied in the Skinner box is

termed operant conditioning. This is because the animals operate on Operant conditioning
their environments to produce an effect. The rat is said to emit the The form of
operant of lever pressing. instrumental
conditioning that
occurs in the Skinner
2.4 The principle of reinforcement box.

Skinner suggested that this behaviour illustrates a very important

general psychological principle, as follows. Behaviour changes as a result
of its consequences: for example, lever pressing is followed by positive
reinforcement, in this case, food. In other words, food ‘reinforces’ the Positive
behaviour that immediately precedes it. The word ‘reinforce’ means ‘to reinforcement
strengthen’ and that is the sense in which the word is used here. In The process of giving
something as a
Skinnerian language, the frequency of instrumental (including operant)
consequence of a
behaviour increases as a result of reinforcement. particular behaviour,
Here was an objective way of describing behaviour, in keeping with the which increases the
frequency of showing
principles of behaviourism. Behaviour that is reinforced will increase in that behaviour.
frequency. To Skinner, there was no need to appeal to events in the
head of the rat to explain behaviour. Psychologists needn’t say that the
rat wanted the food, that it knew that lever pressing would lead to food
or that it expected lever pressing to lead to food or even that it liked the
food. All such terms were avoided since they describe unobservables,
which cannot be measured. Skinner also avoided speculating about
events in the brain of the rat, though he accepted that learning involves
the brain.
Food is not the only thing that can serve as positive reinforcement. If
the animal has been deprived of water, then it can be trained to press a
lever for the reward of small units of water. An animal might also be
trained by the use of sexual reinforcement: a lever press could lead to
brief access to a partner.
A reinforcer is defined simply by its capacity to reinforce. If the
frequency of behaviour that is followed by the ‘thing in question’
Negative
increases, then the ‘thing in question’ is a reinforcer. reinforcement
So far, I have described positive reinforcement. The adjective ‘positive’ The process of
removing something as
refers to the fact that something that was not there before arrives as a a consequence of a
result of the action: for example, food suddenly appears. There is particular behaviour,
another procedure that also strengthens behaviour, known as negative which increases the
reinforcement, where something that is already present is removed. frequency of showing
Suppose that a rat is placed in the Skinner box and a loud sound is that behaviour.

167
Chapter 4 Changing behaviour

present. It might be trained to press the lever to switch the sound off
and the loud sound would be said to act as a ‘negative reinforcer’.

2.5 Some phenomena associated with

reinforcement
It was noted that a reinforcer strengthens behaviour, as indexed by an
increase in the frequency of the behaviour. Once the rat is trained in
this way, what is the effect of the omission of reinforcement? The way
to find out is to remove the food from the Skinner box. At first, the
frequency of lever pressing increases and then gradually declines over
days, until the rat finally quits. Removing food is termed applying
Extinction extinction. After the rat has quit lever pressing, its response is said to
The removal of a ‘have extinguished’.
reinforcer, e.g. taking
the food pellets away At one stage in his study, Skinner ran short of food pellets and needed
from the Skinner box. a way to economise on them. He hit on the following solution. Once a
rat or pigeon is trained, it is not necessary to reinforce every response in
order to maintain behaviour. Rather, say, only every fifth lever press
might be reinforced. Such a schedule of giving pellets is known as
Partial reinforcement partial reinforcement, as opposed to ‘continuous reinforcement’ where
A schedule in which every response earns a pellet.
the rat gets reinforced
only occasionally. The shortage of food pellets had a very significant consequence. When
they had been on partial reinforcement, as compared to continuous
reinforcement, the animals persisted for much longer when applying
extinction before they stopped pressing the lever (Skinner, 1939). This
observation exemplifies the occasional role of chance in scientific
discovery.

2.6 Punishment
Whereas positive or negative reinforcement strengthens the tendency to
show the associated behaviour, there is also a technique, ‘punishment’,
which lowers the probability of a given behaviour. The Skinnerian use
Punishment of the term punishment has some similarities to its lay use, as in to
An event that follows a punish a criminal offender.
response and which
leads to a decrease in Suppose that a rat pokes its nose through a hole in its cage. First, the
the frequency of that frequency with which it does this is recorded. Then, every time it does
response. so, this triggers a loud sound and the frequency of nose poking is again
measured. Suppose that, after introducing the sound, the frequency of

168
 2 B.F. Skinner and the foundations of behaviourist psychology

nose poking decreases. It would be said that the sound is punishing nose
poking. The sound is defined as punishing if there is a decrease in the
frequency of behaviour. Hence, punishment is the mirror image of
reinforcement and is similarly defined in terms of a change in
behaviour.
Punishment again exemplifies the Skinnerian approach. The observation
and measurement of behaviour is fundamental. Punishment is not
defined in terms of its aversiveness as such but rather in respect of its
effect on behaviour. The lay term ‘punishment’ refers to something that
is merely assumed to have an effect on reducing undesirable behaviour,
but the actual measurement is usually not made.

2.7 Stimulus–response psychology

How did Skinner’s discoveries fit into the broader context of
behaviourist psychology? Central to behaviourism would be how
stimulus–response links change as a result of experience, as reflected in
changes in behaviour.
So, how does Skinner’s work fit the approach of stimulus–response
psychology? The response seems clear: the lowering of the lever by the
rat or key pecking by the pigeon. These can be measured objectively
and their precise timing documented. Changes in behaviour over time
can be measured, for example, increased frequency of lever pressing as
the rat becomes more skilled at the task. So, everything seems neat and
tidy for the application of stimulus–response psychology. Or does it?
A serious problem confronted Skinner: the response is clear but what is
the stimulus to this response? You might speculate that the lever is the
stimulus but the lever is present throughout.
Skinner resolved this issue along the following lines. He started by
emphasising the distinction between the kind of conditioning that he
studied and that studied by Pavlov. In Pavlov’s case, there was indeed a
discrete stimulus that triggered the response: for example, the tone
triggered salivation. Behaviour of this kind was said to be controlled by
stimuli that come before the response. Such behaviour was said to be
elicited by stimuli, and the expression elicited behaviour came into Elicited behaviour
use. Behaviour that is
triggered by a stimulus.
By contrast, the behaviour studied by Skinner was of a different kind
and it was argued that it is not elicited by a stimulus that occurred prior
to behaviour. Rather, behaviour was said to be freely emitted and it

169
Chapter 4 Changing behaviour

comes under the control of its consequences. As you have seen, sometimes
these consequences are of a kind that strengthens the behaviour and
they are said to be ‘reinforcing’. At other times, the consequences of
behaviour are punishing: that is, they lower the tendency to perform
this behaviour. Behaviour that is freely emitted in this way and which
comes under the control of its consequences was classified as emitted
Emitted behaviour behaviour. Figure 4.8 summarises the distinction between elicited and
Behaviour that is emitted behaviour.
controlled by its
consequences.
stimulus ANIMAL response

(a)

consequence

reinforcement + - punishment

ANIMAL response

(b)

Figure 4.8 Two types of behaviour (a) elicited behaviour and (b) emitted
behaviour. The plus and minus signs represent reinforcement and
punishment, respectively

Of course animals did not evolve processes of learning in order to get

by in psychology laboratories, so you might like to speculate on their
significance for survival in the wild. Consider a hungry rat exploring
and manipulating its natural environment. Certain actions will provide
access to food: for example, tearing open a sack of grain. The
reinforcement of these actions has a clear benefit for the rat.
So far, the focus of the chapter has been on rats and pigeons. However,
Skinner was concerned that his ideas should be applied to humans and
the discussion will turn to this in the next section.

170
 2 B.F. Skinner and the foundations of behaviourist psychology

Summary
. In instrumental conditioning, the sequence of events leading to
reinforcement depends upon the behaviour of the animal.
. The Skinner box is a piece of apparatus in which operant
conditioning (a form of instrumental conditioning) is studied.
. Whereas reinforcement increases the tendency to show a particular
behaviour, punishment lowers it.
. Skinner drew a distinction between ‘elicited behaviour’, which is
triggered by a stimulus, and ‘emitted behaviour’, which is controlled
by its consequences.

171
Chapter 4 Changing behaviour

3 From the Skinner box to human

behaviour

Activity 4.3
Skinner argued that giving praise for desirable behaviour of a child in
school reinforces that behaviour in the same way that a pellet reinforces
lever-pressing in a Skinner box. Stop and think about this issue. Ask
yourself whether things can really be that simple. Is there any
fundamental difference in these situations?

Superficially, the two behaviours you reflected on in Activity 4.3 might

well be very similar, but praise is surely more complex than a food
pellet. For praise to serve as a positive reinforcer, it is necessary for the
child to understand a spoken language and the meaning of praise
(unlike the rat being reinforced with a food pellet). None of this
undermines the power and efficacy of reinforcement, but it suggests
that complex events going on in the mind of the child help to mediate
the role of the reinforcer. Also, you might think that much human
behaviour (e.g. visiting a friend) starts for reasons that are not
accidental. There is often an intention involved: for example, to bring
cheer to an individual.

3.1 The basic issue

Skinner (1953) argued that the principles of reinforcement and
punishment that he so successfully studied in non-human animals in the
laboratory could be extended to humans. He noted that any species
confronts much the same set of problems in the wild: for example, how
to find food and shelter and how to avoid excesses of heat and cold.
Hence, a similar set of processes of learning might be expected when
species are compared, and humans are just another species.
Skinner argued that human behaviour sometimes goes adrift as a result
of faulty assumptions regarding its causes and a failure to make
scientific observations of it. However, he asserted that behaviourist
techniques can come to the rescue. Take, for example, a child who is
misbehaving. The teacher stops the class at each instance and says ‘Stop
misbehaving or you will get into trouble’. The implicit assumption is
that the threat will be punishing and hence the frequency of

172
 3 From the Skinner box to human behaviour

misbehaviour will come down. From a behaviourist perspective, a

careful measurement of behaviour is needed. This might reveal that, far
from decreasing the frequency of bad behaviour, when it was remarked
upon the behaviour actually increased in frequency. Hence, what was
believed to be punishment was actually positive reinforcement. It might
be speculated that what constituted reinforcement was the attention that
the teacher was giving to the child.
Skinner (1971) suggests that punishment is relatively ineffective as a
control of behaviour. To illustrate this general assumption, consider
again the misbehaving schoolchild. Suppose that punishment had been
used to decrease bad behaviour. Can you see a problem with such a
solution? Punishment, even if effective, teaches only what not to do
rather than what to do. A vacuum can be left within the child’s life. The
child might rebel against the school by playing truant.
If the misbehaving schoolchild’s disruptions are being reinforced by
their teacher’s attention, then omission of reinforcement should be
associated with a decline in the undesirable behaviour. In addition to
imposing extinction conditions, the teacher could try to positively
reinforce behaviour that is incompatible with the undesirable behaviour.
For example, when a child is paying attention or at least not being
disruptive, praise might be offered. These techniques have been tried in
classrooms with success (e.g. Swinson and Harrop, 2005).

3.2 Extrapolation to humans: its validity and

implications
So long as Skinner studied rats and pigeons, few psychologists raised
serious concerns. The notion that animal behaviour is controlled by its
consequences triggered stimulating discussion among psychologists
concerning the limits of this approach. Even some extrapolation from
non-humans to humans might have been tolerable to most, as in
employing techniques of reinforcement in schools, mental hospitals and
prisons. However, Skinner went much further than simply presenting a
toolbox of techniques for altering problematic behaviour. Determinism
Skinner made a fundamental and highly controversial assumption: there The belief that
everything that
is a universal principle of reinforcement that applies to human behaviour and which
happens, including
can account for most of this behaviour. human behaviour, is
This Skinnerian faith in reinforcement is based on the principle of determined by past
events.
determinism and the rejection of ‘free will’. Determinism is the

173
Chapter 4 Changing behaviour

assertion that everything that happens in the world, including human

behaviour, is determined by what went before. So, if only psychologists
knew enough about the events in the life of a person, they would be
able to predict the future behaviour of this person. The counter­
argument to Skinner (the free-will position) is perhaps that which
appeals most broadly to people: humans have the capacity to make free
and spontaneous choices, for which they can be held responsible and
accountable. Their behaviour is not simply at the mercy of their history
of reinforcement.
Skinner’s assumption produced a passionate intellectual battle. His
critics countered with the argument that there are fundamental
differences between humans and other species. Non-humans might be
rigidly determined in their behaviour but humans are different. Humans
speak a sophisticated language and have complex cultures. They have
insight into their behaviour and can make conscious choices. Humans
are capable of articulating the determinants of their behaviour and
reporting them to others, or of remaining inscrutable as to why they did
something. Their behaviour, it was claimed, is characterised by long­
term planning, purpose and intention. The possession of conscious
awareness, language and culture means that humans can consciously
work through plans of action (e.g. to get a university degree by means
of study over years) and then freely execute them. Such a vocabulary
sounds decidedly not ‘rat-like’ or behaviourist in its assertions.

Activity 4.4
Stop and consider the relevance of Skinner’s argument for some of your
behaviour. Could you better use positive reinforcement in dealing with
friends and family? Do you feel that you function better when positively
reinforced for desirable behaviour rather than being negatively reinforced
or punished for wrongdoing? Do you feel that you have free will or that
your behaviour is determined by what happened in the past?
Most people would surely agree that past experiences have a strong
determining influence. For example, someone growing up in a violent
family might be expected to be more prone to exhibit violence, in part
because violent behaviour has been reinforced and in part modelled.
However, it is important to note that not all people exposed to violent
behaviour go on to become violent, and so such an explanation cannot
be the whole story. This is why most psychologists would not go so far
as Skinner and assume total determinism.

174
 3 From the Skinner box to human behaviour

3.3 The social, political and ethical issue

Skinner argued that psychology needed to be used to change the world
into a better place. Indeed, he had a utopian vision of a world that
functioned on the basis of positive reinforcement, while avoiding
aversive forms of control such as punishment. He became a missionary
in the cause of changing the world and felt this would happen only if
human behaviour were changed. This required meeting two closely
related requirements:
1 A science of human behaviour is accepted, based upon behaviour,
avoiding references to what goes on inside the head, such as
intentions, motives, and so on.
2 Societies are designed according to the principles of a scientific
understanding of behaviour, with positive reinforcement at centre
stage.

Therefore, Skinner (1953, 1971) argued that rather than target people’s
minds, social change would occur if it were based upon insights gained
from studying rats and pigeons.
Let us first briefly consider the old way of explaining behaviour.
Traditionally, behaviour is explained in terms of mental events as
reported by introspection. For example, consider a person who puts
money in a charity box to help the poor. It might be said that they did
so with the intention of alleviating suffering. They thought about the plight
of poor people, they felt bad about this and they made a free decision to
give. As free beings, they might equally well have walked by on the
other side of the street, ignoring the poor. To persuade people to give
to charity, one might need to change their hearts and minds.
This type of explanation, of course, appeals to mental events and these
were to be avoided in a behaviourist psychology. Rather, Skinner
(1953, 1971) argued that the charitable person had a history of being
reinforced for performing similar behaviour. Each time in the past that
they showed giving behaviour, they received reinforcement in the form
of, for example, thanks and social acceptance. Hence, their present
behaviour is to be explained historically in terms of the consequences of
similar behaviour in the past.
Consider a judge sentencing a criminal to life in jail for murder. The
summing-up would usually involve something along the lines of ‘You
are an evil person with little in the way of a conscience, who willingly opted
for a career in crime and finally took the life of another. You chose a

175
Chapter 4 Changing behaviour

path of evil and now society requires that you pay the price and feel
remorse.’ It would sound very odd indeed if a Skinnerian judge were to
say, ‘You had a long history of being positively reinforced for violent
behaviour. Now you will receive positive reinforcement only for socially
acceptable actions, while your anti-social actions will undergo
extinction.’
Skinner’s best-known book was entitled Beyond Freedom and Dignity
(Skinner, 1971) and it reached number one in the American bestseller
charts. In 1975, Skinner became the most famous scientist in America
(Guttman, 1977). The title of his book might seem to be a deliberate
provocation and it needs some unpacking. Skinner urged society to
reject the assumption that free will is the main determinant of
behaviour. The ‘dignity’ that society was being asked to move beyond
referred only to a particular kind of dignity: that which derives from
achievement. That is to say, people like to take credit (acquire ‘dignity’)
for those actions of which they are proud, whereas they tend to avoid
responsibility for those of which they are not proud. Extenuating
circumstances can invariably be brought to bear in the latter case.
According to Skinner, rather than accept determinism for bad actions
and free choice for good ones, people need to reject the double
standard. Behaviour is deserving of neither praise nor condemnation;
rather, it needs dispassionate analysis in relation to its determinants.
The book earned criticism like few books before it. Even United States
Vice-President Spiro Agnew (in the presidency of Richard Nixon)
described Skinner as ‘most dangerous’. It is not hard to see why an
American politician would take this stance. The political foundations of
the United States are based on the individualistic assumption that
people are responsible for their actions. They are responsible for the
good and bad that happens to them and can blame no one but
themselves. To take away personal responsibility was seen as basically
anti-American. Skinner was also criticised by those on the left too. He
saw the root cause of social problems as coming from a particular
history of reinforcement of individual behaviour. This was incompatible
with the left-wing view that wrongs are caused by broader social forces,
such as the capitalist economic system, which operate beyond the
individual.

176
 3 From the Skinner box to human behaviour

Summary
. Skinner argued that the principle of reinforcement can be applied
very widely to human behaviour.
. According to Skinner, determinism applies to all behaviour and the
notion of free will is not compatible with scientific thinking.
. Skinner believed that a science of behaviour with positive
reinforcement at centre stage and avoiding aversive controls could
solve the world’s problems.

177
Chapter 4 Changing behaviour

4 Behaviourism after Skinner

Only a few psychologists these days would describe themselves as
behaviourists. The revolutionary zeal of Watson and Skinner to build a
new science of psychology based upon observable behaviour might
appear, then, to have failed.
Although psychology has not adopted behaviourism as its fundamental
framework, it nonetheless owes much to the early behaviourists. The
systematic observation of behaviour and its measurement is implicitly
used by a range of psychologists and it might be argued that this
derives from behaviourism’s early emphasis upon objectivity. The
techniques of reinforcement are widely in use in hospitals, prisons and
schools.
The whole behaviourist enterprise was based upon the assumption of
determinism and the rejection of free will. This argument still rages
back and forth, with passionate advocates of each position. However, it
is unlikely that one side will score a victory. Instead, this is an issue that
cannot be resolved.
In surveying the Skinnerian argument, some have been led to conclude
that what is termed a ‘pragmatic position’ might need to be adopted. At
times psychologists might choose to act as if people have free will,
whereas at other times psychologists might feel that determinism offers
the most insight.

4.1 Different types of learning

Skinner’s emphasis was on the importance of operant conditioning.
However, over the years, there has been an increasing recognition that
there are several different types of learning. For example, learning by
You read about Albert imitating role models has been demonstrated by Albert Bandura. There
Bandura’s research in is also an increasing recognition that classical conditioning occurs much
Chapter 3. more widely than was earlier appreciated. It tended to be thought at one
time that it was applicable mainly to reflexes such as salivation. Now it
is realised that more complex behaviour can reflect such learning. For
example, in drug addicts, an urge to seek and take drugs can be
triggered by being in an environment (e.g. one in which pushers or
other addicts are present) that has been associated in the past with drug
taking (Siegel, 2005).

178
 4 Behaviourism after Skinner

Figure 4.9 Environment acting as a trigger for behaviour

4.2 The value of the Skinner box

The Skinner box is a favourite piece of equipment in a wide range of
laboratories among researchers of many orientations. It is a simple, well­
controlled and very efficient way of observing behaviour in non-human
animals. Consider, for example, a researcher who is studying vision in
birds. He or she can exploit what Skinner termed discrimination
learning. Typically, a bird is first trained to work for food. Then pellets Discrimination
are available only when a red light in the box is on but not when a learning
green light is on. After a time the bird will come to discriminate, so that Learning to respond in
the presence of one
it pecks the key when the red light is on but not when the green light is
event (e.g. a green light)
on. It has shown thereby that it can discriminate between these colours. but not in the presence
Suppose that a new species of bird is discovered and the researcher of another (e.g. a red
light).
wishes to know whether it has colour vision. One way of doing this is
to perform a discrimination test. A member of this species is trained in
the Skinner box to earn food. Food is then made available only if the
bird pecks a disc of a particular colour. If the bird proceeds to peck
only when presented with a disc of a particular colour, then it can be
inferred that it has colour vision.
Research on drugs designed to counter anxiety also uses rats in Skinner
boxes. How could you measure anxiety in a rat? How could you
measure a reduction in anxiety as a result of taking a drug? First, a

179
Chapter 4 Changing behaviour

hungry rat is trained to earn food pellets. Then a tone is followed by a

mild shock to the feet. Rats come to fear the tone, in that it is followed
by a brief lowering of their lever pressing. Then they can be injected
with the drug to see whether this reduces the disruption to lever
pressing that is caused by the tone. When drugs that are used to lower
human anxiety were given to the rats, they reduced the disruptive effect
of the tone (Lane, 1992).
Without doubt, the Skinner box is useful and the procedure of
reinforcement is a highly effective way of studying and changing the
behaviour of an animal. However, disagreement concerns the
interpretation placed upon the results obtained in the Skinner box.
Skinner urged psychologists not to apply mental terms to behaviour but
that has not stopped many psychologists from doing exactly that. For
example, some (e.g. Bolles, 1972) argue that, in the Skinner box, the
Expectancy animal learns to expect food (it forms an expectancy). It learns an
The formation of association between its behaviour and the arrival of food. There is some
mental state where an evidence to support this position, which fits with common sense. When
animal comes to expect
extinction terms are applied, rats are said to show an emotion that is
something to happen
after something else termed ‘frustration’ (Gray, 1987). Why is it called this? The logic is as
has happened. follows. In response to stress, particular hormones are released into the
blood stream. These same hormones are released when extinction is
applied, suggesting that the experience is stressful. Frustration is the
term used to describe the human emotion experienced under
comparable conditions and so psychologists apply this also to rats.

Activity 4.5
If you have a pet, reflect on its behaviour and whether you can recall any
signs of expectancies.
Most cat and dog owners can probably find examples. The prospect of a
walk or food seems to trigger anticipation of good things to follow shortly.
The sound of a can-opener triggers more than just salivation
(e.g. running from a distant room to the site of the food).

4.3 Therapeutic procedures

Contemporary therapy for psychological distress owes much to insights
derived from behaviourism. For example, children who are displaying
self-harm, such as head-banging, have been treated with techniques of
reinforcement for non-harming behaviours (Flora, 2004). Adult patients

180
 4 Behaviourism after Skinner

in mental hospitals who are uncommunicative and uncooperative have

been taught communication and social skills by reinforcing appropriate
behaviours and ignoring inappropriate behaviour (Flora, 2004). One of
the most powerful techniques for treating people with obsessional and
phobic disorders (see Section 1) involves identifying and removing
reinforcement for the behaviour that is excessive. At the same time, any
reductions in the target behaviour can be reinforced with praise.
There are two important qualifications to add and these serve to frame
behaviourist approaches in modern terms:
First, the efficacy of reinforcement can depend upon the patient’s
understanding of what is going on, as mediated by means of a spoken
language. Indeed, reinforcement is likely to be only one aspect of the
therapeutic procedure. Another aspect is likely to focus on the patient’s
thought processes.
Second, closely connected with the notion of behavioural interventions
for change is the explicit recognition that interventions also target
processes of learning that are not based on reinforcement. For example,
consider again the patient with a phobia about going outdoors. The
therapist who escorts the patient on his excursions outdoors is actually
presenting a model of behaviour that can be imitated by the patient.

Summary
. There are relatively few psychologists these days who would describe
themselves as ‘behaviourists’.
. Despite most psychologists rejecting the full Skinnerian message, the
techniques devised by Skinner are in wide use.
. Some researchers speculate that animals learn expectancies.
. There are various types of learning, of which operant conditioning is
only one. The range of therapeutic techniques reflects this diversity.

181
Chapter 4 Changing behaviour

5 The relevance of Skinner to today’s

world
By carefully examining behaviour and its consequences, it is possible to
identify common features of a number of social problems. Some forms
of behaviour that are problematic have an immediate consequence of
positive reinforcement. However, the aversive effects of these same
forms of behaviour arise long after the behaviour. This section will
provide examples of the conflicting nature of the immediate and long­
term consequences.

5.1 Addictions
The original meaning of the word ‘addiction’ was extreme dedication to
something (to ‘give over’) (Alexander and Schweighofer, 1988). Thus,
one could equally well be addicted to, say, praying, love or drugs. For a
period starting in the nineteenth century, the use became confined
largely to drugs, and addiction was seen as essentially a medical problem
needing a medical solution. Recently, there has been a swing back to
something nearer to the original meaning, with awareness that people
can be addicted to a whole range of things including gambling,
shopping, sex, exercise or junk foods, as well as to drugs (Alexander
and Schweighofer, 1988; Alexander, 2008). In the extreme, non-drug­
related addictions can be as destructive as the drug-related ones, leading
to bankruptcy and loss of job, health and family.
In each addiction, it appears that there is a strong and immediate
positively-reinforcing consequence of behaviour, whereas the unpleasant
consequences are long delayed. In order to try to treat addictions, it is
essential to understand how it is that people become addicted.
Let’s consider the example of smoking. Inhaling nicotine is instantly
reinforcing in the sense of encouraging the smoker to repeat the
behaviour. Nicotine arrives in the bloodstream within a second or so of
puffing the cigarette and exerts its effects on the brain. By contrast, the
negative consequences, such as coughing and breathing difficulties,
might arise long after the smoking habit is well established and might
not be linked closely to the act of smoking. Other consequences, such
as lung cancer and heart disease, might occur years into the future and,
at the moment, there is only the threat of them to deter smoking. Such
threats are often ineffective.

182
 5 The relevance of Skinner to today’s world

Gambling is another particularly interesting example from the point of

view of a Skinnerian analysis, since it fits very well with an
interpretation in terms of a special form of partial reinforcement. The
gambler keeps going even when there are only very occasional
reinforcements, which come at unpredictable times. By analogy, the rat
pressing a lever for food on such a schedule takes a long time to quit
when extinction is applied. The person addicted to junk food is under
the control of a reinforcement process that was highly effective in our
early evolution, when food was scarce and much effort needed to be
spent in obtaining it. These days, food is relatively easy to obtain.

Figure 4.10 A form of partial reinforcement?

5.2 Global survival

Today it is often argued that, for human life on earth to survive, people
need to change their ways. This means shifting from our reliance on oil
and implies changing the behaviour of much of the world’s population.
The affluent countries need to consume and pollute much less. How
can people be persuaded to do this? The present chapter describes a
possible contribution arising from psychology. In the context of global
survival, people can learn to change their behaviour.
Skinner was one of the world’s first ‘greens’, if not the first. In 1948, he
published what amounted to a green manifesto, outlining how a society
could be organised without waste or competition to provide for all. His
green manifesto (Skinner, 1948) was extraordinarily prescient. In its
1976 edition (p. xvi), he argued: ‘Not only can we not face the rest of
the world while consuming and polluting as we do, we cannot for long

183
Chapter 4 Changing behaviour

face ourselves while acknowledging the violence and chaos in which we

live.’
Consider the nature of the problem in terms of short-term
consequences, reinforcement and long-term consequences. For example,
cars are quick, comfortable and offer privacy and status (‘positive
reinforcement’). The benefits from car use are personal and immediate.
The costs in terms of climate change and loss of public transport
(‘punishment’) are long-term and diffuse rather than immediate and
personal. By contrast, the use of energy-efficient means of transport
such as walking, bicycles, trains and buses involves immediate exposure
to the elements (‘punishment’). The costs in terms of discomfort are
immediate and personal, whereas the benefits to the environment are
diffuse and long-term. A look at today’s problems suggests the value of
considering the Skinnerian position. For example, regular appeals are
made to change our ways. They usually involve the assumption that first
people’s minds need to be changed and then they will change their
behaviour appropriately. People are urged to think of the future and
therefore to consume less energy. Money is spent on promotion
campaigns urging us to consume less. However, evidence shows that
simply appealing to people to change their ways is relatively ineffective
(Katzev and Johnson, 1987).
Someone adopting a Skinnerian perspective would urge that the solution
is to change the nature of the reinforcers in use. Some of the
behaviours that need to be reduced are inherently positively reinforcing
and so they need to be made less so. According to a Skinnerian
perspective, the situation is analogous to that of the rat in the Skinner
box. In this case, the government, local council or supermarket chain
decides what behaviour will be reinforced and what will be the
reinforcer. For example, the local council might decide that people will
be offered tokens for recycling rubbish. The assumption is that this will
increase the amount of recycling. Some supermarkets already offer a
small discount for reusing bags. This is analogous to the experimenter
deciding the nature of the reinforcer in the Skinner box (for example,
the size of the pellet or whether to use a partial reinforcement
schedule).

184
 5 The relevance of Skinner to today’s world

Box 4.2 A problem with the Skinnerian solution

Skinner’s solution to environmental problems involved extrapolating
findings from the Skinner box and applying them to the complex
issue of environmental change. However, there is a problem with
this extrapolation when it comes to changing behaviour on a global
scale. For the rat in the Skinner box or the prisoner who earns
tokens for good behaviour, it is clear who holds the power to give or
withhold reinforcements and what their intentions are. Similarly,
when consumers are encouraged to recycle more, it is the local
council, the supermarkets or the government that have the power to
reinforce the desired behaviour. In the case of global issues,
however, things are less straightforward. While there is no doubt
that the world as a whole needs to pollute less, or that affluent
countries need to reduce their energy consumption, it is far less
clear who holds the necessary power to offer adequate
reinforcements and shape behaviour to achieve the desired effect.
In other words, if the world is a Skinner box and we are all rats,
who is Skinner? One might argue, therefore, that Skinner’s global
green project does not take adequate account of the power
relations required for global behaviour shaping.

Figure 4.11 Rewards for recycling: reinforcement of environmentally

responsible behaviour

185
Chapter 4 Changing behaviour

5.3 Concluding remark

A general consensus is that behaviour is under a range of controls and
operant conditioning is just one form of learning. In Chapter 3 you
were introduced to social learning and the role of observation and
imitation. You are not expected to decide which is the correct way of
explaining behaviour – in terms of observation or because of
behaviour’s consequences. Both processes exist side by side. Indeed,
behaviour might sometimes occur as a result of imitation and be
strengthened still more by its consequences. Furthermore, many
psychologists would argue that it is not sensible to put operant
conditioning at centre stage. However, it is a control that might well
have been underestimated, and a consideration of Skinner’s approach
could very usefully inform our day-to-day lives as well as public policy
making.

Summary
. Much behaviour that is problematic can be characterised by
immediate reinforcement followed by long-delayed aversive
consequences.
. Addictions, overconsumption and ecologically undesirable behaviours
fit this pattern.
. Skinner urged a reform of society so as to introduce reinforcement
for environmentally responsible behaviour.

186
 References

References
Alexander, B.K. (2008) The Globalization of Addiction: A Study in Poverty of the
Spirit, Oxford, Oxford University Press.
Alexander, B.K. and Schweighofer, A.R.F. (1988) ‘Defining “addiction”’,
Canadian Psychology, vol. 29, no. 2, pp. 151–62.
Bolles, R.C. (1972) ‘Reinforcement, expectancy, and learning’, Psychological
Review, vol. 79, no. 5, pp. 394–409.
Flora, S.R. (2004) The Power of Reinforcement, Albany, NY, State University of
New York Press.
Gray, J.A. (1987) The Psychology of Fear and Stress, Cambridge, Cambridge
University Press.
Guttman, N. (1977) ‘On Skinner and Hull: a reminiscence and projection’,
American Psychologist, vol. 32, no. 5, pp. 321–8.
Katzev, R.D. and Johnson, T.R. (1987) Promoting Energy Conservation: An Analysis
of Behavioral Research, Boulder, CO, Westview Press.
Lane, J.D. (1992) ‘Neurochemical changes associated with the action of acute
administration of diazepam in reversing the behavioural paradigm conditioned
emotional responses (CER)’ Neurochemical Research, vol. 17, no. 5, pp. 497–507
Pavlov, I.P. (1927) Conditioned reflexes: An Investigation of the Physiological
Activity of the Cerebral Cortex, Oxford, Oxford University Press.
Siegel, S. (2005) ‘Drug tolerance, drug addiction, and drug anticipation’, Current
Directions in Psychological Science, vol. 14, no. 6, pp. 296–300.
Skinner, B.F. (1939) The Behavior of Organisms: An Experimental Analysis, New
York, NY, Appleton-Century-Crofts.
Skinner, B.F. (1953) Science and Human Behavior, New York, NY, The Free Press.
Skinner, B.F. (1976 [1948]) Walden Two, New York, NY, Macmillan Publishing.
Skinner, B.F. (1979) The Shaping of a Behaviorist, New York, NY, Alfred A.
Knopf.
Skinner, B.F. (1971) Beyond Freedom and Dignity, New York, NY, Knopf.
Swinson, J. and Harrop, A. (2005) ‘An examination of the effects of a short
course aimed at enabling teachers in infant, junior and secondary schools to
alter the verbal feedback given to their pupils’, Educational Studies, vol. 31, no.
2, pp. 115–29.
Thorndike, E.L. (1898) Animal Intelligence: An Experimental Study of the Associative
Processes in Animals, New York, NY, Columbia University Press.
Watson, J.B. (1913) ‘Psychology as the behaviourist views it’, Psychological Review,
vol. 20, no. 2, pp. 158–77.

187
Chapter 4 Changing behaviour

Watson, J.B. and Rayner, R. (1920) ‘Conditioned emotional reactions’, Journal of

Experimental Psychology, vol. 3, no. 1, pp. 1–14.

188
Chapter 5
Determined to love?
Deborah Custance
Contents
1 Introduction 193
1.1 What is attachment and where does it come from? 193
1.2 Cupboard love or love in the blood? 196
1.3 Love at first sight: imprinting 198
2 Key study: Harlow’s monkeys 201
2.1 Researching animals 203
2.2 Wire versus terry-cloth mothers 205
2.3 Monkey maternal deprivation and abuse 208
3 Evaluating Harlow’s work 211
4 Since Harlow 215
4.1 Sensitive responsiveness in Glasgow 215
4.2 Ainsworth and the Strange situation 216
4.3 Evaluating the use of attachment categories 220
4.4 Strange Situation: research on other animals 221
5 The flexibility of attachment 223
5.1 Is there a critical period in human attachment? 224
5.2 The maternal deprivation hypothesis 225
5.3 Does early attachment have a long-term effect? 227
5.4 Conclusion 229
References 230
Chapter 5 Determined to love?

Aims and objectives

After studying this chapter you should be able to:
. summarise some of the most influential findings of psychological
research on attachment in humans and other species
. describe Harlow’s maternal deprivation experiments with rhesus
macaques and explain their importance with respect to attachment
theory
. outline some of the ethical implications and the scientific value of
using non-human animals in psychological research
. describe Mary Ainsworth’s observational research on children’s
attachment
. recognise that behaviour based upon a biologically inherited system
is not necessarily rigid and determined.

192
 1 Introduction

1 Introduction
The 1989 United Nations’ Convention on the Rights of the Child
declared that: ‘[T]he child, for the full and harmonious development of
his or her personality, should grow up in a family environment, in an
atmosphere of happiness, love and understanding’ (UN, 1989).
The 1959 United Nations Declaration of the Rights of the Child also
stated that, ‘a child of tender years shall not, save in exceptional
circumstances, be separated from his mother’ (UN, 1959).
Is love a basic human right? Can we really dictate love? Is mother-love
especially important for the healthy development of young children?
What is this thing called ‘love’ anyway? It is certainly an extremely
powerful emotion. It is an emotion that draws and bonds us to another
person. We yearn for the presence of a loved one, they preoccupy our
thoughts and we turn to them in times of joy and sorrow. Yet, why do
we love? Is it for the benefit of the other person or is it driven by
selfish motivations? Where does love come from? Do we learn it or are
we born with the capacity to love?
In this chapter, I will explore the nature of love, in particular the love
between a child and its mother. The bond that develops between the
child and the mother, or other primary caregiver, has been called
attachment. I will consider the major scientific theories of attachment Attachment
and some of the most influential studies that have been conducted on A relatively long-term,
the subject. emotionally important
relationship in which
one individual seeks
1.1 What is attachment and where does it come proximity to and
derives security and
from? comfort from the
presence of another.
It seems obvious that all children and their parents should love each
other. They should both derive pleasure and satisfaction from cuddling
and playing together. For a parent, the sometimes overwhelming feelings
of protective love may either come almost instantaneously with the birth
of their child or grow over the first few months of the child’s life.
Most human infants begin to show a preference for one person over
and above everyone else at approximately 7 months of age. If they are
separated from that special person (who is usually, but not always, their
mother) they tend to become very distressed and will do all they can to
be reunited with them. Soon after establishing this primary attachment,
children begin to extend their affections to an inner circle of special

193
Chapter 5 Determined to love?

people, which usually includes other family members, close family

friends, nannies or child minders.
This chapter focuses mostly upon the close emotional bond that
develops between mothers and infants. This is because much research
has focused upon mother–infant attachment. Yet you should bear in
mind that there are lots of different types of attachment: father to child,
grandparent to grandchild, romantic relationships, the dog–human
bond, and even the strong emotional ties that some children form in
response to certain objects such as soft toys or cloths have all been
discussed in terms of attachment. Thus, even though I will focus
primarily on the mother–infant bond, attachment refers to any relatively
long-term emotionally important relationship in which one individual
seeks proximity to and derives security and comfort from the presence
of another.
John Bowlby was the first scientist to use the term ‘attachment’ to refer
to the mutually affectionate bond that develops between a child and his
or her caregivers. In the 1950s Bowlby proposed a theory to explain
how and why attachment relationships develop. In order to appreciate
how radical Bowlby’s ideas were when he first proposed them, it is
necessary to consider the alternative theories that existed at the time.

Activity 5.1
Without reading ahead, spend a couple of minutes considering the
following questions:
. Why do you think children and their caregivers form close, mutually
affectionate relationships?
. Why does a mother or father love their child?
. Why does the child form such strong feelings for his or her parents?

Initially, the questions in Activity 5.1 seem obvious, even easy, don’t
they? However, don’t be surprised if your mind went completely blank
and you couldn’t think of any reasons for why children and parents love
each other. Obvious questions are often devilishly difficult to answer.
Most people take it for granted that parents and infants love each other
and they rarely stop to question why.

194
 1 Introduction

Figure 5.1 Different types of attachment

195
Chapter 5 Determined to love?

1.2 Cupboard love or love in the blood?

Before Bowlby published his work on attachment, there were two main
theories that purported to explain the bond that develops between
infants and their parents: the psychoanalytic theory, and the behaviourist
or learning theory.
You read about Psychoanalytical theorists, such as Sigmund Freud, suggested that we
psychoanalysis in love our parents because they satisfy our basic biological needs. Thus,
Chapter 1, Section 1. as babies, we learn to love whoever feeds us, provides warmth by
holding or cuddling us and alleviates discomfort by changing our
nappies and burping us. According to Freud, infants become distressed
when they are separated from their primary attachment figure because
they are frightened that their basic needs will no longer be satisfied.
Bowlby labelled accounts such as this ‘cupboard-love theories’
(Holmes, 1993).
You learnt about The behaviourist approach to parent–infant bonding is also based upon
behaviourism in cupboard love. Babies love their parents because they have learned to
Chapter 4. associate them with positively rewarding experiences such as being fed
and cuddled. Food, warmth and the alleviation of discomfort are
unconditional stimuli in the sense that children find these things
immediately or intrinsically rewarding. In contrast, infants do not
immediately value their caregivers. Instead, they learn to value them,
especially their mother if they are being breastfed, because she is
habitually associated with the primary reinforcers of food and comfort.
Thus, the infant’s mother constitutes a conditional stimulus.
In many ways psychoanalytic and behaviourist accounts of infant
attachment seem intuitively correct. It seems obvious that a child learns
to love its parents because they satisfy its basic biological needs. Yet,
according to Bowlby’s account, cupboard love does not form the basis
of attachment.
Bowlby’s work on attachment was influenced by Konrad Lorenz and
Niko Tinbergen, two scientists who are widely acknowledged as the
Ethology founding fathers of ethology, namely the study of animal behaviour
The study of animal under natural conditions. In the first half of the twentieth century,
behaviour under natural Lorenz and Tinbergen studied wild birds and noticed that the chicks of
conditions.
some species seemed to possess certain behavioural responses right
from the very moment of hatching rather than having to learn them.
For instance, Tinbergen observed that newly hatched herring gull chicks
would peck at a red spot on the lower mandible of their parents’ beaks.

196
 1 Introduction

In response to this pecking, gull parents regurgitate partially digested

fish for their chick to eat. Tinbergen suspected that the chick’s pecking
response was inborn or innate. To explore the phenomenon further, Innate
Tinbergen collected newly hatched chicks that had never had the Relating to a behaviour,
opportunity to peck at their parents’ beaks and waved various objects in disposition or
characteristic that is
front of them, counting how many times they pecked. He discovered
present from birth.
that they would peck not only at naturalistic adult gull beaks, but also at
objects that moved gently up and down and contrasted with foreground
and background colours. Thus, herring gull chicks seem to be born with
a crude, but nonetheless effective programme for pecking at certain
types of stimuli. Based upon these observations, Tinbergen argued that
the pecking reflex was an evolved predisposition. To understand what Evolved
Tinbergen meant by this, we need to know a little bit about Charles predisposition
Darwin’s theory of evolution by natural selection. A behavioural tendency
that increases an
Darwin argued that biological organisms have evolved or changed over organism’s chances of
time. He proposed a mechanism called natural selection by which this survival and
change might take place. The individual members of all species differ reproductive success.
from one another in various ways. They may be slightly different sizes
or heights, they may be different colours, some may possess slightly
better eyesight or hearing … and so on. Some of these variations may
increase an individual’s chances of survival or their chances of finding a
mate and producing lots of healthy offspring. If an advantageous trait
can be biologically passed on to an organism’s offspring, then more
individuals in the next generation will possess that particular trait and
they will then pass it on to their offspring. Some inherited traits may be
primarily physiological, yet research suggests that a number of them
involve behavioural tendencies that are also capable of increasing an
organism’s chances of survival and reproductive success. Presumably,
possessing from the moment of hatching the tendency to peck at stimuli
with certain features greatly increases a herring gull chick’s chances of
survival.
Influenced by evolutionary theory and ethological research, Bowlby put
forward the radical proposition that human infants possess inbuilt or
innate tendencies that lead them to forge emotionally powerful ties to
stimuli with certain properties (Bowlby, 1979). He argued that these
tendencies are the product of the evolutionary pressure for survival. For
the infant, the need to bond with a caregiver is a matter of personal
survival. The young of many species, including humans, are born weak
and helpless. Without the protection of a diligent caregiver, they would
starve, freeze or be eaten by a predator. It is therefore vital that they

197
Chapter 5 Determined to love?

elicit loving care from and maintain proximity to an adult who can
provide them with food, warmth and protection. Thus Bowlby argued
that infants possess certain inbuilt mechanisms for encouraging care­
giving responses from their parents. Infants cry, particularly if they are
separated from their parents, and smile to indicate pleasure during an
interaction. Bowlby (1979, p. 37) observed: ‘Babies’ smiles are powerful
things, leaving mothers spellbound and enslaved. Who can doubt that
the baby who most readily rewards his mother with a smile is the one
who is best loved and best cared for?’
But what is in it for the parents? Obviously, infants do not provide
food and safety. So why should the parents be primed to be
psychologically manipulated by their children in this way? From an
evolutionary perspective, children constitute the tickets to the parents’
biological immortality or, to put it less romantically, long-term genetic
survival. Individuals can cheat death to a certain extent by leaving a
biological legacy in the form of copies of their genes that are passed on
through the generations. This has led them to develop an innate
predisposition to care for the young.
Although both parents and infants are likely to possess inbuilt
predispositions to bond to one another, Bowlby mostly focused on
attachment from the point of view of the child. He proposed that infant
attachment is based upon three main factors: (a) an inbuilt
predisposition to enjoy social interaction; (b) an innate bias to become
attached to things with certain properties, and (c) an inbuilt fear of the
unknown and unfamiliar. Notice that none of these factors involve
food: this is not cupboard love.

1.3 Love at first sight: imprinting

One of the major influences upon Bowlby’s ideas was, perhaps rather
surprisingly, the behaviour of geese. The Austrian ethologist Konrad
Lorenz was fascinated by the tendency of newly hatched goslings to
follow their mother wherever she went. In order to test the
phenomenon further, he raised a clutch of goose eggs. Practically the
first thing the goslings saw upon hatching was Lorenz, or, to be more
precise, his wellington boots. As you can see in the delightful
photograph (see Figure 5.2), the young birds would follow Lorenz in his
wellington boots wherever he went.

198
 1 Introduction

Figure 5.2 Konrad Lorenz followed by a small flock of imprinted goslings

Lorenz called this process of bonding imprinting (Lorenz and Kickert, Imprinting
1981; Hess, 1958). Several species of bird and even one species of An innate system that
mammal, the shrew, imprint. As soon as the babies are mobile, they allows rapid learning to
occur in animals
tend to follow anything that moves. Under natural conditions, the thing
immediately after birth.
that is most likely to be moving about in the vicinity of the youngster is It involves developing
its parent. However, in laboratory tests with baby birds, they have attachment to a specific
become imprinted upon a toy duck on wheels, yellow rubber gloves and individual or object.
even a geometrical shape moving back and forth behind a plate-glass
window. If newly hatched ducklings follow a moving object for about
ten minutes, imprinting occurs. They will then follow it wherever it goes
and cry piteously if they become separated from the object of their
affection. Early laboratory studies have shown that imprinting occurs
during a specific period in a young bird’s life. Eckhard Hess (1958)
found that if ducklings do not become exposed to a moving object
within the first three days of their life, after that point instead of
imprinting they become frightened and avoid all unfamiliar stimuli.
It would seem clear that imprinting is an innate or evolved system. Yet
ducks and geese are very different from humans. When Bowlby first
suggested that humans might also possess similar inbuilt tendencies to
bond, the scientific community was highly sceptical. He needed more
convincing evidence. That evidence would come from a species closely
related to humans – monkeys.

199
Chapter 5 Determined to love?

Summary
. Psychologists have been interested in attachment, in particular the
bond between a baby and its primary carer.
. Important insight into the study of attachment came from
ethological studies of animal behaviour.
. Drawing on ethological research and evolutionary theory, John
Bowlby challenged the behaviourist and psychoanalytic theories of
attachment, arguing that infants possess innate tendencies that lead
them to forge emotionally powerful relationships with primary
caregivers.

200
 2 Key study: Harlow’s monkeys

2 Key study: Harlow’s monkeys

Harry Harlow was an American psychologist who investigated whether
infants bond with their mother because of cupboard love (i.e. the fact
that their mother provides them with food) or, as Bowlby suggested, an
inbuilt tendency to become attached to stimuli that possess certain
properties (such as being warm and soft to the touch). The problem is
that, under normal circumstances, mothers simultaneously provide food
and tactile comfort for their babies. An observational study alone could
not separate out the influence of these two factors. To test it
experimentally, one group of mothers would have to be prevented from
feeding their babies, while another group would have to be prevented
from holding and cuddling them. Although, for reasons of ethics, such You read about ethics
an experiment is impossible with humans, he could conduct it with in Section 3 of
monkeys. Chapter 2 and
Section 3.2 of
Some of the most influential work on attachment was conducted by Chapter 3.
Harlow on a group of rhesus macaques (Macaca mulatta), a medium­
sized monkey with light brown fur and a pink face. In the wild, they are
found throughout mainland Asia. They are reasonably easy to keep and
breed in captivity and are often used in research.
It is interesting that Harlow came to study attachment in rhesus
monkeys almost by accident. He had hoped to study intelligence in rats,
but when he took up employment at the University of Wisconsin he
discovered, to his great consternation, that the rat laboratory had been
dismantled. However, when he started observing monkeys at the local
zoo, he quickly realised that their intellectual abilities outstripped those
of rats. Harlow secured funding for a primate laboratory at the
university and set out to study intelligence in monkeys.
One of the problems with the wild-caught monkeys that were used in
the laboratory was that they often carried disease. As a result the baby
monkeys were separated from their mothers and housed in separate
cages on their own. During daily cage cleaning Harlow noticed that the
babies would protest as the sanitary pads that lined their cages were
removed. This serendipitous observation intrigued Harlow. Why does
the baby form an attachment to a sanitary pad? To answer this question
Harlow embarked on the famous (some would say infamous) research
into monkey attachment.

201
Chapter 5 Determined to love?

Figure 5.3 Harry Harlow (1905–1981)

Harlow suspected that the infants’ affection for the pads was primarily
based upon ‘contact comfort’. Despite the fact that the baby monkeys
had all their physical needs catered for in terms of food, water and
shelter, they seemed to be bonding with the only soft object in their
otherwise hard and harsh environment. In fact, Harlow observed that,
‘a baby monkey raised on a bare wire-mesh cage floor survives with
difficulty, if at all, during the first five days of life’ (Harlow 1958,
p. 675). Harlow hypothesised that the tactile qualities of stimuli were
more important for infant monkey bonding than the provision of food.
He designed a series of experiments to test this idea further, which will
be described later in Section 2.2.

202
 2 Key study: Harlow’s monkeys

Figure 5.4 Contact comfort

2.1 Researching animals

Activity 5.2
Take a couple of minutes to think about the advantages of using
monkeys to study attachment. Draw up a list of pros and cons for
studying monkeys rather than humans. What can we do and learn by
studying monkeys that would not be possible by studying humans alone?

One advantage of studying monkeys rather than humans is that

scientists can conduct experiments with non-human subjects that for You read about
ethical reasons would not be permitted with human participants. experiments in
Chapters 2 and 3.
A possible disadvantage of using non-human animals in place of
humans is that you can never be completely sure of the degree to which
the findings are applicable to humans. Humans and monkeys are
genetically very similar to one another (rhesus macaques and humans
share 94 per cent of their DNA). But just because monkeys exhibit a
certain pattern of behaviour, it does not necessarily mean that humans
do so as well. You can read more about the issue of cross-species
comparisons in Box 5.1.

203
Chapter 5 Determined to love?

Box 5.1 Why do it this way?

Why do psychologists study monkeys or other animals, if they are
ultimately interested in human psychology?
Non-human animals are often used in psychological research as a
substitute for humans. In Harlow’s case, ethical considerations
prohibited him from conducting his experiments on human infants.
Had he somehow been able to conduct his experiments on humans,
he would no doubt later have been prosecuted for his actions.
However, he enjoyed a career in which he spent over 20 years
socially isolating monkeys in various different ways. As you will see
later, since Harlow completed his research, psychologists have
become increasingly aware of the importance of ethics in animal
research and guidelines have been put in place to protect animal
welfare.
Another advantage of using non-human animals in research is that
they are often thought to provide a simpler model of behaviour than
that found in human beings. Humans are extremely complex
creatures and it can be very difficult to disentangle all the different
factors that influence our behaviour. Non-human animals are often
less sophisticated and their behavioural responses can be easier to
interpret.
Monkeys and apes are often preferred in psychological research on
animals. Human beings are genetically very similar to other species
of primate. We share approximately 98.8 per cent of our DNA with
chimpanzees and 94 per cent with rhesus macaques. Yet one must
be very careful how one interprets this genetic similarity. We share
50 per cent of our DNA with a banana. A small difference in DNA
can make a huge difference in a species’ anatomy and behaviour.
For one thing, despite a 1.2 per cent difference in our DNA, humans
have a brain three times the size of a chimpanzee’s.

204
 2 Key study: Harlow’s monkeys

So, rather than making simple comparisons between species,

animal researchers look for general behavioural principles that can
then be used to draw inferences about human behaviour. For
example, the fact that birds show imprinting does not provide very
convincing evidence that human bonding is based on similar
mechanisms. However, if one looks across many species – from
those genetically distant from humans up to our closest non-human
relatives – and finds bonding processes similar to those exhibited
by humans, the idea of common underlying mechanisms becomes
more convincing.

2.2 Wire versus terry-cloth mothers

To test the idea of ‘contact comfort’, Harlow (1958) constructed two
types of surrogate monkey mother (See Figures 5.3 and 5.5). The first
was called the terry-cloth mother. It was made from a block of wood
covered in a layer of sponge rubber and sheathed in a layer of towelling
(or terry-cloth as it is called in America). A light bulb behind the terry­
cloth mother provided a source of gentle heat. A bottle could be
inserted through the cloth mother’s body, from which the baby could
feed. The second type of mother was made of a cylinder of wire mesh
that was adequate for clinging on to, but lacked any contact comfort. It
too was heated and could have a feeding bottle inserted into it. Harlow
placed these mothers in separate cubicles attached to the babies’ living
quarters. He tested eight newborn monkeys. For four of the monkeys the
terry-cloth mother provided milk, while the wire mother did not, and for
the remaining four monkeys the wire mother provided milk while the
cloth mother did not. In this way, Harlow could directly compare the
variables of contact comfort and food provision.
To describe Harlow’s experiment in more formal scientific terms, the
independent variable (i.e. the aspect of the experiment that was You were introduced to
manipulated) was the type of mother presented to the babies (i.e. wire independent and
or terry-cloth). The dependent variable (i.e. the aspect that Harlow dependent variables in
Section 2.3 of
measured and expected to be affected by the independent variable) was
Chapter 3.
the degree of bonding as revealed by how much time the monkeys
spent clinging to each of the surrogate mothers and the amount they
cried when either of them was removed.
The results of Harlow’s experiment were unequivocal. The babies
showed a clear preference for the terry-cloth mother over the wire

205
Chapter 5 Determined to love?

mother regardless of which of them produced milk. If the wire mother

provided milk, the babies would cling to the terry-cloth mother and
only approach the wire mother to feed, returning to the terry-cloth
mother immediately afterwards. If the wire mother was removed from
the cage, the babies hardly reacted. If the terry-cloth mother was
removed, they became highly distressed.

Figure 5.5 A baby monkey clings to its terry-cloth mother. The wire mother
with its feeding bottle can be seen in the background

Harlow’s findings appeared to confirm John Bowlby’s contention that

the infant–parent bond is based on innate tendencies rather than
cupboard love. Harlow’s systematic and controlled experiment showed
that contact comfort was more important than food in the formation of
attachment. The fact that the babies became upset only when the terry­
cloth mother was removed and showed little concern in response to the

206
 2 Key study: Harlow’s monkeys

removal of the wire mother provides strong evidence that they were
bonded to the former, but not the latter.
Harlow confirmed these results by running three more experiments.
Conducting just one experiment in which the babies were presented
with a wire mother who provided food and a warm soft terry-cloth
mother who provided comfort would probably not have been sufficient
in and of itself to convince people. Since both of the artificial mothers
were placed adjacent to each other in each baby’s cage, it would not be
very surprising if they chose to enjoy the best of both worlds – clinging
to the terry-cloth mother for comfort and popping over to the wire
mother when they were hungry. One becomes more convinced with the
subsequent experimental manipulations.
In one experiment, Harlow separated the babies from their mothers and
placed them into a cubicle where they could lift a hatch to look at
either their terry-cloth mother, the wire mother, or a bare cage. The
babies lifted the hatch most to look at the terry-cloth mother and
looked at the wire mother no more often than the bare cage.
In another experiment, he showed that the babies treated the terry-cloth
mother as a ‘safe base’. The concept of a safe base is important in
infant–mother attachment. Mothers (or other important caregivers) are
thought to provide a safe base, allowing infants to go off and explore
or play, but also to rush back to if they feel threatened. Harlow found
that when the babies were placed in a large room full of toys they
would curl up in a terrified ball if there was no mother or just the wire
mother present (Figure 5.6). However, if the terry-cloth mother were
present they would run and cling on to her and eventually pluck up the
courage to explore, returning to her on a regular basis (Harlow, 1958).
Harlow’s experiments made it absolutely clear that a baby monkey’s
bond with its mother was not based on cupboard love. Even when the
cloth mother provided no food, the babies showed a clear preference
for her over a lactating wire mother. Baby monkeys seem to possess a
predisposition to bond to objects with certain tactile qualities regardless
of their ability to provide food.

207
Chapter 5 Determined to love?

Figure 5.6 The response of a baby monkey placed in an unfamiliar

playroom without the terry-cloth mother

2.3 Monkey maternal deprivation and abuse

Harlow continued studying attachment in monkeys for the next 20
years. The experiments became increasingly dark and disturbing. He
became interested in what it would take to break the bond. Do infants
remain attached even when their caregiver is abusive? Certainly,
observations of abused human children suggest that they will often
continue to cry for and seek proximity to an abusive parent. Harlow
decided to study the phenomenon by constructing a truly vicious
surrogate mother which he called ‘The Iron Maiden’ (Blum, 1994). The
Iron Maiden blew pressurised cold air out of its body so violently that
it would throw the baby monkey against the bars of its cage. In
addition, at random intervals a prod would eject to push the baby
monkey away from the Maiden’s body. Despite this abuse, the baby
monkeys continued to cling and maintain proximity to their monstrously
abusive surrogate mother.
Harlow also began to raise baby monkeys under conditions of partial
and total social isolation (Harlow, 1962; Harlow et al., 1965). Partially

208
 2 Key study: Harlow’s monkeys

isolated monkeys were kept in separate cages, but could see, hear and
smell one another. Totally isolated monkeys were raised in permanently
lit, solid-walled chambers where they had no visual or physical access to
the outside world or to any other living being (either monkey or
human). The monkeys were placed in the chamber a few hours after
birth and kept under these conditions for three, six, twelve or twenty­
four months.
Both partially and totally isolated monkeys exhibited severely disturbed
behaviour. They would compulsively suck their inner cheeks and tongue,
perform repetitive stereotyped behaviour (such as rocking or pacing
back and forth), stare blankly into space as if detached from their
environment, and they would violently attack and bite their own bodies.
When they were eventually allowed access to other monkeys, they were
often overly aggressive and proved unable to form normal social
relationships.
When the isolation-reared monkeys grew up, they developed into highly
socially disturbed adults. The males were often unable to initiate mating
and the females aggressively rebutted the sexual attention of males. If
forced to mate, and if the females became pregnant and gave birth, they
were most often neglectful mothers, but some proved to be horribly
abusive.
Significantly, some of these damaging effects of isolation were shown to
be at least partially reversible. For example, when 6-month-old totally
isolated monkeys were placed with 3-month-old mother-reared infants,
over time they exhibited a marked improvement in their behaviour. So
much so that Harlow described the 3-month-old companions as
‘monkey psychiatrists’ (Suomi et al., 1972). Also, although 75 per cent
of isolation-reared monkeys proved to be inadequate mothers upon
producing their first baby (compared to only 5 per cent of females
raised by their biological mother), they tended to exhibit improved
mothering skills with the production of each subsequent infant
(Suomi, 1978).

209
Chapter 5 Determined to love?

Summary
. Psychologists conduct research on animals to study patterns of
behaviour that are thought to have developed through evolution.
. Because of the evolutionary proximity between humans and non­
human primates, psychologists have used monkeys to study the
evolutionary basis of human behaviour.
. Harry Harlow’s experiments on monkeys provided important
evidence for Bowlby’s claim that infants tend to bond with soft and
warm objects that provide tactile comfort, rather than with those
that provide food.

210
 3 Evaluating Harlow’s work

3 Evaluating Harlow’s work

Many people are shocked by Harlow’s research. Indeed, he is one of the
most controversial figures in the history of psychological science. His
experiments on monkeys seem cruel, even sadistic. Nowadays, the worst
excesses of Harlow’s research would not be permitted. As discussed in
Box 5.2, modern psychological research, on both humans and other
animals, is subject to rigorous ethical scrutiny.

Box 5.2 Ethics of animal research

Often the scientifically best way to study something is far from the
most ethical. For instance, the best way to study attachment in
humans is not to test monkeys, but to experiment on children.
However, it would be entirely unethical to conduct Harlow’s maternal
deprivation or abuse experiments using humans. This raises the
question: is it ethical to use non-human animals in such research?
Animals are used in a wide array of psychological research, from
observing species under wild conditions to conducting vivisection
upon animals in laboratories. When testing humans the welfare of
participants is paramount and no research should be allowed if it is
at all likely to cause long-term discomfort or distress. The same
high standards of protection are not applied to non-human animals.
Nevertheless, there are strict guidelines.
The British Psychological Society has published guidelines for
psychologists working with animals, according to which
psychologists must be able to demonstrate that the benefits of a
study justify the costs to the animals in terms of suffering. Also,
researchers must do all they can to minimise the likely suffering
caused. Ethics committees, comprised of scientific experts,
university officials and responsible members of the general non­
academic community, evaluate written proposals and decide
whether or not a piece of research can proceed.
When Harlow embarked upon his research, none of these ethical
guidelines or procedures were in place. Harlow’s first experiments
with wire and terry-cloth mothers involved only eight monkeys. The
benefits of his findings in terms of improving our understanding of
attachment were profound. Thus, one could argue that although
those eight baby monkeys suffered, the suffering was justifiable
because human society benefited greatly from the knowledge
gained. The problem was that Harlow did not stop there and

211
Chapter 5 Determined to love?

eventually dozens if not hundreds of monkeys were subjected to

extreme levels of distress and physical harm.
Harlow’s experiments have been strongly criticised. Modern ethical
procedures would no doubt prohibit such research. Indeed, one of
Harlow’s colleagues, William Mason, commented that Harlow ‘kept
this going to the point where it was clear to many people that the
work was really violating ordinary sensibilities, that anybody with
respect for life or people would find this offensive’ (Blum, 1994,
p. 96). In some ways, the ethics guidelines are there to protect
animals from psychologists who don’t like animals. Harlow was one
such psychologist. He once said that the only thing he cared about
was ‘whether a monkey will turn out a property I can publish. I don’t
have any love for them. I never have. I don’t really like animals. I
despise cats. I hate dogs. How could you love monkeys?’ (Slater,
2004, p. 137).

Whatever one thinks of the ethics of Harlow’s research, it is undeniable

that the findings from his initial experiments added weight and impetus
to the implementation of some of the most profound changes in
Western child-rearing attitudes and practice. Before Harlow, the main
emphasis in human childcare was placed on providing for bodily needs.
Highly influential theorists, such as Freud and Watson, thought that
children should be treated with a rather cold unemotional detachment.
Freud claimed that maternal over-gratification of an infant’s needs could
be harmful. Similarly, in a book on child rearing, the founder of
behaviourism, J.B. Watson, advised: ‘Never hug and kiss [your children],
never let them sit in your lap. If you must, kiss them once on the
forehead when they say good night’. (Watson, 1928, p. 81–82).

212
 3 Evaluating Harlow’s work

He went on to say:

[W]on’t you then remember when you are tempted to pet your
child that mother love is a dangerous instrument – an instrument
which may inflict a never-healing wound, a wound which will make
infancy unhappy, adolescence a nightmare, an instrument which
may wreck your adult son and daughter’s vocational future and
their chances for marital happiness.
(Watson, 1928, p. 87)

Many child-rearing policies and practices reflected a philosophy of

emotional detachment. Just after birth, babies were briefly held up for
mothers to see and then whisked away from the maternity ward to be
placed straight into cribs or incubators. Mothers were encouraged not to
overindulge their babies and to establish strict feeding and sleep
routines. Mothers were also advised that it was a very bad idea to pick
their babies up on demand, that in fact they would ‘spoil’ them if they
picked them up too often. Breastfeeding was portrayed as inferior to
bottle-feeding. If a baby or older infant had to be admitted into
hospital, parental access was limited to restricted visiting hours. A strict
and unemotional upbringing was thought to toughen children up and
prepare them for the rigours of adult life.
All of these attitudes and practices have been challenged and to a great
extent overturned since the 1950s when Harlow first embarked on his
attachment research. At the time, John Bowlby was at the forefront of
the campaign to change public policy on child-rearing practices. Bowlby
penned the United Nation’s Declaration on the Rights of the Child in
1959 (see the quotes at the start of this chapter). However, when he
insisted that children had the right to love and that they should not be
removed from their mother other than in exceptional circumstances,
many people remained unconvinced. At that stage, Bowlby had very
little scientific evidence to back up his claims. Harlow’s studies are
important because they provided Bowlby with some much-needed
experimental support.

213
Chapter 5 Determined to love?

Summary
. Harlow’s work on the effects of maternal deprivation on behaviour
in monkeys proved controversial as it violated the welfare of
animals.
. Psychological work on animals is today regulated by strict guidelines
issued by professional bodies such as the British Psychological
Society.
. Harlow’s original studies on attachment in monkeys contributed to
important changes in the way society views the relationship between
infants and their primary carer.

214
 4 Since Harlow

4 Since Harlow
Although Harlow’s experiments provided compelling evidence that
bonding in monkeys was based on contact comfort rather than
cupboard love, it does not necessarily follow that the same is true of
humans. One must always bear in mind that, despite the similarities,
monkeys are not humans. Scientific evidence on attachment in humans
was sorely needed. Bowlby and other child researchers set about
designing studies that could reveal the nature of human attachment.
Unlike Harlow, they could not remove human babies and rear them
under different conditions in ways that might damage their emotional
and psychological well-being. Thus, attachment research in humans has,
to a great extent, relied on less intrusive observational studies and
questionnaires.

4.1 Sensitive responsiveness in Glasgow

Rudolph Schaffer and Peggy Emerson (1964) conducted a classic
observational study on the development of attachment in Glaswegian
babies. They visited the homes of sixty working-class Glaswegian
mothers every month for the first year of their babies’ lives and once
more at 18 months. They measured the strength of the attachment by
(a) how upset the babies became when they were temporarily separated
from their mother or other family members, and (b) how nervous they
seemed to be around strangers. Schaffer and Emerson found that
between birth and 3 months of age, the babies did not discriminate
between people. At 4 months of age, they began to show preferences
for certain people. At 7 months, they developed a clear preference for a
single attachment figure (who was often but not always their mother)
and many of them had developed a fear of strangers. Soon afterwards,
at about 9 months, the babies formed multiple attachments with a small
inner social circle, which often included the father, grandparents and
siblings.
Notably, Schaffer and Emerson found that human infant attachment did
not seem to be based upon cupboard love. Just over a third of the
babies (39 per cent) formed their first attachment with someone other
than the person who fed or changed them, such as a grandparent. The
most important factor in determining their primary attachment figure
was not who looked after their basic biological needs, but who
responded most sensitively to their signals. Thus, in support of

215
Chapter 5 Determined to love?

ethological theory, human infant attachment seemed to be based upon

what Schaffer and Emerson (1964) referred to as sensitive
Sensitive responsiveness.
responsiveness
Tendency in children to
bond with persons who 4.2 Ainsworth and the Strange Situation
respond most
sensitively to their No discussion of attachment theory would be complete without
signals. including the work of Mary Ainsworth. Ainsworth joined John Bowlby’s
research group at the Tavistock Clinic in London in 1950. Initially,
Ainsworth was far from convinced about the validity of Bowlby’s
ethological account of attachment. ‘It seemed obvious to her … that a
baby loves his mother because she satisfies his needs’ (Ainsworth,
personal communication in Bretherton, 1992). Yet, when she started
observing the interaction between mothers and babies in detail, the
relevance of Bowlby’s ideas became immediately apparent.
Ainsworth conducted her first observational study of attachment
in 1953. She had moved to Uganda with her husband and they lived
there for two years. During that time, she was able to raise a small
amount of money which allowed her to collect observational data from
twenty-six Ugandan families with unweaned babies between the ages of
1 and 24 months. She observed the families for two hours every
fortnight over a period of nine months.
Despite Freud and Watson’s insistence that ‘mollycoddling’ children is
detrimental to their development, Ainsworth found that the more
responsive the Ugandan mothers were to their babies’ signals, the less
the babies cried and the more confident and explorative they seemed to
be. If the mothers were unresponsive and emotionally detached, in the
way Watson had advocated, their infants tended to cry a lot more and
often seemed clingy and insecure. Thus Ainsworth was able to show
that, just as with Harlow’s terry-cloth mothers, human mothers seem to
act like a ‘safe base’, promoting independent exploration in their infants.
It might strike you as somewhat paradoxical that providing more
attention and affection would encourage a child to be independent. Yet
the idea is that if a child knows that they always have a safe base to
return to, this will give them the confidence to go off and explore. It is
a bit like bungee-jumping from a tall platform. If you are sure that your
bungee cord is dependable and securely attached, you are much more
likely to possess the confidence to jump.

216
 4 Since Harlow

Ainsworth conducted a similar study in the US and found that, just as

with the Ugandan families, maternal sensitivity appeared to promote
confidence and independence in children. For example, she found that
the infants of mothers who had provided extensive tender holding
during the first three months of life sought contact less often from 9 to
12 months of age. Moreover, when contact did occur, it seemed to be
more satisfying and affectionate. She concluded that, ‘an infant whose
mother’s responsiveness helps him to achieve his ends develops
confidence in his own ability to control what happens to him’ (Bell and
Ainsworth, 1972, p. 1188).
Ainsworth’s first two observational studies were extremely work­
intensive and time-consuming. However, based on this work, Ainsworth
developed a highly influential and much-used observational procedure
called the Strange Situation. This involves a simple, albeit somewhat Strange Situation
artificial situation which allows psychologists to study attachment in a A way of assessing
controlled environment and in a more ‘efficient’ (i.e. much less time­ attachment security in a
laboratory setting. It
consuming) manner.
consists of a
The Strange Situation has most often been used to assess infants standardised set of
between the ages of 12 and 24 months. The procedure consists of a episodes involving a
child, their mother and
series of seven consecutive three-minute-long episodes. a stranger in a
Episode 1. The mother and infant enter an unfamiliar room that sequence of separations
and reunions.
contains two chairs and a selection of toys. The infant is allowed to
explore the room while the mother sits and reads.
Episode 2. A stranger enters the room, chats briefly with the mother
and then attempts to interact with the child.
Episode 3. The mother leaves the room and the stranger attempts to
comfort or play with the child. This episode is cut short if the child
becomes very upset.
Episode 4. Reunion occurs. The mother re-enters the room, pauses for
a few seconds by the door to allow the child to initiate approach, and
then interacts with the child as normal. The stranger leaves the room.
Episode 5. The mother leaves the child alone in the room. This
episode is cut short if the child becomes very upset.
Episode 6. The stranger re-enters the room, pauses by the door and
then attempts to comfort and play with the child.
Episode 7. Reunion with the mother occurs and the stranger leaves.

217
Chapter 5 Determined to love?

Figure 5.7 Episodes 2–4 of the Strange Situation. A stranger enters the room and engages with the
child (2); the mother then leaves the child alone with the stranger (3); before returning to be reunited
with the child (4).

As you can imagine, some children became very upset when they were
separated from their mother in Episode 5. It would be completely
irresponsible to leave a young distressed child unsupervised in a room
that was unfamiliar to them. The Strange Situation procedure takes this
ethical issue into account. The children are constantly monitored either
through a one-way mirror or via a wall-mounted video camera that is
linked to a monitor in an adjacent room. Both the experimenter and
mother watch the children during the separation episodes. If the child
becomes too upset the episodes are cut short and the mother returns
before the three minutes are up.
Ainsworth proposed that the Strange Situation reveals different types of
attachment in children. Ainsworth and Bell (1970) classified the majority
of children (70 per cent) as secure. Secure children happily explored the
environment in the presence of their mother. They were upset by their
mother’s absence but they were quickly comforted upon reunion and
would often return to play. Fifteen per cent were classified as anxious­
resistant. These children would not explore even in the mother’s
presence. They were very upset on separation, but they acted in an
ambivalent manner upon reunion. They would often approach, but as
soon as their mother tried to hug them or pick them up, they would
pull away or struggle to be put down. They sometimes seemed angry
with their mother, even striking out at her. Neither mother nor child
appeared to derive much comfort or satisfaction from the reunion. The
final 15 per cent were categorised as anxious-avoidant. These children
seemed rather distant and aloof from the outset. They were not
particularly upset during separation and would snub their mother upon
her return by looking or moving away from her. Since Ainsworth’s
initial observations, a fourth category, disorganised, has been added.
Disorganised infants show signs of indecisiveness and confusion. They
often exhibit rather bizarre responses to separation, such as freezing,

218
 4 Since Harlow

rocking or hair-pulling. Disorganised attachment has been associated

with parental neglect or abuse.

Activity 5.3
Having read about the different categories of attachment, reflect on how
it might feel to be a parent with an infant who has just been observed in
this Strange Situation. How would you respond to your child’s attachment
being put into one of the four categories? Can you think of reasons why
a parent might object to this?

One problem with the Strange Situation research is that inferences are
often made about a child’s attachment style on the basis of a single
observation. In doing so, it may underestimate the importance of other
factors, such as the child’s mood, how well they had slept that day and
the extent to which they are used to situations similar to that to which
they were exposed in the study. The latter proved to be especially
relevant when the Strange Situation was used to test infants from
different countries and cultures. Although the majority tended to follow
a pattern similar to that found in the United States (i.e. 70 per cent
secure, 15 per cent anxious-resistant and 15 per cent anxious-avoidant),
there were some notable exceptions. These exceptions highlight why
some might object to the use of the Strange Situation to classify their
child’s attachment.
Israeli children raised in a residential community or settlement called a
kibbutz showed a much higher incidence (33 per cent) of what
Ainsworth called the anxious-resistant response (Fox, 1977). The
kibbutz children were raised together in large peer groups. They spent
long periods away from their parents under the care of a kibbutzim
nurse called a metapelet. The children were very used to being separated
from their mother, but very unused to seeing strangers.
Data from a study in Japan (Takahashi, 1986) showed an absence of
anxious-avoidant attachments and just as with the kibbutz children
about one-third were categorised as anxious-resistant. However, the
researchers pointed out that in normal circumstances the Japanese
babies were hardly ever left alone in a room by their mother. Most of
the babies became very upset during separation and on reunion. Despite
being instructed to stand motionless for ten seconds by the door, when
the mothers returned nearly all of them went straight over and picked

219
Chapter 5 Determined to love?

their baby up. Thus the Japanese babies had no opportunity to be

anxious-avoidant.
Perhaps the most surprising results came from Germany. Only 40 per
cent of the children were categorised as securely attached, while 49 per
cent of them showed an anxious-avoidant response (Grossman
et al., 1981). The high incidence of avoidant behaviour in German
children was explained in terms of the fact that German children are
encouraged to be independent from a very early age.
The Israeli and Japanese data bring into question the ecological
Ecological validity validity of the Strange Situation study for these cultures. Ecological
The extent to which a validity refers to the degree to which an experiment or controlled
study reflects naturally observational procedure reflects the everyday normal behavioural
occurring or everyday
tendencies of participants outside of testing conditions. Scientific studies
situations.
with low ecological validity are often uninformative or even misleading.
The Strange Situation seemed an inappropriate procedure to use in the
Japanese study since those mothers never left their babies in a room
alone or with a stranger. Thus the babies’ and mothers’ reactions were
to a set of circumstances that never occurred in their everyday lives.
Similarly, in their everyday lives the kibbutzim children were very
unused to seeing strangers; hence this aspect of the Strange Situation
also lacked ecological validity for them. Ainsworth developed the
Strange Situation to test American children from whom she had already
collected many, many hours of naturalistic observations. Thus it has
high ecological validity for Americans, but when it is applied to other
cultures one must be very careful when interpreting the results.

4.3 Evaluating the use of attachment categories

Ainsworth believed that the differences in human infant attachment
types were entirely due to the mother. Secure infants have mothers who
are responsive to their signals. The children within the so-called
anxious-resistant or anxious-avoidant attachment categories have less
responsive mothers and have developed coping strategies to deal with
the fact that their mother will not always sensitively respond to their
needs. Yet Ainsworth has been criticised for placing a culturally specific
value system on the data. She assumed that being more independent
and explorative is a desirable outcome of responsive mothering. This
might be true in American society, with its emphasis on autonomy and
individualism, but it might not be true in other cultures.

220
 4 Since Harlow

In addition, the labels Ainsworth applied to the different attachment

categories (i.e. ‘secure’ and ‘anxious’) are value-laden. The secure
pattern is assumed to be optimal in all situations, but this might not be
the case. There is evidence from baboons to suggest that no particular
mothering style or attachment category is necessarily better or worse
than others. Rather, they are flexible and adaptive responses to
different sets of social circumstances.
Altmann (1980) observed that the wild baboons tend to adopt two
contrasting mothering styles. Some are restrictive, preventing their
babies from wandering away and limiting their access to other baboons.
Other mothers are laissez-faire, allowing their babies to wander where
they like and interact with whom they like. In general, low-ranking
mothers are restrictive, while high-ranking mothers are laissez-faire.
Baboon troops can be dangerous places. There is a strict social
hierarchy and baboons can be very aggressive. If you are low-ranking, it
is dangerous to let your baby interact too much with other baboons
because, should they get into trouble, you will almost certainly lack the
social clout to be able to help them out. In contrast, high-ranking
mothers can much more easily run to their baby’s aid. It is not that one
mothering style is inherently preferable to the other; each is an
appropriate response to a particular context. Human mothering is likely
to be just as flexible and adaptive. Thus, rather than a particular
mothering style or attachment type in humans being inherently
preferable to all others in all circumstances, different styles might suit
different situations.

4.4 Strange Situation research on other animals

You might be surprised to discover that the Strange Situation has also
been used to investigate different species. The attachments of the
chimpanzees who were raised in a nursery by human caregivers
appeared surprisingly similar to those found in human infants (van
Ijzendoorn et al., 2009). The main difference was a high incidence of
disorganised attachment. In this respect, the chimpanzee data seemed
similar to that obtained from children raised in Romanian orphanages
where adults interacted very little with the children beyond tending to
their physical needs. (You can read more about the Romanian orphans
later in this chapter.)
Along with colleagues, I have conducted studies on the dog–human
bond using the Strange Situation (Palmer and Custance, 2008; Prato­

221
Chapter 5 Determined to love?

Previde et al., 2003). We were amazed to see how similar the dogs’
responses were to those of human children. Most of the dogs were
adult, but just like human children, they only liked to play and explore
when their attachment figure (i.e. their owner) was present. When their
owners left the room, the majority of dogs sat by the door, whining,
barking and howling. On reunion, nearly all the dogs approached their
owner, ecstatically greeted him or her and then after a short while most
of them were happy to start playing or exploring again. In other words,
the majority of dogs appeared to be securely attached.

Summary
. Observational studies on human infants have shown that attachment
behaviour in children is not based on cupboard love.
. Mary Ainsworth’s Strange Situation study provided researchers with
a convenient and expedient way of assessing attachment in children.
. Cross-cultural research suggests that attachment should be viewed as
flexible and adaptive. Different attachment styles develop under
different conditions and in response to different situations. There is
no such thing as a universally preferred attachment style.

222
 5 The flexibility of attachment

5 The flexibility of attachment

Despite the findings of researchers such as Harlow, Bowlby and
Ainsworth, many people continued to vehemently resist the suggestion
that any aspect of human behaviour, including attachment, is inbuilt or
innate. As soon as one suggests that behaviour has evolved, there is a
tendency for people to assume that it is somehow ‘set in stone’ or
‘determined by our genes’. There is the fear that if our behaviour is
shown to be the product of a genetically inherited trait, it means that
we have no free will or control over it. Indeed, when Bowlby was first
developing attachment theory, he promoted the view that human
attachment is based on a rather rigid system that determines the course
of all our future relationships. In this final section, I hope to show that
innate predispositions might not be as fixed as one might think.
Earlier in the chapter I mentioned two innate predispositions: herring­
gull-chick beak pecking and imprinting. Early research suggested that
these inherited behavioural patterns are rigid and irreversible. Thus, it
was thought that whatever object a newly hatched bird first imprinted
on, however unsuitable, the process would be irreversible; there would
be no way to replace it with another object. Researchers also thought
that it was impossible for birds to imprint after their critical period,
mainly because after that time they develop a fear of all unfamiliar
stimuli. Finally, it was also thought that imprinting had long-term
consequences in terms of forming a template for all subsequent
relationships.
As ethologists studied these innate behaviour patterns in more detail, it
became clear that they are not as rigid as was first presumed. Hoffman
(1978) exposed 5-day-old ducklings to a moving rectangle. Initially, the
ducklings all attempted to flee, terrified of this unfamiliar stimulus.
Since they were in an enclosed area and could not escape, they huddled
in a corner until eventually their fear subsided. Soon afterwards, they
started to follow the rectangle and emitted distress calls when they were
separated from it. They had imprinted more than two days after their
species’ usual critical period had elapsed. Modern ethologists have
adopted the term ‘sensitive’ rather than ‘critical period’ to reflect the
fact that imprinting, and indeed all innate behaviour patterns, tend to
show some degree of flexibility.
Even after imprinting has occurred, the effects are not necessarily
irreversible. Guiton (1966) exposed baby chickens to yellow rubber

223
Chapter 5 Determined to love?

gloves so that they became imprinted upon them. When they reached
adulthood, they did indeed try to mate with similar gloves. However,
when they were given the opportunity to interact for some time with
members of their own species, they shifted the focus of their amorous
attention to these biologically more appropriate targets.

5.1 Is there a critical period in human attachment?

How does one test whether there is a similarly critical period in human
attachment? It would be unethical to deliberately prevent or hinder
young children from forming an attachment in the first few years of
their life just to see if they would be able to form such a bond after this
period. Tragically, there have been unintentional incidences in which
children have been prevented from bonding during their early years. Jill
Hodges and Barbara Tizard (1989) studied children from a care home
in which there was a huge and constant turnover of staff. By the age of
2, the children had experienced an average of twenty-four carers each.
Thus none of the children had been able to establish a long-term,
meaningful attachment. When the children reached the age of 4, the
care home was disbanded. Twenty-five of the children were returned to
their biological parents, thirty-three were adopted and seven remained
institutionalised with occasional fostering. When the children reached
the age of 16, Hodges and Tizard interviewed and administered
questionnaires to the children, their biological or adoptive mothers and
their teachers. They found that, in comparison to children who had not
experienced institutional care, all the children in the sample tended to
have problems with siblings, poor peer relationships, and teachers found
them more quarrelsome. Despite the fact that their early
institutionalisation seemed to have some long-term detrimental
consequences, the children were still clearly able to make attachments
after the age of 4. Seventeen out of twenty-one adoptive mothers felt
that their child was deeply attached to them. However, only half of the
children restored to their biological families described themselves as
deeply attached. Maybe this was because many of them went back to
conditions of poverty and privation or because some of the parents
indicated that owing to their difficult circumstances they had mixed
feelings about the return of their child. It appears, therefore, that the
quality of attachment depends not so much on the child’s age (in terms
of a critical period) as on the adults, their relationship with the child,
and the broader social context.

224
 5 The flexibility of attachment

5.2 The maternal deprivation hypothesis

A key feature of Bowlby’s work during the 1940s and 1950s is the
emphasis on the importance of the mother. Although Bowlby
sometimes argued that he used the term ‘mother’ to refer to any
primary caregiver, he has been heavily criticised for placing so much
responsibility and pressure on mothers. In the aftermath of the Second
World War, the role of women in society was changing dramatically, as
they increasingly sought to escape their traditional gender roles. It was
felt that Bowlby was preaching that mothers should stay at home and
raise their babies.

Figure 5.8 Escaping the traditional role of woman as homemaker

One of the most controversial of Bowlby’s ideas is the so-called Maternal deprivation
maternal deprivation hypothesis (Bowlby, 1944). He proposed that hypothesis
any separation of a week or more between a child under the age of 5 The idea developed by
John Bowlby that any
and its primary attachment figure, usually the mother, would have a separation of a child
long-term detrimental effect. He derived this hypothesis from his from its mother of a
observations of young offenders. week or more before
the age of 5 will have
Bowlby had been treating a young offender at the Tavistock Clinic in negative long-term
London and he came to suspect that the boy’s problems were related to effects for the child.
his disrupted relationship with his mother. He recruited forty-four

225
Chapter 5 Determined to love?

young men who had been arrested for thieving and compared them
with forty-four non-offending adolescents (Bowlby, 1944). A third of
the offenders were diagnosed with ‘affectionless psychopathology’,
which means they lacked a sense of moral conscience. Bowlby
discovered that, in contrast with non-offenders, most of the young
offenders had been separated from their mother for at least one week
before the age of 5. He suggested that breaking the bonds of
attachment in early life will lead to intellectual, social and emotional
problems in later life that will be permanent and irreversible.
One of the main problems with the data from the ‘forty-four thieves’
You learned about study was that it was correlational. A correlation indicates that as one
correlation in factor increases, another factor is likely to increase or decrease in
Chapter 3, Box 3.2. tandem with it. However, correlations cannot definitively indicate that
two factors are causally linked. There might be one or a number of other
underlying factors that are the true causal agents. The juvenile offenders
that Bowlby studied all came from poor homes. It might be that
poverty leads to a higher incidence of separation between young
children and their parents. For instance, poorer people tend to
experience lower levels of health and are more likely to be hospitalised,
which would involve children and parents being separated. Equally,
poverty alienates people from society and is therefore also likely to
cause higher incidences of anti-social behaviour. Rather than maternal
separation causing delinquency, poverty might independently affect both
whether or not children are separated from their family and the levels of
theft and delinquency.
Michael Rutter (1981) conducted a study into the roots of delinquency
among young people and found evidence to suggest that maternal
separation was in fact not the most important factor. He conducted a
survey of 2000 9–12-year-old boys from the Isle of Wight. He also
studied a group of boys from London whose parents had suffered
mental illness. Rutter found that if parental separation was due to
physical illness or the death of the mother, the boy was unlikely to turn
to crime. However, if separation was due to the mental problems of one
or both parents or to stress and conflict within the family, then the boys
were four times more likely to turn to crime. Rutter concluded that the
cause of anti-social behaviour was the familial stress and conflict that
occurred before separation, rather than separation itself.
Also, Schaffer and Emerson (1964) pointed out that, by 9 months of
age, infants have normally established a small network of attachment
relationships. Thus, even if they are separated from their primary

226
 5 The flexibility of attachment

attachment figure, such as their mother, they often have other

emotionally significant people they can turn to. Hence, separation
should be less detrimental than Bowlby proposed.

5.3 Does early attachment have a long-term

effect?
Bowlby argued that a person’s attachment in childhood lays the
foundation upon which all future relationships are built. If someone’s
first attachment relationship is faulty or unhealthy in any way, then this
will adversely colour their relationships for the rest of their life. Thus, if
the primary attachment relationship was loving and secure, the person
would grow up to be emotionally open, confident and able to establish
healthy adult relationships. In contrast, if the primary attachment bond
was unhealthy and insecure, they are likely to grow up being either
overly clingy or emotionally cold toward others.
Actually, there is rather mixed evidence as to whether early attachments
really do have long-term consequences. A positive association has been
found between a secure attachment and curiosity and problem solving
at age 2, social confidence at 3, empathy and independence at 6 and, in
the case of boys, a lack of behavioural problems at age 6 (Lewis
et al., 1984). Yet, when Peter Zimmermann and colleagues interviewed a
group of 16-year-olds, they found that early attachment patterns at 12–
18 months of age (reported in retrospect) were not a good predictor of
later relationships (Zimmermann et al., 1997). Later events, such as
parental divorce, had a much bigger impact on the nature of the 16­
year-olds’ current attachments.
Michael Rutter and colleagues have also provided evidence to suggest
that early unsuitable attachment experiences can be overcome to a
certain extent (Rutter et al., 2007). After the downfall of the Romanian
Communist regime in 1989, hundreds of orphans were found living in
deprived and squalid conditions within state institutions. The children
had received very little attention or affection within these deplorably
understaffed orphanages.

227
Chapter 5 Determined to love?

Figure 5.9 Children in a Romanian orphanage

Rutter and his colleagues followed the progress of 111 of these orphans
who were adopted in the UK. Thankfully, the children were all adopted
into loving families. At age 6, most of the children had made good
recoveries. However, those children adopted after the age of 2 showed
high levels of disinhibited attachment. Disinhibited attachment is a rather
strange phenomenon in which children behave in an overly familiar and
clingy manner with people they hardly know. Having been starved of
love and attention for the first two years of life, the children seemed
desperate to establish attachments even with complete strangers. In
2007, when the children were 11 years of age, only half of them had
recovered from their earlier disinhibited attachment style. It seems that
children exposed to physical privation and emotional deprivation can
make a full recovery if they are adopted into a loving family, but that
this is more likely to occur if they are adopted at an earlier age.
As the Romanian orphans show, love is an extremely powerful thing.
Once the orphans had been placed in loving families, they showed a
remarkable ability to overcome extreme physical and emotional
deprivation. When they first arrived in Britain, over half of the orphans
exhibited learning difficulties. By their fourth birthday, the majority of
them had an IQ that fell within the normal range. Quite understandably,
some orphans continued to experience problems, particularly if they
were adopted late. However, this does not diminish the huge strides
forward that nearly all of them achieved. So, although absent or severely

228
 5 The flexibility of attachment

inadequate early attachments are clearly harmful, children seem to be

remarkably resilient and, given the opportunity and a great deal of love,
they can prevail.

5.4 Conclusion
The care of a loving adult is so important for an infant’s survival that it
is not surprising that the tendency to form attachments has evolved.
Different sources of evidence from birds, monkeys and humans suggest
that attachment is not based upon cupboard love. Instead, humans and
other animals possess inbuilt tendencies to form attachments with
stimuli that possess certain features. However, human attachment is not
rigidly fixed in early infancy. The past sixty years of attachment research
have shown that, although it may operate according to certain inbuilt
parameters, it is a complex and flexible system that still allows
organisms to adapt to changing circumstances.

Summary
. According to Bowlby’s maternal deprivation hypothesis, even
temporary separation from the mother or primary caregiver early in
life can have negative consequences for subsequent behaviour.
. Subsequent evidence – some of which comes from studies of
children who grew up in conditions of considerable deprivation –
suggests that in the right circumstances, difficulties encountered in
early childhood, including the lack of an attachment, can be
overcome.
. Human attachment is partly determined by innate behavioural
tendencies (such as the tendency to form attachments to soft objects
or primary carers who tend to be most responsive), but it is also
influenced by changing circumstances.

229
Chapter 5 Determined to love?

References
Ainsworth, M.D. and Bell, S.M. (1970) ‘Attachment, exploration, and
separation: illustrated by the behavior of one-year-olds in a strange situation’,
Child Development, vol. 41, no. 1, pp. 49–67.
Altmann, J. (1980) Baboon Mothers and Infants, Cambridge, MA, Harvard
University Press.
Bell, S.M. and Ainsworth, M.D.S. (1972). Infant crying and maternal
responsiveness, Child Development, vol. 43, no. 4, pp. 1171–90.
Blum, D. (1994) The Monkey Wars, Oxford, Oxford University Press.
Bowlby, J. (1944) ‘Forty-four juvenile thieves: their characters and home-life’
(Parts I and II), International Journal of Psychoanalysis, vol. 25, pp. 19–53 and
pp. 107–28.
Bowlby, J. (1979) The Making and Breaking of Affectional Bonds, London,
Tavistock Publications.
Bretherton, I. (1992) ‘The origins of attachment theory: John Bowlby and Mary
Ainsworth’, Developmental Psychology, vol. 28, no. 5, pp. 759–75.
Fox, N.A. (1977) ‘Attachment of kibbutz infants to mother and metapelet’, Child
Development, vol. 48, no. 4, pp. 1228–39.
Grossman, K.E., Grossman, K., Huber, F. and Wartner, U. (1981) ‘German
children’s behavior toward their mothers at 12 months and their fathers at 18
months in Ainsworth’s Strange Situation’, International Journal of Behavioral
Development, vol. 4, no. 2, pp. 157–81.
Guiton, P. (1966) ‘Early experience and sexual object choice in the Brown
Leghorn’, Animal Behaviour, vol. 14, no. 4, pp. 534–8.
Harlow, H.F. (1958) ‘The nature of love’, American Psychologist, vol. 13,
pp. 673–85.
Harlow, H.F. (1962) ‘Development of affection in primates’ in E.L. Bliss (ed.)
Roots of Behavior, pp. 157–66, New York, NY, Harper.
Harlow, H.F., Dodsworth, R.O. and Harlow, M.K. (1965) ‘Total social isolation
in monkeys’, Proceedings of the National Academy of Sciences of the United States of
America, vol. 54, no. 1, pp. 90–7.
Hess, E.H. (1958) ‘“Imprinting” in animals’, Scientific American, vol. 198, no. 3,
pp. 81–90.
Hodges, J. and Tizard, B. (1989) ‘Social and family relationships of ex­
institutional adolescents’, Journal of Child Psychology and Psychiatry, vol. 30, no. 1,
pp. 77–97.
Hoffman, H.S. (1978) ‘Experimental analysis of imprinting and its behavioral
effects’ in Bower, G. (ed.) The Psychology of Learning and Motivation, vol. 12,
pp. 1–37, New York, NY, Academic Press.

230
 References

Holmes, J. (1993) John Bowlby and Attachment Theory, New York, NY, Routledge.
Lewis, M.L., Feiring, C., McGuffog, C. and Jaskir, J. (1984) ‘Prediction
psychopathology in six-year-olds from early social relations’, Child Development,
vol. 55, no. 1, pp. 123–36.
Lorenz, K.Z. and Kickert, R.W. (1981) The Foundations of Ethology, New York,
NY, Springer-Verlag.
Palmer, R. and Custance, D.M. (2008) ‘A counterbalanced version of
Ainsworth’s Strange Situation Procedure reveals secure-base effects in dog–
human relationships’, Applied Animal Behaviour Science, vol. 109, no. 2,
pp. 306–19.
Prato-Previde, E., Custance, D.M., Spiezio, C. and Sabatini, F. (2003) ‘Is the
dog–human relationship an attachment bond? An observational study using
Ainsworth’s Strange Situation’, Behaviour, vol. 140, no. 2, pp. 225–54.
Rutter, M. (1981) Maternal Deprivation Reassessed, New York, NY, Penguin.
Schaffer, H.R. and Emerson, P.F. (1964) ‘The development of social
attachments in infancy’, Monographs of the Society for Research in Child Development,
vol. 29, no. 3, pp. 1–70.
Slater, L. (2004) Opening Skinner’s Box: Great Psychological Experiments of the
Twentieth Century, London, Bloomsbury.
Suomi, S.J., Harlow, H.F. and McKinney, W.T. (1972) ‘Monkey psychiatrists’,
American Journal of Psychiatry, vol. 128, no. 8, pp. 927–32.
Suomi, S.J. (1978) ‘Maternal behavior by socially incompetent monkeys: neglect
and abuse of offspring’, Journal of Pediatric Psychology, vol. 3, no. 1, pp. 28–34.
Rutter, M., Colvert, E., Kreppner, J., Beckett, C., Castle, J., Groothues, C.,
et al. (2007) ‘Early adolescent outcomes for institutionally-deprived and non­
deprived adoptees: I. Disinhibited attachment’, Journal of Child Psychology and
Psychiatry, vol. 48, no. 1, pp. 17–30.
Takahashi, K. (1986) ‘Examining the Strange Situation procedure with Japanese
mothers and 12-month-old infants’, Developmental Psychology, vol. 22, no. 2,
pp. 265–70.
United Nations (UN) (1959) ‘Declaration of the Rights of the Child’, General
Assembly resolution 1386 (XIV), New York, NY, UN; also available online at
www.un.org/cyberschoolbus/humanrights/resources/child.asp (Accessed 13
April 2010).
United Nations (UN) (1989) ‘Convention on the Rights of the Child’, General
Assembly resolution 44/24, New York, NY, UN; also available online at www.
un.org/documents/ga/res/44/a44r025.htm (Accessed 13 April 2010).
van Ijzendoorn, M.H., Bard, K.A., Bakermans-Kravenburg, M.J. and Ivan, K.
(2009) ‘Enhancement of attachment and cognitive development of young
nursery-reared chimpanzees in responsive versus standard care’, Developmental
Psychobiology, vol. 51, no. 2, pp. 173–85.

231
Chapter 5 Determined to love?

Watson, J.B. (1928). Psychological Care of Infant and Child, New York, NY, Harper
and Bros.
Zimmermann, P., Fremmer-Bombik, E., Spangler, G. and Grossman, K.E.
(1997) ‘Attachment in adolescence: a longitudinal perspective’ in Koops, W.,
Hoeksma, J.B. and van den Boom, D.C. (eds) Development of Interaction and
Attachment: Traditional and Non-traditional Approaches, pp. 281–91, Amsterdam,
North-Holland.

232
Chapter 6
Making friends
Charlotte Brownlow
Contents
1 Introduction 237
1.1 What is friendship? 237
1.2 The role of friendships and the influence of peers 240
2 The changing nature of friendship in childhood 242
2.1 Quantitative or qualitative data? 245
3 Researching children’s friendships 248
3.1 Qualitative research methods: interviews 248
3.2 Qualitative research methods: ethnography 250
4 Contemporary explorations of friendships 256
4.1 The social influences of friendships 256
4.2 Cultural influences on the concept of friendship 258
5 A new direction for friendship research? 263
5.1 The changing nature of friendships: the role of
new technologies 263
5.2 A new kind of friendship? 265
5.3 A new kind of method? 267
5.4 Conclusion 268
References 269
Chapter 6 Making friends

Aims and objectives

After studying this chapter you should be able to:
. outline the complexities surrounding the understandings and
definitions of friendships
. reflect on some of the key methodological approaches that have
been used to research friendship
. consider the influences that peers may have on an individual’s
behaviour
. describe the role that cultural experiences may play in our
understandings and experiences of friendship
. outline the possible changing nature of friendships in the light of
new technologies and the possibilities that these may offer for
‘global friendships’.

236
 1 Introduction

1 Introduction
In the previous chapter you were introduced to the work of Harry
Harlow and the way in which psychological research has looked
extensively at children’s relationship with, and attachment to, their
parents and caregivers. There is another important set of relationships
in a child’s life, and that is with their friends. For a long time, however,
the role of friends and the potential influences of friends on the
behaviour of children had been neglected in the literature. This may
seem quite surprising given the frequent references that are made in
everyday exchanges to the influence of friends on an individual. For
example, people may make generalisations regarding a young person
‘being in with the wrong crowd’ or talk about the ‘good influence’ that
a particular new friend has become.
Given the importance of friendship in everyday life, it is worthwhile
looking more closely into psychological research in this area. Friendship
is different from many other relationships that individuals are involved
in. For children in particular, friendship takes on a very different
meaning from other relationships such as those with parents, other
adults and siblings: friendship is usually characterised by both parties
having equal power. This is unlike many other relationships children
have. In a relationship with an adult, the adult is usually the dominant
party. This makes friendship between children an interesting relationship
to focus on. In particular, psychologists have asked questions
concerning the best way to measure and study friendship, the role of
peer influence on behaviour, cultural differences in expressing and
understanding friendship, and the possible changing nature of friendship
given the rising importance of modern technologies in our lives. These
are issues that will be explored in this chapter.

1.1 What is friendship?

Friendship may have different meanings to different people and I would
like to start by thinking about what friendship is and how we can define
it. Before I go on to consider how friendship has been defined by
psychologists, I would like you to pause and think about what
friendship means to you and how you would define the concept of
‘friendship’.

237
Chapter 6 Making friends

Activity 6.1
Spend a few minutes considering the term ‘friendship’. Think about what
this term means to you, and what sort of things you expect from a
friendship. It may be helpful to think about one specific friend and your
relationship with them. Write down some of the features of this
relationship that you think are important.
Now think about what the term ‘friendship’ may mean to children. Do you
think that this will differ depending on what age the children are? Write
down some of the features of friendship that you think may be important
to children.
Now compare your two lists. Have you identified similar features on both
lists or are they different? Which features of friendship are similar and
which are different?

The questions raised in Activity 6.1 concerning the definitions of

friendship and the possible changes as individuals progress through
childhood, adolescence and adulthood will be a theme that I will be
returning to in the chapter. You may want to reflect back on your
thoughts from this activity again once you have completed your reading
of the whole chapter.

Figure 6.1 Early friendships

238
 1 Introduction

The concept of friendship can be difficult to define and may mean

different things to different people at different times. As William
Bukowski and colleagues point out, our language does not necessarily
differentiate between the friendships of infants and toddlers and those
of adolescents and adults despite the differences between these two
types of relationship in respect of qualities and expectations (Bukowski
et al., 1996). Bukowski et al. therefore question whether a single
definition of friendship is possible, one that can adequately describe
close relationships between people in a range of different populations.
In addition to the potential age-related differences in understanding
friendship, the influence of culture should also be considered. This is a
question that I will be returning to later on in this chapter.
A further question for consideration when researching friendship is how
the boundaries of friendship are determined. An important researcher
of friendship, Willard Hartup, notes that researchers have tended to
define friendship in a simplistic way that focuses on the differences
between ‘friend’ and ‘non-friend’. Hartup (1996) argues that friendship
is far more complex than this, and a whole spectrum of relationships is
possible on a continuous scale from best friend to good friend to
occasional friend to non-friend. In addition to the complexities
associated with types of friend, it is also important to remember that
children are involved in a whole range of relationships, including those
with family members and other adults such as teachers, as well as peers.
In any investigation into the understanding of friendship, other social
networks must be borne in mind and friendship should be considered
within the context of the other relationships.
So, researchers investigating this topic need to have an understanding of
what they mean when they talk about friendship. Generally, researchers
agree on three elements as being important in the friendships of
children and adolescents (Bukowski et al., 1996). Each person in the
friendship should:
. gain equal benefits from the relationship – the relationship should
draw on the importance of cooperation and the satisfactory
resolution of potential conflicts
. have a liking for the other(s) and a desire to spend time with them
. have fun and express affection for the other(s).

239
Chapter 6 Making friends

1.2 The role of friendships and the influence of

peers
One of the main issues in friendship research is how peers may exert
an influence on a child or an adolescent. This will be a key
consideration of this chapter too. As you read on, you need to keep in
mind questions concerning the meaning of friendships at various stages
in life and the possible influence of peers, both positive and negative,
on an individual’s behaviour. Friends, for example, may have a strong
impact on the type of music that a person listens to or the clothes that
they wear. In Section 4, I will return to this question and consider the
influence that friends may have on the uptake of smoking. However, it
is also important to consider the potential positive effects that friends
may have on behaviour. Friends can sometime have a beneficial effect
on academic achievement in school, on participation in out-of-school
activities and on avoiding risky behaviours. However, what is the
mechanism by which peers exercise this influence? And is this influence
consistent throughout life?

Figure 6.2 Growing importance of peer influence in adolescence

Phil Erwin (1998) notes that the significance of peer relationships

increases and changes with age. For young children, parents are the

240
 1 Introduction

main source of support and guidance. As children mature into

adolescence, the influence of peers becomes more important. However,
despite occasional disagreements with parents over issues such as
curfews and clothing (common sources of dispute within families), the
values of parents and peer groups are not necessarily opposed
(Erwin, 1998). Indeed, Erwin argues that the peer group can often act
as a powerful force of support for many of the values held by parents,
owing to friendships typically being formed between people of similar
backgrounds. Therefore, even though some groups can be thought to
promote anti-social behaviour, their significance for most adolescents
has frequently been overestimated (Erwin, 1998). The general influence
of peers is typically seen to decline after the early teens, owing to the
increasing focus on romantic relationships, which may serve to move an
individual’s attention outside the friendship circle.
Before moving on to look at the subject of peer influence, however, it
is important to explore in more detail the changing understanding of
friendship in childhood, a topic that was investigated in a key study by
Brian Bigelow and John La Gaipa.

Summary
. Defining friendship is a complex endeavour, as ‘friendship’ can
mean different things to different people at different times.
. The role of friendships and the influence of peers change with age
as parents and carers are increasingly replaced by peers as the main
source of support.
. While there are examples of peers influencing behaviours in a
negative way, the significance of peers in reinforcing positive
behaviours may well be underestimated.

241
Chapter 6 Making friends

2 The changing nature of friendship

in childhood
A pioneering study on children’s friendship was carried out by Brian
Bigelow and John La Gaipa (1975). In the 1970s, at the time when they
began their research, very little was known about children’s friendships.
The focus of relationship research at the time was mainly on emotional
relationships and on understanding what attracts one person to another.
In a bid to explore the under-researched area of children’s friendship,
Bigelow and La Gaipa looked at the differences in children’s
understanding of friendship at various stages of development. They
devised a novel means of investigating the gradual changes in the
interpretation of friendship as children grow older. In doing so they
helped to shed new light on the important role that friends play in
children’s lives.
In order to investigate children’s understanding of friendship, Bigelow
and La Gaipa asked children to think about their best friend of the
same sex and write an essay about what they expected of their best
friend and how this might be different from their expectations of other
friends and acquaintances. Bigelow and La Gaipa compiled a large
sample of 480 essays written by children aged between six and fourteen
years who came from upper-working-class and lower-middle-class
homes in Windsor, Ontario, Canada. Children who took part in the
study were recruited from eight schools – thirty girls and thirty boys
from each.
The research approach adopted by Bigelow and La Gaipa was very
different from those that you have previously encountered in this book.
Rather than conducting observations of children’s behaviour, carrying
out a well-designed experiment or administering a questionnaire,
Bigelow and La Gaipa collected data in the form of written essays. This
meant that all the data that they collected was in a written, text format.
Their decision to use essays as a source of data raises a number of
questions that I would like you to think about as I go on to consider
Bigelow and La Gaipa’s work.
Before the study began, the researchers came up with a list of different
characteristics of a best friend that they anticipated the children might
include in their essays. Particular expectations are made of best friends
that are not necessarily true for other friends and acquaintances,
including notions such as sharing common activities, engaging in

242
 2 The changing nature of friendship in childhood

organised play, admiration, loyalty and commitment, genuineness, and

similarity in attitudes and values. Bigelow and La Gaipa arrived at
twenty-one different categories of friendship expectations, which were
then used to analyse the children’s written accounts of their best friend.
Bigelow and La Gaipa used an approach called content analysis: they Content analysis
took all of the 480 children’s essays and compared them against their An approach to the
list of twenty-one friendship expectations in order to count how many analysis of written,
audio and visual
times each particular expectation was mentioned (which is usually
material that involves
referred to as a ‘frequency count’). Once the occurrences of the identifying and
different expectations had been highlighted in the essays, Bigelow and counting pre-selected
La Gaipa were able to use the frequency counts to look for any patterns features relevant to the
in the data that would provide information about the changing nature research question.
of children’s friendships. They carried out comparisons, for example
between younger and older children in their sample, and also looked for
differences in the understanding of friendship between boys and girls.

Figure 6.3 Sharing common activities as a basis for friendship

Bigelow and La Gaipa found some important differences in the

expectations that children have of a best friend at different ages. Sixteen
of the original twenty-one friendship expectations were more frequent in
the older children’s descriptions of their best friend compared to the
younger children’s. This suggests that expectations of a best friend
become increasingly complex and sophisticated as children get older.
There was a clear progression from a focus on shared activities to

243
Chapter 6 Making friends

prioritising intimacy and the importance of similarity in attitudes and

values.
Bigelow and La Gaipa also looked at gender differences. Recall that
each child was asked to write about a friend of the same sex, so the
comparison was made between boy–boy and girl–girl friendships. They
found that generally the differences between boys and girls were
relatively minor. However, interestingly, there were differences regarding
expectations of activities concerned with organised play, with boys being
more likely to cite this as an expectation. Bigelow and La Gaipa
explained that this was largely owing to competitive group sport being
more of a male activity in their particular sample.
Bigelow and La Gaipa proposed that the findings of the study could be
considered in terms of a three-stage model of the development of
friendship expectations. In the first stage there is an emphasis placed on
shared activities and concrete expectations such as the ability to see
each other and the importance of geographical closeness. At this age,
parents are also likely to be important in facilitating the meeting of
children and hence the development of friendships. The second stage
sees a transition from the focus on individual needs to an emphasis on
sharing, loyalty and commitment. In the final stage, expectations draw
upon the importance of similarity in attitudes, values and interests and
the potential for the development of an intimate and confiding
relationship. Friendship expectations therefore become more complex as
children develop.

Activity 6.2
The method used to collect data about children’s friendship is an
important area for consideration. Pause for a few minutes and think
about the method used by Bigelow and La Gaipa. The researchers asked
children aged between six and fourteen years to write essays about what
they would expect from a best friend of the same sex.
What do you think about Bigelow and La Gaipa’s method? Do you think
that it is a good way of tapping into children’s friendship expectations?
What might be some of the problems with this approach?

Remember that this research was undertaken at a time when there was
not a strong research focus on children’s friendships, and its importance
lies in the fact that it drew attention to what was at the time an under­
researched area. However, you may want to consider how easy it is,

244
 2 The changing nature of friendship in childhood

particularly for young children, to express themselves in written form.

Bigelow and La Gaipa comment in their research paper that this may be
an important issue for consideration, in that children, especially younger
ones, may lack the verbal and written skills necessary to effectively
discuss expectations in this form.

2.1 Quantitative or qualitative data?

Let’s now consider in a little more detail Bigelow and La Gaipa’s
approach. Children were asked to write an essay about what they felt
were important expectations of their best friend. This novel method
enabled the children to write whatever they felt was individually
important to them, and the data that was collected is in the format of
written text. Therefore it was qualitative in nature. You read about the
distinction between
However, Bigelow and La Gaipa then compared each of the children’s qualitative and
essays with a list of expectations that they had created before collecting quantitative data in
the data and made a frequency count of how many times each of the Box 1.1, Chapter 1.
expectations appeared in the children’s essays. They therefore
transformed qualitative data (written text) into quantitative data
(frequency counts). In doing this, it could be argued, they lost the
individual and personal dimension of the children’s accounts. On the
other hand, the numerical summary enabled them to draw inferences
about the differences between the age groups and make more
systematic comparisons between boys and girls. This decision to convert
the written descriptions into numerical summaries is further examined
in Box 6.1.

Box 6.1 Why do it this way?

Bigelow and La Gaipa’s approach allowed each child to write their
own individual account of friendship expectations. However, these
individual accounts were then compared and scored in relation to a
predetermined set of friendship expectations, ultimately changing
the nature of the research data from descriptions to frequency
counts. Bigelow and La Gaipa, therefore, ultimately transformed
written, textual data into the kind of data that could be quantified.
Several issues are raised by this decision.
First, given the total number of children participating in the study,
the decision to code essays according to an agreed set of
categories enabled the researchers to analyse the large number of
essays in a consistent manner. It would be extremely difficult and

245
Chapter 6 Making friends

time-consuming to preserve the individual voices and opinions

within such a large sample. Working with a large sample was,
however, crucial to the study because Bigelow and La Gaipa
wanted to make generalisations about children’s friendships, and
develop a model of the developing nature of friendship that could be
applied to a wider population.
Remember also that Bigelow and La Gaipa came up with a number
of different expectations that they looked for in all the essays. This
means that they were looking for aspects of friendship that are
common across the essays, rather than at individual features of
each child’s friendship experience. Had they been interested in
individual accounts, a much smaller sample of essays would have
been collected and studied in more detail.
Finally, the fact that the essays were explored in relation to
categories generated by the researchers is important. Children’s
expectations of friendship may not be the same as those that
Bigelow and La Gaipa expected to find. It is possible that there
were expectations other than those that the researchers looked for
in the essays, which would have been overlooked by the content
analysis.

The work of Bigelow and La Gaipa is important for our understanding

of friendship for several reasons. First, it placed the concept of
children’s friendship on the research agenda. Second, it provided some
interesting findings concerning the changing nature of friendship and
how expectations of friends develop as children get older. The latter
issue is linked to the question that I posed in Section 1 concerning the
difficulties with defining friendship, and the differences that may exist
between friendships of young children and those of adolescents and
adults. Finally, Bigelow and La Gaipa introduced an interesting and
novel research method for studying relationships, namely using written
accounts and descriptions as data.

246
 2 The changing nature of friendship in childhood

Summary
. The work of Bigelow and La Gaipa is important as one of the first
studies of children’s friendship.
. Bigelow and La Gaipa (1975) provided evidence of the changing
nature of friendship expectations between the ages of six and
fourteen.
. Their work illustrates how qualitative data can be transformed into
quantitative data, through content analysis.

247
Chapter 6 Making friends

3 Researching children’s friendships

In this section I would like to focus more closely on the subject of
method, and contrast the approach taken by Bigelow and La Gaipa with
some other ways of researching friendship. I would particularly like to
focus on the use of more qualitative research approaches, which are
frequently used in this area.

3.1 Qualitative research methods: interviews

One way of researching children’s friendships would be to interview
You were introduced to children about them. This involves the researcher having a list of
interviews as a way of questions that they want to ask children, although these questions are
collecting data in used flexibly so that the interviewer can follow up on any interesting
Chapter 1, section 2.3.
comments that the children make. This approach ensures that all those
taking part in the study are asked more or less the same questions,
while at the same time allowing each child being interviewed to bring up
issues that the researcher may not have otherwise considered.
However, when using interviews in research on children, it is important
to consider the language skills of the children and how well they would
be able to verbalise their thoughts and feelings to an (often unfamiliar)
adult researcher. Interviews may therefore be more suited for research
involving older children and teenagers.
An example of using interviews for researching friendship can be found
in the work of William Damon. Here is an extract from his interview
with Jack, an American thirteen-year-old, about his friendship with
Jimmy:
Interviewer: Why is Jimmy your best friend?
Jack: I don’t know, I guess it’s because we talk a lot and stuff.
Interviewer: What do you talk about?
Jack: Secret stuff, you know, what we think of him or her or whoever.
And sports, things we both like to do.
Interviewer: How did you meet Jimmy?
Jack: I don’t know; hanging around, I guess. We just sort of got
friendly after a while.
Interviewer: When did you get friendly?

248
 3 Researching children’s friendships

Jack: After we found out we didn’t have to worry about the other guy
blabbing and spreading stuff around.
Interviewer: Why would you worry about that?
Jack: Well, you need someone you can tell anything to, all kinds of
things that you don’t want to spread around. That’s why you’re
someone’s friend.
Interviewer: Is that why Jimmy is your friend? …
Jack: Yes, and we like the same kinds of things. We speak the same
language. My mother says we’re two peas in a pod.
Interviewer: What would you say you like best about Jimmy?
Jack: Well, you know, we can say what we want to around each other,
you don’t have to act cool around him or anything. Some of the older
kids are always pretending to be big shots, acting real tough. That kind
of stuff, it … turns me off.
Interviewer: How do you know who to become friends with and who
not to?
Jack: Well, you don’t really pick your friends, it just grows on you. You
find out that you can talk to someone, you can tell them your problems,
when you understand each other.
(Damon, 1977, pp. 163–4)

Thinking back to the work of Bigelow and La Gaipa, we can see many
similarities between Jack’s account and what they identified as children’s
expectations in a best friend, especially in relation to loyalty, closeness
and trustworthiness. Note for instance that Jack mentions ‘not blabbing
and spreading stuff around’, ‘saying what we want around each other’
and talking about ‘Secret stuff ’ as key aspects of friendship. So,
although Damon (1977) was using a very different method from
Bigelow and La Gaipa, he was touching on similar issues. A key
difference that remains, however, is that Damon’s work is focused at an
individual level. Damon did not compare Jack’s responses to
predetermined categories or try to make generalisations with regard to
developmental age.

249
Chapter 6 Making friends

Figure 6.4 Loyalty, closeness and trustworthiness: the basis of friendship

among teenage boys

3.2 Qualitative research methods: ethnography

William Corsaro argues for yet another approach to studying children’s
friendships, one that is very different from those introduced previously.
Corsaro is interested in how children talk to each other and believes that
research on children’s friendships should focus on children’s individual
understandings of the word ‘friend’. Corsaro’s approach is therefore
different from that of Bigelow and La Gaipa in that he is not interested
in translating the children’s words into numbers and looking for general
patterns. Rather, like Damon (1977), he is interested in an individual
child’s understanding of friendship. Also, Corsaro seeks to explore what
friendship means in particular places at particular times, and how it is
described in communication between children, rather than in a formal
interview with an adult researcher.
In his work on children’s friendships, Corsaro makes detailed notes of
Ethnography
children’s activities and their interactions with others, and also video­
A research approach
where the researcher records them. This is an example of an ethnographic approach to
carries out extensive research. In order to carry out this work, the researcher needs to
observations of a group become a member of a particular group and carry out observations
through being involved from within the group for months or sometimes even years. Because
in their activities over a the researcher is also a member of the group that they are studying,
period of time.
they will become familiar with important cultural influences and values

250
 3 Researching children’s friendships

that are commonly shared by the group, thus enabling them to develop
a very complex picture of what is happening within the group. If the
researcher chose to adopt experimental methods (e.g. Bandura’s Bobo
Doll study), or conduct observations from outside the group (for
instance Ainsworth’s Strange Situation study), some of these often You read about
complicated and subtle influences might be missed. Bandura’s research in
Chapter 3 and
Ainsworth’s research in
Activity 6.3 Chapter 5.
What are some of the issues that may be raised by such an approach?
Think particularly about the advantages and disadvantages of using
ethnographic methods.
Consider the use of this type of research method when studying
children’s friendships. Can a researcher become an active participant in
a child’s social world?

The questions raised in Activity 6.3 are important when deciding

whether to conduct an ethnographic study. First, one of the strong
benefits of using this method is the rich and complex data that can be
generated. Rather than asking a child to verbalise their thoughts and
feelings on the topic of friendship, which, depending on the child’s age,
may be quite a challenging task in itself and something they have never
done before, the approach enables the researcher to observe a first-hand
account of the child’s experiences.
However, there remains an important question concerning how the
researcher can become part of the social grouping, particularly when
researching with children and adolescents, where the researcher’s
typically adult status will be all too conspicuous. Corsaro (2006) notes
that it can be especially hard to ‘blend in’, given that physically adults
are larger than children and are perceived by children as being ‘in
charge’. In order to achieve the status of the child’s friend, the
researcher would need to overcome such problems. From experience of
using ethnography in his own research, Corsaro suggests an approach in
which the researcher initially watches and observes from a distance for
a while and then waits for the children to approach him and invite him
into their social world. He has found that this has worked well in the
research that he has carried out.

251
Chapter 6 Making friends

Figure 6.5 William Corsaro with a group of children

Using this ethnographic approach, Corsaro has highlighted many

interesting aspects of children’s understandings of friendship and how
they may be developed and maintained. The following is an example of
the kind of data that was collected during Corsaro’s observation of a
group of children aged three playing in a box. Martin and Dwight are in
a box and Denny runs by:
Martin to Denny: Denny!
Denny to Martin: Martin!
[Denny climbs in box next to Martin.]
Martin to Denny: Dwight is here.
Denny to Martin: Can I come?
Martin to Denny: Yeah, you can come.
Denny to Martin: Yeah, ‘cause I’m your friend, right?
Martin to Denny: ‘Cause I’m your friend.
(Corsaro, 1985, p. 163)

In the example above you can see how children are using the concept
of friendship in order to gain access to play activities. Corsaro also

252
 3 Researching children’s friendships

found that children referred to themselves as friends because they were

playing together.
Using such methods to investigate children’s friendships has provided
examples that show that children can talk about friendship in quite
abstract and complicated terms, at a younger age than previous research
suggested. For example, Jenny and Betty are playing in a large wooden
box. Betty has just come back to Jenny having played with another
child, Linda:
Betty to Jenny: I do like you, Jenny … I do.
Jenny to Betty: I know it
Betty to Jenny: Yeah. But I just ran away from you. You know why?
Jenny to Betty: Why?
Betty to Jenny: Because I –
Jenny to Betty: You wanted to play with Linda?
Betty to Jenny: Yeah.
Jenny to Betty: I ranned away with you. Wasn’t that funny?
Betty to Jenny: Yes.
Jenny to Betty: Cause I wanted to know what happened.
Betty to Jenny: I know you wanted – all the time – you wanna know
because you’re my best friend.
Jenny to Betty: Right.
(Corsaro, 1985, p. 166)

Jenny and Betty are only three years old, yet are clearly showing that
they are concerned about how the other felt when they played with
someone else. Betty reassures Jenny that she knew why Jenny wanted to
know where she was and that they are still best friends. This showing of
concern and regard for others is something that Bigelow and La Gaipa
argued becomes an important aspect of friendship much later in a
child’s life, not at the age of three.
In order to explain the discrepancy between Corsaro’s finding and that
of Bigelow and La Gaipa, we need to consider the issue of method, that
is how researchers explored children’s understandings of friendships. It
may be that subtleties such as the ones apparent in the examples above

253
Chapter 6 Making friends

can be tapped into by engaging with children in their environment on

their level. By contrast, a complex task such as writing about
expectations of friendships, or explaining experiences to an adult
interviewer, might yield an incomplete picture of a child’s understanding
of the social world. What Corsaro did was to engage with children on
their own terms. Then, by reflecting on his notes and video recordings,
and by analysing the conversations between children, he explored and
illuminated the world of children’s friendship.

Box 6.2 Working with textual data

A researcher who adopts a more qualitative approach in their
research needs to think about the ways in which interview or
ethnographic data can be analysed. Recall that Bigelow and La
Gaipa transformed their qualitative data into quantitative data. They
looked for the occurrence of previously decided-upon categories,
and created numerical summaries in the form of frequency counts.
However, it is possible to analyse textual data without transforming
it into numbers. One way of doing so would be to look for themes
across the transcripts or written accounts (without having a
predetermined set of categories against which the accounts would
be compared) and then summarise these themes using the
children’s actual words and utterances as illustrative examples.
Analysing qualitative data in a way that retains the individual view of
the participants is an important challenge for qualitative
researchers. Generally, qualitative approaches have moved away
from collecting data from lots of participants, with the view of
generalising findings to the population as whole. Instead, research
involves collecting data from fewer participants, which is then
analysed in more detail. Examples from personal accounts are used
to illustrate broader ideas and conclusions, which are rooted in the
data, namely the experiences and opinions voiced by participants.
There are different ways of analysing textual data, but a common
starting point is to read and reread transcripts of interviews,
recorded conversations or written accounts and look for common
and recurring themes. Unlike in the case of Bigelow and La Gaipa,
the aim is not to look for categories that have been chosen in
advance by researchers, but to pull out, from what participants have
actually said, the main themes.

254
 3 Researching children’s friendships

Because analysis of textual data aims to preserve the individual

perspective, it allows a more detailed and rich picture of a person’s
understanding and expectation of friendship to emerge.

Summary
. There are different approaches to studying friendships and these will
generate different data, which can be analysed in different ways.
Each approach has distinct advantages and disadvantages.
. Qualitative approaches seek to understand friendship from an
individual’s perspective. These approaches draw on personal
accounts of friendship formation and friendship experiences.

255
Chapter 6 Making friends

4 Contemporary explorations of
friendships
In recent years, psychologists have built on the early research on
friendship in childhood and adolescence and have explored a number of
more specific aspects of this type of relationship. Among them are
positive and negative social influences exerted in the context of
friendship, and the effects of culture on interaction between peers.

4.1 The social influences of friendships

Friendships have been cited as being an important influence, both
positive and negative, on a range of behaviours. For example,
friendships influence how children experience starting school, which
means that as well as being a source of companionship, friends are also
important in giving support to a child’s learning. The importance of
peer relationships continues across the lifespan. Researching friendships
in adults has the advantage that participants can better articulate and
reflect upon their experiences, and talk about how peers have influenced
them.

Activity 6.4
Let’s pause briefly to consider the question of influence by our peers.
Think about whether you have ever been influenced by or have
influenced your peers and in what way – for example with regard to
fashion, activities, hobbies, etc. Was this influence positive or negative?
Are we inevitably influenced by our friends? Can we actively choose to
follow or resist peer influence?

Working in Australia, Kim McLeod and her colleagues focused on the

influence of friends on the uptake of smoking among adolescents.
McLeod et al. (2008) used telephone interviews to collect qualitative
data, which enabled them to explore the social context of smoking
behaviour, with a particular focus on the role of friends. McLeod et al.
argue that early work on the influence of friendships tends to view
adolescents as passive recipients of peer influences. So, for example,
they may go along with and behave in a similar way to their peers
without much reflection. An alternative view is that adolescents in fact

256
 4 Contemporary explorations of friendships

play an active role in selecting the peers that they associate with and
that peer influence involves a more complex dynamic.
In order to investigate the influence of friends on smoking behaviour,
McLeod et al. studied identical twins, of whom one smoked and the
other didn’t, and explored whether this difference was related to the
friendship groups of the twins. McLeod et al. argued that qualitative
methods are more appropriate when studying the often-complicated
relationships between friends because they enable an in-depth
understanding of the issue of peer influence from the perspective of the
individual. McLeod et al. interviewed fourteen pairs of identical twins,
nine female and five male pairs, who differed in their smoking status.
Their ages ranged from 27 to 33 years.
In their analysis McLeod et al. found that the different friendship
groups of the twin pairs reflected the twins’ decision to be a smoker or
not. While some of the twins reported sharing friends at secondary
school, after this all the participants reported having their own
friendship circles that were different from their twin’s. The participants
reported that having friends who smoked, or didn’t smoke, had an
effect on their smoking status. Smokers in this study accounted for their
behaviour in terms of social mobility, for example using smoking as a
way of adopting a rebellious image, or gaining access to a particular
group or scene. It also enabled a sense of group acceptance to be
nurtured. Similarly, the non-smoking twins reported the dominance of a
non-smoking image within their chosen social circles. McLeod et al.
therefore concluded that both smokers and non-smokers are aware of
the role that smoking can play in creating a social image and developing
and maintaining a sense of collective identity.
However, it is important to note that, of the participants in this study,
few smokers discussed their smoking uptake as a direct outcome of
peer influence. They were not encouraged or made to smoke by their
peers. Rather, they reported that they became smokers because the
people that they wanted to associate with happened to be smokers. The
participants described their friendship circle as consisting of people
whom they wanted to spend time with, and smoking was just one of
the many things that members of this social circle engaged in. The
findings of McLeod et al. therefore point towards a complex influence
of friends on an individual’s behaviour, which can depend on wider
issues of fitting in and social approval.

257
Chapter 6 Making friends

Figure 6.6 Growing importance of image, group acceptance and perhaps

rebellion: friendships among young adults

4.2 Cultural influences on the concept of friendship

So far I have considered friendship in very broad terms and have not
examined any potential differences between different groups of people
with regard to how they perceive and interact with friends. Indeed, this
is reflected in early studies of children’s friendships, which were
primarily conducted within North American and UK contexts. What I
would like to do now is consider issues of culture and how they may
influence friendship formation and expectations.
The need for a focus on the influence of culture has become
increasingly important in psychological research. In Chapter 5 you read
about Mary Ainsworth’s Strange Situation study and how the
descriptions of attachment behaviours put forward by Ainsworth
reflected Western assumptions about what constitutes appropriate
attachment style, without taking into account cultural differences in
child-rearing practices. This is also the case when researching friendship.
Being aware of the many similarities and differences between cultural
understandings of friendship may lead us to a different understanding
of what the term ‘friendship’ means.
Yorkys González, Digno Moreno and Barry Schneider (2004) –
researchers working at two universities, one in Cuba and one in Canada

258
 4 Contemporary explorations of friendships

– sought to explore the influence of culture through the comparison of

friendships within collectivist and individualistic societies. The Collectivist cultures
authors proposed that individualistic cultures are characterised by the Cultures characterised
valuing of individual goals and achievements, and this could therefore by a focus on values
such as concern for
be considered to be inconsistent with the responsibilities of a close
others and harmonious
friendship. In contrast, the authors characterised collectivist cultures as relationships.
those that focus on values such as concern for others and the
maintenance of harmonious relationships. These are features that are
conducive to the development of close friendships. In order to explore Individualistic
the potential differences between friendships in two different cultures, cultures
González et al. compared the friendships of adolescents in Canada, Cultures characterised
by a focus on valuing
which they drew on as an example of a culture of individualism, and individual goals and
Cuba, which they drew on as an example of a collectivist culture. achievements.
González et al. used the method developed by Bigelow and La Gaipa.
They asked 300 Cuban and 294 Canadian adolescents to write an essay
about their best friend of the same gender and consider what they
expected of their best friend. Specifically, the adolescents were asked to
consider their best friend rather than other friends and think about the
features that made a best friend different from other schoolmates.
González et al. also analysed the essays by looking for and counting the
occurrences of a set of categories that were very similar to those used
by Bigelow and La Gaipa.
González et al. found that there were some important differences as
well as similarities between the two groups of participants. Authenticity,
loyalty and acceptance featured in essays that were collected in both
countries. However, the authors found that essays by the Cuban
adolescents were more likely to also discuss issues concerning giving
and receiving help. The Cuban essays were also more likely to focus on
character admiration as a criterion for a close relationship. In contrast,
Canadian essays were more likely to focus on issues concerning the
sharing of common interests and a shared history of social interaction.

259
Chapter 6 Making friends

Figure 6.7 Friendship in Cuba

From their study, González et al. concluded that the essays from the
Cuban and Canadian adolescents reflected many aspects of the cultural
values of the society within which the adolescents grew up. The authors
proposed that the Canadian responses, with their focus on common
interests, reflected the values of a society that prioritises individual
preferences and choices. In contrast, the Cuban essays frequently
mentioned mutual assistance, which again reflected dominant ideals
within the society, where the welfare and interests of other group
members are considered to be especially important. González et al.

260
 4 Contemporary explorations of friendships

were, however, keen to point out that there were important similarities
between the two cultures, and, more importantly, that no judgement
should be made about whether individualist or collectivist cultures
foster ‘better’ styles of friendship.
While González et al.’s study provides some important insights into the
importance of considering culture when investigating friendship, other
studies have shown that the categorisation of cultures as individualist or
collectivist may be too broad. A subsequent project led by Doran
French examined the proposition that friendships in collectivist cultures
are less extensive but more intimate than those found in cultures
shaped by individualism. French et al. (2006) compared the friendships
of Indonesian, South Korean and US college students. They used a
research approach that required participants to keep a record, over a
two-week period, of social interactions with friends. Participants were
also asked to rate the quality of their relationship with their two best
friends. The responses from students in the US, Indonesia and South
Korea were then compared using eight different criteria, such as
‘intimate disclosure’ (the extent to which they confided in the friend) or
‘exclusivity’ (the extent to which they would rather spend time alone
with the friend). French et al. proposed that if the concepts of
collectivism and individualism are useful for understanding friendship,
then the friendship patterns of Indonesian and South Korean college
students should be similar, as both South Korea and Indonesia are
characterised as collectivist cultures. In turn, these should be different
from friendship patterns among American college students owing to the
cultural focus on individualism in the United States.
Their findings, however, suggested something different. Indonesian and
South Korean students differed from each other on seven of the eight
criteria that were used for comparison. Moreover, both were more
similar to the American students than to each other. This suggests that
categories such as ‘collectivist’ and ‘individualist’ may be too general to
adequately capture the subtle and varied ways in which culture
influences friendships. French et al. therefore concluded that friendship
research needs to be more sensitive to the complex ways in which
cultures differ from each other and how these differences may affect
expectations of friendship.

261
Chapter 6 Making friends

Summary
. Findings from studies on friendship suggest that the influence of
peers is both subtle and complex.
. Cultural influences may have an important role in shaping our
understandings of friendship.
. Concepts such as individualism and collectivism that are sometimes
used when exploring cultural differences may be too general to
capture the subtle influence of culture on human behaviour and
social interaction.

262
 5 A new direction for friendship research?

5 A new direction for friendship

research?
In this chapter I have considered the concept of friendship, and have
paid close attention to the methods used by researchers in investigating
people’s experiences and expectations of friendships. While keeping
methods in mind, I would now like to broaden our focus and look at
the different ways that people may experience interactions and
friendships, paying particularly close attention to the role of new
technologies.

5.1 The changing nature of friendships: the role of

new technologies
The development of new technologies has had a significant impact on
the way that people choose to interact with each other. The availability
of mobile phones with texting facilities, for example, makes it easier to
send short messages to other people without the need for what could
be a lengthy telephone conversation. Similarly, the development of
internet technologies has meant that people can communicate online
through a range of tools such as email, web forums, real-time chat and
webcam video conferencing. Such developments mean that it has
become easier to maintain and possibly develop relationships over long
distances. The use of such tools therefore questions some early
assumptions regarding the nature of friendship and a requirement for
geographical closeness in the development and maintenance of a
relationship. Given the increasingly important role that the internet has
come to play over recent years, I would like to pay particular attention
to the role of the internet in the development and maintenance of
online friendships.

263
Chapter 6 Making friends

Figure 6.8 Interacting with friends, online

Activity 6.5
Spend a couple of minutes considering the following questions. Have you
used internet technologies for either friendship formation or friendship
maintenance? Do you feel that people can be as intimate in online
relationships as they can be in face-to-face relationships? Do you feel
that certain types of people would be drawn to this type of friendship?
What are some of the positives and drawbacks offered by the use of
such new technologies?
I will explore these issues in the rest of this section.

Following the rapid uptake of internet technologies and the dominant

role that they play in some people’s lives, there have been debates
concerning whether the use of the internet will isolate individuals or act
as a way of connecting them to others. If such technologies do act as
ways of developing and confirming social ties, then will online
relationships be as intimate and close as those offline? Major questions
have been raised concerning the quality of online friendships owing to
the lack of non-verbal social cues that many of us rely on in the
development and maintenance of relationships. For example, common

264
 5 A new direction for friendship research?

sources of confusion for online interactions are messages employing

humour or sarcasm. However, while it could be argued that the absence
of non-verbal cues such as nodding in agreement, smiling and so on
could be a problem for online friendships, there is a whole system of
contextual cues that have been developed in order to help facilitate
online relationships and overcome the potential difficulties that may be
caused by the lack of face-to-face non-verbal cues. These cues are
referred to as emoticons and serve to provide further clues about the
intended meaning of the message. For example, any potential
misunderstanding about the meaning of a message can be avoided with
the use of a smiley :-) or a winking smiley ;-). These symbols appear to
be an important resource for internet users in the absence of non-verbal
cues and voice tones that guide interactions in face-to-face settings.
Other resources include a common pool of acronyms, such as LOL
(laugh out loud), which also compensate for the lack of non-verbal cues.

5.2 A new kind of friendship?

The first decade of the twenty-first century saw the emergence of a new
form of online communication: social networking websites. These sites
have created new ways of developing and maintaining friendships over
the internet. Social networking sites facilitate contact between people
who share interests, and enable them to interact with each other and
exchange information and images. They have a whole host of features
that enable users to communicate with each other. For example, users
can post information about themselves, upload photographs, and
receive and send messages to friends. They can also locate other users
of the site with whom they may have been friends in the past but have
lost touch. Such sites therefore make it possible for people to develop,
rekindle and maintain friendships without the need for geographical
closeness.
Social networking sites have seen a rapid growth in use since their
emergence in the early 2000s. According to a press release issued by
Facebook (one of the more popular social networking sites), in
September 2009 there were an estimated 300 million active users of
Facebook, with more than 150 million users logging on to their account
every day (Facebook, 2009).
In a study that examined the uses of social networking sites, John
Raacke and Jennifer Bonds-Raacke (2008) found that US college
students reported spending a significant amount of their time on these

265
Chapter 6 Making friends

sites, leading the authors to conclude that the sites must therefore be
meeting the personal and social needs of users. Participants in Raacke
and Bonds-Raacke’s study reported that online social networking is
especially important for keeping in touch with both current and old
friends, and making new friends, indicating that the users of such
technologies considered such online relationships to fit within the
category of ‘friendship’. The increase in the use of new technologies
may therefore lead us to question our earlier understandings of
friendship, in that geographical proximity and face-to-face interaction
are clearly no longer as important a factor in friendship as they once
were.
Users of online social networking may report having several hundred
‘friends’ with whom they interact. However, can we really be friends
with so many people in the same way that we can be with a smaller
social group? Maybe our language cannot differentiate between online
social networking friends and other friends? This leads to a potential
problem for the social networking sites. What do they call people who
interact online? One possibility is for sites to grade friendships, whereby
users can opt for a certain type of communication and higher level of
disclosure with close friends while having a different, more ‘shallow’,
interaction with people who are more likely to be referred to in face-to­
face interactions as ‘acquaintances’. In any case, the possibilities that
online networking provides for forming and maintaining relationships
require us to think very carefully about what the terms ‘friend’ and
‘friendship’ mean.
There have also been several questions posed, particularly in the media,
concerning the potentially damaging consequences of excessive
engagement with social networking sites. Does the increased
communication with friends online lead to a loss of the ability to
communicate in face-to-face situations? Recent research by Raacke and
Bonds-Raacke (2008) suggests that users of social networking sites
consider their online relationships as an important form of friendship.
At the same time, these do not preclude or impair the development and
maintenance of offline relationships. In fact, individuals use new
technologies in very different ways. Some may choose to tap into the
benefits of online developments as a way of maintaining a current
relationship over geographical distance, while others may initially meet
online and then, as their friendship develops, meet in the ‘offline world’
as well.

266
 5 A new direction for friendship research?

5.3 A new kind of method?

Online friendships also lead us to return to two other important themes
that I have focused on in this chapter: those of method and the
importance of culture in our understandings of friendship. New ways of
expressing and experiencing friendships may require new methods
through which to investigate friendships. Researching on the internet
has many advantages, including the opportunity to interact with and
gather data from people from a wider range of geographical locations
than would be possible with more traditional data collection methods.
This means, for example, that researchers could extend the work of
Bigelow and La Gaipa or González et al. and ask lots of people from
very different places around the world to write an essay about their best
friend and email it to them. Researchers could also use a more
qualitative ethnographic method inspired by, for example, the work of
Corsaro, and become members of online groups and chat to people
online about their friendships. However, any work online raises the
difficulty that researchers cannot always be certain about who their
participants are, given that people can adopt whatever online identity
they choose.
It is also necessary to consider the important cultural variations in
understanding friendship and how online friendships reflect these. Much
of the communication on the internet is still conducted in English, and
research in this area tends to be limited to the English-speaking world.
What about the parts of the world that do not communicate in English
or which currently do not have widespread access to the internet? If
research were limited to online methods, these populations would be
excluded.
These issues cannot be easily resolved, but should nevertheless be
acknowledged and borne in mind by researchers. Nevertheless, given the
increasingly important role that the internet plays in the lives of many
individuals, the development of global online friendship networks
remains a possible exciting new avenue for future friendship research.

5.4 Conclusion
In order to understand the complexities involved in the formation and
maintenance of friendships in children and adolescents, psychologists
need to adopt a sophisticated research approach. They need to take into
account the limits to children’s abilities to express themselves through

267
Chapter 6 Making friends

written language or interviews, and pay greater consideration to the way

culture can influence what is expected of friends. Also, when
considering the influence of peers on behaviour, it is important to view
children and adolescents as active parties in the friendship rather than
as passive recipients, easily influenced by others.
The pioneering work of Bigelow and La Gaipa has played a central role
in highlighting the developmental shifts in friendship expectations and
the changing understandings of friendship as children mature. However,
at the time when Bigelow and La Gaipa conducted their research, they
could not have anticipated that three decades later children would be as
likely to interact with friends online as they are in the playground and
that the meaning of the word ‘friend’ would be extended to cover these
online relationships. This is yet another example of the way in which
technological advancement and the changing nature of human
interaction constantly throws up new challenges for psychologists to
address.

Summary
. New technologies have become more widely available in recent
times, and may necessitate a revision of the concept of friendship.
. The popularity of online networking opens up the possibility that
some of the traditional aspects of friendship, such as geographical
closeness, may become less important.
. These new ways of interacting with friends may mean that new
methods for investigating friendship will need to be developed.

268
 References

References
Bigelow, B.J. and La Gaipa, J. (1975) ‘Children’s written descriptions of
friendship: a multidimensional analysis’, Developmental Psychology, vol. 11, no. 6,
pp. 857–8.
Bukowski, W.M., Newcomb, A.F. and Hartup, W.W. (1996) ‘Friendship and its
significance in childhood and adolescence: introduction and comment’ in
Bukowski, W.M., Newcomb, A.F. and Hartup, W.W. (eds) The Company They
Keep. Friendship in Childhood and Adolescence, New York, NY, Cambridge
University Press.
Corsaro, W.A. (1985) Friendship and Peer Culture in the Early Years, New Jersey,
NJ, Ablex Publishing Corporation.
Corsaro, W.A. (2006) ‘Qualitative research on children’s peer relations in
cultural context’ in Chen, X., French, D.C. and Schneider, B.H. (eds), Peer
Relationships in Cultural Context, New York, NY, Cambridge University Press.
Damon, W. (1977) The Social World of the Child, San Francisco, CA, Jossey-Bass.
Erwin, P. (1998) Friendship in Childhood and Adolescence, Psychology Focus Series,
London, Routledge.
Facebook (2009) Pressroom [online], http://www.facebook.com/press/info.php?
statistics (Accessed 25 September 2009).
French, D.C., Bae, A., Pidada, S. and Lee, O. (2006) ‘Friendships of
Indonesian, South Korean and US college students’, Personal Relationships,
vol. 13, no. 1, pp. 69–81.
González, Y.S., Moreno, D.S. and Schneider, B.H. (2004) ‘Friendship
expectations of early adolescents in Cuba and Canada’, Journal of Cross-Cultural
Psychology, vol. 35, no. 4, pp. 436–45.
Hartup, W.W. (1996) ‘The company they keep: friendships and their
developmental significance’, Child Development, vol. 67, no. 1, pp. 1–13.
McLeod, K., White, V., Mullins, R., Davey, C., Wakefield, M. and Hill, D.
(2008) ‘How do friends influence smoking uptake? Findings from qualitative
interviews with identical twins’, The Journal of Genetic Psychology, vol. 169, no. 2,
pp. 117–31.
Raacke, J. and Bonds-Raacke, J. (2008) ‘MySpace and Facebook: applying the
uses and gratifications theory to exploring friend-networking sites’,
CyberPsychology and Behaviour, vol. 11, no. 2, pp. 169–74.

269
 Conclusion

Conclusion
In Part 2 of Investigating Psychology you read about the work of B.F.
Skinner on learning, about Harry Harlow’s experiments on attachment
in monkeys and about Brian Bigelow and John La Gaipa’s study on
children’s friendships. All three chapters addressed a similar aim:
understanding various influences on human behaviour. The chapters
examined how behaviour is shaped through specific reinforcement
schedules, how it is influenced by innate factors that determine the
relationship with – and attachment to – caregivers in infancy, and how
our lives are shaped through the experience of a different kind of
attachment which is formed in and throughout childhood.
An important point to emerge from Part 2 is that research conducted
by psychologists does not just involve human participants but may also
centre on non-human animals. Psychologists study animals for different
reasons. The research described in Chapter 4 investigated a very basic
form of learning which is thought to apply to most species. The
principles of behaviour shaping, uncovered in the Skinner box using
rats and pigeons, are thought to apply beyond the laboratory and to
humans too. Primary attachment investigated by Harlow was also
believed to constitute an evolved behavioural pattern, which meant that
inferences about human attachment could be drawn from animal
research.
It is worth noting, however, that psychologists sometimes study animals
also for practical reasons. Harry Harlow’s experiments described in
Chapter 5 could not have been conducted on human infants for reasons
of ethics. Although even today psychologists sometimes use animals as
substitutes for humans, animal research is regulated by strict ethical
guidelines that take into account issues of animal welfare. So, many of
Harlow’s experiments would be deemed unethical today, just as it is now
no longer possible to replicate the work of Stanley Milgram described in
Chapter 2.
Like Part 1 of Investigating Psychology, Part 2 underlines the fact that
researchers are often influenced by the broader social and intellectual
context. Skinner’s work was influenced by the findings of Ivan Pavlov,
John B. Watson and Edward Thorndike, and especially by Watson’s view
that psychology should be an objective science and concern itself only
with observable and measurable behaviour. Harlow’s research, on the
other hand, should be viewed in the light of John Bowlby’s notion that

271
Part 2 Conclusion

the attachment between a child and its parents is based on the provision
of comfort, while Bowlby in turn was influenced by the work of
ethologists on innate predispositions in animals. The research by
Bigelow and La Gaipa was, admittedly, not influenced by any particular
key idea of their time. Instead it was the lack of research on children’s
friendship that prompted the two psychologists to carry out their
pioneering research. However, the more recent studies on online
friendships are very much influenced by a dramatic development in
contemporary society, namely the arrival of computers and the internet
in homes in the early twenty-first century.
In Part 2 you also learnt that research is sometimes influenced by
chance circumstances. Skinner decided to test a partial reinforcement
schedule because he ran short of food pellets. Harlow only came to
study attachment in rhesus monkeys because the rat laboratory had
been dismantled at the university where he worked, and he could not
study intelligence in rats as he had planned. What is more, had he not
noticed, purely by chance, that infant monkeys were distressed when the
soft lining of the cage was removed during cleaning, he probably never
would have embarked on the study of attachment.
The three chapters in Part 2 of Investigating Psychology offered you further
tasters of the different sub-disciplines of psychology. Chapters 5 and 6
(just like Chapter 3 in Part 1) looked at research that belongs to the
domain of developmental psychology, in that the studies looked at children’s
behaviour and their psychological development. Chapter 5 also
introduced you to evolutionary psychology, which focuses on the application
of evolutionary principles to explain human behaviour. In Chapter 4
you encountered an approach in psychology called behaviourism, which
was one of several dominant schools of thought in the first half of the
twentieth century, and one that involved animal research.
The chapters also offered further insight into the range of methods used
by psychologists. In Chapters 4 and 5 you read about the use of
laboratory experiments, but this time involving animals rather than
human participants. In Chapter 5 you also learned about the importance
of interviews and observations when studying attachment in humans.
These two methods of collecting data have proved especially useful
when exploring behaviour that cannot be studied experimentally, for
ethical and other reasons. Chapter 6, on the other hand, highlighted the
important role that qualitative data plays in psychological research.
Although Bigelow and La Gaipa coded their qualitative data, thereby
transforming children’s textual responses into numbers, other

272
 Conclusion

approaches to studying friendship involve analysing children’s accounts

in ways that allow the individuality and uniqueness of each child’s
experience to be preserved.
The fact that psychologists may use very different approaches and
methods when exploring the same topic is important to bear in mind,
as this diversity within the discipline helps to build a more complete
understanding of human behaviour and its complexity. For example, the
exploration of the similarities between humans and non-human animals
with regard to learning and attachment led to important insights into
the evolved, or innate, aspects of human behaviour. At the same time,
the work on human participants highlighted the importance of culture
and environmental factors on behaviour, including attachment. Cross­
cultural research on attachment, for instance, revealed that the evolved
aspects of human behaviour invariably interact with outside influences,
including culture. Specific attachment patterns (as well as assumptions
about what constitutes optimal attachment) are products of culture and
the environment, and not just of evolution. Also, rather than being
simply determined by innate factors, human behaviour has been shown
to be flexible and adaptive. It is therefore the interaction between the
two influences – between nature (evolved or inherited predispositions)
and nurture (the culture and the environment) that determines who we
are and what we do.
Finally, there is something else that brings together many studies
presented in Part 2. Research on conditioning animals, studies on the
evolution of attachment and examinations of children’s understanding of
friendship all focused either on visible behaviour or on language
produced during interactions or play. None of the studies discussed in
the three chapters were concerned specifically with what happens in a
person’s (or an animal’s) brain or what mental processes are taking place
as they perform the various behaviours or interact with others. This is
the focus of Part 3, where you will read about how psychologists
investigate the basic mental functions and processes without which
interaction with the environment would be difficult, if not impossible.
This poses a broader question, namely how do psychologists explore
something that they cannot directly observe? How do they investigate
the inner workings of things such as language, attention and memory?

273
Part 3
 Introduction

Introduction
In several places in Parts 1 and 2 of Investigating Psychology you
encountered the notion of cognitive processes. The word ‘cognition’ refers
to a whole range of internal mental processes that allow us to take in,
store and use information or knowledge about the world around us.
These processes include attention, perception, memory, reasoning and
language. Studying these processes can help us understand a range of
everyday experiences, some of which you will read about in Part 3. In
Chapters 7–9 you will learn about the processes involved in using
language to communicate, why it is not a good idea to drive while using
a mobile phone, and how it is that some might remember an event that
never actually happened. As cognition is the focus of Part 3, let us look
back at what you have read about it already, in Parts 1 and 2.
In Chapter 1 you were introduced to the notion of cognitive style – a
particular way of structuring and processing information. Individuals
high in dogmatism were said to possess a rigid and closed-minded
cognitive style which means that they are swayed more by who is
presenting the message (an authority figure, for example) than by what
information it actually contains. So, a personality attribute (dogmatism)
was associated with a particular way of thinking and reasoning.
In Chapter 3 you read that one mechanism which is thought to
underpin social learning involves a change in a person’s attitudes, beliefs
and judgements about aggression and its appropriateness. This
mechanism, which concerns a change in a person’s thinking processes as
a result of experience (for instance, desensitisation), was referred to as a
cognitive mechanism.
Also, in Chapter 4 you read that behaviourism advocated a psychology
that looked at what was observable and could properly be measured,
namely behaviour. Behaviourism was therefore not interested in mental
processes, or cognition, at all. However, other researchers looking at
animal behaviour found it difficult to avoid using mental terms to
explain learning. Furthermore, the complexity of human behaviour is
difficult to account for without any recourse to mental processes. In
humans, what counts as ‘reinforcement’ or ‘punishment’ depends on the
way a person perceives and thinks about their environment.
From these three examples alone, you will have a sense of the
importance of things that cannot be directly observed. What this

277
Part 3 Introduction

suggests is that psychologists need to investigate not just what an

individual does, but also what happens ‘inside their head’.
Part 3 will begin to explore cognition by looking literally at what
happens ‘inside the head’. Cognitive processes are all located in a single
organ: the brain. The specific focus of Chapter 7, by Frederick Toates,
is the role that different brain regions play in controlling the ability to
comprehend and use language. You will read about the discoveries made
by two important nineteenth-century figures, Pierre Paul Broca and Carl
Wernicke, who studied damaged brains and who identified two
particular areas of the brain, one involved in the production of speech
and the other in the comprehension of language. Since their pioneering
work, a more complex model of how language is regulated by the brain
has emerged. You will read about the technological advances that made
it possible for researchers to ‘see’ inside the working brain, leading to
new insights into how the brain controls language, and the ability of the
brain to recover from damage.
Chapters 8 and 9 then move on to explore research that used
experiments to look at two different cognitive functions: attention and
memory. Chapter 8, by Helen Edgar and Graham Edgar, introduces the
work of Donald Broadbent, who is recognised as one of the founders
of modern cognitive psychology. Broadbent was interested in the
cognitive process of attention and the fact that our ability to attend to
stimuli in the environment is limited: we cannot attend to everything
around us. Broadbent sought to develop a model that would describe
and explain how the human mind processes information. Put simply, he
was interested in the ‘software’ of the mind rather than the ‘hardware’
of the brain. In the chapter you will read about how Broadbent and
other psychologists used experiments to create, test and revise a model
of attention. You will also read about how this research can be applied
to a practical real-world issue, namely the question of whether or not it
is dangerous to use a mobile phone while driving.
The final chapter of Part 3 looks at research on human memory.
Chapter 9, by Graham Pike and Nicola Brace, focuses on one particular
aspect of memory, memory for events, and considers research that has
relevance for those witnessing a crime. It is difficult to imagine a
situation where remembering accurately is as important as when one is
summoned to testify as an eyewitness to a crime. Understanding what
sorts of factors influence memory of events can provide valuable
information for the legal profession. What is more, it can also
contribute to our understanding of how memory operates more

278
 Introduction

generally. The chapter discusses the work of Elizabeth Loftus and

colleagues, and describes some widely cited experiments on leading
questions and false memories. The chapter also introduces an important
study by Sir Frederic Bartlett, tutor of Donald Broadbent and another
important forefather of modern cognitive psychology.
Part 3 therefore takes you through different approaches to studying
what happens ‘inside the head’, from different techniques for studying
the brain to experiments testing ideas about attention or memory. In
this part of the book, you will also learn about the relevance of research
on brain function and cognition to a selection of real-life problems.

279
Chapter 7
Language and the brain
Frederick Toates
Contents
1 Introduction 285
1.1 The case of stroke 285
1.2 Understanding and misunderstanding 288
1.3 Phrenology 289
2 Two dominant figures: Broca and Wernicke 292
2.1 Pierre Paul Broca 292
2.2 Carl Wernicke 298
3 The assumption of localisation of function 300
3.1 Support for localisation of function 300
3.2 Limits to localisation of function 302
3.3 The use of analogies 303
4 Recent developments in neuropsychology 305
4.1 Techniques for studying the brain 306
4.2 Tracing the links between brain regions that contribute
to language 308
4.3 Language beyond speech 310
4.4 Investigating flexibility 311
5 Contemporary relevance 315
5.1 A two-way street 315
5.2 Conclusion 317
References 318
Chapter 7 Language and the brain

Aims and objectives

After reading this chapter you should be able to:
. describe what is meant by the term ‘stroke’ and understand the
disturbances to language that often follow it
. explain what is meant by the assertion that a region of the brain is
associated with a particular psychological function
. describe Broca’s and Wernicke’s contributions to our understanding
of the relationship between the brain and language
. distinguish between Broca’s aphasia and Wernicke’s aphasia, with
regard to brain regions and disruption to language
. explain what is meant by flexibility and plasticity of the brain
. outline how technical advances have enabled new insights into the
psychology of language.

284
 1 Introduction

1 Introduction
This chapter is about understanding the brain and its role in language.
On hearing the word ‘language’, most of us probably think first of
listening to and understanding spoken words, or retrieving the right
words from memory and uttering them using appropriate sounds.
Listening and speaking are, however, not the only means of using
language. There is written language of course, and sign language. Sign
language, as used in communication with deaf people, is a form that
does not involve either spoken or written words. An important question
for psychologists has been whether the relationship between the brain
and language is the same for any system of language, irrespective of its
means of expression, and this is a question to which I shall return later
on in the chapter.
A valuable source of insight into the link between language and the
brain has been the study of what happens when things go wrong; that
is, when language impairment occurs as a result of brain damage. Such
damage can occur for various reasons, including traffic accidents and
war wounds. But the most common cause in the UK is stroke.

1.1 The case of stroke

Suddenly the light went out. Seven hours later I woke up in

hospital. I couldn’t move my right side, and my speech had gone.
(David Diston, stroke survivor,
quoted in The Stroke Association, 2009, p. 3)

Stroke
In the UK, about 150,000 people suffer from a stroke each year. The A loss of supply of
term stroke refers to injury to the brain caused by an interruption of blood to a region in the
the blood supply to the cells in the brain. The body is made up of tiny brain, resulting in
components called cells and the brain itself contains billions of disruption of brain
specialised cells called neurons. Each cell throughout the body, function.
including neurons, requires a supply of nutrients and oxygen in order to
survive and to perform its function. These are brought to each cell by
Neuron
means of blood travelling through the network of blood vessels. If A type of cell involved
deprived of this supply, there is a danger that a cell will die. in transmitting
messages between
different parts of the
brain.

285
Chapter 7 Language and the brain

FRONT

arteries

RIGHT LEFT
BRAIN BRAIN

BACK

Figure 7.1 A drawing of the human brain (viewed from below) showing the
blood vessels running through it

Figure 7.1 shows the network of blood vessels that supply nutrients and
oxygen to the cells of the brain. If there is a loss of supply of blood to
a region of the brain, the cells there will cease to function and can soon
die. The resulting disruption of brain function is called a stroke.
A stroke can be caused by the blockage of a blood vessel, as when fatty
substances build up on its wall, or when the wall of a vessel breaks (see
Figure 7.2). Loss of function of the cells in a brain region means
disruption to the feature of behaviour controlled by that region. So, if
the region controls an arm, loss of its blood supply will disrupt the
ability to raise that arm. If the region controls language, then there will
be difficulty in understanding spoken words, or speech will be lost or
slurred. In some cases, the disruption is so extensive that several such
features of behaviour or functions will be disturbed.

286
 1 Introduction

brain stagnant
cells blood

vessel fatty
rupture
deposit
of
wall
blood flow blood flow blood flow
(a) (b) (c)

Figure 7.2 The basis of a stroke: (a) intact vessel; (b) a blood vessel
blocked by fatty substance; (c) rupturing of vessel wall

Although not all victims of stroke will exhibit the same symptoms, most
will experience some disturbance to language and to the movement of
the arm and/or facial muscles. This is why The Stroke Association
(2009, p. 6) gives the guidance to act FAST when you suspect that
someone has suffered from a stroke, where FAST is not only advice on
speed but also an acronym for the Face-Arm-Speech Test:
F Facial weakness: Can the person smile? Has their mouth or an eye
drooped?
A Arm weakness: Can the person raise both arms?
S Speech problems: Can the person speak clearly and understand what
you say?
T Test these symptoms.

Activity 7.1
If you see someone having a stroke, the priority is to get them to a
hospital as soon as possible. The doctor will seek to maximise the
chances that the individual affected survives and returns to a functioning
life. But where do psychologists come in? Spend a couple of minutes
thinking about why psychologists might be interested in stroke and how
they might become involved in the recovery process.

First, on recovery, stroke patients might find themselves disabled in

terms of not being able to move limbs or speak. They might suffer

287
Chapter 7 Language and the brain

psychological changes, including depression. Psychologists often help the

rehabilitation of stroke patients by informing their therapy. Second, by
studying the effects of a stroke, or another form of brain injury that
affects behaviour or performance, psychologists can develop a better
understanding of how behaviour and cognitive processes such as
language and memory depend upon what goes on in the brain.
This chapter is primarily concerned with the latter issue and addresses
two specific questions using the example of language. First, it examines
how researchers arrived at the view that the brain forms the basis of
psychological processes and, second, it explores how knowledge of the
brain helps us to understand how these processes are controlled.

1.2 Understanding and misunderstanding

The effect of a stroke, in terms of disruption to speech and movement,
is widely known. Other disorders also point to the brain. Disruption of
memory, as in dementia, follows deterioration of the brain. The
expression ‘brain dead’ seems to be synonymous with the end of life.
The idea that the brain forms the basis of our thought processes and
behaviour is so much a part of our understanding of the human
condition that it is hard to imagine anyone could ever have believed
otherwise.
However, the ancient Egyptians thought that the ‘soul’ was located in
the bowels and heart (Finger, 2000). When preserving the bodies of
their dead leaders, with all the reverence that was appropriate for their
journey to the afterlife, the brain was unceremoniously extracted
through the nostrils. Strangely enough, in spite of taking such a ‘dim
view’ of the brain, the Egyptians recognised that injury to it could have
various consequences, such as disruption to coordination between the
eyes and the hands. Quite brilliantly, they accurately observed that
damage to one side of the brain usually had more serious consequences
for the opposite side of the body.
To associate the mind with the brain requires taking a journey of several
hundreds of years and to the other side of the Mediterranean. Some
400 years BC in ancient Greece, the followers of Hippocrates (the man
widely regarded as the father of medicine) hit upon the truth. By
carefully observing human bodies in illness and health, and scrutinising
the development, spread and course of different illnesses, they
challenged the notion that illness was a punishment sent by the gods.
Rather, they explained illnesses in natural terms. ‘Madness’ and epilepsy

288
 1 Introduction

were, for instance, described as manifestations of abnormality of the

brain.
If abnormalities of behaviour are seen as afflictions sent by the gods,
people can argue that these are punishments for wrongdoing. If they are
instead understood as the product of unusual brain activity, then
perhaps people are more likely to show tolerance. In seventeenth­
century America, however, uncontrolled twitching and jerking was seen
as evidence not of divine punishment but of possession by demons.
Those with this condition, particularly women, risked execution as
witches. It is now believed that these unfortunate people suffered from
Huntington’s disease (Vessie, 1932). This disorder of brain regions
involved in controlling the muscles is inherited. Those suffering from
the condition in America apparently all derived from a family group
who emigrated in 1630 from Bures in Suffolk.

1.3 Phrenology
In modern times, the interest in the brain dates back to the late
eighteenth century, when the Viennese doctor Franz Joseph Gall
proposed that different parts of the outer region of the brain serve
different roles or functions. He argued that each part of the brain has
responsibility for the control of a particular behaviour or feature of
mind. He was not the first to state this belief, the credit going to a
Swedish Christian mystic, Swedenborg, but Gall was the first to gain
wide recognition for advancing it. Phrenology
The study of the
According to this view, memory, for example, would be stored in one contours of the head
brain region, whereas a different region would be responsible for and trying to link this
romantic love. This theory was called phrenology. Gall devised a to the function of the
various regions of
phrenology map which proposed the functional anatomy of the brain.
brain. In large part, it is
The term ‘anatomy’ refers to the structure of the body, so the anatomy now discredited.
of the brain is a description of its structure involving its appearance and
the location of such things as contours and blood vessels. The word
‘function’ refers to what different bits of the brain do, or the role they Functional anatomy
play, with regard to behaviour, cognitive processes or individual A description of the
characteristics. So, the phrenology map charted the brain regions and regions of the brain in
terms of the function
their roles, identifying, for example, parts involved in ‘language’ or that they serve in the
‘benevolence’ (see Figure 7.3). control of cognitive
processes,
characteristics and
behaviour.

289
Chapter 7 Language and the brain

Figure 7.3 A phrenology map

Gall challenged two prevailing views. First, the religious authorities in

Austria objected to his attempts to link personality characteristics to a
material substance, namely the brain. They preferred to see them as
features of an immaterial soul. Second, there was a disagreement with
those who did accept that the brain was the basis of mental processes,
but who argued that the whole brain contributed equally to every
psychological activity (Graves, 1997).
Gall’s idea contained a considerable amount of truth. Distinct bits of
the brain do serve different functions. A given region of the brain tends
to be associated with the same function in different individuals. This
means regions can be labelled in terms of what they do. However, a
region is rarely responsible for just one function. Rather, it is now known
that regions of the brain interact or work together in determining
psychological processes.

290
 1 Introduction

Gall also went ‘to excess’ in applying his approach. He found people
whose behaviour deviated from the norm (e.g. geniuses, the mentally ill
and murderers), compared their heads with those of so-called normal
individuals, and claimed to have found differences that explain deviance.
Gall asserted that unusual visible features of the head corresponded to
unusual structures of the brain and thereby to the production of
unusual behaviour. So, for example, a particularly amorous lady friend
was said to have an abnormally large ‘amorous region’ of her brain,
visible on the contours of the skull. Such unwarranted speculation led
phrenology into disrepute. Nevertheless, the idea that different brain
regions serve different roles was a sound one. Gall had rightly noted,
for example, that loss of speech can arise from damage to the front half
of the brain (Schiller, 1979). This leads us to the important
contributions of Pierre Paul Broca and Carl Wernicke to our
understanding of the relationship between the brain and language. Their
work is described in the next section.

Summary
. It is now widely accepted that the brain forms the basis of cognitive
processes and behaviour.
. In stroke, a disruption to behaviour, including speech, can arise
from damage to a part of the brain.
. The study of brains damaged by stroke can give insights into the
normal functioning of the brain.
. Phrenology was a theory that associated particular parts of the brain
with specific psychological functions. It stood in opposition to two
views: (a) that psychological processes are not rooted in the brain;
and (b) that all parts of the brain have equal responsibility for
different psychological processes.

291
Chapter 7 Language and the brain

2 Two dominant figures: Broca and

Wernicke
Understanding the links between brain and language owes much to two
dominant figures, Broca and Wernicke. Although neither of them was a
psychologist, they were a major influence on the development of
Neuropsychology neuropsychology, which is a branch of psychology concerned with the
The sub-discipline of relationship between the brain and psychological processes.
psychology concerned
with studying the
structure and function 2.1 Pierre Paul Broca
of the brain, and
therefore the Pierre Paul Broca was born in 1824 in the small French town of Sainte­
relationship between Foy-la-Grande. On leaving school, he studied medicine in Paris. Broca
the brain and made contributions in a wide range of medical and related fields but I
psychological processes. shall concentrate only on his contribution to the study of language.
Influences on his scientific career included François Leuret, one of
Broca’s teachers of medicine. Leuret studied the folds (‘grooves and
ridges’) that are clearly visible on the outer surface of much of the brain
(see Figure 7.4). He advocated the view that their pattern is not
haphazard. A study of the brains of humans and other species revealed
a pattern that was consistent for and characteristic of each species.

FRONT

BACK

Figure 7.4 The outer surface of the brain (shown from above)

292
 2 Two dominant figures: Broca and Wernicke

Another of Broca’s teachers reiterated the belief of Gall that speech was
impaired by damage specifically to the front half of the brain. This was
based upon the frequent observation that damage to the front of the
brain was associated with disruption to speech while damage to the rear
tended to leave speech unaffected. However, the evidence for the latter
proved to be less sound – some researchers found that disruption to
speech sometimes occurred even when damage was to the rear part of
the brain.
Those who were sceptical that the front part of the brain is involved in
speech cited the case of Phineas Gage, a man who possessed what
might be the most famous brain in all history (the candidate for second
place will be described in a moment). In 1848, Phineas Gage was
working on the construction of a railway in Vermont, USA when an
explosion sent an iron bar through part of the front of his brain (see
Figure 7.5). Miraculously Gage survived and, although he suffered from
changes of personality, his capacity for language, in terms of
understanding and articulation, was left intact.

Figure 7.5 A reconstruction of the damage to the brain of Phineas Gage

293
Chapter 7 Language and the brain

Based upon such evidence, a more nuanced claim emerged: the effect of
brain damage depends upon the exact site of damage within the brain. So,
speech could be disrupted while such actions as walking or non-speech­
related movements of the mouth and tongue would remain intact. This
was believed to be because damage to certain parts of the brain led to
specific impairments related to speech, while other functions were left
unimpaired.
Broca and other researchers studied the brain by dissecting the brains
of deceased people. They were especially interested in the outer layer of
Cortex the brain, termed the cortex (see Figure 7.6).
The outer layer of the
brain.
cortex

Figure 7.6 The brain, in part cut away to reveal the cortex. The cortex is
formed of grey matter with white matter lying underneath

Broca made his most famous observation, which is the focus of this
chapter, in 1861 while examining the brain of a 51-year-old man, Mr
Leborgne, who died while under the care of Broca at the Bicêtre
hospital in Paris (Broca, 1861). This unfortunate man had been without
the capacity of speech for the last twenty-one years of his life and there
was impairment in his ability to move his right arm. To any question,
he gave the spoken response ‘tan’, and thereby acquired the name of
‘Tan’ in the hospital. By means of the gestures that the patient made
with his left hand and its individual fingers, Broca concluded that Mr

294
 2 Two dominant figures: Broca and Wernicke

Leborgne was able to understand speech. At autopsy, Broca noted

disruption to tissue in a region on the left side of the brain towards the
front (see Figure 7.7). The left and right sides of the brain (as shown in
Figure 7.1) are commonly referred to as the left and right
hemispheres. Broca concluded that this specific location in the left Hemisphere
hemisphere was responsible for speech production. The left and right
hemispheres are the left
Observation of other patients with loss of the ability to speak followed, and right sides of the
and the autopsies demonstrated a similar disruption to the same area of brain.
the left side of the brain. In several cases studied by Broca, there was
also disruption of the ability to move the right arm, the significance of
which will become apparent later. Then, as now, it was rare that damage
to the right hemisphere of the brain led to major disruption of speech.
Underpinning Broca’s conclusion was a very basic principle, the
principle of localisation. According to this principle, different regions Principle of
of the brain have responsibility for different psychological functions. localisation
Also, Broca was suggesting that the brain is asymmetric in respect of The principle that
psychological functions
function: speech is controlled by a region in the left hemisphere only.
can be associated with
particular regions of
the brain.

Figure 7.7 The brain of Mr Leborgne

Activity 7.2
Take a moment to look at Figures 7.5 and 7.7 in more detail. Compare
the site of damage to the brains of Phineas Gage and Mr Leborgne.As

295
Chapter 7 Language and the brain

you can see, the damage to the brain of Phineas Gage is to an area
further to the front, compared to the hole in the brain of Mr Leborgne.
This might explain why Phineas Gage’s language ability was intact.

Between 1861 and 1865, Broca became more confident about his
findings concerning the localisation and asymmetry in the functioning of
the brain, especially as more and more evidence from deceased patients
confirmed that there was asymmetry in the control of speech, and that
disruption of speech production was associated with damage to the left
brain. However, Broca’s hypothesis of asymmetry went against the then
widely held view that the brain had to be symmetrical in both its
structure and function (Buckingham, 2006). From a religious
perspective, a perfect act of creation demanded perfect symmetry, so the
brain had to be symmetrical too.
Subsequently, the brain region that Broca identified as being involved in
speech production acquired the name of ‘Broca’s area’ (see Figure 7.8).
A disruption to language caused by damage to the brain is generally
Aphasia referred to as aphasia, so impairment resulting from damage to Broca’s
Impairment to language area has become known as Broca’s aphasia. This particular type of
or language disorder. aphasia consists of an inability to articulate speech fluently, so that
language may consist only of disjointed words. Normally, writing is also
disrupted. Understanding of language can, however, be near to normal.

Broca’s area Wernicke’s area

Figure 7.8 The left hemisphere of the brain, indicating Broca’s and
Wernicke’s areas

296
 2 Two dominant figures: Broca and Wernicke

Box 7.1 The nature of discovery

Progress in science and personal fame depend not just on making
original observations and having clever ideas. It is also important to
be in the right place at the right time, to have a university position
and to get one’s findings published in a reputable journal.
Although Broca’s name is universally associated with the formal
discovery of the principle of asymmetry, one of his fellow­
countrymen appears to have made this observation considerably
earlier than Broca. Marc Dax was a doctor in Sommières, a small
community about twenty miles from Montpellier. Dax died in 1837,
some twenty-four years before Broca made his groundbreaking
observations.
Much later, it was claimed that, in the year before his death, Dax
had presented a paper to a conference of doctors from the south of
France, in Montpellier. In this paper, he apparently described his
twenty-five years of observations of an association between loss of
speech and disruption of movement in the right arm. Dax had
concluded that in all such patients damage was to the left side of
the brain. He observed that patients who lost speech did not lose all
tongue movements, as in paralysis of the muscles controlling the
tongue. They could move their tongues but could not exploit such
movement in the service of speech. The problem, as Dax correctly
reasoned, was an inability to organise the sequence of muscular
actions that underlie speech.
Alas, Montpellier is ‘not Paris’ and Dax was a small-town doctor, so
the paper did not attract wide attention. In fact, it has proved
impossible to verify that Dax actually presented his paper at the
conference in Montpellier (Buckingham, 2006). He might have
shown it to just a few colleagues. So, how did he finally come to
national attention? In 1863, twenty-six years after his death, his
son, Dr Gustave Dax, brought his father’s paper to the Academy of
Medicine in Paris, where somewhat belatedly it received attention.
However, by the time its contents were publicised two years later,
Broca’s name had already become associated with the discovery.
Hugh Buckingham (2006, p. 618) wrote: ‘Once again in science, we
have likely witnessed another case where a previous investigator
actually hit on a concept first, only to lose credit to a later
researcher who more properly “framed” the hypothesis in a detailed,
all-encompassing publication.’

297
Chapter 7 Language and the brain

2.2 Carl Wernicke

Carl Wernicke was born in 1848 in Germany and studied medicine in
Breslau. He later served in the psychiatric department of Breslau
university hospital. Like Broca, he made original contributions in a
number of areas of medicine and psychiatry. Wernicke was interested in
the new discoveries on language coming from Paris. From the careful
study of patients, he identified another region of the brain, the
disturbance of which also led to a disruption of language. However,
while Broca was interested in patients who exhibited impairment in
speech production, but who could understand language, Wernicke
focused on those who experienced a breakdown in the ability both to
understand speech and to formulate coherent sentences. He discovered
that this impairment was associated with damage to a different area,
located towards the middle of the outer surface of the left hemisphere
of the brain. Predictably, this region of the cortex became known as
‘Wernicke’s area’ (see Figure 7.8), and impairment associated with
damage to this area is referred to as Wernicke’s aphasia.
Wernicke’s aphasia manifests itself as a breakdown in the ability to
understand speech and to formulate coherent, meaningful sentences.
Patients are often able to utter words, the grammar can often be
correct, but sentences are deficient in meaning.
Wernicke’s name has gone down in history for his account of the nature
of disruption to speech. But there are indications that, as with Broca,
others had earlier realised that speech could be disrupted by damage to
the area later to take Wernicke’s name. Indeed, one such person appears
to have been none other than Dr Gustave Dax, the man who had
advanced his father’s claim to have discovered the asymmetry of the
brain (Finger and Roe, 1999).
Wernicke’s fame did not arise only from the discovery of the role of a
specific region of brain, important as this was. In addition, he suggested
that the control of behaviour, such as language and speaking, depended
upon the interaction of identifiable regions of the brain (Pillman, 2003).
Wernicke proposed that the area bearing his name and Broca’s area
must be linked and interact with one another to produce language.
Wernicke’s area is near to the part of the cortex that analyses sounds
and is involved in understanding speech. Broca’s area, on the other
hand, is involved with the process of articulation of speech. Only
together do they act to produce normal language processing.

298
 2 Two dominant figures: Broca and Wernicke

Summary
. Broca articulated the principle of localisation.
. Broca found that damage to a region in the left hemisphere of the
brain caused disruption to the production of speech. This area came
to be called Broca’s area and the resulting speech deficit became
known as Broca’s aphasia.
. Wernicke discovered that an area of the left brain nearer the back
disrupted the comprehension of speech. This area became known as
Wernicke’s area and the resulting language deficit became known as
Wernicke’s aphasia.

299
Chapter 7 Language and the brain

3 The assumption of localisation of

function
The fundamental assumption underlying the work of Broca and
Wernicke was that identifiable parts of the brain can be associated with
particular aspects of cognition and behaviour. Closely associated with
this is the idea that one can gain useful information from studying
damaged brains, including how ‘normal’, undamaged brains function.
Also, Wernicke put forward the view that, in order to understand how
the brain works, researchers need to study how its various parts interact.

3.1 Support for localisation of function

Following Broca and Wernicke, an increasing amount of evidence from
people with brain damage pointed to particular parts of the brain being
associated with particular psychological functions. When the damage is
in the left hemisphere, for example, there is likely to be a disruption to
the production of speech or comprehension of the spoken word, or
both. Similarly, disruption to the control of limbs resulting from brain
damage will affect the opposite side of the body to the site of brain
damage. The same applies to the loss of sensation (e.g. of touch). All of
these findings provide support for the notion of localisation of function.
Ever since Wernicke’s and Broca’s pioneering work, researchers have
worked on identifying areas of the brain associated with particular
functions. For instance, they identified a specific area of the cortex
involved in motor (muscular) movement. This research confirmed what
Ancient Egyptians had already suspected, namely that muscle
movements on the left side of the body are controlled by the right
hemisphere of the brain and vice versa. This is why Broca’s patients,
who had damage to the left side of the brain (Broca’s area), often had
difficulty moving the right arm.
The area of the brain that controls muscle movement is known as the
Motor homunculus motor cortex (see Figure 7.9(a)). Yet even within this area, there is
A functional map of
further specialisation. Figure 7.9(b) is an illustration of what is known as
the motor cortex
showing the regions of a motor homunculus – a functional map of the motor cortex. It
the body over which shows each part of the body and the corresponding part of the brain
each region of motor that exerts some control over that body part. It exemplifies the principle
cortex exerts some of localisation.
control.

300
 3 The assumption of localisation of function

motor cortex fingers

es kl kne hip
to an trunk
e e
wrist

shoulder
han little

elbow
rin
g

d
m
th idd
um le
bro b in
ey w nec de
fac elid k x

vocalization
lips e an
de
ye
g ball

salivatio
jaw gue owin
ton wall

ma n
s

sti
ca
tio
n
(a) (b)

Figure 7.9 (a) the left side of the brain, highlighting the motor cortex; (b) a
section of the motor cortex and the motor homunculus

Activity 7.3
Look at Figure 7.9(b) and examine the motor homunculus and the
relationship between brain region and body part.
. Consider each region of the brain and each body part in terms of the
amount of brain devoted to it. What could be the possible significance
of this?
. What is the significance of the location of Broca’s area relative to the
parts of the motor homunculus?

Some regions of the body (e.g. the fingers) have disproportionately large
regions of the cortex devoted to their control, whereas other parts
(e.g. the hips) have relatively small areas. This reflects the relative degree
of sensitivity of control that is exerted over the different body parts.
Broca’s area is adjacent to that part of the motor cortex concerned with
controlling the mouth and tongue. Broca’s area interacts with this part
of the motor cortex, sending ‘ instructions’ for movements of the
mouth and tongue in articulating speech.

301
Chapter 7 Language and the brain

3.2 Limits to localisation of function

The evidence from brain damage and studies of the motor homunculus
provide support for the principle of localisation of function. However,
this evidence largely relates to control over parts of the body and
certain sensations. Could the principle of localisation be extended to the
whole range of behaviour and cognitive processes? Was Gall right in
that areas of the brain responsible for complex emotional processing, as
in compassion or romantic love, could be similarly identified?
Researchers did indeed subsequently find that parts deeper than the
cortex were involved in the regulation of sensations such as anger or
erotic arousal (remember that the cortex is just the top layer). However,
brain surgery raised some doubts about localisation of function
regarding the control of human emotions, and this will be outlined next.
The brains that provided so much evidence for Broca and Wernicke
received their damage from such causes as sword wounds, tumours or
stroke. Another source of information relevant to localisation arose later
in cases where parts of the brain were surgically cut or even removed.
There were two reasons why this was done. The medical need to
remove diseased brain tissue, as when trying to halt the progress of a
cancerous tumour, is surely uncontroversial. By contrast, what is highly
controversial is the technique of psychosurgery, which was practised
extensively between the 1930s and the 1950s as a treatment for
different types of mental illness. The rationale behind this was that by
removing selected brain tissue or simply making targeted surgical cuts
(usually to the region at the front of the brain), the working of the brain
would alter with subsequent benefits for the person receiving the
operation.
The results of psychosurgery were mixed. Human emotions and
psychological functioning are subtle and doubtless arise from numerous
brain regions acting in concert. It has proved very difficult to associate
surgical intervention at a particular region with a particular desired
outcome, such as to lift chronic depression or anxiety. In fact,
undesirable outcomes such as total apathy were far more common. So,
the results of psychosurgery suggested that there was no simple
relationship between specific areas of the brain and emotions such as
anxiety and anger. This is why it is rarely used today.
The poor results of psychosurgery therefore cast some doubt over the
claim that there is always a close link between a particular part of the
brain and a specific feature of behaviour. The link between brain and

302
 3 The assumption of localisation of function

psychological processes is more complex than localisation of function

suggests. In the late twentieth century, researchers sought to understand
how the brain areas worked together as a whole to control behaviour
and cognition, as you will learn in Section 4.

3.3 The use of analogies

Psychologists have suggested that analogies can prove useful in
understanding brain and behaviour. An analogy is when something
unfamiliar or difficult to understand is explained by comparison to
something that is familiar. Analogies of different forms are very
commonly used to convey ideas to others.
Which analogy you find useful will depend upon what you are most
familiar with already. People who are familiar with computers might like
to draw analogies between the brain and a computer. Someone else
might prefer an analogy with another type of electronic device, such as
a radio. Every radio set consists of different components: a battery, a
receiver, a speaker, the dials and buttons, as well as a whole host of
circuits and wires. Each component corresponds to a specific function:
power supply, capturing a signal, emitting sound, switching between
stations, etc., but all components interact or work together to capture
the message transmitted from a radio station far away.
While you might think that this analogy captures the essence of
localisation of function, the British psychologist Richard Gregory used
this same analogy to highlight the need to be very cautious when
speculating about what damage to individual brain parts can tell us
about the brain (Gregory, 1961). Suppose that we carefully remove a
component from a radio that was hitherto functioning normally.
Immediately, the radio starts to emit an awful howling sound. No one
would conclude that the normal role of the component within the radio
was to suppress howling. Rather, the howling arises from the way that
all the remaining components interact.
The same applies to the brain. The fact that damage to a part of the
brain causes a particular symptom does not imply a direct
correspondence between that brain part and the symptom. Damage to
one part also interferes with the working of other regions of the brain
and upsets a complex system. This is why drawing conclusions from
studies of patients with brain damage is difficult.

303
Chapter 7 Language and the brain

The analogy with the radio, although illustrative, is of limited usefulness.

Not only is the brain more complex than a radio, but it is also a living
organism. If you remove a component from a radio, the rest of the
radio presumably remains the same, even after weeks. You could replace
the component and expect the radio to work the same as before.
However, if part of the brain is missing, damaged or removed, there is
a tendency for the remainder of the brain to compensate for the
missing part. This flexibility or plasticity of the brain will be described in
Section 4.4. First, some relatively new technological developments in
brain research will be examined, which have allowed psychologists and
those from other, related disciplines to ‘see’ what is going on in the
living brain.

Summary
. Following Broca and Wernicke, further evidence pointed to
localisation of function: particular parts of the brain having
responsibility for particular behaviour. Disruption following brain
damage exemplifies this.
. The motor cortex is a strip of brain that is involved in the control
of movement of the various parts of the body. The motor
homunculus shows the relationship between the regions of the
motor cortex and the parts of the body over which they exert some
control.
. The results of psychosurgery suggested a more complex relationship
between the brain and certain psychological processes than implied
by localisation of function.
. Caution is needed when using findings from research on damaged
brains to explain normal brain functioning.

304
 4 Recent developments in neuropsychology

4 Recent developments in
neuropsychology
Recent advances in brain surgery have provided significant new avenues
for examining brain structures and their role in language. I am not
talking about psychosurgery, a practice that is somewhat out of favour,
but other forms of surgery to remove tumours or blood clots, or to
relieve the symptoms of epilepsy. In most instances, brain surgery
involves opening a section of skull, which enables the surgeon to
operate on a piece of brain tissue. The procedure can be performed
under local anaesthetic. The brain functions normally during the
operation and the patient can hear and speak to the surgeon (Penfield
and Roberts, 1959). This is useful because during surgery different areas
can be stimulated electrically and the effect on the patient’s behaviour
or abilities noted. This enables the surgeon to ‘work around’ the key
areas of the brain and keep the adverse effects of surgery to a
minimum.
So, how has this new type of brain surgery contributed to our
understanding of language? During surgery, if the brain region
underlying the production of speech is stimulated in a particular way,
the patient will be unable to speak. You might intuitively think that
stimulation equals action and that the patient would be triggered to
speak excessively and utter random expressions. But this is not the case.
Speech is the product of complex patterns of activity in neurons in,
among other regions, Broca’s area. Artificial stimulation disrupts these
patterns of activity and inhibits speech. By contrast, stimulation of parts
of the motor cortex does indeed trigger such reactions as involuntary
jerking of a limb.
Brain surgery enabled scientists to develop a more sophisticated map of
the brain, and confirm many of the earlier findings derived using less
sophisticated methods. But it also demonstrated that it is not actually
possible to map the brain precisely, or to define the borders of Broca’s
or Wernicke’s area in a way that would apply to every patient. Brains are
not just complex, but to some extent each patient requires his or her
own functional map.
In addition to learning from brain surgery, there are a number of other
ways of exploring the relationship between the brain and psychological
processes. Relatively recent advances in technology and medicine have
provided a rich array of techniques that have enabled psychologists to

305
Chapter 7 Language and the brain

gain a much better understanding of how the brain controls language.

Techniques for studying neurons give new insights into how the brain
can reorganise itself following damage. Next, I will describe a number
of these techniques and what they reveal about language.

4.1 Techniques for studying the brain

One technique developed just after the Second World War is the ‘Wada
technique’ (named after the Canadian neurologist Juhn A. Wada who
invented it). It involves an anaesthetic being injected into the artery that
supplies blood to one hemisphere of the brain. The anaesthetic
temporarily impairs the activity within just this hemisphere. Hence,
during the short period when the anaesthetic is effective, the activity of
the non-injected hemisphere acting alone is revealed. Disruption to
speech can be observed in response to requests, for example, to name
objects shown to the individual. Each hemisphere is usually tested in
turn. This technique reveals that usually, but not always, the left
hemisphere is dominant for language, just as Broca and Wernicke had
shown. Sometimes, however, both hemispheres share responsibility for
language, and very occasionally the right hemisphere is dominant. The
test is not really a research tool as it is not very precise. Instead, it is
done prior to brain surgery, so that the surgeon can get an estimate of
the likely disruption of the operation to the patient’s cognitive processes.
A more recent technique is magnetic resonance imaging (MRI), which
involves looking at the detailed structures of the brain. A person is
placed in an MRI scanner, and a detailed image of their brain is taken
using magnetic waves. It is similar to an X-ray but a lot more
sophisticated. The technique can show up contrasts between different
components of the brain, such as blood vessels, fluid-filled spaces and
different kinds of underlying collections of neurons, as well as any
diseased tissue. It can also be used by researchers to look in fine detail
at the deeper structures of the brain without surgery. If there is damage
to deep brain structures, there is usually a build-up of fluid and these
fluid-filled spaces are visible on the scan. Without using a knife,
different sections through the brain can be selected and viewed, rather
as if the brain had been literally sliced.
MRI can be applied to deceased brains, as well as living brains. Recall
that Broca identified a clearly visible lesion on the cortex of the brain of
Mr Leborgne (see Figure 7.7). A second patient, Mr Lelong, showed a
similar lesion at the same location. Traditionally, Broca’s area was

306
 4 Recent developments in neuropsychology

defined in terms of damage to this outer layer of the brain. Other

researchers suggested that Broca’s aphasia could also arise from damage
to deeper structures of the brain. Broca said little about the possible
role of lesions to deeper structures in his two patients, and there was a
good reason for this: he did not subject the brains to dissection. Rather,
they were placed in alcohol to be preserved intact for posterity in a
Paris museum. After well over a hundred years of resting in alcohol, the
two brains have been subject to penetrating examination using an MRI
scanner (Dronkers et al., 2007).

Figure 7.10 MRI scan of the brain of Mr Leborgne (from above) showing
several sections through the brain at different depths. (Source: Dronkers
et al., 2007, p. 1437, Figure 4)

Figure 7.10 shows the results of the MRI scan of Mr Leborgne’s brain.
As you can see, there is clear damage, visible in all five sections, to the
area to the front of the left hemisphere, which has come to be known
as Broca’s area. Crucially however, the scan, especially the third section,
also shows that the damage to Mr Leborgne’s brain extended below the
cortex. What this suggests is that the extent of the damage cannot be
determined simply by observing the outer surface. The main benefit of
the MRI scanner is that it allows researchers to look below the outer
surface without actually dissecting the brain. Dronkers et al. (2007,
p. 1441) who scanned the two brains wrote: ‘Fortunately, Broca had
great foresight in preserving these historic brains and in some ways,
Leborgne and Lelong can speak to us more eloquently now than they
could over 140 years ago’.
Whereas magnetic resonance imaging looks at the structure of the brain,
a different technique known as functional brain imaging looks at the
activity of the brain. As described in Section 1, the brain is made up of
billions of cells or neurons. When a region of the brain is active
(i.e. processing information), there is a high level of activity in the
neurons in that area. This means that these areas will require more
oxygen and energy and therefore an increased blood flow. So, the more

307
Chapter 7 Language and the brain

activity there is in a region of the brain, the greater will be the need for
blood. The adjustment in blood supply occurs naturally and
automatically. Each region signals its need for blood and the vessels
supplying each region dilate or constrict in diameter accordingly.
By monitoring the amount of blood flow to each region of the brain, it
is possible to identify which areas are most active under different
conditions. So, suppose an experimenter sets a task of listening to
spoken language and repeating what is heard. Greatest activity should
be seen in the areas underlying the processing of speech or the
production of spoken words. By contrast, if the task were to consist of
examining a picture, a high level of activity should be seen in regions of
the brain concerned with processing information coming from the eyes.
Triggering anger in the participants (and, yes, this is done occasionally!)
should trigger a flow of blood to brain regions underlying emotion.
One way of monitoring blood flow is to undertake functional MRI or
fMRI fMRI. Here, an individual is placed in the MRI scanner and asked to
Functional magnetic undertake a specific task or undergo a specific sensory process. For
resonance imaging is a example, they can be asked to perform a variety of language tasks, each
technique that allows
involving a different aspect of language. Differences in blood flow to
the blood flow in the
brain to be monitored brain regions can then be monitored.
while the individual In another functional imaging technique, investigators inject into the
undertakes a particular
task.
blood a small amount of a harmless radioactive tracer, known as a
‘label’. The apparatus can detect the tracer for a short period following
injection. The increase in blood flow in regions where there is high
activity will manifest itself on the image as high density of the tracer.

4.2 Tracing the links between brain regions that

contribute to language
The development of techniques for tracing pathways in the brain has
enabled a clearer understanding of how different brain regions
contribute to language, and how they work together.
Figure 7.11 shows some of the interactions between the brain regions
involved in understanding and performing speech (Geschwind, 1972).
Suppose that a person is asked to listen to words and then repeat them
aloud. First, auditory information arrives in the auditory cortex, where
analysis of sounds occurs. This information is passed to Wernicke’s area
where sounds are linked to words, and their meaning is processed; that
is, language comprehension takes place. For the person to repeat the

308
 4 Recent developments in neuropsychology

sounds, appropriate information must then be conveyed to Broca’s area.

Note the pathway termed the arcuate fasciculus. This consists of a bundle
of neurons along which information about the words that the person
has heard is conveyed. At Broca’s area, information on the sounds
heard is processed further and instructions relevant to speech
production are conveyed to the face area of the motor cortex, where
commands to the facial muscles are made. From the motor cortex,
signals are sent to those muscles that control the mouth and tongue,
etc., so that the sound is produced. This flow of information is
summarised in Figure 7.11b.

sound

arcuate
process sounds
fasciculus
auditory cortex

extract
language sounds

Wernicke’s area

organise
speech

Broca’s area

command
movements

motor cortex
Broca’s Wernicke’s
area area
effect
action

mouth/tongue etc.

(a) (b)
sound

Figure 7.11 Flow of information involved in language comprehension and

production: (a) the left side of the brain highlighting the arcuate fasciculus;
(b) the processes involved in repeating a word

In describing the interactions between the brain regions, I have spoken

of signals (‘information’) being transmitted from one region to another;

309
Chapter 7 Language and the brain

for example, from Wernicke’s area to Broca’s area. However, you have
also read such phrases as ‘Wernicke’s area where sounds are linked to
words, and their meaning is processed’. Psychologists would say that a
brain area such as Wernicke’s performs ‘information processing’: it
receives information on raw sounds and, acting with other brain areas,
extracts meaningful speech from this. Exactly how this happens, no one
is entirely clear about, but the following analogy might be useful. A
computer performs information processing. Suppose that you ask it to
add up a series of numbers. You feed in the numbers, press the right
button and the computer does information processing which provides
you with the answer: the sum. The brain’s processing is a bit like that,
except it is vastly more complex than a computer! You might be able to
determine what parts of the computer were involved in the computation
but without knowing exactly how it is done.
The description in this section of how the regions in the brain work
together to control language is, however, somewhat simplistic. Recall
that magnetic resonance imaging of the deceased brains of Mr Leborgne
and Mr Lelong revealed that damage was not limited to Broca’s area,
but extended to other structures, some of which were located below the
cortex. This suggests that the pathway involved in language is likely to
be more complex and to involve even more brain regions than the
above description suggests.
Functional magnetic resonance imaging has also revealed that the flow
of information between the areas of the brain involved in language
processing is more complex than Figure 11b suggests. Broca’s area, for
instance, is active not just during speech production, as Broca had
suggested, but also during comprehension of speech. Also, electrical
stimulation of Broca’s area does not just disrupt speech production but
can also impair speech comprehension (Fadiga et al., 2006).

4.3 Language beyond speech

Most of the discussion of language and the brain so far has focused on
spoken language, the kind that manifests itself as sound coming through
the ears (input) and sound from the mouth (output). There is more to
language than speech, however, and this issue raises some interesting
questions (Bellugi et al., 1989). Are the brain processes that underlie
language necessarily tied to sound? Is the ability to hear spoken language
necessary for the development of these brain regions in childhood? In
Section 3.1, I drew your attention to the proximity of Broca’s area to

310
 4 Recent developments in neuropsychology

the regions of the motor cortex involved with movement of the mouth.
This is because in most people there is correspondence between
language and speech. But in sign language, for instance, language is
linked to movements of the arms and hands. So, would the brain region
of the left hemisphere termed ‘Broca’s area’ also underlie the
organisation of sign language? Alternatively, would there be a functional
equivalent of Broca’s area adjacent to the regions of the motor cortex
that control the arms and hands?
Functional magnetic brain imaging conducted to address this question
shows that Broca’s area is activated in a similar way for individuals who
use only sign or only spoken languages. In people who are able to
express themselves in both spoken and sign language, there is a similar
activation of Broca’s area regardless of which of the two they use
(Emmorey, 2006). Similarly, electrical stimulation of Broca’s area
disrupts expression of sign language, in the same way that it disrupts
spoken language. Finally, damage to this region through stroke or injury
disrupts the expression of sign language (Bellugi et al., 1989).
What about Wernicke’s area? As noted in Section 2.2, Wernicke’s area is
adjacent to the region of the cortex processing sounds. This raises the
question of whether damage to this area would also disrupt the
understanding of sign language, and indeed it does (Hickok et al., 2002).
All this evidence points to common features in the organisation of
spoken and sign languages. In both cases, there is a predominant
involvement of the left hemisphere and similar regions, among them
Broca’s and Wernicke’s areas.

4.4 Investigating flexibility

Figure 7.11b shows in simplified form how language understanding and
speech are organised in most people. That is to say, given normal life
circumstances, these are the processes involved in language
comprehension and production. However, this diagram does not mean
that the processes are inevitably fixed and unchanging. Sometimes there
can be radical reorganisation within the brain, relative to the ‘basic
design’. As a general principle, disruption to the normal process
underlying speech can be the trigger to reorganisation within the brain. Plasticity
Furthermore, undergoing speech therapy, even in patients with serious The property of the
brain that it can alter
aphasia, can often restore part of the lost function. This ability of the
its structure in the light
brain to adapt is known as plasticity. You can apply another analogy of experience.

311
Chapter 7 Language and the brain

here: a plastic object is not rigid in its structure but, rather, its shape
can be changed according to the pressure placed upon the object.
Consider the following example. While conducting his research on
language impairment, Broca reported the case of a woman who
appeared, from her autopsy, to have been born without the region of
brain that later came to be known as ‘Broca’s area’. Nonetheless, she
showed normal speech. Broca reasoned that, in her case, the intact right
half of the brain had taken over responsibility for speech.

Box 7.2 Why do it this way?

People whose brains are different in some way have always been a
source of great interest to researchers. You have already seen the
insight Broca and Wernicke gained from investigating brain damage.
The interest is no less these days, even with the advent of powerful
technology to aid the understanding of brain function. Of course, it
hardly needs saying that investigators cannot inflict deliberate
damage on human brains in order to see what the consequences
are. Therefore, they must be confined to accidents of nature, such
as people born with parts of the brain missing and losses of brain
parts as a result of war, accidents, etc. In neuropsychology, detailed
explorations of single patients are described as ‘case studies’. Also,
unlike in the nineteenth century, today the patients’ anonymity is
protected and they are usually identified simply by their initials. Of
course, the results of research involving a single patient cannot be
generalised, but they nevertheless provide useful insight into the
workings of the brain, especially when researchers stumble on an
unusual case, such as when the brain compensates for a missing
region and so function is not as badly impaired as it might have
been.

Modern techniques of imaging, which were unavailable to Broca, have

proved invaluable in studying the flexibility of the brain. If Broca’s
patient were alive today, she could be studied using an fMRI scanner to
see which part of the brain is active when she is processing language. It
could be that she indeed developed an equivalent of Broca’s area in the
right hemisphere, or alternatively that intact parts of the left hemisphere
might have taken over some of the role of the hemisphere’s missing
regions (Crosson et al., 2007). The fact that the woman was born
without Broca’s area is important, as the brain is better able to

312
 4 Recent developments in neuropsychology

reorganise at a younger age than in older years. Young brains are still
developing and have a greater capacity to show plasticity.
Figure 7.12 shows the connections within a group of neurons. Suppose
that some of the neurons in the brain are damaged (for example neuron
A in Figure 7.12b) and wither as a result (Figure 7.12c). This damage
can trigger growth in other intact neurons, which leads to the
development of new links that compensate for the loss (Figure 7.12d).
This growth and the emergence of a new connection represent a
fundamental reorganisation.

Cut
neuron A neuron

(a) (b)

neuron neuron
A

(c) (d)

Figure 7.12 Changes in connections between neurons: (a) set of

connections in which A is intact; (b) damage done to A; (c) A withers; (d) a
new link is formed (link B)

313
Chapter 7 Language and the brain

Summary
. Temporary disruption of particular brain regions can reveal which of
them are involved in language.
. Imaging techniques reveal structures of the brain and which parts of
the brain are most active under different conditions.
. The same brain regions are involved in controlling different forms
of language, such as sign language and spoken language.
. The term ‘plasticity’ refers to the ability of the brain to adapt and
alter its structure in the light of experience.

314
 5 Contemporary relevance

5 Contemporary relevance
5.1 A two-way street
The chapter started by describing stroke and its consequences in terms
of disruption to language and other functions. So what is the relevance
of the research described in Sections 2–4 to understanding stroke? Over
the years, stroke patients have provided psychologists with valuable
evidence about the link between the brain and psychological functions.
But have these discoveries helped people with stroke in any way?
The basic principle of localisation first described by Broca means that
we can understand why strokes that affect the left hemisphere are more
likely to disrupt language compared with strokes that affect the right
hemisphere. Understanding that the left hemisphere has major
responsibility for the right half of the body allows us to understand why
strokes that disrupt speech are often associated with disruptions to
movement of limbs or sensations corresponding to the right side of the
body. Stroke patients have proved to be a valuable source of
information in confirming the validity of these early observations. At
the same time, knowing the basic localisation of function enables
doctors not only to diagnose stroke more easily, but also to determine,
on the basis of visible symptoms (speech impairment, paralysis in the
arm, etc.), which brain region is likely to be affected.
Research on plasticity, or the brain’s capacity to reorganise, has also
yielded results. By studying the survivors of stroke, researchers have
been able to map the reorganisation of brain function and monitor its
progress using brain-imaging techniques. For example, changes in brain
activity that happen as one hemisphere acquires a new capacity after
damage to the other hemisphere can be followed over a period of time.
Shifts within a single hemisphere from the damaged regions to intact
regions can similarly be detected (Crosson et al., 2007).
This research has proved valuable in developing novel techniques for
treating stroke symptoms. There is a wide armoury of different
techniques to help the rehabilitation of stroke patients, and they need to
be employed with an understanding of the processes that underlie the
plasticity of the brain. As a general principle, gradual retraining is
necessary in order for the brain to be able to reorganise (Crosson
et al., 2007). In the case of loss of speech, such retraining would
normally be specific to speech.

315
Chapter 7 Language and the brain

Computer technology has provided a technique in which patients are

encouraged to interact with a ‘virtual environment’ on a computer
screen (Rand et al., 2009). For example, they visit a virtual supermarket,
select items, load a trolley, ask for information and listen for
announcements on bargain offers. They are provided with instant
feedback on each action that they take (e.g. a spoken request for
information yields the appropriate answer). It enables a graded
transition back to real-world activities, which might otherwise at first
appear daunting to the patient.
Another technique is based on the understanding that the right half of
the brain is particularly involved in the control of the left half of the
body (Crosson et al., 2007). If speech is impaired by damage to the left
brain, steps can be taken to encourage the right brain to take over
control of speech. One way to attempt to do this is to get the patient to
point with his/her left hand while trying to utter expressions. The logic
is that activation of the right hemisphere by pointing with the left hand
could spread to activation of potential speech processes in the right
hemisphere.
Music can play a vital role in helping stroke patients. One recent study
found a particular improvement in the use of the limbs as a result of
learning the basics of playing a musical instrument, such as hitting a
drum with drumsticks (Altenmüller et al., 2009). This gives immediate
feedback of sound for an action. Possibly the technique could prove
useful for the rehabilitation of speech.
The technique of ‘melodic intonation therapy’ (MIT), to help stroke
patients, consists of daily sessions of therapy over seventy to eighty days
(Schlaug et al., 2009). The rationale for MIT is the observation that, if
asked to sing a familiar song, patients who suffered a stroke in the left
brain and who have seriously disturbed speech can often produce a
correct sequence of words. They would not otherwise be able to speak
these same words. The assumption is that either the right hemisphere or
both left and right hemispheres underlie the ability to sing. So, patients
are trained to sing very short and carefully selected sentences.
Simultaneously with singing, they are asked to pace this with rhythmic
tapping by their left hand. The assumption underlying the tapping is
that this serves as a pacemaker to areas of the right hemisphere with
some basic potential for spoken language. It could be that there is a
natural coupling between brain areas underlying the control of hand
movements and speech. Hand gestures tend to accompany speech –
even when speaking on the telephone! The sequence of words to be

316
 5 Contemporary relevance

sung is gradually made more difficult. The expectation is that this will
develop speaking capacity, which indeed it tends to do (Schlaug
et al., 2009).
Research on the brain has made a contribution not only to our
understanding of how the brain works and the role it plays in normal
functioning, but also to helping stroke patients on their road to
recovery. When dealing with patients recovering from stroke, however, it
is important to bear in mind that disruption to specific functions such
as speech and movement is one part of a bigger picture. The caring
professions are increasingly accepting the need for holistic approaches
to treatment. This is where psychologists have a crucial role to play.
Stroke patients sometimes suffer from apathy, depression and problems
with attention. This is bound to have repercussions for their
rehabilitation. This is why techniques designed to facilitate the
reorganisation of the brain must take into account the patient’s
motivation, their mood and their psychological well-being more
generally (Altenmüller et al., 2009). Treatment and therapeutic
intervention is therefore never just about the brain, but also about the
person as a whole.

5.2 Conclusion
With a focus on humans, the chapter has demonstrated where links can
be formed between the brain and psychological processes. A number of
interacting brain regions that play a role in producing and understanding
language can be identified. However, caution needs to be exercised in
identifying a particular brain region with a particular function. The brain
exhibits the property of plasticity such that, to some extent, it can
reorganise itself following brain damage.

Summary
. Stroke patients can not only be helped by new research but they can
also contribute to understanding how the brain works.
. Recovery of language ability in stroke patients is associated with
plasticity in the brain that can be identified by using imaging
techniques.
. Different therapeutic techniques have been devised to assist this
recovery.

317
Chapter 7 Language and the brain

References
Altenmüller, E., Marco-Pallares, J. Münte, T.F. and Schneider, S. (2009) ‘Neural
reorganization underlies improvement in stroke-induced motor dysfunction by
music-supported therapy’, Annals of the New York Academy of Sciences, vol. 1169,
pp. 395–405.
Bellugi, U., Poizner, H. and Klima, E.S. (1989) ‘Language, modality and the
brain’, Trends in Neurosciences, vol. 12, no. 10, pp. 380–8.
Broca, P. (1861) ‘Perte de la parole, ramollissement chronique et destruction
partielle du lobe antérieur gauche du cerveau’, Bulletins de la Société d’
Anthropologie, vol. 2, pp. 235–8.
Buckingham, H.W. (2006) ‘The Marc Dax (1770–1837)/ Paul Broca (1824–
1880) controversy over priority in science: left hemisphere specificity for seat
of articulate language and for lesions that cause aphemia’, Clinical Linguistics and
Phonetics, vol. 20, no. 7–8, pp. 613–19.
Crosson, B. et al. (2007) ‘Functional MRI of language in aphasia: a review of
the literature and the methodological challenges’, Neuropsychology Review, vol. 17,
no. 2, pp. 157–77.
Dronkers, N.F., Plaisant, O., Iba-Zizen, M.T. and Cabanis, E.A. (2007) ‘Paul
Broca’s historic cases: high resolution MR imaging of the brains of Leborgne
and Lelong’, Brain, vol. 130, no. 5, pp. 1432–41.
Emmorey, K. (2006) ‘The role of Broca’s area in sign language’ in Grodzinsky
and Amunts (eds) (2006).
Fadiga, L., Craighero, L. and Roy, A. (2006) ‘Broca’s region: a speech area?’ in
Grodzinsky and Amunts (eds) (2006).
Finger, S. (2000) Minds Behind the Brain: A History of the Pioneers and their
Discoveries, New York, NY, Oxford University Press.
Finger, S. and Roe, D. (1999) ‘Does Gustave Dax deserve to be forgotten? The
temporal lobe theory and other contributions of an overlooked figure in the
history of language and cerebral dominance’, Brain and Language, vol. 69, no. 1,
pp. 16–30.
Geschwind, N. (1972) ‘Language and the brain’, Scientific American, vol. 226, no.
4 (April), pp. 76–83.
Graves, R.E. (1997) ‘The legacy of the Wernicke–Lichtheim model’, Journal of
the History of the Neurosciences, vol. 6, no. 1, pp. 3–20.
Gregory, R.L. (1961) ‘The brain as an engineering problem’ in Thorpe, W.H.
and Zangwill, O.L. (eds) Current Problems in Animal Behaviour, Cambridge,
Cambridge University Press.
Grodzinsky, Y. and Amunts, K. (eds) Broca’s Region, New York, NY, Oxford
University Press.

318
 References

Hickok, G., Love-Geffen, T. and Klima, E.S. (2002) ‘Role of the left
hemisphere in sign language comprehension’, Brain and Language, vol. 82, no. 2,
pp. 167–78.
Penfield, W. and Roberts, L. (1959) Speech and Brain Mechanisms, Princeton, NJ,
Princeton University Press.
Pillman, F. (2003) ‘Carl Wernicke (1848–1905)’, Journal of Neurology, vol. 250,
no. 11, pp. 1390–1.
Rand, D., Weiss, P.L.T. and Katz, N. (2009) ‘Training multitasking in a virtual
supermarket: a novel intervention after stroke’, The American Journal of
Occupational Therapy, vol. 63, no. 5, pp. 535–42.
Schiller, F. (1979) Paul Broca: Founder of French Anthropology, Explorer of the Brain,
Berkeley, CA, University of California Press.
Schlaug, G., Marchina, S. and Norton, A. (2009) ‘Evidence of plasticity in
white-matter tracts of patients with chronic Broca’s aphasia undergoing intense
intonation-based speech therapy’, Annals of the New York Academy of Sciences,
vol. 1169, pp. 385–94.
The Stroke Association (2009) ‘What is a stroke’ [online],
www.stroke.org.uk/document.rm?id=699 (Accessed 22 March 2010).
Vessie, P.R. (1932) ‘On the transmission of Huntington’s chorea for 300 years
– the Bures family group’, Journal of Nervous and Mental Disease, vol. 76, no. 6,
pp. 553–73.

319
Chapter 8
Paying attention
Helen Edgar and Graham Edgar
Contents
1 Introduction 325
1.1 What is attention? 327
1.2 Attention and driving: the effect of using a mobile phone 329
2 The psychology of attention 332
2.1 Donald Broadbent and the beginnings of experimental
cognitive psychology 332
2.2 The experiments underlying Broadbent’s model 336
3 The development of Broadbent’s model 340
3.1 The work of Colin Cherry: early indications that meaning
may be important in allocating attention 340
3.2 Cherry’s experiments 343
3.3 Implications of Cherry’s findings for Broadbent’s model 346
4 Later developments in theories of attention 348
4.1 Confirmation that meaning has an effect on the
allocation of attention 348
4.2 Limited processing resources are why we need attention 349
4.3 The neurophysiology of attention 350
5 Attention in context 354
5.1 Automatic and controlled attention 354
5.2 Attention and experience 355
5.3 Back to the real world 356
5.4 Conclusion 357
References 359
Chapter 8 Paying attention

Aims and objectives

By the end of this chapter you should be able to:
. begin to appreciate the application of research on attention to real­
world tasks, specifically driving behaviour
. describe Donald Broadbent’s research on the cognitive processes of
attention
. outline the experimental approach of Broadbent and other
psychologists, and discuss how these experiments informed the
development of theories of attention
. recognise how experimental studies of attention have produced
theoretical advances that can be used to understand real-world
problems.

324
 1 Introduction

1 Introduction
We would like to begin this chapter with an anecdote from one of the
authors (Graham), as it illustrates rather nicely why it is important to
understand what ‘attention’ is, and how it works:

A few years ago I was cycling through Cardiff. The road was clear
and straight, the weather was fine, and visibility was excellent. On
my left was a lay-by with a row of shops (see Figure 8.1). I was
dimly aware of a car (a red Ford Fiesta) coming from the opposite
direction and positioning to turn right into the shops’ lay-by. I
assumed that the driver was waiting for me to pass before turning,
as I was moving quite fast. I then became acutely aware that they
weren’t waiting, they were turning in to me and there was nowhere
to go on my bicycle. Actually, there was somewhere for me to go
(without the bicycle), and that was straight over the bonnet of the
car in a graceful arc, landing in the gutter on the far side. As well
as being a personally painful experience, this is also one of the
most common forms of accident (the so-called ‘right-turn’
accident), and I can vouch for that, as it has happened to me
twice! What is also not uncommon is the explanation given by the
driver of the car that they ‘Didn’t see me at all’. On the ‘plus’ side,
I am proud to be one of a select band of psychologists that has
succeeded in denting a car using only their head and a bicycle.

325
Chapter 8 Paying attention

Figure 8.1 The road that left a lasting impression on one of the authors of
this chapter

Activity 8.1
Plainly, the driver of the car that Graham crashed into was not aware of
everything that was going on around them. You would probably agree
that they were not ‘paying attention’. Is that unusual? Are you aware of
everything that is going on around you as you are reading this? Is the
television or the radio on? Can you recall anything of the programmes
that have been on while you were reading this? Are you aware of noises
from outside (assuming you are indoors of course!)? Passing traffic? Try
concentrating on noises and/or objects around you. Can you read this at
the same time as paying attention to noises and/or objects, or to the TV
or the radio? Probably not, which does suggest that, like the Fiesta
driver, people have difficulty maintaining an awareness of all that is
happening around them.

The activity you have just done involved the conscious allocation of
attention. You were asked to attend to what was going on around you
whereas, before that, you were (we hope) attending to the material in
this chapter. You are most likely to be aware of how you allocate
attention when you are consciously shifting it from one thing to

326
 1 Introduction

another, as you were asked to do in Activity 8.1. But what is attention

and do we always know what we are attending to?

1.1 What is attention?

One of the most widely used definitions of attention was provided by
William James, more than a century ago. James is regarded as one of
the founders of experimental psychology, who established one of the
first ever experimental psychology laboratories in 1875. In 1890, James
wrote:

Every one knows what attention is. It is the taking possession by

the mind, in clear and vivid form, of one out of what seem several
simultaneously possible objects or trains of thought. Focalization,
concentration, of consciousness are of its essence. It implies
withdrawal from some things in order to deal effectively with
others …
(James, 1890, pp. 403–4)

James’s definition is interesting in that it openly acknowledges that

attention is a term in general use and not just an element of Attention
psychological jargon (although it may be that as well!). People often talk A cognitive process
about someone being ‘absent-minded’, or of being told to ‘pay involving the selection
of information for
attention’ to something. These terms suggest that attention is regarded
further processing, such
as a capacity that can be directed to prioritise one aspect of the world as the extraction of
over another (as you were doing in Activity 8.1). meaning.
James’s definition also emphasises that people cannot attend to
everything at the same time. If there are many competing stimuli,
people tend to focus on only one or a few of them and are then
conscious (that is, aware) of those few stimuli. This provides a clue to
how some lapses in attention occur. If your attention is directed
elsewhere, you may miss some of what is going on around you.
Reading this chapter is (or should be!) an example of ‘sustained
attention’. You maintain your focus of attention on the material in the
chapter and attend relatively little to what is going on around you. As
demonstrated in Activity 8.1, however, you can consciously shift your
attention to other aspects of the world if you wish. But are all shifts of
attention consciously controlled? Are shifts of attention always driven by
the individual, or can certain stimuli ‘grab’ our attention? Indeed, it
does seem that certain stimuli, such as loud noises and sudden

327
Chapter 8 Paying attention

movements, naturally attract our attention. Other stimuli, on the other

hand, might attract our attention as a result of experience. For example,
the attention of parents of young children will almost always be
attracted by a child’s voice saying, ‘Oops’.
Designers of cars, aircraft and other complex systems are aware that it
may be necessary to ‘direct’ the user’s attention to certain places. Have a
look at Figure 8.2. The cockpit of a modern aircraft can provide pilots
with a phenomenal amount of information, most of which they need to
be aware of only when something goes wrong. Designers try to use
warnings to make sure that attention is directed to certain systems only
when it is absolutely necessary. The same principle holds in driving.
Most of the time you can ignore the petrol gauge in your car – except
when the warning light comes on and then you really should attend to
it. Given the fact that people can attend to only a limited number of
things at any one time, it is necessary to be extremely cautious in the
use of warnings, and it is easy to have too many. Indeed, one modern
fighter aircraft has so many spoken warnings, which sound so often,
that the warning system has become known among pilots as ‘Nagging
Nora’.

Figure 8.2 The cockpit of a modern civil airliner. All of the information
provided by the displays (and there is quite a lot of it) may be needed at
some point – but thankfully not all at the same time!

328
 1 Introduction

1.2 Attention and driving: the effect of using a

mobile phone
You should have the idea by now that attention is a precious resource.
If you attend to one thing, then you have fewer attentional resources to
spare for other things. This can cause problems if resources are
squandered on some tasks at the expense of other, more safety-critical,
tasks. For example, improvements in technology have significantly
increased the number of tasks that you can do while driving. Perhaps
the most obvious example of this multitasking is the use of mobile
telephones while at the wheel of a car. Once the technology for mobile
phones had advanced to the stage where they were relatively small, they
quickly became very popular. Indeed, a report on mobile phone use
while driving estimated that in 2004, at any given moment in the day, 8
per cent of drivers on US roads were using their mobile phone
(Glassbrenner, 2005).
So, does using a mobile phone while driving impair driving
performance? The discussion so far suggests that it might, and the law
in the UK agrees. Using a hand-held mobile phone while driving was
made illegal at the end of 2003. Looking again at James’s definition of
attention, it seems that using a mobile phone will almost certainly take
precious cognitive resources that could perhaps be better directed
elsewhere to improve the user’s driving. But does using a mobile phone
really make that much difference to driving? Donald Redelmeier and
Robert Tibshirani (1997) looked at the mobile-phone records of 699
drivers who had been involved in a driving accident. They found that
24 per cent of these drivers were using their phone within the ten
minutes prior to the accident. Given Glassbrenner’s findings described
earlier, this suggests that drivers on mobile phones are over-represented
in the accident statistics – whilst 8 per cent of the general driving
population were estimated to be using their phone at any one time, this
group comprised 24 per cent of the accident population. These figures
suggest that there is a strong association, or correlation, between
mobile-phone use and the likelihood of having an accident. You were introduced to
the term ‘correlation’ in
One has to be cautious, however, about drawing firm conclusions from Box 3.2, Section 3,
correlational data. Just because there is a correlation between mobile Chapter 3. Issues
phone use and driving accidents does not necessarily mean that the relating to correlational
phone use contributed directly to the accident. As David Strayer and data were also
colleagues (2006) point out, it may simply be that the people who use discussed in Section 5,
Chapter 5.

329
Chapter 8 Paying attention

phones while driving are simply more prone to risky behaviour, so that
their driving is more dangerous with or without a mobile phone.
Although correlational studies cannot establish a causal link between
mobile-phone use and decrements in driving performance, experimental
studies could look at the effect of using a mobile phone while driving;
for example, by comparing two situations that are matched as closely as
possible except for the fact that the participant is using a mobile phone
in one situation and not in the other. This also allows an investigation
of just what aspects of driving performance might be affected by phone
use. One such study (Brookhuis et al., 1991) found that driving while
using a mobile phone actually decreased the amount of swerving by the
driver, but increased the driver’s reaction time to a change in speed of the
car in front. In the context of the discussion so far, it is possible that
the drivers directed their attention to the vigilance task (maintaining lane
position) but this was at the expense of detecting a sudden change in
the situation (e.g. the car in front braking). This study emphasises that
attention has a number of interacting components, including (in the case
of driving) maintaining vigilance and detecting changes.
Are we consciously aware of where we are allocating attention and how
it is being ‘shared out’? A study entitled ‘Sorry, can’t talk now… just
overtaking a lorry: the definition and experimental investigation of the
problem of driving and hands free car phone use’ (Boase et al., 1988),
found that mobile-phone use significantly impaired the performance of
the nine drivers who took part in the study. The interesting aspect of
this study is that interviews with these nine drivers revealed that none
of them believed their performance had been affected by the use of a
mobile phone. They appeared to be unaware that the attention they
were paying to their phone may have impaired their driving.
The title of the Boase et al. paper was remarkably prescient of an actual
accident that took place in the UK some years later. A Royal Society for
the Prevention of Accidents (RoSPA) report cites the case of a driver
overtaking another vehicle at up to 90 miles per hour in heavy rain.
The overtaking driver lost control, hit a tree, and was fatally injured.
When the police checked the dead driver’s phone they found a text
message three pages long that had been received two minutes before the
crash. The coroner’s report included the phrase, ‘A message had been
accessed, it may have diverted his attention’ (RoSPA, 2002, p. 11).

330
 1 Introduction

Summary
. Attention is a complex cognitive process which involves focus or
concentration on certain stimuli in the environment.
. There are limits on how much we can attend to at any one time.
. These limits have implications for a range of real-world tasks,
including driving.

331
Chapter 8 Paying attention

2 The psychology of attention

2.1 Donald Broadbent and the beginnings of
experimental cognitive psychology
The previous section outlined what attention is and discussed some
studies on driving that were relevant to a real-world problem. This
applied stream in psychological research is part of a tradition that is
strongly associated with one of the founders of modern cognitive
psychology, Donald Broadbent. This section will consider the early
contribution of Broadbent and others to our understanding of attention.

Figure 8.3 Donald E. Broadbent (1926–1993)

As well as being a famous cognitive psychologist, Broadbent was a pre­

eminent figure in the development of the discipline of applied
psychology. In 1943, aged just 17, Broadbent joined the Royal Air Force
(RAF). It was this experience in the air force that sparked his interest in
psychology (Weiskrantz, 1994). During flight training and operations,
incidents occurred that, in Broadbent’s opinion, could have been
avoided if the capabilities of the operators had been considered in the
design of the equipment. Specifically, Broadbent was interested in how
the design could be driven by an understanding of the capacities and

332
 2 The psychology of attention

limitations of the operator, rather than designing the equipment and

then hoping the operator could get to grips with it somehow.
One specific problem that piqued Broadbent’s interest was the design of
the AT6 aircraft, which had two levers close together under the seat;
one of them pulled up the flaps (a manoeuvre needed when landing)
and the other pulled up the undercarriage. Several pilots made the
mistake of pulling the wrong lever during a landing manoeuvre, and
lifted the wheels instead of the flaps, thus damaging a very expensive
aircraft!
Broadbent’s experiences in the RAF inspired him to read psychology at
Cambridge, after he was demobilised in 1947. When he graduated he
started work on a project sponsored by the Royal Navy to look at the
effects of noise on performance. Broadbent went on to explore, using
experiments, what happens when ‘listeners characteristically are
confronted with a multitude of signals arriving more or less
simultaneously, but can react to only a limited sub-set’ (Weiskrantz,
1994, p. 36). How is this choice made? How do they preferentially select
only some incoming information? In 1958, he published a book titled
Perception and Communication where he described his limited capacity
model of attention. Broadbent used a diagram, a simplified version of
which is shown in Figure 8.4, to describe human information processing
as a system with a finite limit for the intake and storage of information.
Broadbent’s model was relevant to many of the everyday problems
discussed in Section 1. Indeed, the notion of the limited capacity of our
attentional system is present in James’s original definition of attention
(1890, p. 4), in that the process of attention ‘implies withdrawal from
some things in order to deal effectively with others’.

333
Chapter 8 Paying attention

S S
h e
S o l F
S
e r e i
t
n t c l Limited capacity Further
o
s t t channel processing
r
e t i e
e
s e v r
r e
m

Feedback

Figure 8.4 A simplified version of Broadbent’s information-flow diagram

(Source: modified from Broadbent, 1958, p. 299, Figure 7)

According to Broadbent’s model, information about our environment

enters the first system (labelled ‘Senses’) in parallel via the different
senses (e.g. the eyes and ears). These ‘inputs’ are then stored briefly
before being filtered through the selective filter. The information may
be filtered on the basis of the physical properties of the stimulus, such
as the intensity of the sound. Alternatively it may be filtered using
feedback from memory stores, prior experience and expectations. For
example, sounds recognised as warning signals would be selected for
further processing. Information then exits the filter in series (only one
Limited capacity piece of information at a time), through the limited capacity channel,
channel to be processed by other systems. The boxes in the diagram represent
A channel through the different systems that process information, and the arrows indicate
which filtered and
the flow of information through those processing systems.
attended-to information
passes for further Broadbent suggested that things such as multitasking (i.e. performing
processing, such as more than one task at the same time) could be achieved only by rapid
storage in memory. It is
‘limited’ as regards the
switching amongst inputs, as there was a limit to how much information
amount of information could flow through the systems. Broadbent’s model was rather vague as
it can deal with at any to what lies after the filter, although it indicates that there are processes
one time. that govern which inputs should be selected and when to change the
selection. This information-processing model constituted the first
testable model of attention, and Broadbent’s approach actively
Information­
encouraged others to conduct experiments that would test it.
processing model
A hypothetical model
depicting how
information taken in by
the senses might be
processed in the brain.

334
 2 The psychology of attention

Box 8.1 Influences on the work of Donald Broadbent

In 1947, Broadbent was accepted by Pembroke College, Cambridge
to study experimental psychology. At that time the department head
was Sir Frederic Bartlett, who had been involved in wartime
research into human performance and was ideally placed to
encourage Broadbent’s interest in applied psychology. Bartlett
strongly believed that psychology should strive to address real-life
problems, but also that psychological theories should be derived
from experimental research. Broadbent strongly supported this view
and went on to champion it throughout his career.
After graduation, Broadbent was employed by the Royal Navy to
study the effects of noise on performance. This research was
carried out at the Applied Psychology Unit at Cambridge (APU),
because the noise involved was considered to be too distracting for
other staff in the naval research laboratory. Broadbent then moved
on to study the effects of noise on vigilance tasks in gunnery and
air traffic control systems, where operators were bombarded by
many simultaneous incoming messages (signals) but could react to
only a limited subset of those signals.
An engineering approach to psychology had been established at
Cambridge by Kenneth Craik, the first director of the APU, who died
just before Broadbent joined the university. Broadbent was
nevertheless able to study Craik’s manuscripts, which compared
human information processing with artificial information-processing
systems such as computers.
Broadbent’s use of a flow diagram to depict his model of attention
clearly demonstrated the influence of Craik’s ideas. He used ideas
and terminology from emerging fields concerned with computer
systems and information processing. Broadbent liked this way of
depicting processes that went on in the brain and which he could
not observe directly.

335
Chapter 8 Paying attention

In his book Perception and Communication (1958), Broadbent laid

down a framework for the understanding of attention, memory and
comprehension that formed the basis of what is now called
‘cognitive psychology’. At the time when Perception and
Communication was first published, psychology, especially in the
See Chapter 4 for a USA, was dominated by behaviourism. Experiments designed to
discussion of study operant conditioning took no account of prior knowledge or
behaviourism experience and were therefore far removed from the complex real­
life situations that Broadbent aimed to address.

Broadbent’s abstract representation of human mental processes, the

information-flow diagram, was to prove enduringly popular in what
became known as cognitive psychology. His experiments showed how
the new sub-discipline of cognitive psychology could address complex,
real-world problems. The use of information-processing terminology
and flow diagrams with boxes and arrows quickly became the accepted
way to describe cognitive processes that could not be observed directly,
and it is still used today. Note that Broadbent was not suggesting that
the actual brain structures and systems could be ignored as irrelevant
‘hardware’. It is just that he did not undertake physiological or
neuropsychological studies of real brain systems. His information­
processing theory did not seek to identify the brain structures
underlying the attentional systems but to describe the processes
involved and develop a theoretical model of attention.

2.2 The experiments underlying Broadbent’s model

Broadbent’s model of attention was developed and refined from a large
body of empirical work. Perhaps the most influential experiments
Dichotic listening involved ‘dichotic listening’. In a dichotic listening task, different
When two different stimuli are presented to the two ears simultaneously and the participant
messages are presented may be asked, for example, to recall all the material presented to one
simultaneously, one to
ear first, before recalling material presented to the other. The general
each ear.
principle of dichotic listening is illustrated in Figure 8.5.

336
 2 The psychology of attention

328 194

Right ear first

3,2,8,1,9,4

Figure 8.5 An example of a dichotic listening task. The three numbers

shown on each side are presented simultaneously, and the task here is to
report those presented to right ear first. The correct response is therefore:
3, 2, 8, 1, 9, 4

Dichotic listening experiments were designed to be analogous to the

complex tasks Broadbent worked on early in his career. For example,
gunnery and air traffic control personnel received many, often
simultaneous, messages and had to select the appropriate ones to
respond to. Dichotic listening simplifies these tasks to allow the
isolation and investigation of one variable at a time in a controlled
environment. Different aspects of the stimulus (a message or a task)
that might be expected to affect performance (making it easier or more
difficult to perform to a required standard) were manipulated in a series
of studies. In addition to the similarity to complex, real-world tasks, the
use of auditory stimuli had another practical advantage: while we can
move our eyes to control what we see, it is much harder to move our
ears (especially if the stimuli are delivered via headphones!). From a
theoretical point of view, dichotic stimulus presentation provides a
simple and elegant way of taxing the human information-processing
system so that filtering becomes necessary and unavoidable. Broadbent
and one of his contemporaries, Colin Cherry, made extensive use of
dichotic listening in their experiments. Some of these experiments will
be described in detail as they demonstrate how the model shown in
Figure 8.4 was developed and refined by a series of linked experiments,
each one providing a little extra information to help complete the
theoretical jigsaw.
Previous studies had shown that when participants were presented with
three digits to one ear and another three digits to the other ear
simultaneously, they could repeat all six digits from memory, but
normally all three from one ear were recalled before those from the
other. These results led Broadbent to deduce that there is a mechanism,

337
Chapter 8 Paying attention

which cannot be observed directly, that can only pass information

serially (one item after another) and another mechanism which can store
information from two (or more) channels simultaneously (i.e. can handle
information ‘in parallel’), but that the latter is limited with respect to
how much information it can store and how long for.
In order to test his theory, Broadbent (1954) undertook a series of
experiments designed to see whether, for example, expectations had an
effect on the allocation of attention and subsequent recall. Participants
were presented with two sets of digits simultaneously, one set to each
ear. They were then asked to write down the digits from memory,
beginning with those in an ear chosen by the experimenter. More
mistakes were made on the second set of digits when participants knew
in advance which set they would have to report first (left or right). This
suggested that attention was directed to the chosen ear and the material
from the other ear was not always retrieved accurately as it had not
been selected for attention.
The results from these experiments showed that participants reported
the material they processed first more accurately than the later material,
which may have been lost from the brief memory store by the time the
participant got around to reporting it. This finding was true for
simultaneous presentation of stimuli (as shown in Figure 8.5) as well as
for successive presentation (where one set of stimuli follows on from
the first set of stimuli). Broadbent (1957, pp. 2–3) wrote: ‘We can listen
to only one voice at once, and the first words we hear are the best
recalled.’

Activity 8.2
Try dichotic listening for yourself.
Listen to two conversations at the same time; one on the telephone (do
warn the person on the other end of the line first!) and the other with
someone in the same room as you (someone talking on the radio or
television will do).

You are likely to have found it difficult to follow both conversations at

the same time. This is what would be predicted by Broadbent’s model.
Other experiments using dichotic listening sought to investigate the
effects of delay on recall. The results were consistent with Broadbent’s

338
 2 The psychology of attention

prediction that any delay in recovering items from the limited-capacity

storage mechanism reduced the efficiency of recall.
Irrelevant stimuli were also considered. The presentation of irrelevant
material (for instance, letters) to the same ear as the ‘target’ stimuli
(numbers) made the task of reporting the target stimuli almost
impossible. Broadbent noted that this is in stark contrast to a finding by
Cherry (1953) that participants were able to ignore irrelevant material
when the target stimuli were presented to one ear and the irrelevant
material to the other. Cherry’s experiment also provided support for the
notion that filtering can be done on the basis of simple physical features
(for example, according to the location or intensity of the sound) but
that, if irrelevant items cannot be discriminated on the basis of such
physical features, they will then interfere with the processing of the
target items. (We will return to this experiment in Section 3.)
The experiments conducted by Broadbent and also by other
psychologists were used by Broadbent to test and refine specific aspects
of his information-processing model. By systematically manipulating
different variables, a slightly different aspect of the process was
investigated. Taken in isolation, the results of individual experiments
could be open to at least two different interpretations, but each
successive experiment narrowed down the possible number of
interpretations, thereby increasing the confidence with which
conclusions could be drawn about the workings of human attention.

Summary
. Donald Broadbent proposed a very influential model of attention.
. His research and his interest in real-world problems were informed
by his experience with the RAF.
. For his model, Broadbent drew on the then emerging fields
concerned with information and computer systems.
. His use of information-processing terminology and flow diagrams to
describe cognitive processes that could not be observed directly is
still popular today in cognitive psychology.
. The model was developed and refined on the basis of the results
from dichotic listening experiments.

339
Chapter 8 Paying attention

3 The development of Broadbent’s

model
3.1 The work of Colin Cherry: early indications that
meaning may be important in allocating attention
A contemporary of Broadbent was Colin Cherry. Like Broadbent’s
work, Cherry’s focused on how psychological theory and the insight it
provides could be applied to practical problems. Cherry agreed that
incoming stimuli can be filtered on the basis of simple physical
characteristics, such as voice intensity. However, after carrying out a
number of studies, he suggested that sometimes the incoming stimuli
(e.g. the messages presented to each ear) is analysed more fully before a
decision is made about which one is attended to. Specifically, stimuli are
sometimes filtered according to their meaning. Cherry drew this
conclusion from the observation that people can perform tasks such as
listening to one speaker when others are speaking at the same time. He
used the term ‘cocktail party problem’ to describe this ability to follow
what one person is saying, when others are also speaking.
Could Broadbent’s model be refined to incorporate Cherry’s findings?
Before looking at this and at Cherry’s studies in more detail, it would be
Hypothesis useful to clarify the role of the hypothesis when conducting
A prediction or experiments (see Box 8.2).
statement that is tested
in an experiment or
correlational study.
Box 8.2 Why do it this way?
Experiments are generally conducted to test a hypothesis. A
hypothesis is basically a prediction, phrased as a simple statement.
Crucially, this statement may not be true, but it should be logical; it
is set up as a starting point and an experiment is then conducted to
see whether or not the statement is true. It’s important to note that
the aim of an experiment is not to ‘prove the hypothesis true’, but to
gather evidence to decide whether it is likely to be true or not. A
Theory negative result can be as informative (although often not as
A set of propositions
satisfying for the experimenter) as a positive one.
about a psychological
phenomenon The hypothesis is usually devised in the first place by looking at an
(e.g. attention) that existing theory. Theories allow one to make predictions that can be
forms the basis of a tested using a suitable method. These predictions are phrased as
testable explanation. hypotheses. For example, you can do an experiment to test a

340
 3 The development of Broadbent’s model

hypothesis, revise the theory from which it was derived, (or not,
depending on how the experiment turned out) and generate further
hypotheses. This is known as the cycle of enquiry. Cycle of enquiry
The way in which the
One of the very first studies of the effects of using a phone while questions that research
driving provides an example of hypothesis testing. This was a study addresses are often
conducted by Ivan Brown et al. (1969) at the Applied Psychology derived from theories
Unit in Cambridge while Donald Broadbent was the director. This or explanations, and the
was ‘cutting edge’ research at that time, as demonstrated by the findings of that
fact that the mobile phone had to be simulated for the study, as research then generate
there were none in general use. Interestingly, the simulation used new questions or
the equivalent of a hands-free mobile phone. refinements to theory
or explanation.
Brown et al. proposed that performing two tasks simultaneously,
driving and telephoning, would be challenging for a person. Not only
might using a telephone interfere with steering or gear changing,
but attention would need to be switched between the demands of
the two tasks. You may remember that Broadbent suggested that
multitasking could be achieved only by rapid switching amongst
inputs. The hypothesis Brown et al. tested was similar to: ‘Using a
mobile phone will affect performance on a driving task’ (the actual
wording is more precise and beyond the scope of this box!).
Participants were required to drive round a test track and judge
whether or not they would be able to pass through gaps of varying
sizes, some smaller and some bigger than the actual width of the
car.
Two measures of performance (dependent variables) used in the
study were:
speed (time taken to complete the circuit)
accuracy (how well the driver judged the width of the gaps).
There were two conditions of particular interest in this experiment.
In the experimental condition the driver drove round the test track
while answering questions on the mobile phone, and in the control
condition the driver drove the same route without using the phone.
These two conditions resulted from the manipulation of the
independent variable – the use or not of a mobile phone.
Other variables – for example, age, driving experience and gender
– were controlled by having the same participants take part in both
conditions (phone and no phone). Controlled variables are the You read about the
factors that are kept constant in the experiment, so that only the controlled nature of
variable being investigated in the experiment influences the experiments in
observed result. If other variables that might have an effect are not Chapters 2 and 3 of
Part 1.
controlled (i.e. they are not kept the same across both conditions),

341
Chapter 8 Paying attention

then it is not possible to say whether the observed effect is due to

the manipulation of the independent variable and the results of the
study will not be valid. An alternative approach is to try to ensure
that any variables that cannot be controlled act randomly across the
conditions. For example, it is possible that some things might
distract the drivers (birds flying across the track, sunlight, etc.), but
as long as they don’t distract them more in one condition than
another it should not affect the overall result.
With all these precautions in place, you now have an experiment
that can provide a valid test of the hypothesis, ‘Using a mobile
phone will affect performance on a driving task’. The independent
variable is whether or not the driver is using a mobile phone and
the dependent variables are aspects of driving performance. If there
is a difference in the dependent variables measured (speed and
accuracy) between the two conditions, then you can say that the
findings provide support for the hypothesis. In the Brown et al.
study, driving time increased when participants were telephoning,
and judgements regarding whether the width of the gaps were
sufficient were impaired.
This is where it becomes important to control, as much as possible,
for all variables other than the one manipulated (use of a phone). If
another variable happened to differ across the conditions (e.g. you
had all novice drivers in one condition and all experienced drivers in
the other), then you could not be sure that a difference in the
conditions was really due to the effect of your independent variable,
or the other, confounding variable. If you did not control for driving
experience, for example, you could not be sure that any difference
was due only to phone use and not to driving experience.
Once you know whether your hypothesis is supported or not, you
can then go back to the theory from which the hypothesis was
derived, modify the theory if necessary, then generate another
hypothesis and test it. For example, you might hypothesise that
using mobile phones has a larger detrimental effect on the
performance of drivers who are tired or stressed, and then design
another experiment to test that. The more experiments you do, the
more comprehensive the theory becomes (you hope!).

Cherry’s (1953) work is a good example of using a series of experiments

to develop a theory. The approach taken by Cherry of using spoken
language in his experiments was very different from the majority of
studies that had preceded his. Using natural speech as a stimulus

342
 3 The development of Broadbent’s model

required a different way of thinking about experimental design. Cherry

approached this work by thinking about how a machine or filter might
be designed to discriminate between continuous speech when more than
one person is talking at once. First, he identified some of the factors
that might aid this process:
1 the direction from which the voice originates
2 the presence of additional cues such as movement of the lips and
gestures
3 characteristics of the voice such as pitch, speed and whether the
voice is male or female
4 different accents and dialects
5 the ease with which it is possible to predict what the next item of
information is likely to be, based on what has gone before; for
example, this is much easier with continuous prose, such as
complete sentences, compared with random words.

By recording two (interwoven) messages spoken by the same speaker,

Cherry was able to control for factors one to four above, allowing him
to isolate the effect of the fifth factor, namely meaning. The findings of
the experiments looking at the importance of meaning are described
next.

3.2 Cherry’s experiments

Cherry (1953) carried out a series of experiments designed to determine
exactly what information allows participants to separate out two spoken
messages that are presented together. Remember that messages were
spoken by the same speaker, so participants would not be able to
distinguish between these messages on the basis of the physical
characteristics described in points 2–4 above, such as the characteristics
of the voice. The two messages were presented either as ‘mixed’ or as
‘simultaneous’ speech as described next (also illustrated in Figure 8.6):
. The presentation of mixed messages involved two different spoken
messages about different subjects presented to both ears. That is,
both ears heard both messages at the same time, giving the
impression, as Cherry notes, of ‘a babel’.
. The presentation of simultaneous messages involved two messages
about different subjects, one presented to one ear and one to the

343
Chapter 8 Paying attention

other so that each ear heard only one of the messages (a basic
dichotic listening task).
Mixed speech
A chain is only as strong A chain is only as strong
as its weakest link as its weakest link

A bird in the hand is A bird in the hand is

worth two in the bush worth two in the bush

A chain is only as strong as its weakest link

Simultaneous speech

A chain is only as strong A bird in the hand is

as its weakest link worth two in the bush

A chain is only as strong as its weakest link

Figure 8.6 Mixed and simultaneous presentation of natural speech

In the experiment, participants in the simultaneous speech condition

were asked to repeat whatever message was presented in the right ear,
while in the mixed speech condition they were asked to pick out and
repeat one of the messages.
Cherry obtained the following results from the mixed speech condition:
. The different messages could be identified and separated by the
listener even when they were not allowed to write anything down;
that is, when verbal repetition of each message was the only
response allowed. Participants found this task difficult, however, and
the tapes were played up to twenty times.
. Writing the responses down made the task much easier. It was easier
to separate out meaningful messages than nonsense ones (even if the
nonsense ones obeyed normal grammatical rules), implying that the

344
 3 The development of Broadbent’s model

content of a message influenced participants’ ability to attend to, and

separate out, the two messages.

Using simultaneous speeches, Cherry found that:

. As regards to the message presented to the right ear (i.e. the
message that they were asked to repeat), participants could easily
report this message whilst another was being played in the other ear.
(This finding was described earlier, towards the end of Section 2.)
. As for the message presented to the left ear (i.e. the message that
they were instructed to ignore), the participants were unaware of its
content to the extent that they did not ‘notice’ if the language
spoken in the unattended message changed during presentation (for
example, from English to German). Participants accurately reported
whether or not the unattended message consisted of human speech,
although specific words, phrases or the language used were not
reported. If the speech was reversed (i.e. played backwards), this also
was difficult to identify, with some participants reporting it as
normal speech and others being able to say only that there was
‘something queer about it’.

Activity 8.3
Imagine an experiment where exactly the same message is presented to
both ears, but with a slight delay so this message arrives at the right ear
before the left ear. Participants are asked to attend to the message
presented to the right ear. Would you expect participants to notice that
the same message was presented to the left ear?
Reflect on this for a moment before reading on.

This is precisely what Cherry did in a variation of his experiment. He

presented participants with the same message to both ears, but with one
following slightly after the other. The results showed that when the
delay between the two messages was between two and six seconds, most
participants did notice that the two messages were the same. This
suggests that they were extracting meaning from the information that
was directed to both ears, even the one to which they were not
attending.

345
Chapter 8 Paying attention

3.3 Implications of Cherry’s findings for

Broadbent’s model
The results of these experiments provided support for the notion that
material can be filtered on the basis of simple physical features, such as
what ear the sound is played to. This was what Broadbent’s model
suggested. The findings from Cherry’s experiment where a different
message was presented to each ear (simultaneous speech condition) are
consistent with Broadbent’s model. Participants attending to one ear
were often unaware of the language spoken in the other ear.
If, however, the material to be discriminated was all presented to both
ears at the same time (mixed speech condition), then cues other than
simple physical characteristics appeared to aid in attending to a
particular message. Meaningful messages were easier to separate out
than nonsense ones.
Also, if the messages to the two ears were actually the same message
but one was delayed by a small amount, then most participants reported
being aware that the unattended message was the same as the attended
one.
Thus, Cherry’s findings, while generally supporting Broadbent’s notion
of the filtering of material on the basis of physical characteristics, also
imply that the content or meaning of a message can influence how
attention is allocated. Findings such as these presented a paradox for
Broadbent’s model in that they suggested that some processing of the
stimulus (such as content and meaning) could take place before it is
selected for attention.
As Broadbent progressed his own research, he adapted his model of
attention to accommodate as much of the new findings as possible. The
model gradually became more complex. The version published in 1958
and shown in Figure 8.4 incorporated the idea that feedback from
memory, prior experience and expectations can influence the workings
of the selective attentional filter.
The modification of Broadbent’s theories in the light of experimental
evidence is a good example of the process that Broadbent believed
experimental psychology should follow, namely that ‘Data should
determine theory rather than theory being used to search for particular
data’ (Weiskrantz, 1994, p. 35).

346
 3 The development of Broadbent’s model

Summary
. Experiments are designed to test hypotheses that are derived from
theory or previous experiments. The results of the experiments can
then inform refinements to existing theories or help in the
formation of new hypotheses.
. The experiments conducted by Cherry (1953) provided evidence that
information may be filtered on the basis of its physical
characteristics, as suggested by Broadbent.
. The findings of Cherry also suggested that meaning may be
extracted before filtering takes place, and these results were
incorporated into Broadbent’s model.

347
Chapter 8 Paying attention

4 Later developments in theories of

attention
The work of Cherry provided early indications that meaning might also
under certain circumstances be important with regard to which
information was selected and attended to. This finding was confirmed
by other studies reported next, which in turn generated alternative
approaches to explaining attention.

4.1 Confirmation that meaning has an effect on the

allocation of attention
Jeffrey Gray and Alexander Wedderburn (1960) presented different
messages simultaneously, one to each ear as Cherry had done. The
difference was that the messages were interspersed so that one
particular message switched from one ear to the other, as illustrated in
Figure 8.7.
Switched speech
A chain is the hand is A bird is only as strong
as its in the bush worth two weakest link

A chain is only as strong as its weakest link

Figure 8.7 Switching of messages from one ear to ear

Gray and Wedderburn found that if participants were instructed to

follow the meaning of the message, then they could do so even though
this involved switching attention from ear to ear. Furthermore, Anne
Treisman (1960) found that participants would switch attention from ear
to ear to follow the meaning of a message even if they were instructed
not to.
These studies thus suggest that filtering may, in some circumstances,
occur much later in the processing system, after meaning has been
extracted. But this filtering, on the basis of meaning, was taking place
because there was not enough attentional capacity to properly process

348
 4 Later developments in theories of attention

both messages. With so much information falling on our senses, early

filtering on the basis of physical characteristics is desirable as this
reduces the amount of information being passed along for further
processing. However, later filtering is also desirable as there is a limit to
how much information can be properly and fully processed, or in other
words, to how much of the incoming information we can properly make
sense of.

4.2 Limited processing resources are why we need

attention
Regardless of whether filtering occurs early or late, what appears to be
important is that there is a limit to the amount of information we can
attend to. Daniel Kahneman (1973) suggested that our brains can
process only a certain amount of information at any one time. That is,
our cognitive resources are limited, and thus can be used to process
only a limited amount of material. The implications of this assumption
are clear and fit well with the studies we have already discussed; if you
allocate cognitive resources to one task, then you will have fewer to
spare for other concurrent tasks. Thus, filtering, whether early or late,
acts as a way of reducing the drain on the central cognitive resources
that are used to fully process information, for example, when learning
something to remember later on.

Activity 8.4
Let’s demonstrate the notion of a limited capacity, i.e. that there is a limit
to our cognitive processing resources. Pick two pages from a book that
are of similar length and that you have not yet read. Read the first page
and time how long it takes you to read it. Now try to read the second
page but, while you are doing it, try saying out loud the seven times
table. See how long it takes to read the second page.

Did you find that it took you longer to read the second page? Can you
remember as much from it as from the first page? According to
Kahneman, the second page should have taken longer and you should
remember less of it because saying the seven times table takes cognitive
resources that you would otherwise use to process the material you are
trying to read – and the task becomes more difficult.
The studies described so far in this chapter (including those of Cherry)
can be explained on the basis of the allocation of limited-processing

349
Chapter 8 Paying attention

resources and, indeed, Kahneman based his limited-capacity theory on

the research available at the time. The development and testing of the
theory did not end there, however. At the time of writing this chapter,
Kahneman’s 1973 book had been cited more than 4000 times by other
authors, many presenting data supporting, refining, or challenging
Kahneman’s original ideas. As you have seen in the discussion of
hypothesis testing, this is how theories develop and are refined or
rejected. A theory such as the limited-capacity theory of Kahneman is
not an end point of research, but simply a point in a cycle of enquiry
whereby theories are constantly tested and refined.

4.3 The neurophysiology of attention

Kahneman’s theory has stood the test of time. More evidence has been
provided as different techniques have been developed that were not
available to the likes of Broadbent, Cherry and Kahneman. One widely
used technique is that of functional magnetic resonance imaging (fMRI).
Functional MRI is of great interest to psychologists because it can be
used to detect localised changes in blood flow in the brain. This, in
itself, would be of limited use except for the finding that changes in
blood flow appear to immediately follow, and are closely associated
You read about fMRI with, increases in brain activity. Thus, fMRI can be used to give an
in Chapter 7, almost real-time picture of brain activity.
Section 4.1.
Functional MRI has been widely used in the study of attention and, in
particular, how changes in attention are reflected by changes in brain
activity. A study by Sabine Kastner et al. (1998) compared the brain
activity generated by viewing multiple visual stimuli with brain activity
generated by viewing one of those stimuli in isolation. Examples of the
stimuli are shown in Figure 8.8.

350
 4 Later developments in theories of attention

Figure 8.8 An example of the stimuli used by Kastner et al. (1998). Four
complex images were presented either sequentially (A) or simultaneously (B)

According to Kahneman’s theory, viewing multiple stimuli

simultaneously should place a greater load on the individual’s
information-processing system than presenting a single item. Kastner
et al. found that the level of brain activity in response to the lone
stimulus (the sequential condition) was higher than when the same
stimulus was viewed as part of a group (as in the simultaneous
condition). Thus, it appears that processing resources are limited and, if
spread across a number of stimuli, then each stimulus has less
processing devoted to it than when one stimulus is presented in
isolation. Kastner et al. also found that, if participants viewed all the
stimuli together but were instructed to direct attention to only one of the
stimuli in the group, then brain activity associated with this ‘attended-to’
stimulus increased, and that activity associated with the other (‘non­
attended-to’) stimuli decreased. The results of this study are certainly
consistent with the notion that attention involves the allocation of a
limited pool of resources (as suggested by Kahneman), and the idea that
if more resources are directed to one location (i.e. attention is ‘paid’ to

351
Chapter 8 Paying attention

that location), then fewer are available for use elsewhere. In other
words, the more tasks you perform simultaneously, the fewer resources
are available for each one. Attention refers to the process by which
those resources are allocated.
Functional MRI studies, in conjunction with other research techniques
for investigating brain function, have helped to show how different
areas of the brain are involved in different aspects of attention (Kastner
and Pinsk, 2004). For example, there is a posterior system (so called
because the relevant areas are mostly at the back of the brain) that
appears to be responsible for selecting, and hence attending to, objects,
based on features such as position, shape or colour – very much the
sort of physical properties that Broadbent and Cherry thought were
used for early filtering of incoming stimuli. In addition to this system,
there is also an anterior system (involving areas at the front of the
brain) that appears to control which of the features mentioned above will
be used as a cue to direct attention. There is a third region (mostly to
the front of the right hemisphere) that appears to be involved in
allowing attention to be maintained on a particular set of stimuli (Coull
et al., 1996). Studies such as these have provided some evidence for
which brain regions are responsible for controlling attention.
So, is the notion of an attentional ‘filter’ still valid? The first thing to
note is that Broadbent used the term ‘filter’ simply as a metaphor for
what was happening in the brain. It is an elegant and convenient way of
conceptualising what is happening when we try to process incoming
information. The term ‘filter’ simply encapsulates the notion that some
information gets through and some does not. What later studies have
done is to increase our understanding of why filtering may be necessary,
how filtering is done and which brain regions are responsible. Thus, we
can still say that, especially when there is a lot of it, some incoming
information is filtered out by our attentional systems. What has changed
since Broadbent’s time is that we now have a far greater understanding
of the mechanisms and complexities of that filtering process.
As usual, a word of caution is necessary. Breaking the attentional system
down into its component parts is a useful and necessary way of
studying it and, indeed, is an inherently ‘scientific’ way of doing so. But
labelling parts of the process (e.g. ‘filter’ and ‘channel’ as in Figure 8.4)
somehow makes them appear more solid and discrete. It is important to
emphasise that these processes interact and overlap, not only with each
other, but also with other processes within the brain, such as those
concerned with perception, knowledge and emotion. Broadbent

352
 4 Later developments in theories of attention

explicitly addressed these interactions, and later studies have helped to

show how complex they can be, and how they can influence
performance in the real world.

Summary
. Further evidence has supported the notion that meaning may be
extracted before material is selected for further conscious
processing, and hence that filtering may occur later on in the
information-processing system.
. Kahneman’s limited-capacity theory proposed that there are limited
cognitive processing resources to make sense of incoming
information.
. There is evidence that several different brain regions are responsible
for controlling attention.

353
Chapter 8 Paying attention

5 Attention in context
5.1 Automatic and controlled attention
From the studies discussed so far in this chapter, you may be
wondering how anyone ever manages to do anything at all because their
cognitive resources appear to be so limited. There is, however, another
aspect of attention that, so far, we have not really considered.
Throughout this chapter, we have been developing the idea that, when
you pay attention to something, you become consciously aware of it.
But, as we will consider in this section, it appears to be possible to
process information and make responses without being conscious of the
underlying processing.

Activity 8.5
Think about a journey you make regularly, or a task that you perform
every day. You will probably find that you have a general recollection of
the journey, for example, to the extent that you remember the route, and
things along it, very well. This is a memory built up over many repetitions
of the journey. But how about details of a specific journey? For example,
do you remember how many red cars you saw? Can you remember if
you saw any red cars?

Did you find that you could not remember the details of a specific
journey? Perhaps you could recall a journey where you set off on a
route that is supposed to deviate at some point from a regular route,
such as the trip to work, only to find that you followed the regular
route anyway! You can probably recall examples where you reached the
end of a regular journey, almost without any recollection of that
journey. It was as if you were working automatically and your cognitive
resources were allocated elsewhere (perhaps thinking about work or
what to cook for dinner later).
This apparent distinction between processes that we are aware of and
those that appear to happen automatically has led to the development
of ‘two-process’ theories of attention, such as that proposed by Richard
Controlled attention
The conscious Shiffrin and Walter Schneider (1977). Two-process theories suggest that
allocation of resources there are, broadly, two types of attention. The first is the one that we
to the processing of have been discussing throughout this chapter: conscious (or controlled)
specific stimuli. attention. Controlled attention, as the name suggests, can be directed

354
 5 Attention in context

and controlled by the individual who is aware of the information that is

being attended to. The cost of this type of processing is that it draws
on cognitive resources and, as has already been demonstrated, this
limits the amount of information that you can consciously attend to.
The second type of attention is referred to as automatic processing. Automatic processing
Again, as the name suggests, this occurs automatically, below the level The automatic
of conscious awareness. Many aspects of a driving task, including route processing of stimuli
that occurs without
finding over familiar routes, can be controlled by automatic processes.
conscious awareness.
The advantage of using automatic processes is that large amounts of
material can be processed without drawing on scant cognitive resources.
The cost is that automatic processes are difficult to stop, precisely
because they are not ‘controlled’. This is how you can end up in the
wrong place at the end of a journey – your driving is being controlled
by an automatic process while your conscious attention is directed
elsewhere.

5.2 Attention and experience

Odmar Neumann (1984) suggested that an automatic process requires
appropriate stimuli from the environment that then trigger the skills
required to complete the task at a level below that of conscious
awareness. Thus, when driving a familiar route, the familiar stimuli
along the route trigger the appropriate automatic responses that have
been learned over many repetitions of the journey. Little conscious
control is required and, as a result, at the end of the journey we may
have little conscious recollection of it.
Furthermore, Neumann suggested that automatic processes are not
entirely uncontrolled, but are controlled at a level below conscious
awareness. Individuals can allocate attention to specific information; for
example, they can determine which part of their environment to allocate
attention to by moving the eyes. Daniel Gopher (1993) argued that
appropriate allocation of attention is a cognitive skill in itself, and one
that can be improved by training and experience.
Studies of driving support the notion that experience may affect the
allocation of attention. Geoffrey Underwood (2007) looked at the visual
scanning patterns of drivers and found that, whereas experienced
drivers adapted their scanning patterns to road complexity, novice
drivers did not. The change in scanning patterns was even more
pronounced in trained police drivers. Experienced drivers may not be

355
Chapter 8 Paying attention

aware that they are doing it, but they are changing the way in which
they allocate attention, depending on the driving conditions.
If experience affects the way in which people gather information, this
Top-down processes implies a possible involvement of so-called top-down processes,
When experience, whereby allocation of attention may be influenced by (for example)
expectations and/or expectancies and knowledge. Indeed, this interaction between incoming
stored knowledge
information (flowing bottom-up) and higher-level influences acting top­
influence the allocation
of resources to certain down in the control of attention was addressed by Broadbent in the
stimuli. work we have already discussed. You may remember from Section 2
that Broadbent’s model included the possibility that feedback from
information stored in memory, prior experience and expectations could
influence the selective filter.

5.3 Back to the real world

Returning to the notion of applying theories of attention to real-world
problems, the authors of this chapter conducted a study (Edgar and
Edgar, 2007) that demonstrated the influence of expectations on the
allocation of attention. A war game was conducted on an extensive
1:300 scale terrain model, as shown in Figure 8.9, using physical scale
models of tanks and other forces. Participants’ awareness of aspects of
the situation was measured. For example, one participant demonstrated
an impressive awareness of the broad strategic sweep of the battle but
no awareness at all of some basic physical features, such as the fact that
his own forces were coloured grey while those of the enemy were
coloured green. Such a lack of attention to basic physical characteristics
resulted in this participant attempting to open fire on his own troops in
a ‘friendly fire’ incident. This player appeared to allocate attention
according to what he thought should be there, based on his expectations
(enemy troops), rather than allocating resources to what actually was
there (his own troops).
This finding demonstrates that although on many occasions our
expectations may usefully direct us to the more relevant stimuli in our
surrounding environment, at times our expectations can also result in us
overlooking important aspects of the situation that are relevant to the
task at hand.

356
 5 Attention in context

Figure 8.9 Using a 1:300 scale terrain model to reveal the attentional basis
of friendly fire

So, what was going on in the head of our errant Fiesta driver described
in the anecdote that began this chapter? It is likely that the driver was
following a route that he had done many times before and was
therefore on ‘automatic pilot’. Furthermore, he almost certainly was not
expecting to see a cyclist. Thus, although the driver’s visual system may
well have detected a cyclist, the stimulus was not salient enough to attract
his attention and so he did not become aware of the danger and carried
on as if there were no cyclist. Rather like the friendly-fire incident just
described, it appears that the driver was responding to what he believed
should be there, not what actually was there. The failure of the Fiesta
driver to attend to an oncoming cyclist demonstrates how little of the
world around us we are really aware of at any one time.

5.4 Conclusion
The evidence presented throughout this chapter suggests that people
have limited attentional resources and can rarely process all of the
information impinging on their senses. This is why using a mobile
phone in a car is a bad idea. Using a phone requires vital cognitive
resources that could be better used elsewhere when driving.

357
Chapter 8 Paying attention

This chapter also suggested that there are two broad aspects to trying
to understand attention. The first is the theoretical interest in how the
human brain takes in and processes (or doesn’t process) information,
which helps us gain a better understanding of how the human mind
works. The second aspect relates to the use of this understanding to
explain how things work in the ‘real world’ and, crucially, to make
things work better, such as improving the way information is presented
to pilots and drivers, and designing effective early warning systems.
The development and testing of a theory alongside the application of that
theory to real-world problems was the approach advocated and
practised by Donald Broadbent. After all, a theory that cannot be
applied in practice needs to be treated with extreme caution! We hope
that, after reading this chapter, you can appreciate why Broadbent’s
systematic approach to experimental research and theory development
has formed the foundation for much later work and why he is,
justifiably, regarded as one of the key figures in the development of
modern cognitive psychology and in the establishment of psychology as
a scientific discipline.

Summary
. As well as conscious attention, there can be automatic processing of
information, which operates below the level of conscious awareness.
. Automatic processing does not draw on our limited cognitive
resources.
. Top-down processes may operate so that expectations, knowledge
and experience can influence allocation of attention.

358
 References

References
Boase, M., Hannigan, S. and Porter, J.M. (1988) ‘Sorry, can’t talk now … just
overtaking a lorry: the definition and experimental investigation of the problem
of driving and hands free car phone use’ in Megaw, E.D. (ed.) Contemporary
Ergonomics, London, Taylor and Francis.
Broadbent, D.E. (1954) ‘The role of auditory localization and attention in
memory span’, Journal of Experimental Psychology, vol. 47, no. 3, pp. 191–6.
Broadbent, D.E. (1957) ‘Immediate memory and simultaneous stimuli’, The
Quarterly Journal of Experimental Psychology, vol. 9, no. 1, pp. 1–11.
Broadbent, D.E. (1958) Perception and Communication, London, Pergamon.
Brookhuis, K.A., de Vries, G. and de Waard, D. (1991) ‘The effects of mobile
telephoning on driving performance’, Accident Analysis and Prevention, vol. 23,
no. 4, pp. 309–16.
Brown, I.D., Tickner, A.H. and Simmonds, D.C.V. (1969) ‘Interference between
concurrent tasks of driving and telephoning’, Journal of Applied Psychology,
vol. 53, no. 5, pp. 419–24.
Cherry, E.C. (1953) ‘Some experiments on the recognition of speech with one
and two ears’, Journal of the Acoustical Society of America, vol. 25, no. 5,
pp. 975–9.
Coull, J.T., Frith, C.D., Frackowiak, R.S.J. and Grasby, P.M. (1996) ‘A fronto­
parietal network for rapid visual information processing: a PET study of
sustained attention and working memory’, Neuropsychologia, vol. 34, no. 11,
pp. 1085–95.
Edgar, G. and Edgar, H. (2007) ‘Using signal detection theory to measure
situation awareness: the technique, the tool (QUASATM), the test, the way
forward’ in Cook, M., Noyes, J. and Masakowski, Y. (eds) Decision Making in
Complex Environments, Aldershot, Ashgate.
Glassbrenner, D. (2005) Driver Cell Phone Use in 2004: Overall Results, Research
note DOT HS 809 847, US Department of Transportation, National Highway
Traffic Safety Administration (NHTSA), Washington, DC; also available online
at www-nrd.nhtsa.dot.gov/Pubs/809847.PDF (Accessed 25 March 2010).
Gopher, D. (1993) ‘The skill of attentional control: acquisition and execution
of attentional strategies’ in Kornblum, S. and Meyer, D.E. (eds) Attention and
Performance XIV: Synergies in Experimental Psychology, Artificial Intelligence and
Cognitive Neuroscience, Cambridge, MA, MIT Press.
Gray, J.A. and Wedderburn, A.A.I. (1960) ‘Grouping strategies with
simultaneous stimuli’, Quarterly Journal of Experimental Psychology, vol. 12, no. 3,
pp. 180–4.
James, W. (1890) Principles of Psychology, Vol. 1, New York, NY, Holt.
Kahneman, D. (1973) Attention and Effort, Englewood Cliffs, NJ, Prentice-Hall.

359
Chapter 8 Paying attention

Kastner, S., De Weerd, P., Desimone, R. and Ungerleider, L.G. (1998)

‘Mechanisms of directed attention in the human extrastriate cortex as revealed
by functional MRI’, Science, vol. 282, no. 5386, pp. 108–10.
Kastner, S. and Pinsk, M.A. (2004) ‘Visual attention as a multilevel selection’,
Cognitive, Affective, and Behavioral Neuroscience, vol. 4, no. 4, pp. 483–500.
Neumann, O. (1984) ‘Automatic processing: a review of recent findings and a
plea for an old theory’ in Prinz, W. and Sanders, A.F. (eds) Cognition and Motor
Processes, Berlin, Springer-Verlag.
Redelmeier, D.A. and Tibshirani, R.J. (1997) ‘Association between cellular­
telephone calls and motor vehicle collisions’, New England Journal of Medicine,
vol. 336, no. 7, pp. 453–8.
Royal Society for the Prevention of Accidents (RoSPA) (2002) The Risk of Using
a Mobile Phone While Driving, Birmingham, RoSPA; also available online at
www.rospa.com/roadsafety/info/mobile_phone_report.pdf (Accessed 24
March 2010).
Shiffrin, R.M. and Schneider, W. (1977) ‘Controlled and automatic human
information processing: II. Perceptual learning, automatic attending and a
general theory’, Psychological Review, vol. 84, no. 2, pp. 127–90.
Strayer, D.L., Drews, F.A. and Crouch, D.J. (2006) ‘A comparison of the cell
phone driver and the drunk driver’, Human Factors, vol. 48, no. 2, pp. 381–91.
Treisman, A.M. (1960) ‘Contextual cues in selective listening’, Quarterly Journal
of Experimental Psychology, vol. 12, no. 4, pp. 242–8.
Underwood, G. (2007) ‘Visual attention and the transition from novice to
advanced driver’, Ergonomics, vol. 50, no. 8, pp. 1235–49.
Weiskrantz, L. (1994) ‘Donald Eric Broadbent, 6 May 1926–10 April 1993’,
Biographical Memoirs of Fellows of the Royal Society, vol. 40, pp. 33–42.

360
Chapter 9
Witnessing and remembering
Graham Pike and Nicola Brace
Contents
1 Introduction 365
1.1 Types of memory 366
1.2 What are memories? 368
1.3 Reconstructing memories 369
2 The Loftus and Palmer (1974) study 374
2.1 Reconstruction of automobile destruction 374
2.2 Experiment 1: estimating speed 375
2.3 Experiment 2: did you see any broken glass? 378
2.4 Integrating memories 380
3 Method: applied experiments 382
3.1 Ecological validity, consequentiality, practicality and
research ethics 382
3.2 Evaluating Loftus and Palmer (1974) 385
3.3 Elizabeth Loftus 387
4 False memories 389
4.1 The ‘lost in the mall’ study 389
4.2 False memories and food preferences 393
5 Questioning memory 397
5.1 Police interviewing 397
5.2 Courtroom practices 398
5.3 Conclusion 400
References 401
Chapter 9 Witnessing and remembering

Aims and objectives

After studying this chapter you should be able to:
. describe the different types of memory and provide examples of
episodic memory
. discuss the reconstructive nature of episodic memory
. appreciate the methodological and ethical challenges facing those
who research eyewitness memory
. describe how research on the misinformation effect has moved on
into research on false memories
. appreciate the impact of psychological findings on police practice
and interviewing guidance.

364
 1 Introduction

1 Introduction
The last chapter introduced you to the notion of limited capacity,
particularly that there are limits to our capacity to attend to the world
around us. As you will see in the current chapter, there are also limits
to our capacity to remember. Even if we are able to direct all our
attention to an event, it is not necessarily the case that we will be able to
remember all that happened.
Like attention, memory is important in all aspects of everyday life. This
chapter, as well as exploring memory more generally, looks specifically
at the question of eyewitness memory – people’s ability to remember
crimes, car accidents and other important events. As you will see,
eyewitnesses experience great difficulties when it comes to remembering
a crime. Not only do they forget a lot of what happened, but they
sometimes also remember things that didn’t happen at all. For instance,
the 2010 TV series called Eyewitness, which the BBC produced in
collaboration with The Open University, studied the memories of
witnesses for staged, but very realistically enacted, crimes such as a
violent attack in a pub and a heist from a security van. The differences
between their accounts and what actually took place were astounding.
For example, they disagreed on who perpetrated the attack in the pub;
indeed, one witness confused the attacker and the victim. In the case of
the heist, one witness reported seeing a car reverse round a corner and
the criminals leaping out, although the incident involved no so such
event. Another witness described in some detail the sunglasses that one
of the gunmen was wearing, even reporting the brand. However, not
only was the man in question not wearing any sunglasses, he was in fact
wearing a balaclava.

365
Chapter 9 Witnessing and remembering

Figure 9.1 From balaclava to sunglasses: the gunman wearing the dark
balaclava (image A) may have been misremembered by the witness as
having dark rather than light eye patches (image B), leading them to recall
the perpetrator as having worn sunglasses (image C)

To understand why eyewitnesses tend to make such fundamental errors,

we first need to find out more about how human memory works. This
chapter will explore memory and examine why our memory of events is
often incomplete, and also how it is that we sometimes report
remembering events that never happened at all.

1.1 Types of memory

We have a friend called Jim who has often been described as having a
very good memory. If you want to know who sang an obscure pop
Semantic memory song from the 1980s or the name of a character from a television
Remembering factual series, Jim is your man. You probably know someone like Jim, but even
information. if you don’t you probably equate the phrase ‘good memory’ with
someone who can readily bring to mind a large number of facts.
Procedural memory Memory for facts, which is referred to by psychologists as semantic
Remembering how to memory, is actually only a small part of our memories, which also
perform actions. include:
. procedural memory: remembering how to perform certain actions
Prospective memory (e.g. ride a bicycle)
Remembering actions . prospective memory: remembering that you need to do something
that need to be in the future (e.g. that tomorrow you need to pick up the dry­
completed in the
cleaning)
future.

366
 1 Introduction

. episodic memory: remembering an event that happened in the past Episodic memory
(e.g. a football game or a play) Remembering events
that happened in the
. autobiographical memory: a type of episodic memory that relates past.
to events that are of personal relevance to the person doing the
remembering (e.g. first day at school).
Autobiographical
Another way of categorising memory is to look at the type of task that memory
remembering actually involves. For example, exams that are designed to Remembering events
test the memory of a student routinely use two different types of that are of personal
question. The first type consists of essay questions or any other format relevance.
that asks a question without providing the answer. These questions
require the student to recall some of the material that they learned and
memorised. The second type is multiple-choice questions, which both
ask a question and provide the answer (among a series of incorrect
answers). This type of question requires the student to recognise the
answer. In everyday life we use recognition memory almost constantly.
As long as we can perceive objects, places and people, our mind
automatically (at least most of the time) recognises what we are seeing.
For instance, we can tell from memory that the person standing in front
of us is our mother and that she is standing in front of a tree.
Memories that are recalled tend to occur mostly in conversations,
particularly when we are describing something from memory. Witnesses
are routinely asked to use both recall and recognition memories as part
of a police investigation. For example, recognition memory is used if
the witness is asked to identify the perpetrator of the crime from a line­
up, and recall memory is used when the witness is asked to give a
statement of what they saw.
Encoding
As well as there being different types of memory and remembering, it is Processing information
also possible to describe memory as happening through three distinct received from one of
stages. The first is encoding, which involves processing information the senses into
from the physical senses into memory. The second stage is storage, memory.
which involves the information that was encoded being stored in
memory until it is needed again. The final stage is retrieval, which is
Storage
when the information is brought from memory to be used again. For
Storing information in
example, when studying you try to encode the information you read in memory.
a textbook, store it until your exam and then retrieve it when answering
a question.
Retrieval
Another way of looking at these three stages is by using computers as Recovering information
an analogy. When you enter some text into your word-processing that is stored in
software, the computer encodes the information. Once you select the memory.

367
Chapter 9 Witnessing and remembering

‘save’ icon it stores it so that it can be retrieved at a later date,

whenever you chose to ‘open’ that file.
This chapter will concentrate on how people recall an episodic memory,
in that it will look at witnesses describing a crime from memory. This
also means that subsequent sections will focus on the retrieval stage.
There is a practical reason for this. The police interview eyewitnesses
only after a crime has taken place. By that time they have no influence
on how the memory of the crime was encoded or stored. The police
can only assist the witness to recall the event as accurately as possible.
This is why applied research on eyewitness memory has focused
extensively on the retrieval stage.

1.2 What are memories?

The different ways of classifying memory, described in Section 1.1,
provide part of the answer to this question. Regardless of the specific
type, memory essentially involves information from one of the senses
(e.g. vision or hearing) being encoded and stored so that it can later be
retrieved. But in relation to human memory, this is only part of the
answer.
One way of describing human memory is to look at what it is not. Let’s
return to the computer analogy. To some extent computer and human
memory are similar, as both involve the encoding, storage and retrieval
of information. However, computer memory:
. is based in one or more discrete components, usually the computer’s
hard drive and memory chips
. is perfect, in that (barring such things as viruses and spilt coffee) the
information that is encoded, stored and retrieved is exactly the same
. does not require the computer to be able to understand the
‘meaning’ of the information being stored.

By comparison, human memory:

. is stored throughout the brain. Although it is possible to identify certain
brain structures that are important to particular parts of the memory
process, the ‘memories’ themselves seem not to be stored in any one
particular region, but are instead distributed throughout the brain.
. is imperfect. If a computer read this chapter, if it was either scanned
as an image or typed in as a word-processor file, it would be able to
reproduce every word in this chapter in the correct order. Imagine

368
 1 Introduction

the results if I asked you to do the same task! Critically, the

computer would reproduce the same information no matter what
the conditions were under which it was asked to do so. As you will
see, human memories seem to change significantly as a result of how
they are retrieved.
. often requires understanding the meaning of the information. Although a
computer would be able to reproduce every word in this chapter,
there is no real sense in which it could be considered to have
understood the information contained in the words. However, if I
asked you to recall this chapter, it is likely that most of what you
would remember is what the words and sentences meant. Indeed,
without being able to deduce this meaning it is likely you would
remember very little at all. To think of it another way, imagine this
chapter was written using ancient Egyptian hieroglyphics: how much
do you think you could remember? Of course, it matters little to the
computer what language is used, because the computer just
remembers the characters and their order, not what they mean.

The differences in the performance of computer and human memory

provide hints as to the extremely complex role that memory has to play
in human life. Unlike the computer, humans need to make sense of
their environments and that requires creating meaning of the world.
Most of the time any inaccuracies in our memories will go unnoticed
and are unlikely to have any particularly negative consequences. On
occasion, though, people encounter a situation where errors of memory
can have adverse consequences. One such situation is in an exam,
where failure to remember facts can lead to a poor mark. Another
situation is where the police rely on the memory of an eyewitness to
discover what happened in a crime. As you will see in Section 1.3,
people not only forget information but sometimes also remember things
that didn’t happen.

1.3 Reconstructing memories

To say that when presented with a page of text people remember the
meaning of the words rather than every actual word seems obvious.
However, deciding exactly what the meaning of the text is requires that
it is first interpreted, and how it is interpreted depends on who the
person doing the interpreting is. You can imagine that different people
may well interpret the same basic information, whether this information
is a page of text, a picture or a real-life scene, in very different ways.

369
Chapter 9 Witnessing and remembering

That information can be interpreted differently poses a particular

problem in the case of eyewitness memory, as the police need witnesses
to describe what happened accurately. To demonstrate some of the
errors people make when trying to remember, we will take a look at a
famous study conducted in 1932 by Sir Frederic Bartlett. Bartlett taught
Donald Broadbent at Cambridge and, like Broadbent, is considered by
many to be an important figure in the history of cognitive psychology.
The study involved asking a group of British participants to read and
remember a Native American folk tale called ‘War of the Ghosts’.

Figure 9.2 Frederic Bartlett (1886–1969)

370
 1 Introduction

Activity 9.1
The 'War of the Ghosts' story used by Bartlett is reproduced below. To
experience the task faced by Bartlett’s participants, read the story
through once, attempt to recall as much of it as you can, and then
answer the specific questions that we have provided after the story. Do
not reread the story before you have tried to recall it and attempted the
questions.

Box 9.1 War of the Ghosts

One night two young men from Egulac went down to the river to
hunt seals and while they were there it became foggy and calm.
Then they heard war-cries, and they thought: ‘Maybe this is a war­
party’. They escaped to the shore, and hid behind a log. Now
canoes came up, and they heard the noise of paddles, and saw one
canoe coming up to them. There were five men in the canoe, and
they said:
‘What do you think? We wish to take you along. We are going up
the river to make war on the people.’
One of the young men said, ‘I have no arrows.’
‘Arrows are in the canoe,’ they said.
‘I will not go along. I might be killed. My relatives do not know
where I have gone. But you,’ he said, turning to the other, ‘may go
with them.’
So one of the young men went, but the other returned home.
And the warriors went on up the river to a town on the other side of
Kalama. The people came down to the water and they began to
fight, and many were killed. But presently the young man heard one
of the warriors say, ‘Quick, let us go home: that Indian has been
hit.’ Now he thought: ‘Oh, they are ghosts.’ He did not feel sick, but
they said he had been shot.
So the canoes went back to Egulac and the young man went
ashore to his house and made a fire. And he told everybody and
said: ‘Behold I accompanied the ghosts, and we went to fight. Many
of our fellows were killed, and many of those who attacked us were
killed. They said I was hit, and I did not feel sick.’

371
Chapter 9 Witnessing and remembering

He told it all, and then he became quiet. When the sun rose he fell
down. Something black came out of his mouth. His face became
contorted. The people jumped up and cried.
He was dead.
(Bartlett, 1932, p. 65)

Now write down on a separate sheet all you can recall of the story. Once
you’ve done this, try to answer the questions below before looking at the
story again.
1 How many men from Egulac went down to the river?
2 When they heard war-cries, what did they hide behind?
3 What was the name of the place near the town where the warriors
went? Was it:
(a) Katana
(b) Lakaka
(c) Malaka
(d) Kalama
4 Had the men from Egulac caught any fish before they heard the war­
cries?

Now you can look back at the story to see how much you recalled and
whether you answered the questions correctly.

You may well have found that you could answer the questions even if
you did not recall that part of the story earlier, particularly question 3,
where you were also presented with the answers (and therefore had to
recognise rather than recall the answer). Specific questions such as those
above can help us remember extra information because the question
itself contains information which prompts our memories, or in other
words leads us to remember more. Question 4 demonstrates one danger
Leading questions with such leading questions, as the information contained in the
Questions that suggest question is false, as the story states that the men were hunting seals,
or imply an answer. and not fish. As you will see, this type of false information can
influence what we remember.
When Bartlett studied how well people could remember the 'War of the
Ghosts' story, he discovered that not only did people fail to recall the

372
 1 Introduction

whole story, but they often recalled elements of the tale either
incorrectly or indeed recalled something that was not part of the story
at all. For example, some participants recalled there being a boat, rather
than a canoe, and one reported that the death at the end was the result
of a fever and that the person was foaming at the mouth. Bartlett
suggested that these changes were a result of the participants
interpreting the story, and, as the participants came from a Western
culture, they tended to interpret the events from a Western viewpoint.
So, for example, to the participants in Bartlett’s study, the event in the
final paragraph resembled an epileptic seizure, so ‘something black’
coming from a person’s mouth was remembered as the culturally more
familiar phenomenon of ‘foaming at the mouth’. In other words,
participants were not remembering exactly what they had read, but were
reconstructing the meaning of the text in the light of their own culture.
The task used by Bartlett is fairly unusual, particularly as it involves
people from one culture trying to remember a story from another
culture. For that reason we cannot necessarily conclude that memory is
always subject to the distortions reported by Bartlett, particularly when
we are dealing with everyday scenarios. Another way of saying this is
that the task used by Bartlett was not particularly ecologically valid in You read about
relation to everyday memory. Nonetheless, Bartlett’s work is extremely ecological validity in
important in demonstrating that the memories we retrieve may not be a Section 4.2 of
Chapter 5.
particularly good match for the original event. Also, it shows that we
remember information in ways that make it meaningful to ourselves.
Most of the time, people are unaware that memory involves
reconstruction and most of the time it matters little if an event is
recalled accurately or not. However, as you will see in the next section,
there are occasions when errors in recall can have serious consequences.

Summary
. There are different types of memory, including semantic, procedural,
prospective, episodic and autobiographical memory. Recalling an
event, such as a crime, involves episodic memory.
. Memory can involve recall or recognition, and is often described as
involving the stages of encoding, storage and retrieval.
. Remembering can involve reconstructing what was experienced in
the light of our expectations, knowledge and experiences, which can
lead to errors and even to memories of things that never happened.

373
Chapter 9 Witnessing and remembering

2 The Loftus and Palmer (1974) study

Errors in our memory often tend to go unnoticed in everyday life and,
even if noticed, they tend not to be that important. It is also rare for a
situation to arise that requires us to remember a past event in particular
detail, as usually all we need to recall are a few meaningful elements.
Last week I (Graham) celebrated my friend Hayley’s birthday. I can
remember that I picked up an Indian takeaway on the way, that I took
my two collies (as my friends love dogs), and that another friend,
Gemma, spent quite a bit of the evening looking for badgers in the
garden. I cannot remember who was in the Indian restaurant and,
beyond the fact that Gemma had a straw hat, I could not tell you what
anyone was wearing. Such detail is just unimportant. However, what if it
had turned out that the Indian restaurant had been robbed just after I
was there, or one of my friends had gone missing that night? The police
would want to know as much as possible and I would suddenly be in a
position where I was expected to remember very specific details.
Just as trying to recall the details of the 'War of the Ghosts' story can
be considered to be outside the everyday experience of the participants,
witnessing and trying to recall what happened in a crime is also outside
the everyday experience of most of us. One key difference between
recalling an event as part of everyday life and recalling a crime in a
police interview is that the latter requires us to remember as many
details as possible in response to questions. In 1974, Elizabeth Loftus
and John Palmer published two experiments that examined what effect
the specific wording used in questioning had on the information
remembered by a participant. The research, which was supported by the
US Urban Mass Transportation Administration, and has become one of
the most famous and often-cited psychological studies ever conducted,
is described next.

2.1 Reconstruction of automobile destruction

The first thing to note about this study is the title chosen by the
authors, ‘Reconstruction of automobile destruction: an example of the
interaction between language and memory’, and particularly that they
used the word ‘reconstruction’ prominently. That memories are
‘reconstructed’ rather than remembered exactly as the original event
occurred was the conclusion Bartlett reached following his research
conducted forty years previously. Loftus and Palmer were building on

374
 2 The Loftus and Palmer (1974) study

the work of Bartlett, and examining how the reconstructive nature of

memory interacts with the specific language used in questions designed
to probe recall.
The study conducted by Loftus and Palmer is an example of applied
research, in that it seeks to address a particular issue that has arisen in
the ‘real world’ rather than an issue arising from academic theory. As
the results of applied research need to be relevant to the ‘real world’, it
is important that the studies conducted use stimuli and tasks that are a
good match for real-world situations. The applied problem being
studied by Loftus and Palmer (1974) concerns how accurately a witness
can describe what happened in an automobile collision, particularly how
the questions they are asked might affect what they remember
happening.
If you have ever been involved in or witnessed an automobile collision,
you will know that the speed of the vehicles tends to be a key part of
any subsequent investigation, whether by the police or by insurance
companies. The problem is that very often the only information about
the speed of the vehicles comes from those that witness the collision,
and people tend not to be particularly good at estimating vehicle speed.
As an example of this, Loftus and Palmer cite evidence presented by
James Marshall (1969) that involved asking Air Force personnel to
estimate the speed of a car that was moving at 12 miles per hour
(mph). Even though the participants knew in advance that they would
be asked about the speed of the car, estimates ranged from 10 mph to
50 mph.
Loftus and Palmer suggest that, given this very large range in estimates,
it is likely that there are variables that could influence the estimate given
by an individual and that the specific phrasing of the question used to
elicit the estimate could be one such factor. Loftus and Palmer’s
rationale for choosing the phrasing of the question came from legal
studies and rules relating to leading questions.

2.2 Experiment 1: estimating speed

To study this issue, Loftus and Palmer conducted an experiment in
which forty-five students were each shown seven film clips of vehicle
collisions taken from driver education films.

375
Chapter 9 Witnessing and remembering

Figure 9.3 A recent replication of Loftus and Palmer (1974)

Following each clip, the participants were asked to complete a

questionnaire, which first asked them to give an account of the accident
and then asked specific follow-up questions including a question on
what speed the cars were going.
Four of the clips were staged collisions at particular speeds (20 mph, 30
mph and two at 40 mph), and these were first used to determine how
Mean accurate the participants’ estimates of speed were. The mean estimated
An average that is speed for each of these four clips is presented in Table 9.1.
calculated by adding
together all the items to
be averaged and then Table 9.1 Mean estimates of speed in Loftus and Palmer’s (1974)
dividing the total by the experiment
number of items.
Actual speed of vehicles Mean estimate of vehicles’ speed
(mph) (mph)

20 37.7
30 36.2
40 39.7
40 36.1

376
 2 The Loftus and Palmer (1974) study

Activity 9.2
Have a look at the data in Table 9.1. Does anything strike you as
unusual about the mean estimates of speed? Try looking for trends in the
data. For example, do the numbers increase or decrease in each
column?

The actual speed of the vehicles (left-hand column) increases towards

the bottom of the table, and in fact doubles from 20 mph to 40 mph.
However, the mean estimates of speed do not increase and instead do
not vary much at all. This suggests that the estimates of speed given by
the participants did not tend to increase as the speed of the vehicle
increased and instead participants gave an estimate of about 37 mph
whether the vehicle was travelling at 20 mph or 40 mph! In other
words, the participants were not very good at estimating speed, just as
Marshall had previously noted.
Loftus and Palmer examined the effect that the specific phrasing of the
question had on estimates of speed by varying the key verb that was
used in the question. The question asking participants to provide an
estimate of speed was: ‘About how fast were the cars going when they
hit each other?’ The question was constructed so that the key verb ‘hit’
could be replaced with an alternative. In all, five different verbs were
used, namely:
. hit
. smashed
. collided
. bumped
. contacted.

Each participant was asked the same question using the same verb for
each clip they saw. So, each verb was used with a total of nine
participants. If we were to describe this experiment in terms of the
variables used, we could say that the independent variable (the variable
manipulated) was the phrasing of the question, and that the dependent
variable (the variable measured) was the estimate of speed provided by
the participant. The authors calculated the mean speed reported for
each of the five questions, and these are presented in Table 9.2.

377
Chapter 9 Witnessing and remembering

Table 9.2 Mean speed reported by key verb used in Loftus and Palmer’s
(1974) experiment

Key verb Mean speed reported

hit 34.0 mph

smashed 40.5 mph
collided 39.3 mph
bumped 38.1 mph
contacted 31.8 mph

So, although the participants saw exactly the same collisions, using the
verb ‘smashed’ resulted in speed estimates that were on average 8.7
mph higher than when the verb ‘contacted’ was used.

2.3 Experiment 2: did you see any broken glass?

The results of Experiment 1 show that although the actual speed of the
vehicles seems to have little effect on the estimates given, the particular
phrasing of the question does have a marked effect. Loftus and Palmer
identified two possible interpretations of this finding.
. Participants are uncertain whether to say 30 mph or 40 mph, for
example, and are swayed in their decision by the verb used, so that,
for example, ‘smashed’ biases them toward 40 and ‘contacted’
toward 30.
. The phrasing of the question causes a change in the participant’s
memory of the collision, so that using the verb ‘smashed’ results in
their memory being of a faster collision than when ‘contacted’ is
used.

In many respects the second explanation is the more interesting. If

correct, it tells us something profound about human memory – namely
that a memory may be reconstructed in accordance with the wording of
the question.
To explore which one of the two interpretations of the findings was
more likely, Loftus and Palmer conducted a second experiment. They
wanted to find out whether the wording of a question could lead
participants to reconstruct a memory of an event so as to include
entirely novel elements. In other words, they wanted to see if they could
get participants to remember details that never actually happened.

378
 2 The Loftus and Palmer (1974) study

In this experiment 150 participants were shown a short film depicting a

fairly low-speed vehicle collision that did not result in any broken glass.
Each participant then completed a questionnaire similar to the one used
in Experiment 1. Again, the critical question was the one relating to the
speed of the vehicles, but this time only three variations were used, with
each variation being given to fifty participants:
. About how fast were the cars going when they smashed into each
other? (The ‘smashed’ condition)
. About how fast were the cars going when they hit each other? (The
‘hit’ condition)
. No question about the speed of the vehicles. (The control condition)

A week after viewing the film, the participants returned and, without You were introduced to
seeing the film again, were asked a series of ten questions about the the definition of a
collision. The key question was ‘Did you see any broken glass?’, to control condition in
Section 2.3 of
which the participant had to answer either ‘yes’ or ‘no’. This question is
Chapter 3.
misleading, as the film did not show any broken glass. The authors
expected that those participants previously given the question containing
the verb ‘smashed’ would be more likely to report broken glass than
those in either the ‘hit’ or control condition.
The results of this experiment revealed that participants in the
‘smashed’ condition provided higher estimates of speed (mean estimate
10.46 mph) than those in the ‘hit’ condition (mean estimate 8.0 mph).
In Table 9.3 you can see how the participants in each condition
responded to the question about the presence of broken glass.

Table 9.3 Distribution of ‘yes’ and ‘no’ responses to the question ‘Did you
see any broken glass?’ in Loftus and Palmer (1974), Experiment 2

Response Verb condition

Smashed Hit Control

Yes 16 7 6
No 34 43 44

Looking at Table 9.3, it is clear that more than twice as many

participants in the ‘smashed’ condition erroneously reported broken
glass as participants in the other two conditions (16 in the ‘smash’
condition and only 6 or 7 in the other conditions). One important thing
to note about the data presented in Table 9.3 is that in every condition

379
Chapter 9 Witnessing and remembering

far more participants responded ‘no’ than ‘yes’, which shows that most
participants’ memories were accurate even when asked a question that
was intended to mislead.

2.4 Integrating memories

As using the verb ‘smashed’ led to both higher speed estimates and
more participants remembering broken glass, one might be tempted to
conclude that this experiment simply shows that people with higher
estimates of speed are also more likely to remember broken glass. To
test whether this is the case, Loftus and Palmer conducted additional
analysis on their data and found that, while it may be the case that
people in the ‘smashed’ condition gave higher estimates of the car’s
speed, the estimate of speed does not fully explain the effect of the
question on the recall of broken glass. In other words, use of the word
‘smashed’ seems to have had an effect above and beyond that of
increasing the estimate of speed.
In interpreting these findings, Loftus and Palmer suggested that a
memory can consist of both information encoded while watching the
actual event and information encoded when retrieving the memory.
These two sources of information can be integrated into one memory,
so that a person is unable to tell the two sources apart. In this case,
some participants were integrating the suggestion that the collision
involved two vehicles ‘smashing’ into one another and that there might
have been broken glass, into their memory of the actual collision.
Since Loftus and Palmer (1974) published their research, many further
experiments have been conducted in this area and the term post-event
Post-event information has been developed to refer to information that may be
information incorporated into a memory of an event after it has happened. Likewise,
Knowledge provided the term misinformation effect has come to be used to refer to the
about an event after it
errors that can occur in memory as a result of post-event information,
has taken place.
such as being asked a leading question that suggests something that
didn’t happen. As you will see, providing post-event information can
Misinformation effect have a powerful effect on memory, and in regard of misinformation, it
Errors in memory can result in people remembering things that never happened!
caused by incorrect
post-event information.

380
 2 The Loftus and Palmer (1974) study

Summary
. Loftus and Palmer (1974) conducted two experiments to explore
factors that influence memory for the details of a car accident.
. They found that the way in which a question was phrased influenced
what was recalled.
. They proposed that information recalled about an event may contain
the original information encoded plus information provided at the
time of recall.

381
Chapter 9 Witnessing and remembering

3 Method: applied experiments

The experiments conducted by Loftus and Palmer (1974) are examples
of applied research, in that they were clearly designed to investigate an
issue of relevance to accident investigation and witness interviewing
more broadly. However, simply attempting to answer an applied
question is only part of what constitutes applied research. In addition, it
is important to design the study in such a way that the results obtained
can be usefully and meaningfully applied to the area in question.
Constructing experiments that achieve this is quite a complex and
difficult task.

3.1 Ecological validity, consequentiality, practicality

and research ethics
One of the differences between applied and theoretical research is that
the former needs to try to match the conditions of the real-world
situation being studied as closely as possible. Psychologists use the term
‘ecological validity’ to refer to this matching of experimental and real­
world conditions. Consider two experiments designed to test eyewitness
memory. In the first experiment a ‘live’ crime is staged in front of
unsuspecting participants, who are then taken to a police station to be
interviewed about what they saw, and several weeks later attend an
identification parade to try to identify the perpetrator. As this
experiment is a good match for what would happen to witnesses in a
real crime and real investigation, it can be said to have high ecological
validity. The second experiment involves showing participants
photographs of a scene at a restaurant and instructing them to
remember as much of the scene as they can, and then giving them
multiple-choice questions to test their memory. As the conditions in this
second experiment are a poor match for those of a real crime and
investigation, it can be said to have low ecological validity.

382
 3 Method: applied experiments

Figure 9.4 Two ways of researching eyewitness memory: participants in a

laboratory viewing a video of a staged crime and participants observing the
staged event live in a side street

Ecological validity is just one of the factors an applied researcher needs

to bear in mind. Consider the high ecological validity eyewitness
experiment described in the previous paragraph. The participants see a
live crime and are interviewed by the police, just like a witness, but their
experience is the same as that of a witness only if they are not told that
they are taking part in an experiment until after the experiment is
finished. If at any point during the experiment the participants learn
that the crime and the investigation are staged, then they will realise that
their actions will have no real consequences. A real witness knows that
any evidence they provide could lead to someone being either
prosecuted or released, but a participant in an experiment knows that
any consequences are limited to the results of the experiment. Thus,
consequentiality is a key difference between experimental research and the
real world, even if the research has high ecological validity.
Of course, an experiment can have both high ecological validity and
consequentiality if the participants do not realise that they are taking
part in an experiment. The problem with this suggestion is that it
involves deceiving the participants and, particularly if the study is one of
witness memory, exposing them to a potentially stressful and therefore
harmful situation. Such deception is unethical as it violates the ethics
principles that participants must give their informed consent and have You read about
the right to withdraw from the research at any point. informed consent and
the right to withdraw in
When designing an applied experiment, the task of the researcher is to Chapter 2, Section 3.1.
do their best to achieve high ecological validity and adhere to ethics
guidelines, while also considering what is practical. This last factor,
practicality, tends to play a crucial role in applied research, if only

383
Chapter 9 Witnessing and remembering

because there tends to be only a very limited amount of time and

research funding available. For example, to achieve high ecological
validity in eyewitness memory research, it would seem sensible for
researchers to hire actors and use live staged crimes. However, live
staged crimes are not always practical and tend to be very expensive.
The other problem with live events is that if you have to restage them
(to add more participants to your experiment, for example), it is very
hard to replicate the first event exactly – and it is important that all
participants see exactly the same thing. It is for these reasons that most
eyewitness memory research, like that of Loftus and Palmer (1974), uses
video rather than live events, and has a researcher to test or interview
the participants rather than an actual police or insurance investigator.
Likewise, a lot of applied research is conducted in the laboratory or
classroom, rather than ‘in the field’.

Activity 9.3
As research ethics require participants taking part in an eyewitness
memory study to be informed that they are taking part in research, they
cannot be made to think they are participating in a real police
investigation. Can you think of any way that some degree of
consequentiality could be built into such a study?
If you’re stuck for ideas, consider ‘who’ the participants are, ‘how’ they
are recruited and ‘what’ they are told about the research.

What type of thing did you come up with? Below we have listed some
of the techniques eyewitness memory researchers have employed:
. telling participants that the results of the study will inform expert
testimony given by the researcher in court
. telling participants that the results of the study will be used to
inform changes to police guidelines and practice
. using real police officers, possibly in a police station, to interview
the eyewitnesses.

Did you think of any of the ideas in the list above? One thing you may
have considered is rewarding the participants according to how well
they performed. One problem with this is that it is difficult to design a
reward structure that would mimic the actual experiences of being
involved in a real investigation. Another problem is finding sufficient
funds to give rewards substantial enough to influence behaviour.

384
 3 Method: applied experiments

3.2 Evaluating Loftus and Palmer (1974)

Consider the two experiments conducted by Loftus and Palmer with
regard to their ethics, practicality, ecological validity and consequentiality.
It would have been impractical (not to mention downright dangerous)
to stage actual vehicle collisions, so the use of videos of collisions is as
ecologically valid as possible. The only deception used in the study was
not telling the participants that the verb in the key question would be
manipulated, which is a very minor detail that would not affect the
experience of the participant, so would seem irrelevant in terms of
informed consent. That leaves consequentiality, which as we have seen
is almost impossible to include in an experiment without seriously
contravening ethical guidelines.
One way in which Experiment 1 did deviate from the experience of a
real witness was that each participant saw seven collisions and had to
answer questions about each one. In reality, a witness is likely to
provide details about only one collision, so this element of the study
can be considered to lack ecological validity. Another potential problem
with the experiments is that of demand characteristics, a term that Demand
refers to the tendency of participants to behave in an experiment in the characteristics
way they think the researcher wants them to behave. Recent research in The cues in an
experiment that
this area (Ost et al., 2008) has drawn a distinction between false reporting
participants can use to
and false memories – between whether post-event information actually work out how the
alters what a person remembers or just what they say they remember – experimenter expects
and has estimated that the rate of false memory is approximately only them to behave.
one-third of the rate of false reporting. In other words, it could be the
case that twice as many participants say they remember something that
didn’t happen as the number who actually do remember it. It is possible,
then, that at least some of the participants in Experiment 2 were simply
saying they remembered broken glass because (even if unconsciously)
that’s what they thought the researcher wanted, but that their memories
had not actually been changed.
Although demand characteristics are undoubtedly a problem, it is
important to remember that Loftus and Palmer were conducting applied
research and that a real witness in a real investigation may also behave,
even answer questions, in the way they think the police officer wants
them to behave, so in this case demand characteristics are replicating
pressures that would be experienced by a real witness.

385
Chapter 9 Witnessing and remembering

Box 9.2 Why do it this way?

Although there are disadvantages to the type of experiment
conducted by Loftus and Palmer, this technique also has several
important advantages over other methods of exploring memory for
vehicle collisions. One possible alternative would be to interview
witnesses to real collisions. Although this does away with problems
relating to ecological validity and consequentiality it has the
disadvantage that the researcher would not know what actually
happened in the collision, including what speed the cars were
travelling at. As Loftus and Palmer have shown, just because a
witness says the speed was 30 mph and that there was broken
glass does not mean that this is accurate. Of course, if there were
multiple witnesses you could compare accounts, but how would you
ever know which was the accurate one? One method that would be
ecologically valid and would have a record of what actually
happened would be to interview spectators who had witnessed a
crash in a motorsport event (so that accounts could be compared
with video footage of the crash). The problem with this approach is
the lack of control over the spectators. For example, different
people would see different things if they saw the crash from
different positions and there would be no way of preventing
spectators from discussing the crash or even watching it on TV.
This would mean that their memories would already have been
contaminated by post-event information.
The type of applied experiment conducted by Loftus and Palmer
therefore has the advantage that the researchers know exactly what
happened in the events presented, and can ensure that all
participants see the same thing and are not exposed to post-event
information outside the experiment. In other words, an applied
experiment allows the researchers to have considerable control
over and knowledge of what happens in the study. Another big
advantage of an applied experiment over, say, interviews or surveys
of real witnesses is that an experiment allows the researcher to
establish cause and effect, while other techniques are often limited
to seeing whether two different variables are related or associated.
As Loftus and Palmer used an experiment, they were able to
conclude that the verb used in the question affected participants'
recall of the event and their judgement of speed.

386
 3 Method: applied experiments

3.3 Elizabeth Loftus

Since the publication of the Loftus and Palmer (1974) study, Elizabeth
Loftus has become a world-renowned psychologist, contributing a great
deal to our understanding of human memory (despite being voted ‘least
likely to succeed as a psychologist’ by her fellow graduate students!) and
conducting many more experiments relevant to witness memory. Her
PhD research concerned semantic memory and how people retrieved
information such as the names of animals and, although she continued
to research in this area for a few years, after moving from Stanford to
the University of Washington (Seattle) she decided she wanted to
conduct research with more social relevance. She managed to combine
her expertise in memory with her interest in crime and legal issues by
beginning research into witness memory. The 1974 article that she
wrote with John Palmer brought her work to the attention of the legal
community and led not only to invitations to address legal seminars but
eventually to appearances as an expert witness. This initially proved to
be a battle as such testimony was generally considered to be
inadmissible in the USA in the 1970s.

Figure 9.5 Elizabeth Loftus

Memory is relevant to many criminal investigations and trials; indeed,

any investigation involving a witness will also involve witness memory.
One type of case that became of particular interest to Loftus involved
what is known as ‘recovered’ memories, namely memories of childhood

387
Chapter 9 Witnessing and remembering

abuse that the victims remember only many years later after
participating in counselling sessions. The debate centres on the accuracy
of these memories. This debate intrigued Loftus and was to become the
focus of a great deal of her research over the next few decades. In
particular she was interested in whether it was possible to implant a
memory of an event that never happened. The next section explores
some of the research conducted by Loftus on this issue.

Summary
. The key focus of applied research is to address issues of real-life
relevance, and therefore it needs to mimic real-world conditions as
far as possible.
. Practical considerations and research ethics limit the extent to which
research can achieve both ecological validity and consequentiality.
. Elizabeth Loftus is a leading figure in research on eyewitness
memory.

388
 4 False memories

4 False memories
The debate as to whether recovered memories are real or false
memories has been as polarised as it has been vehement, and has
become known as the ‘memory wars’. One of the problems for
psychologists studying this phenomenon is that, given the trauma and
timescale involved, it is very hard to conduct an experiment to ‘test’
such memories. Loftus was able to find a way of investigating the issue
of false memories that has become a standard tool in applied cognitive
psychology, and which first occurred to her while she was driving past a
shopping mall.

4.1 The ‘lost in the mall’ study

The pioneering study that followed was conducted by Elizabeth Loftus
together with colleague Jacqueline Pickrell (Loftus and Pickrell, 1995),
and is frequently referred to as the ‘lost in the mall’ study because
participants were led to believe that they had been lost as a child when
out shopping, either in a department store or in a shopping centre.
The study involved three phases. In the first phase, twenty-four
participants worked through a booklet that contained four short stories
about events from their childhood that had previously been provided by
a parent, sibling or other older relative. Participants were asked to write
down what they could recall, if anything, about each event in the
booklet. If they did not remember the event, then they were to write ‘I
do not remember this’.
Of the four events depicted in the stories, three were true and one was
false. The false story involved the participant getting lost, when aged
about five years, during a shopping trip. The relative supplied details
about a plausible shopping trip – for example, where this might have
taken place – and also verified that the participant had not actually
experienced such an event around that age. Each fabricated false event
contained common elements such as ‘lost for an extended period’,
‘crying’ and ‘found by an elderly woman’.

389
Chapter 9 Witnessing and remembering

Figure 9.6 It is easy to get lost in a shopping centre

One to two weeks later, in phase two, the participants were reminded of
the four events and then asked to recall as much as they could about
them, regardless of whether or not they had previously provided that
information in the booklets. They were also asked to rate the clarity of
their memory for each event, and their confidence that if given more
time they would remember more details. After a further one to two
weeks, in phase three, the participants were again interviewed using the
same procedure.
What did the findings reveal? The results showed that across all three
phases the participants remembered about 49 out of the 72 true events
(68 per cent). Of key interest, though, was whether or not they would
claim to remember anything about the false event. In phase one, seven
of the twenty-four participants wrote something about the false story in
the initial booklet. In phase two, six of these seven continued to report
something about being lost and they also went on to report either a bit
or all of the false event in phase three.
In summary, then, 75 per cent of participants resisted the suggestion
False memory
that they had been lost as child, but 25 per cent were led to believe that
A memory for an event
that did not actually an entire event had happened to them. In other words, 25 per cent of
take place but was participants formed a false memory. A key question is whether these
instead only suggested false memories were in any way different from the real memories.
to have occurred. Loftus and Pickrell (1995) found that participants used fewer words to

390
 4 False memories

describe false memories than they did true memories. In addition, the
clarity and confidence ratings provided by the participants were also
lower for false memories than for true memories. Below is an extract
from the Loftus and Pickrell paper, outlining a false memory:

During the second interview, she said,

I vaguely, vague, I mean this is very vague, remember the lady
helping me and Tim and my mom doing something else, but I
don’t remember crying. I mean I can remember a hundred times
crying … I just remember bits and pieces of it. I remember
being with the lady. I remember going shopping. I don’t think I,
I don’t remember the sunglasses part.

She went on to remember that the elderly lady who helped her was
‘heavy-set and older. Like my brother said, nice.’
(Loftus and Pickrell, 1995, p. 723)

This extract nicely illustrates the sparse nature of many false memory
reports. Other studies have also found that reports of real memories
generally contain more detail, in particular sensory detail such as sounds
and smells, than reports of false memories do (e.g. Schooler
et al., 1986). The point, though, is that participants believe that the
event did actually occur in their past, even though they report very little
in relation to this false memory. Furthermore, with repeated suggestive
interviews – that is, interviews that include leading questions –
participants will recall the event in quite a bit more detail (Ost
et al., 2005).

Activity 9.4
Reflect for a moment on why Loftus and Pickrell picked being lost as a
child for the false event. Are there any ethical considerations?

Remember that Loftus was interested in researching whether memories

recovered from childhood were real or false memories, so it was
necessary to find something that adult participants might have
experienced, in their childhood. Being lost as a child for a brief period
of time is something that many might have experienced so this is a
relatively plausible event. It was chosen because it was deemed to be

391
Chapter 9 Witnessing and remembering

‘safe’ in that, although reasonably serious, it would not be traumatic,

and so implanting such a memory should not cause harm to anyone.
Other researchers in the mid 1990s conducted similar studies,
convincing adults that as children they had been hospitalised, or had
experienced an accident at a family event, for example. In one particular
study, some children were persuaded that when younger they had been
to hospital to have a mousetrap removed from their hand (Ceci
et al., 1994). Typically, around 25–30 per cent of adults who participated
in such studies reported false memories, and if children were the
participants then the percentage was higher. The rate of false-memory
reporting has varied across studies and this is owing to differences in
how plausible the suggested event was considered to be by the
participants and also how the suggested event was presented to them.
One of the more unusual methods of planting a false memory is to
show doctored photographs (see Figure 9.7). The photograph on the
left is the original, which has been doctored in the image on the right to
show a different background. The doctored photograph could be used
to plant a memory of a trip on a boat.

Figure 9.7 Doctoring photographs: a means of implanting false memories

392
 4 False memories

4.2 False memories and food preferences

Loftus and Pickrell found that false memories were less detailed and
less clear, but in what other ways may a false, implanted memory be
different from a real memory? One way of exploring this question is to
consider whether a false memory can influence behaviour in the same
way that a real memory can. Several studies have considered whether a
false memory can influence attitudes and actions, and explored whether
being told of a bad experience with some object meant that the object
was avoided later on. As memories about food are relatively easy to
manipulate and are not traumatic, several studies have implanted the
false suggestion that participants had become ill after eating a certain
food when they were children and then looked to see if that food was
avoided in the future.
One such study was undertaken by Elke Geraerts et al. (2008). It
involved 180 first-year undergraduates at a university in the
Netherlands, who were told they were taking part in a study called
‘Food and personality’. In phase one, they were asked about their
experiences and preferences regarding food. This included finding out
whether they had got sick after eating egg salad before their tenth
birthday, and also how much they liked to consume different foods,
including egg salad.
In phase two, one week later, 120 of these participants were given the
false suggestion that as a child they had been ill after eating egg salad.
The remaining sixty participants acted as a control group and were not Control group
given this false suggestion. The experimenters implanted the false Participants who are
memory by giving the participants false feedback about their responses allocated to the
‘baseline’ or control
in phase one; among other things, the ‘false suggestion’ group were told
condition.
‘You got sick after eating egg salad’.
Following this they were again asked about food preferences, allowing
Geraerts et al. to see if there had been changes in how confident
participants were as to whether they had been sick after eating egg salad
as a child. Participants were also asked to indicate whether they had a
specific belief or memory from before age ten about a number of
different events, one of which was being sick after eating egg salad.
The key question, though, was whether participants would act
differently as a result of the false suggestion and start to avoid eating
egg salad. To explore this, participants were taken to another room
where they received a bogus debriefing, and as a treat were offered

393
Chapter 9 Witnessing and remembering

drinks and sandwiches. The sandwiches contained five different fillings,

one of which was egg salad. Which type of sandwich was eaten by each
participant was recorded by someone who was not aware of whether
the participant was in the false suggestion group or in the control
group.
Four months later, the same participants were asked to participate in
what they thought was a separate study but in reality comprised the
third phase of the same study. Many volunteered and during this study
they were invited to eat up food that would otherwise be thrown away.
You won’t be surprised to learn that the food comprised five different
types of sandwich, one of which was egg salad. After the food was
removed they were again asked about their food preferences and
whether they had a specific belief or memory about being sick after
eating egg salad.
The results showed that none of the participants had realised the study
was about false memories. Of those participating in all three phases of
the study:
. 35 were classified as believers as their confidence that the event had
happened increased and they reported a memory or belief that the
event had taken place.
. 54 were non-believers as their confidence about the critical event
remained the same or decreased.
. 50 were control participants.

The findings are summarised in Figure 9.8.

Activity 9.5
Inspect the graph for a moment. Did these three groups (believers, non­
believers and control) differ in the number of egg-salad sandwiches they
had eaten? Which group changed most from Phase 2 to Phase 3?

394
 4 False memories

0.5
Mean number of egg-salad sandwiches consumed

Believers
0.45
Nonbelievers
0.4 Control group

0.35

0.3

0.25

0.2

0.15

0.1

0.05

0
Phase 2 Phase 3

Figure 9.8 Mean number of egg-salad sandwiches consumed in the second

and third phase

In the second phase, both believers and non-believers ate fewer egg­
salad sandwiches compared with control participants. In the third phase,
believers ate fewer egg-salad sandwiches than did both non-believers
and control participants, and it was the non-believers group who
changed most across the two phases. Providing false feedback acted in
the short term to deter both believers and non-believers from eating
egg-salad sandwiches, probably because the notion of being sick after
eating egg salad was fresh in their minds. This false feedback had no
longer-lasting effect on non-believers, who ate a similar number of egg­
salad sandwiches as the control group in phase 3. Geraerts et al.
suggested that believers, having contemplated the egg-salad event, had
created memories about being ill after eating egg salad as a child and
this false belief influenced their eating behaviour four months later, so
that this group continued to eat far fewer egg-salad sandwiches than the
control group.
So, there is evidence to suggest that false memories relating to food can
have repercussions for later eating behaviour. Research is now looking
at the influence of other types of false memory on other behaviours. If
false memories behave like real memories in influencing behaviour, you
might be wondering whether there is a way of telling whether a
particular memory is true or false. This is one of the biggest challenges

395
Chapter 9 Witnessing and remembering

facing memory researchers. Although on average false memories tend to

be less detailed and contain less sensory information, this and other
more complex methods for distinguishing between true and false
memories have not been found to be sufficiently reliable for use in the
courtroom. Moreover, Bernstein and Loftus (2009) suggest that:

In essence, all memory is false to some degree. Memory is

inherently a reconstructive process, whereby we piece together the
past to form a coherent narrative that becomes our autobiography.
In the process of reconstructing the past, we color and shape our
life’s experiences based on what we know about the world. Our
job as memory researchers and as human beings is to determine
the portion of memory that reflects reality and the portion that
reflects inference and bias. This is no simple feat, but one worthy
of our continued investigation.
(Loftus and Bernstein, 2009, p. 373)

Summary
. Studies have found that some participants report memories that
have been implanted by researchers; these are called false memories.
. Loftus and Pickrell (1995) pioneered research on false memories.
. Recent work suggests that false memories about being sick from egg
salad affected some participants’ eating behaviour, even after four
months.

396
 5 Questioning memory

5 Questioning memory
The research conducted by Loftus and colleagues highlights how fallible
and malleable memory can be. However, it is worth bearing in mind
that memory is not always prone to inaccuracies. In any one study of
false memory, only a minority of people were shown to be susceptible
to suggestions made by researchers or to leading questions. This is
important because it suggests that memories can be and often are
accurate. This is why psychologists have conducted a great deal of
research into developing methods for helping people to remember
accurately. This chapter concludes by considering how best to help
someone recall as much and as accurately as possible, and, as you will
see, this is another area where psychology has had a significant impact.

5.1 Police interviewing

The most obvious piece of advice arising from research is to avoid the
use of leading questions, and this is something that is taught to police
officers during their training in many countries. The training provided
to police officers in the UK is grounded in research, and guidance is
provided on how best to interview and what tools might assist witness
recall.
One such interviewing tool is that witnesses are asked to recall the
event they saw without interruption. The police interviewer may start
with what is called an open-ended question, which is a question that Open-ended question
does not lead the witness in any way; for example, ‘What do you A question that does
remember about the event?’ The idea behind this ‘report everything’ not suggest or imply an
answer and usually
approach is that allowing witnesses to remember one thing, even
invites the witness to
something possibly irrelevant, such as the colour of the carpet, might provide as much detail
cue or trigger the witness to recall something else about the event. Also, as they can.
this encourages the police interviewer to hear the full story from the
witness, and then later to ask questions that are compatible with the
witness’s own version of events.

397
Chapter 9 Witnessing and remembering

Figure 9.9 Interviewing technique is crucial for assisting witness recall

Interview training also covers a number of other tools designed to assist

memory, which are known as mnemonic techniques. For example, police
interviewers are instructed on how useful it can be to ask the witness to
mentally reinstate the context of the event. This includes recalling the
physical environment, such as the location in which the crime took
place, as well as lighting conditions, smells, emotional reactions and
feelings. You may have used this technique yourself when trying to find
something you have misplaced by thinking back to when you last had
seen or used the lost item.
This approach to interviewing is based on memory research, and the
psychologists who brought together the relevant research to apply it to
police interviewing were Ron Fisher and Ed Geiselman (e.g. Fisher and
Geiselman, 1992). They proposed an interviewing procedure called the
Cognitive interview cognitive interview, which has not only been incorporated into the
A procedure for police training undertaken in England and Wales but has also informed
interviewing witnesses NHS guidance on interviewing those involved in patient safety incidents.
that draws on research
on memory and
incorporates a number 5.2 Courtroom practices
of retrieval strategies.
The finding that our memories are prone to suggestion has also
influenced courtroom practices. It is deemed acceptable to pose leading
questions in the courtroom during cross-examination; for example, ‘The
car was dark blue, wasn’t it?’ This is because it is considered crucial by

398
 5 Questioning memory

the legal profession that the accuracy of evidence obtained be probed,

to allow any unreliable or dishonest witness to be exposed. The Youth
Justice and Criminal Evidence Act of 1999 has, however, recognised
that certain witnesses may be more susceptible to the effects of leading
questions than others. This Act acknowledges the problems facing
witnesses deemed vulnerable – because of their age or because of
learning difficulties, for example – and who might find giving evidence
at a criminal trial to be a particularly difficult experience.

Figure 9.10 Evidence via video link

Special measures have been introduced for such witnesses, one of which
is to allow the police in England and Wales to record the initial witness
evidence on video, which can then be presented to the court as
evidence-in-chief. Guidance on how to interview witnesses for the
purposes of criminal proceedings was provided, among others, by
psychologists Ray Bull and Helen Westcott in the Memorandum of Good
Practice (Home Office, 1992). This guidance was revised and expanded
in Achieving Best Evidence (Home Office, 2002). The guidance promotes
the use of open-ended questions (e.g. ‘Please tell me what happened.’)
followed by specific, non-leading questions (e.g. ‘What colour was the
car?’ if the witness has already mentioned a car), and to minimise or
avoid forced-choice questions (e.g. ‘Was the car black or blue?’) or
questions that have several parts. In Scotland, a steering committee
worked with psychologists Amina Memon and Lynn Hulse to put
together guidance on interviewing children and other vulnerable

399
Chapter 9 Witnessing and remembering

witnesses (Scottish Executive, 2003). This collaboration between

psychologists and policymakers has helped to bring about improvements
in the way evidence is elicited, in particular from vulnerable witnesses.

5.3 Conclusion
This chapter has focused on episodic memory, namely memory for
events. You were introduced to the early work of Bartlett that
highlighted the reconstructive nature of memory and the way in which
people’s knowledge, general beliefs and expectations may shape their
recall of events. The malleability of memory has been explored further
through Elizabeth Loftus’s work on eyewitness memory. Together with
other researchers’ work, Loftus’s research has shown not only that
leading questions can lead to inaccuracies in recall but also that some
participants can even be led to believe that an entire fictitious event
took place earlier in their life.
In spite of the fallibility and malleability of memory, there are many
episodes from our past that we will remember with a high degree of
accuracy, and any errors are unlikely to have any significant impact.
However, very occasionally people find themselves in a situation – such
as witnessing a crime – when remembering accurately becomes
especially important. For this reason, knowing how memory works in
such unusual and often stressful circumstances, and understanding how
interviewing techniques can help improve the accuracy of recall, has
significant practical implications. Witness testimony represents an
important area of applied research, in that psychological theories and
experimental findings have made a significant contribution to law
enforcement and the criminal justice system more generally.

Summary
. Research by psychologists on memory has been applied to witness
interviewing, and the cognitive interview was designed for use in
police interviews.
. Psychologists have also contributed to guidance on how to help
vulnerable witnesses give evidence.

400
 References

References
Bartlett, F.C. (1932) Remembering: A Study in Experimental and Social Psychology,
Cambridge, Cambridge University Press.
Bernstein, D.M. and Loftus, E.F. (2009) ‘How to tell if a particular memory is
true or false’, Perspectives on Psychological Science, vol. 4, no. 4, pp. 370–4.
Ceci, S.J., Huffman, M.L.C., Smith, E., and Loftus, E.F. (1994) ‘Repeatedly
thinking about a non-event’, Consciousness and Cognition, vol. 3, nos 3–4,
pp. 388–407.
Fisher, R.P. and Geiselman, R.E. (1992) Memory Enhancing Techniques for
Investigative Interviewing: The Cognitive Interview. Springfield, IL, Charles C. Thomas.
Geraerts, E., Bernstein, D.M., Merckelbach, H., Linders, C., Raymaekers, L.
and Loftus, E.F. (2008) ‘Lasting false beliefs and their behavioral
consequences’, Psychological Science, vol. 19, no. 8, pp. 749–53.
Home Office (1992) The Memorandum of Good Practice on Video Recorded Interviews
with Child Witnesses for Criminal Proceedings, London, Home Office.
Home Office (2002) Achieving Best Evidence in Criminal Proceedings: Guidance for
Vulnerable and Intimidated Witnesses, Including Children, London, Home Office
Communication Directorate.
Loftus, E.F. and Palmer, J.C. (1974) ‘Reconstruction of automobile destruction:
an example of the interaction between language and memory’, Journal of Verbal
Learning and Verbal Behavior, vol. 13, no. 5, pp. 585–9.
Loftus, E.F., and Pickrell, J.E. (1995) ‘The formation of false memories’,
Psychiatric Annals, vol. 25, no. 12, pp. 720–5.
Marshall, J. (1969) Law and Psychology in Conflict, New York, NY, Anchor Books.
Ost, J., Foster, S., Costall, A. and Bull, R. (2005) ‘False reports of childhood
events in appropriate interviews’, Memory, vol. 13, no. 7, pp. 700–10.
Ost, J., Granhag, P.A., Udell, J. and Roos af Hjelmsäter, E. (2008) ‘Familiarity
breeds distortion: the effects of media exposure on false reports of real life
traumatic events’, Memory, vol. 16, no. 1, pp. 76–85.
Schooler, J.W., Gerhard, D., and Loftus, E.F. (1986) ‘Qualities of the unreal’,
Journal of Experimental Psychology: Learning, Memory, and Cognition, vol. 12, no. 2,
pp. 171–81.
Scottish Executive (2003) Guidance on Interviewing Children Witnesses in Scotland,
Edinburgh, The Stationery Office.

401
 Conclusion

Conclusion
In the last three chapters of Investigating Psychology you read about the
work of Pierre Paul Broca and Carl Wernicke on the relationship
between the brain and language (Chapter 7), about Broadbent’s
experimental work on attention (Chapter 8), and about the applied
research of Elizabeth Loftus and colleagues on eyewitness memory
(Chapter 9). Research described in the three chapters has one thing in
common: it explores cognitive processes – such as language, attention
or memory – that take place in the mind and are therefore not directly
observable. However, while Chapter 7 explored how a particular
psychological function, namely language, is controlled by the brain,
Chapters 8 and 9 focused largely on experimental work that was more
interested in examining the limits of the human information-processing
system, without looking into the underlying brain structures.
Another important point that emerges from the chapters is the
relationship between an explanation, often formulated as a psychological
theory, and evidence. As you have learned, theory (even if only a
preliminary one) about how some process works informs a research
hypothesis. By testing the hypothesis a researcher gathers evidence
which then feeds back into theory. A theory is refined in the light of the
new evidence, and new hypotheses are generated to test it further. This
continuous process that links theory and research underlies the cycle of
enquiry. For example, in Chapter 7, you read about Broca’s and
Wernicke’s early findings about the neurological basis of language, which
focused on two areas of the brain. Over the subsequent century their
account of the control of language was refined (and continues to be
refined even today), on the basis of evidence from a variety of studies. In
Chapter 8, you read about Broadbent’s model of attention developed on
the basis of experiments using dichotic listening, and about how this
model was tested and refined by the findings of further experimental
work. Finally, as you read in Chapter 9, Loftus and Palmer’s work
consisted of a series of experiments that were necessary in order to
assess and revise their own notions about how memory for events
operates.
The cycle of enquiry raises once again an issue that you encountered in
some of the earlier chapters: namely, that the work of a researcher in
psychology is rarely complete. There are always theories to be tested,
gaps in research to be filled, new questions to be answered and new
insights revealed by new technologies and new techniques. Psychology

403
Part 3 Conclusion

has to be based on evidence and this evidence (as well as the theories
that it helps to build) is subject to continuous evaluation and
interpretation.
Another message conveyed in this last part of the book is that theories
and research can have real-world relevance. In Chapter 7, you saw how
research on the neurological basis of language revealed the plasticity of
the brain. Understanding that there is the possibility of some recovery
of function has led to the development of a number of therapeutic
techniques designed to maximise recovery and alleviate some of the
problems faced by those suffering from brain damage. Research on the
limits of human attention (Chapter 8) and on how and why performing
two tasks simultaneously causes interference has been used to address a
number of real-life issues, from the effects of mobile-phone use while
driving, to cockpit design and understanding the cognitive dynamic
behind friendly-fire incidents on the battlefield. Finally, in Chapter 9,
you read about the fallibility of memory. You were probably already
aware of how easy it is to forget things, but may have been less aware
of the possibility that memories could be so easily (and often
unintentionally) distorted, or under certain circumstances even
implanted. This has important implications for those rare situations
when it is especially important to remember something we saw or
experienced as accurately as possible.
These final three chapters also offered you a taster of further sub­
disciplines of psychology. Chapter 7 introduced a particular area of
biological psychology, namely neuropsychology, which is concerned with the
relationship between the brain and different psychological functions.
Chapters 8 and 9 introduced you to cognitive psychology, which explores
how information is processed, and how knowledge is stored and
retrieved.
What is particularly challenging to psychologists investigating cognitive
processes is that they are dealing with phenomena, such as perception,
attention, memory, thinking and language, that do not lend themselves
to direct observation. One option, which you learned about in
Chapter 7, is to look at how these processes are controlled by the brain.
As you read in the chapter, psychologists have to be cautious about
how they interpret the results of brain damage. The correspondence
between damage in one area of the brain and the loss of a specific
psychological function does not imply that this area is directly and solely
responsible for regulating that function (remember the radio analogy).
Only with the discovery of more recent techniques, such as brain

404
 Conclusion

imaging, has it been possible to gain a more accurate understanding of

the complex nature of the relationship between the brain (and indeed
different areas of the brain) and cognitive functions.
An alternative to looking at the brain – the ‘hardware’ behind cognition
– is to devise models of the ‘software’ – namely, the human
information-processing system. This is where analogies can be useful.
Broadbent, for instance, drew on terms from the world of computers,
and these terms provided him with a way of thinking about the
relationship between different cognitive processes in humans. However,
it is important to be aware of the limitations of analogies. As Chapter 9
highlighted, there are important differences between how a computer
processes information and how a person does so.
Crucially, as you have read in Chapter 8, creating abstract models to
represent cognition and examining the brain are not mutually exclusive
approaches to studying how humans process information about the
world around them. Experiments using brain-imaging techniques can
help psychologists build a richer picture of how cognitive processes
might operate.
Finally, it is worth noting that cognition does not exist in isolation from
the issues discussed in Parts 1 and 2 of Investigating Psychology. Cognitive
processes interact not only with one another but also with other aspects
of the person and situation. In Chapter 7, you read about how
therapeutic intervention to assist aphasia needs to acknowledge the
other problems the individual may have, such as limited attention or
apathy. The notion of a finite amount of processing resources,
introduced in Chapter 8, does not preclude that our performance is
likely to be influenced by factors such as stress or fatigue, or even by
the mere presence of others. In Chapter 9, you saw that what we
remember is influenced not only by factors within us but also by what
is around us – by the way a question is phrased, by the information
another person provides, and by the culture to which we belong. So,
although psychologists seek to isolate and focus on a particular process
or behaviour, they are aware that in reality this process or behaviour
does not operate in isolation, but is part of the infinitely complex
mechanism that is the human mind.

405
Final note
Nine chapters and several hundred pages after you embarked on this
voyage of discovery, the time has come to take stock, reflect on the
experience as a whole and consider what you have learned about
psychology.
The main aim of the book has been to initiate you into the world of
psychology. By exploring some of the most important and influential
studies conducted by psychologists, the preceding chapters gave you a
taster of the main areas of research and introduced you to the different
methods used by psychologists to study the human mind and behaviour
(and in some cases animal behaviour). You learned how a number of
pioneering figures within the discipline formulated research questions
and created innovative ways of answering them; how they developed,
tested and refined theories about human behaviour, performance,
personality characteristics, and mental states; and, most importantly,
how they inspired successive generations of researchers to push the
boundaries of human knowledge and further enhance our understanding
of psychology. It is important to note, however, that what you read
about in the chapters is just a snapshot of all the work carried out by
psychologists to date, but one that nevertheless introduced you to a
number of broader issues which are important for understanding what
psychology is about.
So, when all the work examined in the preceding chapters is pulled
together, what picture emerges of psychology?
One of the main tasks of Investigating Psychology has been to introduce you
to what psychologists do and how they go about the tasks of discovering
how the mind works or why humans (or other animals) behave the way
they do. In reading the material presented in the chapters you probably
noticed that psychology places important emphasis on research.
Psychology is an evidence-based discipline. In everyday language, the
term ‘psychologist’ is often used to describe someone who helps people
address personal problems and issues. Having read the nine chapters of
Investigating Psychology, you will now realise that there is more to
psychology than this. Much of what we know about psychology has
come from painstaking research using a whole range of methods:
experiments, observations, questionnaires, interviews, case studies, to
name but a few. Psychology, as a perpetually developing body of
knowledge, has come about as a result of endless testing and refining of

409
Final note

theories, ideas and methods, with the help of available technology – and
occasionally luck. So, an important thing that distinguishes psychology
from a purely philosophical enquiry about the human condition (of the
kind that has been around since antiquity), or from everyday, lay
theories about ‘human nature’, is this emphasis on research and evidence.
The emphasis on evidence and research makes psychology an inherently
dynamic discipline. Every hypothesis tested and every prediction
confirmed only yields more questions to be answered, and provides
opportunities for further investigation. The cycle of enquiry that drives
psychological research is a never-ending process. However, the dynamic
nature of the discipline stems also from the fact that developments in
society as a whole are constantly creating new challenges as well as new
opportunities for psychological investigation. For example, the invention
of complex electronic instruments, or even just the emergence, over the
past hundred years, of the car as an everyday mode of transport, has
highlighted the limitations of human attention which, all of a sudden,
acquired importance that they did not have a century or two ago. The
rising popularity of the internet since the 1990s has changed the way
younger generations in particular interact with others, which in turn
meant that many assumptions about social development and children’s
interactions with their peers needed to be revised. The rise of modern
warfare has similarly posed new questions about obedience to authority
and its potentially tragic consequences. All of these developments
presented psychology with new challenges. At the same time, the
evolution of science and technology – especially the invention of brain­
imaging techniques, computers, video and audio recording devices, etc.
– has provided psychologists with new opportunities, and has made
psychology an even more fascinating, varied and sophisticated discipline.
Important in this respect is also the increasing awareness of the role of
ethics in psychological research. The realisation over the past half a
century or so that research must at all times preserve the dignity and
safety of human participants and consider the welfare of animals used
in experiments has imposed limits on what psychologists can do. The
work of Harry Harlow, for example, is a clear example of why drawing
these limits was both necessary and welcome. And yet, the emphasis on
ethics in research has not brought psychological research to a halt.
Instead, researchers have developed innovative ways of studying those
aspects of behaviour that it would be unethical to explore in a
traditional laboratory setting. Technology has proved useful here too,

410
with computer simulation and virtual worlds providing exciting new
opportunities for ethically responsible research.
So what has been the main finding of all this research and discovery?
Part 3 of Investigating Psychology ended with a reference to the mind as an
‘infinitely complex mechanism’. This sentence captures what is probably
the most important message of this book. Our behaviour, personality
characteristics, mental states and abilities, all of which define who we
are and how we interact with the world around us, are the outcome of a
complex interplay of a variety of different factors, processes and
influences. We are as much a product of what we might refer to as
nature – things like brain structures and innate aspects of behaviour
which are passed on through the genes – as we are of nurture – for
instance, learning, the influence of parents or peers, and more generally
the effect of the environment. All of these different internal and
external factors come together to create who we are and what we do.
So, when trying to explain the human mind and behaviour, looking at
any one of these factors is seldom enough. Answers are not to be
sought just inside the brain, in abstract models of the human
information-processing system, in some aspect of personality or in the
influence of the environment. Rather, they are to be found in the
interplay between the different elements that interact to produce the
complex mechanism that we are. Although psychologists sometimes
disagree about the relative importance of the different influences, there
is general recognition that who we are and what we do is a result of a
complex interplay of biological, cognitive and social factors. We are the
product of an interaction between nature and nurture, internal and
external influences. It is also worth noting that, in this context, debates
and disagreements are not necessarily a bad thing. On the contrary –
they are an intrinsic part of scholarly endeavour and an important
driving force behind innovation and scientific advancement.
The increasing awareness of the complexity of psychology’s subject
matter and the development of the wide variety of methods that are
used to study the human mind and behaviour have led to the
emergence of more specialised sub-disciplines within psychology. In the
chapters of Investigating Psychology you encountered a number of them:
social psychology, cognitive psychology, neuropsychology, developmental
psychology and the psychology of individual differences. This is not an
exhaustive list. There are also many others, including clinical psychology
and occupational psychology. There are also more complex
combinations – developmental neuropsychology, evolutionary social

411
Final note

psychology, etc. Within many of these sub-disciplines there are further

distinctions on the basis of, for instance, the preference for qualitative
or quantitative methods, or adherence to a specific theoretical
perspective or tradition.
The existence of so many different divisions within psychology can
leave one feeling somewhat overwhelmed. The important thing to
remember is that none of the sub-disciplines exists in isolation. Imagine
for a second that you are a developmental psychologist embarking on a
project to study attention in very young children. To address this
question you would need to acquaint yourself with the more general
cognitive psychological research on attention in adults, of the kind you
read about in Chapter 8. You might also find it useful to consult
literature from neuropsychology that looked at the development, in early
childhood, of specific brain regions that are implicated in regulating
cognitive functions such as attention. You could also consult the work
of clinical or educational psychologists and see whether findings from
studies with children who are unable to concentrate for long (for
example, those suffering from Attention Deficit/Hyperactivity Disorder)
can help you shed light on normal development of attention. In this
example, as is often the case, the different sub-disciplines work together
to provide a more complete picture of a highly complex phenomenon.
Another theme running through the pages of Investigating Psychology is that
psychological research is often informed by the desire to understand
some real-life, everyday aspect of human (or animal) behaviour; or
indeed by the need to shed light on and even solve some sort of
practical problem. From understanding how the brain works and
helping patients recover from brain damage, through working with the
police on developing ways of obtaining accurate witness accounts, to
understanding the psychology of perpetrators of genocide or the role of
the media in aggression, researchers have always been responsive to the
issues and concerns of the society and era that they inhabited. This is
an important point because it shows that psychologists are not
interested in the mind or behaviour for its own sake. In fact, a number
of sub-disciplines within psychology have a distinctly applied focus, in
that they look at how the understanding of psychology can be made
relevant to what happens in courtrooms and prisons (forensic
psychology), in the classroom (educational psychology), on the sports
pitch (sports psychology) or in the workplace (occupational psychology).
And yet, it is also worth noting that when psychologists address
practical concerns, they rarely do so in isolation from the broader world

412
of academic enquiry. Psychologists are not the only researchers
interested in human behaviour or psychological processes, nor is what
they do always distinct from the work carried out in other disciplines or
professions. Neuropsychologists, for example, work closely with the
medical profession, and the lines that delineate the two disciplines are
often blurred. Cognitive psychologists are working increasingly closely
with computer scientists in pursuit of a more accurate understanding of
the human information-processing system. In fact cognitive science, which
is today a research area in its own right, emerged in recent decades as a
product of this collaboration. Psychologists interested in innate aspects
of human behaviour will often work together with zoologists and
ethologists. You also learned that Milgram’s work was not only
influenced by the desire to explain a historical event, but that it also had
an impact on how some historians interpret the conduct of perpetrators
of Nazi crimes. Equally, it is today common for books or
documentaries about famous historical figures to reflect on the person’s
childhood and their relationship with parents or parental figures. This is
a manifestation of the subtle but enduring influence in contemporary
culture of Sigmund Freud and his theories about the role of childhood
relationships on adult personality.
This collaboration with other disciplines offers yet another illustration
of psychology’s importance in contemporary society. It shows that
psychological research, together with other disciplines and professions,
makes a difference and, as well as addressing practical problems, also
influences the way in which people, and society as a whole, view and
interpret events in the world. This continuing relevance of psychology
goes a long way towards explaining why the interest in it remains so
strong in the world today.
We hope that you enjoyed the material covered in Investigating Psychology
and that it whetted your appetite for the discipline. Most importantly,
we hope that it instilled in you the desire to pursue an interest in all
things psychological!
Jovan Byford and Nicola Brace

413
 Acknowledgements

Acknowledgements
Grateful acknowledgement is made to the following sources:

Figures
Figure 1.1: Copyright © Chris Barham/Daily Mail/Rex Features;
Figure 1.2: Mary Evans Picture Library/SIGMUND FREUD
COPYRIGHTS; Figure 1.3 left: Copyright © Popperfoto/Getty Images;
Figure 1.3 right: Getty Images; Figure 1.4: Copyright © Imagno/Getty
Images; Figure 1.5: Copyright © Janine Wiedel/Photofusion; Figure 1.6:
Copyright © Underwood & Underwood/Corbis; Figure 1.7: Copyright
© dpa/Corbis; Figure 1.7: Copyright © Bettmann/Corbis; Figure 1.8:
Copyright © Yevgeny Khaldei/Corbis; Figure 1.9: Copyright © Farrizio
Bensch/Corbis; Figure 2.1 top left: Copyright © Time & Life Pictures/
Getty Images; Figure 2.1 top right: Copyright © Steve Drew/EMPICS
Sports/Press Association Images; Figure 2.1 bottom: Copyright ©
LOOK Die Bildagentur de Fotografen GmbH/Alamy; Figure 2.2:
Courtesy of Alexandra Milgram; Figure 2.3: Copyright © Popperfoto/
Getty Images; Figures 2.4, 2.5, 2.6 and 2.7: Milgram, S (1974)
Obedience to Authority, Harper & Row. Copyright © 1974 by Stanley
Milgram; Figure 2.8: Slater, M et al (2006) ‘A Virtual Reprise of the
Stanley Milgram Obedience Experiments’, PLoS ONE, No 1(1).
Copyright © 2006 Slater et al; Figure 2.9: Courtesy of Yad Vashem –
Holocaust Martyrs’ and Heroes’ Remembrance Authority; Figure 3.1:
Copyright © picturesbyrob/Alamy; Figure 3.2 left: Copyright ©
Universal Pictures/Ronald Grant Archive; Figure 3.2 right: Copyright ©
Rockstar Games; Figure 3.3: Copyright © Time & Life Pictures/Getty
Images; Figures 3.4 and 3.5: Copyright © Albert Bandura; Figure 3.6:
Copyright © 2005 Bubbles Photo Library; Figure 3.7: Copyright ©
MBI/Alamy; Figure 3.8: Byron, T (2008) Safer Children in a Digital
World: The Report of the Byron Review, Department for Children,
Schools and Families. Crown copyright material is reproduced under
Class Licence Number C01W0000065 with the permission of the
Controller, Office of Public Sector Information (OPSI); Figure 3.9:
Copyright © NetPics/Alamy; Figure 4.1: Courtesy of Frederick Toates;
Figure 4.7: Copyright © Nina Leen/Time & Life Pictures/Getty
Images; Figure 4.9: Copyright © Denkou/Alamy; Figure 4.10: Copyright
© Matt Rourke/AP/Press Association Images; Figure 4.11: Copyright
© Ashley Cooper/Corbis; Figure 5.1 top left, top right and bottom
right: Copyright © John Birdsall/Press Association Images; Figure 5.1
bottom left: Copyright © Andres Rodriguez/Alamy; Figures 5.2, 5.3
and 5.5: Copyright © Nina Leen/Time & Life Pictures/Getty Images;

415
Acknowledgements

Figures 5.4 and 5.6: Copyright © Science Source/Science Photo Library;

Figure 5.8: Copyright © Underwood & Underwood/Corbis; Figure 5.9:
Copyright © Steve Back/Daily Mail/Rex Features; Figure 6.1: Copyright
© Jerry Monkman/Narurepl.com; Figure 6.2: Copyright © Janine
Wiedel Photolibrary/Alamy; Figure 6.3: Copyright © Sally and Richard
Greenhill/Alamy; Figure 6.4: Copyright © Picture Partners/Alamy;
Figure 6.5: by permission of William Corsaro; Figures 6.6 and 6.7:
Copyright © John Birdsall/Press Association Images, photos posed by
models; Figure 6.8: Copyright © Image Source/Alamy; Figure 7.3:
Copyright © Mark Strozier/iStock; Figure 7.5: Damasio, H, Grabowski,
T et al (1994) ‘The return of Phineas Gage: clues about the brain from
the skull of a famous patient’, Science, Vol 264, No 5162; Figures 7.7
and 7.10: Dronkers, N F (2007) ‘Paul Broca’s historic cases: high
resolution MR imaging of the brains of Leborgne and Lelong’, Brain,
Vol 130, May 2007, Oxford University Press. Copyright © 2007 by the
Guarantors of Brain; Figures 8.1 and 8.9: Courtesy of Graham Edgar;
Figure 8.2: Copyright © Justin Lane/epa/Corbis; Figure 8.3: Copyright
© UK Medical Research Council used by kind permission; Figure 8.4:
adapted from Broadbent, D E (1958) Perception and Communication,
Pergamon Press; Figure 8.8: Kastner, S (1998) ‘Mechanisms of directed
attention in the human extrastriate cortex as revealed by functional
MRI’, Science, Vol 282, 2 October 1998. Reprinted with permission
from AAAS; Figure 9.2: Copyright © UK Medical Research Council
used by kind permission; Figure 9.5: Copyright © Jodi Hilton/Press
Association Images; Figure 9.6: Copyright © Dave Thompson/PA
Archive/Press Association Images; Figure 9.7: Courtesy of Nicola
Brace; Figure 9.9: Copyright © Bubbles Photolibrary/Alamy;
Figure 9.10: Copyright © David R Frazier Photolibrary Inc/Alamy.
Every effort has been made to contact copyright holders. If any have
been inadvertently overlooked the publishers will be pleased to make
the necessary arrangements at the first opportunity.

416
 Index

Index
A Loftus and Palmer on eyewitness memory 374–
81, 382, 403
accidents
and witness testimony 400
and driving 325–6, 357
Apter, Brian 5
and mobile phone use 329–30
Arendt, Hannah 64–5, 66
acquiescence response bias 39–41
attachment
addiction, and behaviourism 179, 182–3
different types of 194, 195
adolescent friendships 238, 239, 251
flexibility of 223–9
cultural influences on 258–61
see also infant attachment
peer influences on 240, 241, 256–7
attention 7, 277, 278, 279, 325–58, 403, 404
adoption, and human attachment 224
automatic and controlled 354–5
Adorno, Theodor 12, 24, 26–45, 48, 51, 139, 140,
Broadbent's theories of 332–9, 340–7, 350, 358
141
Cherry's work on 337, 339, 340–7
see also authoritarian personality
the ‘cocktail party problem’ 340
aggressive behaviour 11–13
lapses in 325–6, 327
and Bobo doll study 109–22
limited capacity of 333, 334, 339, 349–50, 351–2
different psychological approaches to study of
neurophysiology of 350–3
140
and real-world problems 356–7, 358
and social learning 13, 103–8, 123–8, 131–2,
sustained 327
134–5, 139
see also driving and attention; filtering
Agnew, Spiro 176
attitudes
agreeableness, and personality 21
measuring personality and 28–32
Ainsworth, Mary, 216–19, 220–1, 251, 258
and social learning 124–5, 126
see also Strange Situation studies
authoritarian aggression 48
Altemeyer, B. 47–50, 103
authoritarian personality 12, 23–53
Alton, David 129
Adorno et al. study of 26–7, 27–45, 48, 51, 139,
American Psychological Association, and Milgram's
140, 141
obedience studies 77, 82
and the causes of authoritarianism 33–6
Anderson, Craig A. 127
evaluation of 39–45
animal research
and the F-scale 31–3, 34, 39–41
and behaviourism 141, 146, 158–61, 162–70,
measuring personality 27–33
175, 179–80, 271
see also right-wing authoritarianism
ethics of 201, 204, 211–12, 410
authoritarian submission 48
and infant attachment 196–7, 198–9, 201–9,
autobiographical memory 367
211–12, 221–2, 223–4, 271–2, 273
automatic attention 354–5
anti-Semitism 26, 29, 31, 32, 46, 52
automatic processing 355
anti-social behaviour, peer influences on 241
avatars, and obedience research 86–7
anxiety, and psychosurgery 302
anxious-avoidant attachment 218, 219, 220 B
anxious-resistant attachment 218, 219, 220
aphasia 405 babies see infant attachment
and brain plasticity 311–12 baboons, and infant attachment 221
Broca's 296, 307 Bandura, Albert 109, 178
Wernicke's 298 Bartlett, Sir Frederic 279, 400
applied research 382–8 ‘War of the Ghosts’ story and memory 370–3,
ecological validity of 382–4 374–5
baseline conditions, in experimental procedures 113

417
Index

Bass, Bernard 40–1 and cognitive functions 404–5

Baumrind, Diana 79–80, 82 and the complexity of human behaviour 411
behaviourism 146, 158–86, 271, 272 cortex 294, 298, 300
and addictions 182–3 and emotions 302
and cognition 277 functional anatomy of 289, 290
and consequences of behaviour 153, 154–7, 158, and language 278, 285, 291, 292–9, 305, 307–10,
162, 169–70, 186 403
emergence of 161–2 left and right hemispheres 295, 306, 312, 315,
and environmental change 183–5 316
and human behaviour 172–7 localisation of function 295–6, 300–4, 315
and infant attachment 196, 197, 205 and memory 368
and instrumental conditioning 162–7 and the mind 288–9
and reinforcement 145–6, 167–9 motor cortex 300, 301, 305, 309, 311
and social learning 186 motor homunculus 300–1, 302
and stimulus–response psychology 162, 169–70 and neuropsychology 292, 312–13, 404
and therapeutic procedures 181 and phrenology 289–91
see also animal research; Skinner, B.F. plasticity of 304, 311–13, 404
beliefs, and social learning 124–5, 126 brain damage 285, 404
Bernstein, D.M. 396 ancient civilisations and understandings of 288–9
bias and dementia 288
and acquiescence response 39–41 individuals with, case studies 312–13
confirmatory 42 Mr Leborgne 294–6, 306–7
and double blind procedures 42 Mr Lelong 306–7, 310
Bigelow, Brian 146, 241, 242–7, 249, 250, 253–4, Phineas Gage 293, 295–6
259, 268, 272 lesions in the brain 306–7
‘Big Five’ personality factors 21 speech impairment 291, 293–6, 297, 298, 300, 301,
Billig, Michael 45 302–4, 305, 307, 309, 311, 312
biological psychology, and violent behaviour 140 see also stroke
birds and infant attachment, imprinting 198–9, 205, brain-imaging techniques 305–8, 404–5
223–4 functional brain imaging (fMRI) 307–8, 310,
Blass, T. 64 311, 350–3
Boase, M. 330 MRI scanning 306–8
Bober, Magdalena 105 brain surgery 305, 306
Bobo doll study 13, 109–22, 139, 140, 141, 251 Braverman, Morris 81
design of the experiment 112–14 British National Party 51
ethics of 121–2 British Psychological Society
findings 114–17 ‘Guidelines for Psychologists Working with
interpreting results of 118–22 Animals’ 211
Bond, Michael 84 Parliamentary Officer 129
Bonds-Raacke, Jennifer 265–6 and research ethics 82
Bosnia 11 Broadbent, Donald 278, 279, 332–9, 350, 358, 370,
Bowlby, John 194, 196, 197–8, 199, 213, 216, 223, 403
225–6, 271–2 see also attention
see also infant attachment; maternal deprivation Broca, Pierre Paul 278, 291, 292–6, 297, 298, 300,
hypothesis 403
the brain 278, 279, 283–317 Broca's aphasia 296, 307
the arcuate fasciculus 309 Broca's area 296, 300, 301, 305, 309
cells and neurons 285–6 and brain imaging 306–7, 310

418
 Index

and brain plasticity 312 cognitive psychology 332, 358, 404–5, 413
and spoken and sign languages 311 see also attention; memory
Brown, Gordon 130 cognitive science 413
Browning, Christopher 90–2 cognitive style 277
Brown, Ivan 341–2 and the authoritarian personality 47, 49
Buckingham, Hugh 297 collectivist cultures, adolescent friendships in 259–
Bukowski, William 239 61
Bull, Ray 399 communism, and the authoritarian personality 46–7
Burger, Jerry M. 87–8 computer games and social learning 103–4, 105–8,
Bushman, Brad J. 126, 127, 129 126, 130–2
Byron Review, Safer Children in a Digital World 130– computers
4 analogies between brains and 303, 310
and rehabilitation of stroke patients 316
C and memory
Cambridge University, Applied Psychology Unit comparing human and computer memory
(APU), and Broadbent's work on attention 335, 368–9
341 retrieval stage of 367–8
Canada study 241, 242–7, 249, 250, 261 see also internet; virtual worlds
case studies, of brain-damaged individuals 312–13 conditional stimulus 161
cats, behaviourist experiments on 159–61 conditioning 161–2
chat rooms, childrens' use of 105 classical 162
Cherry, Colin 337, 339, 340, 343–5, 346, 349, 350 operant 167, 178, 186
see also attention confirmatory bias 42
children confounding variables
as witnesses 399, 400 in experimental procedures 114
personality development in 23 in experimental research on driving and mobile
and the authoritarian personality 34–6, 49 phone use 342
and media violence 13, 103, 107–28, 129–35 conscientiousness, and personality 21
see also infant attachment; schoolchildren conservative attitudes, and the authoritarian
children's friendships 146, 237, 242–55 personality 29, 31, 32, 46
and age 238, 239, 243–4 consequentiality, and applied research 383, 384,
Bigelow and La Gaipa study of 241, 242–7, 249, 385, 386
250, 259, 268, 272 content analysis 243
and ethnographic research 253–4 continuous reinforcement 168
cultural influences on 258–61 control conditions, in Bobo doll study 113, 116–17
and gender 243, 244 control groups, research on false memories and food
peer influences in 240–1, 256–7, 268 preferences 393, 394
qualitative and quantitative data on 245–6 controlled attention 354–5
and social networks 239 controlled environment, and the Skinner box 165
see also adolescent friendships controlled nature of experiments 73–4, 112
classical conditioning 162, 178–9 controlled variables, experimental research on driving
clinical psychologists 140 and mobile phone use 341–2
‘cocktail party problem’ 340 correlations
coding in Bobo doll study 114–15 between mobile phone use and driving accidents
cognition 277–9 329–30
cognitive interviews 398 and the maternal deprivation hypothesis 226
cognitive mechanisms, in social learning 124–5, 277 in studies of media effects on violence 120, 127,
cognitive processes 7, 277, 278, 404–5 131–2

419
Index

Corsaro, William 250–4 dogmatism

Costa, Paul 21 and the authoritarian personality 47, 49
cost-benefits, and research ethics 78, 79–80, 121 and cognitive style 277
courtroom practices, questioning witnesses 398–400 see also Rokeach, Milton
Craik, Kenneth 335 dogs
crime see eyewitness memory; witnesses and behavioural research, Pavlov's dogs 158–9,
criminals, sentencing of and behaviourism 175–6 161, 162, 169
critical period, in human attachment 224 studies on the dog–human bond 221–2
Cuba 258–61 double blind procedures 42
cultural norms driving and attention
and the authoritarian personality 44, 52 automatic processing 355
reconstructing memory according to 373 effect of using a mobile phone 8, 277, 278, 329–
cultural variations, and obedience studies 84 30, 341–2, 357
cupboard love, and infant attachment 196–8, 206–7, and experience 355–6
215, 229 lapses of attention 325–6, 357
cycle of enquiry 403–4, 410 warning systems in cars 328
and hypothesis testing 341, 403 Dronkers, N.F. 307
dropped letter technique (Milgarm) 94
D drug addiction 179, 182
Damon, William 248–9, 250
Darwin, Charles 197
E
Dax, Gustave 297, 298 ecological validity
Dax, Marc 297 of applied research 382–4
debriefing, and Milgram's obedience studies 81 of eyewitness memory research 385, 386
demand characteristics, in eyewitness memory research and reconstruction of memories 373
385 of the Strange Situation study 220
dementia 288 education
dependent variables behaviourism and 146
in Bobo doll study 115 see also schoolchildren
in behaviourist research 165 educational psychology 412
in experimental research 341, 377 Egypt, ancient 288, 300
in infant attachment research, Harlow's monkeys Eichmann, Adolf 64–6, 90, 139
205 elicited behaviour 169, 170
depression Elms, Alan 89, 139
and psychosurgery 302 Emerson, Peggy 215–16, 226–7
and stroke patients 288, 317 emitted behaviour 170
and time spent online 3, 4–5 emoticons 265
desensitisation, and exposure to media violence 124– emotional detachment, and child-rearing practices
5, 130 212–13, 216
determinism 173–4, 176, 178 emotional stability, and personality type 21
developmental psychology 140, 272 encoding memories 367, 380
dichotic listening, and Broadbent's model of attention environmental change, and behaviourism 183–5
336–9, 403 environmental triggers, for human behaviour 178–9
direction of effect, in studies of media and violence epilepsy and the brain 289, 305
120 episodic memory 367, 368, 400
discrimination learning 179–80 Erwin, Phil 240–1
disinhibited attachment 228 ethics of psychological research 77, 410–11
disorganised attachment 218–19, 221 animal research 201, 204, 211–12, 410

420
 Index

and behaviourism 175–6 in Nazi Germany 12, 23–4, 26, 42–4, 51

Bobo doll study 121–2 see also F-Scale (fascism)
Burger's obedience research 88 FAST (Face-Arm-Speech Test) 287
and children's exposure to media violence 130 film, and media violence 105–7, 126
and eyewitness memory research 384 filtering of stimuli
and Harlow's research 201, 204, 211–12, 271 dichotic listening 339
Milgram's obedience studies 77–83, 141, 271 and limited processing resources 349–50
Nuremberg Code 78 and meaning 348–9
ethnocentrism 28, 29, 30, 31, 32 mixed and simultaneous messages 343–5, 346
ethnographic research, on children's friendships and neurophysiology 352–3
250–5, 267 in real-world situations 356–7
ethology, and infant attachment 196–7, 216, 271–2 Fisher, Ron 398
evidence 3, 409–10 fMRI (functional MRI) 308, 310, 312
and psychological theory 403–4 and attention 350–3
evolutionary psychology 272 food preferences, and false memories 393–6
and infant attachment 197–8 forensic psychology 140, 412
evolved predisposition 197 French, Doran 261
examinations, and different types of memory 367 Frenkel-Brunswik, Else 12, 24, 26–7
expectancy, and the Skinner box 180 frequency counts, Bigelow and La Gaipa study of
experimental social psychology 140 children's friendships 243, 245
experiments 4, 6, 272 Freud, Sigmund 12, 22–3, 196, 212, 216, 413
conditions in 110–13, 116–17 friendship 145, 237–68
controlled nature of 73–4, 112 and age 238, 239
hypothesis testing 340–2 cultural influences on 239, 258–61, 267
extinction, and reinforcement 168 determining boundaries of 239
extraversion, and personality type 21 new technologies and the changing nature of
eyewitness memory 278–9, 374–88, 400, 403 263–7, 268, 272
and applied experiments 382–3, 384 social influences of 256–7
and ecological validity 382 see also children's friendships
errors in 365–6, 369 F-Scale (fascism) and the authoritarian personality
ethics of research on 384 31–3, 34, 43, 46, 48
Loftus and Palmer study of 374–81, 384, 387 and acquiescence response bias 39–41
misinformation effect 380 and obedience research 64, 89
post-event information 380, 385 functional brain imaging 307–8, 310, 311
recall and recognition 367 fundamental attribution error 19–20
reconstructing 370
Eysenck, Hans 21 G
Gage, Phineas 293, 295–6
F Gall, Franz Joseph 289–91, 293, 302
Facebook 265 gambling, and partial reinforcement 183
false information, and leading questions 372 Geiselman, Ed 398
false memories 279, 385, 389–96 gender
and food preferences 393–6 and children's friendships 243, 244
and leading questions 397 and Milgram's obedience studies 74, 84
‘lost in the mall’ study 389–92 and social learning 109, 113, 114, 116, 117, 125
familiar strangers study (Milgram) 93–4 generalisation and social learning 118
fascism genocide 11, 51
and the authoritarian personality 51–3, 148 Geraerts, Elke, 393–6

421
Index

Goldberg, Lewis R. 21 and Bowlby 194, 196, 197–8, 199, 213, 216, 223,
Gopher, Daniel 355 271–2
Gore, Helen 5 critical period in 224
Gray, Jeffrey 348 cross-cultural studies of 219–20, 273
Greece, ancient 288 and cupboard love 196–8, 206–7, 215, 229
Gregory, Richard 303 ethics of research on 201, 204, 211–12, 271
Guiton, P. 223–4 Harlow's monkey studies 201–9, 211–13, 215
and imprinting 198–9, 205, 223–4
H and innate behaviour patterns 223–4
Hagell, Ann 127–8 longterm consequences of 227–9
happiness, and time spent online 3, 4–5 maternal deprivation hypothesis 225–7
Harlow, Harry 146, 201–9, 211–12, 213, 215, 223, multiple attachments 193–4, 226–7
271, 272, 410 primary attachment figures 193, 215, 225–7, 271
see also attachment research on humans 215–22
Harris, Eric 106 and sensitive responsiveness 215–16
Hartup, William 239 Strange Situation studies 216–22, 251, 258
Hess, Eckhard 199 information processing, and the brain 310
Hippocrates 288 information-processing model of attention 333–6,
Hitler, Adolf 12, 43–4, 52 337, 339, 405, 413
Hodges, Jill 224 informed consent 78, 79, 121, 385
Hofling, Charles 85–6, 92 innate responses, and infant attachment 197
Huesmann, Rowell 126, 127, 129 instrumental conditioning 162–7
Hulse, Lynn 399–400 intellect, and personality 21
human behaviour internet
and behaviourism 172–7, 178–9, 181, 182–5 Byron Review on the protection of children 130,
complexity of 139–41, 411–12 132–4
consequences and change in 153, 154–7 children's use of 104–5
influences on 145–6, 153–7 depression and surfing the web 3, 4–5
peer influences on 240–1 impact on social development 410
role of personality in 19–20 online friendships 263–7, 268, 272
Huntingdon's disease 289 interviews 6, 262
hypothesis testing 340–2, 403, 410 authoritarian personality study (Adorno et al.)
34–5, 37, 42, 141
I cognitive interviews 398
police interviewing of witnesses 397–8
imitative aggression, in Bobo doll study 115, 117
researching children's friendships 248–9, 254
imprinting 198–9, 205, 223–4
independent variables J
in Bobo doll study 112–13, 114, 121
in behaviourist research 165 James, William, 327, 329, 333
in experimental research 342, 377 Jews
in infant attachment research 205 Nazi atrocities toward 64–6
individual differences, approach to personality 20–1, and Milgram's obedience research 63–4, 90–2,
140 139
individualistic cultures, adolescent friendships in see also anti-Semitism
259–61
Indonesia 261
K
infant attachment 8, 146, 193–229, 271–2, 273 Kahnemann, Daniel 349–50, 351–2
Kastner, Sabine 350–1

422
 Index

Kinnock, Neil 20 hypothesis 225–7

Klebold, Dylan 106 and monkeys 208–9
media representations of violence
L reducing the impact of 129–35
La Gaipa, John 146, 241, 242–7, 249, 250, 253–4, and social learning 13, 103–8, 123–8
259, 268, 272 see also Bobo doll study
language 7, 277, 403 melodic intonation therapy (MIT) 316–17
and the brain 278, 285, 291, 292–9, 403 Memon, Amina 399–400
and brain surgery 305 memory 7, 277, 278–9, 365–400, 403
and functional imaging techniques 307–8 and attention 338–9, 354
and interactions between brain regions 308–10 errors in recall 373
and sign language 285, 311 fallibility of 400
and stroke 286, 287, 288 false reporting 385
leading questions 372, 379–80 malleability of 400
in the courtroom 398–9 questioning 397–400
and false memory 397 reconstruction 369–73
learning repressed or recovered memories 387–8
discrimination learning 179–80 three stages of 367–8, 380
and human behaviour 156–7, 178–9 types of 366–8
through reinforcement 146 see also eyewitness memory; false memories
see also social learning mental hospitals, changing behaviour of patients
Leblanc, Warren 106 173, 181
Lederer, Gerda 44 mental illness
Levinson, Daniel 12, 24, 26, 27 and the brain 289, 291
limited capacity channel 334 psychosurgery for 302–3
limited capacity model of attention 333, 334, 339, meta-analysis 126, 127, 129
349–50 Milgram, Stanley 47, 62–83, 413
Little Albert, and behaviourist research 159, 161 see also obedience research
Livingstone, Sonia 105 misinformation effect, and eyewitness memory 380
localisation of brain function 300–4, 303–4 Mixon, Don 84–5
limits to 302–3 mnemonic techniques, and police interviewers 398
principle of 295–6 mobile phone use
and stroke 315 and friendships 263
support for 300–1 while driving 8, 277, 278, 329–30, 341–2, 357
Loftus, Elizabeth 279, 387–8, 396, 400 monkeys
longitudinal studies, of media effects on violence and infant attachment research 201–9, 211–13
120 monkey maternal deprivation and abuse 208–9
Lorenz, Konrad 198–9 reasons for using 203–5
love see infant attachment mothers see infant attachment
Lueret, François 292 motor homunculus, in the brain 300–1
MRI (magnetic resonance imaging) 306–8
M fMRI (functional MRI) 308, 310, 312, 350–3
multitasking
McCrae, Robert 21
and driving 329–30, 341–2
McLeod, Kim 256–7
and the limited capacity model of attention 334
Mandel, David 91, 92
music, and stroke patients 31–17
Marshall, James 375, 377
Mason, William 212
maternal deprivation

423
Index

N right to withdraw 78, 80, 121, 122

Pavlov, Ivan 158–9, 161–2, 169, 271
natural selection, and infant attachment 197
perception 277
nature–nurture debate 273, 411
personality
Nazi Germany
approaches to 19, 20–3, 140
and the authoritarian personality 12, 23–4, 26,
development of differences in 12
42–4, 51, 52, 139
and the fundamental attribution error 19–20
neo-Nazi groups 51, 53
influences on aggressive and violent behaviour 8,
and obedience research 63–6, 77–8, 80, 90–2
12
negative reinforcement 167–8
measuring 12, 27–33, 34, 37
Neumann, Ohmar 355
and obedience studies 89–90, 139
neurons and brain damage 285–6, 312, 313
and social learning 126
and language comprehension and production 309
see also authoritarian personality
neuropsychology 292, 404, 411, 412, 413
personality theory 21
case studies in 312–13
Pettigrew, Thomas 43, 44, 52
Newburn, Tim 127–8
phobias, changing behaviour 155–6, 181
Nixon, Richard 176
phrenology 289–91
Nuremberg Code, and Milgram's obedience studies
Pickrell, Jacqueline 389–92, 393
78, 80
pigeons, behaviourist experiments on 162, 164
nurses, and obedience research 85–6, 92
police, interviewing of witnesses 397–8
O positive reinforcement 167, 173, 174, 175–6
and addictions 182–3
obedience research 12–13, 61–95, 139, 140, 141 and environmental change 184, 185
Milgram's experiments 12–13, 62–83, 89–95, 139 post-event information, and eyewitness memory 380,
and modern warfare 410 385
and the presence of others 61–2 practicality, and applied research 383–4
replications of 84–6, 87–8 primary attachment 193
in virtual worlds 86–7 prisons, behavioural reinforcement in 146, 155, 173,
observation 6, 13, 141, 272 178
Strange Situation studies 216–22 procedural memory 366
see also Bobo doll study prospective memory 366
occupational psychology 412 psychoanalytic theory
open-ended questions, and police interviewing 397– and the authoritarian personality 34–6, 44, 45,
8, 399 46–7, 49
operant conditioning 167, 178, 186 and infant attachment 196
oral stage, in Freudian psychoanalysis 23 and personality 12, 223
orphanages, adoption of children from 227–9 psychology
collaboration with other disciplines 411–12
P sub-disciplines of 411–12
Pakeerah, Stefan 106–7 psychosurgery 302–3
Palmer, John 374–81, 384, 385–6, 387, 403 punishment 145–6, 168–9, 170
partial imitative responses, in Bobo doll study 115, and human behaviour 172, 173, 277
117
partial reinforcement 168, 272 Q
and gambling 183 qualitative data 37
participants in research studies 73 ethnography 250–5
and informed consent 78, 79, 121, 385 interviews 248–9, 254
protecting the welfare of 78, 79–80, 121, 122 written accounts 146, 245–6, 254

424
 Index

quantitative data 37, 146 scale, measures of personality

frequency counts 243, 245 and the authoritarian personality 28–33, 34, 37
questionnaires 5, 6 right-wing authoritarian scale 48–9
measuring personality and attitudes 29–30, 31–2, see also F-scale (fascism)
34, 141 Schaffer, Rudolph 215–16, 226–7
and acquiescence response bias 39–41 Schermerhorn, Richard 53
questions and witness memory 397–400 schoolchildren
see also leading questions behavioural reinforcement in 155, 172–3, 178
improvement in behaviour 3, 5–6
R obedience to authority 62
Raacke, John 265–6 see also education
racial discrimination, and the authoritarian personality scripts 125
43 Second World War Nazi atrocities
rats, behaviourist experiments on 162–7, 168–9 and the authoritarian personality 11, 23–4, 26,
reasoning 277 42–4, 51, 139
recognition recall 367 and obedience research 63–6, 77–8, 80, 90–2
recovered memories 387–8 secure attachment 218, 221, 222, 227
false memory research 389–96 semantic memory 366, 387
Redelmeier, Donald 329 sensitive responsiveness, and infant attachment
reflex 158–9 215–16
reinforcement 145–6, 167–9, 170 sentencing of criminals, and behaviourism 175–6
and human behaviour 172–7, 178, 181, 271 sexual violence, and mechanisms of social learning
and addictions 182, 183 125
and cognition 277 Shiffrin, Richard 354–5
and environmental change 184, 185 sign language 285, 311
and infant attachment 196 situational influences
and the Skinner box 180 on aggressive behaviour 8, 12 119–20
replication of obedience studies 84–6, 87–8 on human behaviour 50, 89–92, 103, 139
repressed memories 387–8 six degrees of separation study (Milgram) 94
retrieval, stage of memory 367–8, 380 Skinner, B.F. 145–6, 157, 158–71, 172, 173–7, 178,
right-wing authoritarianism, 183–4
Altemeyer's study of 47–50, 103 see also behaviourism
and authoritarian movements 53 Skinner box 146, 163–7, 271
and communism 46–7 and discrimination learning 179–80
and Nazi atrocities 11, 23–4, 26 and environmental change 184, 185
and obedience research 62, 63, 89 Slater, Mel 86–7, 88
and ordinary people 53 Smith, Brewster 26–7
Rokeach, Milton 46–7, 48, 49 Smith, Peter 84
role-play, and obedience research 84–5 smoking behaviour
Rosenthal, Robert 42 peer influences on 256–7
Ross, Dorothea 109 quitting smoking 154, 155, 182
Ross, Lee 19 social learning 13, 103–35, 139, 141, 153
Ross, Sheila 109 and behaviourism 186
Rutter, Michael 226, 227–8 mechanisms behind 123–8
Rwanda 11 attitudes and beliefs 124–5
cognitive mechanisms 124–5, 277
S and media representations of violence 107–8
multiple influences 126–7
Sanford, Nevitt 12, 24, 26, 27

425
Index

priming 123–4 and hypothesis testing 340–1

scripts 125 psychological theory and evidence 403–4
and right-wing authoritarianism 48–9 real-world relevance of 404
see also Bobo doll study Thompson, Robert 106
social networking websites 265–6 Thorndike, Edward 159–61, 162, 271
South Africa 43 Tibshirani, Robert 329
South Korea 261 Tinbergen, Niko 196–7
Soviet Union 46–7 Tizard, Barbara 224
speech 285 top-down processes, of attention 356
and brain development 310–11 twin studies, of smoking behaviour in adolescence
interactions between brain regions involved in 257
308–10
and speech impairment 293–9, 300, 301 U
retraining of stroke patients with loss of 315–17 unconditional stimulus 161, 196
speech therapy 311 unconscious aspect of personality 22–3
sports psychology 412 and the authoritarian personality 36, 44
stimulus United Nations
conditional 161 Convention on the Rights of the Child (1989)
and reflex 158–9 193
unconditional 161 Declaration of the Rights of the Child (1959)
stimulus–response link 160–1 193, 213
stimulus–response psychology 162, 169–70 United States 43, 106, 176, 217, 219, 261, 265–6
storage stage in memory 367
Strange Situation studies 216–22, 251, 258 V
attachment categories 217–19, 220–1
violence
cross-cultural comparisons 219–20, 273
explaining human 11–13
on non-human animals 221–2
and far-right fascist movements 51
Strayer, David 329–30
violent offenders and media violence 127–8
stroke 285–8
see also aggressive behaviour; media representations
and arm weakness 287
of violence
and Broca's area 311
virtual worlds
causes of 285–6
interactive gaming 103–4
FAST (Face-Arm-Speech Test) 287
obedience research in 86–7
holistics approaches to treatment of 317
vulnerable witnesses, questioning of 399–400
and language disturbances 286, 287, 288
psychologists and the recovery process 287–8, W
315–17
sub-disciplines of psychology 411–12 Wada technique, for studying the brain 306
surgery Walker, Matthew 4
brain surgery 305 warfare, rise of modern 410
psychosurgery 302–3 war games, attention and `friendly fire' incidents
356–7
T ‘War of the Ghosts’ story 370–3, 374–5
Watson, John B. 159, 161, 178, 212–13, 216, 271
technological development
See also behaviourism
challenges to psychology of 410
Wedderburn, Alexander 348
and ethically responsible research 410–11
Wernicke, Carl 278, 291, 292, 300, 403
textual data 254–5
Wernicke's aphasia 298
theories
Wernicke's area 296, 298, 305, 308, 309–10

426
 Index

and spoken and sign languages 311

Westcott, Helen 399
witnesses
courtroom practices 398–400
police interviewing of 397–8
see also eyewitness memory

Y
young offenders
and the maternal deprivation hypothesis 225–6
and media portrayal of violence 13, 127–8

Z
Zimmermann, Peter 227

427