The Evolution of Organization Analysis in ASQ, 1959-1979

Author(s): Richard L. Daft

Source: Administrative Science Quarterly, Vol. 25, No. 4 (Dec., 1980), pp. 623-636
Published by: Sage Publications, Inc. on behalf of the Johnson Graduate School of
Management, Cornell University
Stable URL: https://www.jstor.org/stable/2392284
Accessed: 30-05-2019 07:41 UTC

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/2392284?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Sage Publications, Inc., Johnson Graduate School of Management, Cornell University

are collaborating with JSTOR to digitize, preserve and extend access to Administrative
Science Quarterly

This content downloaded from 123.63.98.246 on Thu, 30 May 2019 07:41:55 UTC
All use subject to https://about.jstor.org/terms
The Evolution of Or- Articles in the Administrative Science Quarterly are
analyzed by (1) complexity of organizational models and
ganization Analysis in
(2) the variety of language used to transmit observations
ASO, 1959-1979 about organizations. A sharp trend toward low-variety
statistical languages has taken place, which may repre-
Richard L. Daft sent an organizational mapping phase in which simple,
quantifiable realtionships have been formally defined and
measured. A broader scope of research languages will be
needed in the future to enable researchers to analyze
organizational models of greater complexity.

Where do we stand as a field of inquiry? What are the

theoretical bases of organizational research? What tech-
niques should be used, what direction taken, to achieve sig-
nificant new knowledge? Professional meetings and informal
gatherings of organization researchers often reflect uncer-
tainty and discontent about the field. Arguments such as
qualitative versus quantitative research methods arise regu-
larly, and talk about a radical approach to some topic or
other is frequent. Perhaps, as an academic discipline, or-
ganization studies is simply in a pre-science stage of devel-
opment. Whatever the explanation, uncertainty often seems
more prevalent than either pride in past accomplishments or
anticipation about future new knowledge.

One way to gain perspective on the evolving field is to

examine publications on organizations over a period of time
to see where we are and how we arrived here. My purpose
here is to develop and report a framework for analyzing
changes in the content of articles in the Administrative Sci-
ence Quarterly over the past twenty years. ASQ seems an
appropriate focal point because of its position in the fore-
front of organizational studies and because most articles
published in it are devoted to organizations.

THEORETICAL FRAMEWORK

The framework used to organize ASQ articles is adapted

from ideas reported by Pondy and Mitroff (1978) and Daft
and Wiginton (1978). The analysis was based on (1) the re-
search "language" used in the articles and (2) the assump-
tions about the complexity of the organizational system as
revealed in the model used.

Research Languages

Language was used in the broadest sense to encompass

the many types of language used in research. These differ
in their ability to transmit information (Daft and Wiginton,
1978; Pondy and Mitroff, 1978). If a language has a large
pool of symbols, which are not very restrictive and which
can communicate a wide range of ideas, then the language
would have high variety and would be considered quite
complex. If a language has few symbols, however, which
are restrictive and which can communicate a narrow range
of ideas, then the language would be low in variety and
would be considered simple. Low-variety languages often
convey exact, unequivocal meaning to users, and symbols
are considered quite precise; high-variety languages often
? 1980 by Cornell University.
0001 -8392/80/2504-0623$00.75 have a range of meanings and may be perceived as some-

623/ Administrative Science Quarterly

This content downloaded from 123.63.98.246 on Thu, 30 May 2019 07:41:55 UTC
All use subject to https://about.jstor.org/terms
 High variety

6 Nonverbal, imaginative -art, music, painting

5 Nonverbal, realistic - photographs, lifelike illustrations

4 General verbal expression - open-ended interviews, reports of

observations

3 Categorization - frequencies, percentages, cross-tabulations

2 Linear statistics - correlations, regression, general linear models

1 Analytical mathematics - calculus, dynamic equilibrium models

Low variety

Figure 1. Continuum of language variety.

what ambiguous. Figure 1, which lists languages according

to the degree of variety, tends to be oversimplified, but
shows the general pattern. Movement up the scale from
analytical mathematics to linear statistics to simple
categories, for example, represents decreasing precision in
the information conveyed. Language thus may represent re-
search method to some extent, as when statistical analyses
report findings from a survey or when general verbal ex-
pression describes case-study results.

The reason the research language used is important to

analyze is that it determines the kind of information that can
be transmitted. The language of mathematics and statistics
can transmit information about countable variables and rela-
tionships; the language of art and music can transmit a
broad scope of ideas and emotions which cannot be
counted, or even verbalized. Consider, for example, using
different languages to analyze Shakespearean plays. One
might choose a statistical or mathematical language to count
the frequency of certain words, or the length of clauses, or
perhaps the construction of ratios relating one part of
speech to another. Such a quantitative language provides a
very narrow insight into the plays. Another person might rely
on verbal discussion, analyzing specific passages, quoting
liberally from the plays, describing what they mean, and
providing different interpretations. Finally, one might be
moved to write a poem or a symphony to share one's feel-
ings or to paint a picture to communicate an image. Each
language reveals different aspects of Shakespeare's plays.
Thus the pattern of language selection used in ASQ may
reveal something about which languages have been domi-
nant in organization research, and hence the kind of things
we have been telling one another about organizations, or -
perhaps even more important - the kind of things we have
not been telling one another.

Research Models

A conceptual model is only a crude, inadequate description

of organizations, but the choice of variables and expected
624/ASQ

This content downloaded from 123.63.98.246 on Thu, 30 May 2019 07:41:55 UTC
All use subject to https://about.jstor.org/terms
 Evolution of Organization Analysis

relationships in the model reveal the investigator's assump-

tions (often implicit) about organizations. An investigator may
view the organization as a geographical territory, with static
elements to be located and charted and mapped, or as a
flowing, dynamic, unpredictable cluster of human interac-
tions that have to be experienced first hand. The pattern of
organization models reported in ASQ may indicate the view
of organizations held in the past, how that view has
changed, and perhaps the view of organization that may
become important in the future.

Boulding's (1956) scale (Figure 2) is used to analyze system

complexity (Pondy and Mitroff, 1978). The advantage of
Boulding's scale is that it captures and describes qualitative
changes in system complexity. Simply adding more variables
to an organization study would make it more complex in
number of relationships, but as long as the basic view of
organization is the same, the study will not move up Bould-
ing's scale. Each increment on the scale requires a substan-
tive change in models, such as moving from static to
dynamic relationships, or from nonhuman to human charac-
teristics.

METHOD

ASQ spans 24 years. Articles included in the first two to

three years of a journal might not reflect typical submission
procedures or selection practices, and so the first three
years were omitted, leaving a twenty-year span from 1959
Complex

8 Social organizations

7 Symbol-processing systems

6 Differentiated systems

5 Growth systems

4 Open systems

3 Control systems

2 Clockworks

1 Frameworks

Simple

Figure 2. Boulding's scale of system complexity.

625/ASQ

This content downloaded from 123.63.98.246 on Thu, 30 May 2019 07:41:55 UTC
All use subject to https://about.jstor.org/terms
 to 1 979 for analysis. Since changes in research languages
and models could be expected to occur slowly, the articles
in every fifth year were chosen for analysis. A breakdown of
these articles by type and year appears in the Table. Not all
papers from these years were included in the analysis.
Boulding's scale and the research language scale apply to
theoretical models of an empirical world. Literature reviews
(Pondy, 1969b), papers on methodology (Pettigrew, 1979;
Salancik, 1979), critiques of other papers (Mayhew and Gray,
1969; Weick, 1969), responses (Walter, 1969), and news
and notes for information (Fromkin, 1969; Hayes, 1969) do
not convey new knowledge about organizations and do not
fit into the language or system scales. The pool of papers
for analysis thus consisted of the 119 empirical and theoret-
ical papers represented in the first two rows of the Table. Of
these papers, an additional 15 were omitted because they
were not congruent with the scale dimensions. These in-
cluded surveys of strictly individual phenomena not tied to
an organizational context (Glaser, 1959; Alderfer, Kaplan, and
Smith, 1974; Tichy, 1974) or topics not concerned with or-
ganizational dimensions (Guetzkow, 1959).

Table

Type of ASQ Article by Year

Type of
Article 1959 1964 1969 1974 1979

Empirical 12 13 31 25 20
Theory 6 5 3 4 0
Method 0 0 0 1 14*
Other 1 1 l0t 1 1

Total 1 9 1 9 44 31 35

*13 of these in the December 1979 special issue on qualitative methods.

tIncluded criticisms and responses, literature reviews, and 4 descriptions of
psychological laboratories that appeared as part of a special issue on labora-
tory research.

Procedures

Research model. Each article was read to determine the

type of research question being asked. Most articles fell
clearly into one of Boulding's scale levels; however, the
boundaries of the system categories were not precise, and
the assumptions underlying an article were not always obvi-
ous. Thorough reading was required, for system level was
sometimes different from what the title or abstract might
suggest. As an illustration, Pfeffer and Salancik (1974) inves-
tigated university decision making. Decision and political pro-
cesses fall into Boulding's scale level 6 (differentiated sys-
tem), but the dynamics of coalition formation or other inter-
nal decision processes typical of level 6 were not assessed.
The study did explore power differences, and it operational-
ized measures of influence that were related to other
characteristics, such as percentage of budget allocated to
departments, and so it was assigned to level 5.

Boulding's scale (Figure 2) was originally designed to em-

brace all types of systems in all of the sciences, but he did
626/ASQ

This content downloaded from 123.63.98.246 on Thu, 30 May 2019 07:41:55 UTC
All use subject to https://about.jstor.org/terms
 Evolution of Organization Analysis

not consider many possible models in any single field, such

as organization studies. The scale levels and organization
analogs for each level were as follows.

Scale 1. Frameworks. The static, structural properties of

the system are considered, as in a map, including physical
dimensions, and arrangement and location of elements. This
level includes snapshot studies of organizations including,
studies of organization charts, construction of typologies,
and analyses of size and administrative ratio. The basic as-
sumption is a closed system, with no analysis of output,
goals, human properties, or internal information and decision
processes. Examples include Pondy's (1969a) analysis of
administrative intensity, the analysis (Pugh et al., 1969) and
taxonomy (Pugh, Hickson, and Hinings, 1969) reported by
the Aston group, Ouchi and Dowling's (1974) definition of
span of control, Murphy's (1969) examination of geograph-
ical distribution of federal funds, and Beyer and Trice's
(1979) reanalysis of size and complexity.

Scale 2. Clockworks. This is basically the same as the

framework, except that the system has dynamic properties.
The purest example of organizational clockworks are the
structural control models of organizational change-that ap-
peared in the American Sociological Review (Hummon,
Doreian, and Teuter, 1975; Daft and MacMillan, 1977).
Clockwork models of organization in ASQ include models of
personnel flows into and through organization systems
(White, 1969; Rosenbaum, 1979) and studies of turnover
patterns (Smith, 1979).
Scale 3. Control systems. This is essentially a closed sys-
tem, which can adapt to an externally provided target. Sim-
ple flow of information, and interaction between controller
and operator in reaction to system target, distinguishes this
system level from lower system levels. Examples are simple
studies of communication frequency and direction (Simpson,
1959), of participant responses to leader style (Lowin, Hrap-
chak, and Kavanagh, 1 969) or reward structure (Evan and
Simons, 1969), and the response or compliance of organiza-
tion participants to organizational conditions (Dewar and
Werbel, 1979; Oldham and Brass, 1979).

Scale 4. Open systems. This is characterized by throughput

(input, transformation, output). Systems at this level have
the property of self-maintenance, have the goal of survival,
and can adapt to the environment to maintain their exis-
tence. Organization models at this scale would include out-
put and performance relationships, such as input-output
analyses, production functions, and tests of efficiency and
effectiveness. Examples include studies of performance ef-
ficiency (Becker and Baloff, 1969; Carzo and Yanouzas,
1969) and system effectiveness (Osborn and Hunt, 1974;
Turcotte, 1974; Coulter, 1979), and studies that examine in-
trusion of the external environment into the organization in
simple ways (Smith et al., 1969; Franko, 1974; Sutton,
1974).

Scale 5. Growth systems. At this scale level, differentiation

appears, with different subparts performing interdependent
functions to achieve system goals. The form of the system
can change over time, growing, evolving, and adapting to

627/ASQ

This content downloaded from 123.63.98.246 on Thu, 30 May 2019 07:41:55 UTC
All use subject to https://about.jstor.org/terms
 the environment. The system has purpose and direction.
Key notions are differences and interplay among subunits,
changes in the organization over time, organizational incep-
tion and growth, and goal formation. Examples are the ex-
planation of organizational configuration by examining the
history of the organization and/or the external environment
(Janowitz, 1959; Rosengren, 1964; Stern, 1979), the process
of innovation (Lynton, 1 969), and internal differences that
develop over time in both horizontal and vertical directions
and the effects on the organization (Walton, Dutton, and
Cafferty, 1969; Hinings et al., 1974; Pfeffer and Salancik,
1974).

Scale 6. Differentiated systems. This is highly elaborated

with specialized subparts that have refined information re-
ceptors and processors. Subparts import information about
the environment which the system uses to adapt to the
environment. A nervous system of sorts exists, in which
diverse information is stored and organized. Multiple goals
can be pursued and the organization has a choice of re-
sponses. Studies of information utilization, coalition forma-
tion, decision processes within the organization, processes
of role and goal conflict, and coordination processes would
typically be included. Examples are the studies of role and
goal conflict dynamics by Cressey (1959) and Grusky (1959),
Keegan's (1974) study of multinational scanning, Driver and
Streufert's (1969) models of information-processing sys-
tems, and Alexander's (1 979) study of decision-making
processes.

Scale 7. Symbolprocessing systems. Here elements

within the system have the ability to communicate, to pro-
duce and observe symbols, to deal with abstract concepts,
and to be aware of time, history, and themselves. This sys-
tem has characteristics which are uniquely human and
which would not be found in lower-level systems. Organiza-
tion studies at this level would investigate complex aspects
of humans as elements in a larger organization system -
the individual's concept of organization (cause maps), politi-
cal processes, affective dimensions of organizations, the use
of language, information content, and mechanisms of
shared meaning. Only a few organizational studies explore
research questions at this level. Examples are participants'
cause maps of organizations (Bougon, Weick, and Binkhorst,
1977), friendship ties (Lincoln and Miller, 1979), and the rela-
tionship between emotional needs and the organization (Ar-
gyris, 1959).

Scale 8. Social organizations. This scale level pertains to

human beings in groups and to the unique qualities that
emerge only in group settings. Value systems, religion, mor-
als, art and music, shared symbols, and other collective man-
ifestations of complex human emotion appear here. Organi-
zation studies of the language, symbolism, shared mean-
ings, and cultural aspects of organization would be included
in this category. This system level represents the most
complex system, but no research at this level has been
reported in ASQ.

Research language. Each article was studied to determine

which language was the primary vehicle for conveying data

628/ASQ

This content downloaded from 123.63.98.246 on Thu, 30 May 2019 07:41:55 UTC
All use subject to https://about.jstor.org/terms
 Evolution of Organization Analysis

and observations. Most articles included general verbal ex-

pression along with other, often statistical, languages. For
this classification the primary language used to summarize
and transmit the author's data and findings was used in the
analysis. All studies contained verbal language, but if tables
of correlation coefficients were the primary means for
summarizing and communicating observations, then the arti-
cle was considered level 2 - linear statistics. In a few arti-
cles, two language forms were used almost equally, so the
article was put midway between those two points on the
scale. Pondy and Birnberg's (1969) paper on resource alloca-
tion in task groups, for example, used response categories
(level 3) and direct quotations from respondents (level 4) in
almost equal amounts to illustrate the research findings, so
it was rated 3.5 on the language scale.
Examples of articles and their primary research language
that appeared in sample years of ASQ are as follows.

Level 1. Analytical mathematics. Pondy's (1 969a) calculus

models of administrative intensity and White's (1969) model
of personnel flows.

Level 2. Linear statistics. Hage and Aiken's (1969) correla-

tion study of organization structure, Salancik and Pfeffer's
(1974) correlation and regression analysis of power, and
Lynch's (1 974) factor analysis of technology dimensions.
Level 3. Simple categories. Cafferata's (1979) use of per-
centages to assess satisfaction with a professional associa-
tion, and Simpson's (1959) frequency tabulations of com-
munication episodes.

Level 4. General verbal expression. G rusky's (1 959) report

of observed role conflict in a prison and Stern's (1979)
analysis of changes in the NCAA.

Levels 5 and 6. Nonverbal. No examples of research

studies using these languages appeared during sample
years. However, during 1969 a special issue describing
psychology laboratories used photographs (level 5), and re-
productions of paintings and cartoons (level 6) were inserted
between articles.

RESULTS

Each article from the five sample years and the average for
each year are reported in Figure 3. Publications in 1959 tend
to be located above and to the right of publications in other
years. Investigators were exploring complex dimensions of
organizations, such as the effects of cultural differences on
administrative processes, or the pressures and interactions
arising from conflicting constituencies in a prison system,
and they often used case descriptions to analyze and report
their observations.

From 1959 through 1964 to 1969, the average ASQ publica-

tion showed a definite shift toward a more simplified view
of organizations and toward the use of low-variety lan-
guages. Comparative studies of organization structure (scale
level 1) were reported, as were several studies on the ef-
fects of structure or leader behavior on participants (scale
level 3), and systematic categorization and linear statistics
replaced case-type discussions. This period represents a

629/ASQ

This content downloaded from 123.63.98.246 on Thu, 30 May 2019 07:41:55 UTC
All use subject to https://about.jstor.org/terms
 High 1959-*
1964 - A
6.. 1969--
1974 - 0

1979-A
Average -
5

*OA *OOA

4 +00 -* AmIL> ELL

1959

X) - AA~~~A
> 1964
mu 60MM *AMMO
:, 3 * mm. LO AO OG.,
* 1974
1969 A *m
1979
Auuuuuo AAE0OOO moo
2 OOOOA OGA

1 . .

Low l
Simple Complexity of organization models Complex

Figure 3. ASQ publications by variety of language and complexity of

organization models.

dramatic change; organizational researchers communicating

to one another through ASQ literally were speaking a new
language in 1969 compared to 1959 and were sharing new
views of organization. These were views of structure rather
than process, of systematic laboratory and survey investiga-
tions rather than detailed case analyses. The research values
which emerged during this period, and which pointed to-
ward future work, were reflected in the appearance of the
first questionnaire-correlation study of organization structure
(Hall, 1962), Blau's (1965) call for comparative organization
research to replace case studies, and ASQ 's June 1969
issue devoted to laboratory studies of organization.
The period from 1969 to 1979 showed a continued trend
toward low-variety, statistical languages (downward direction
in Figure 3), although the shift was much less pronounced
than from 1959 to 1969. By 1979, correlation, multiple-
regression, and factor-analysis techniques were commonly
used; linear statistics was by far the most frequently used
research language. This movement toward low-variety re-
search languages was also reflected in other journals that
published organization studies (Fennell, 1975).
Organization models reversed direction after 1969 toward
greater complexity. Researchers were reaching out to an-
swer more difficult questions. Longitudinal data and
operationalization of a range of nonstructural elements were
more frequently published in 1974 and 1979 than in 1969.

630/ASQ

This content downloaded from 123.63.98.246 on Thu, 30 May 2019 07:41:55 UTC
All use subject to https://about.jstor.org/terms
 Evolution of Organization Analysis

Linear statistics as a research tool was probably becoming

better understood and was being extended to analyze new
variables and more difficult research problems.

An additional pattern in Figure 3 is the moderate positive

slope for all scatter points. The band of points in Figure 3 is
fairly broad, but it reflects a positive relationship between
language variety and model complexity. Moreover, no article
appeared in the upper left-hand or lower right-hand corners
of the graph.

DISCUSSION

Looking back, the first few issues of ASQ contained high

hopes and optimistic expectations for a new science of ad-
ministration. The early articles were almost all qualitative, but
the notion of a science gradually developed. By 1969 most
articles were systematic analyses of organizations, and by
1979 most studies used linear statistics of some form. The
movement toward linear statistics may represent an impor-
tant phase in organization theory. A mapping phase may
have been necessary in this period to define and measure
formally the characteristics of organizations. Statistical tech-
niques were learned and utilized that made possible new,
comparative views of organizations.

The changes in ASQ from 1959 to 1979 probably reflect

discipline-wide changes taking place in research methods
and views of organization. A possible alternate explanation
might be ASQ editorial policy, implicit or explicit. For exam-
ple, new editors took over before 1969 and 1979, and their
views might have influenced the type of articles published.
Furthermore, editorial board members, though at the fore-
front of a discipline, may not reflect the diversity within the
discipline. However, editors and board members do not con-
trol the type of papers submitted for publication, and their
goal is to publish articles that report new knowledge. Their
perception about how to achieve that goal is probably con-
gruent with recent thinking in the discipline. Thus, the
changes in Figure 3 probably generalize beyond the confines
of ASQ to reflect at least partly the changes taking place
within the discipline.

The other pattern in Figure 3 is the moderate positive rela-

tionship between research models and languages. Do simple
languages cause simple models, or vice versa, or are they
two independent variables? The languages used to describe
organizational phenomena may limit the type of organiza-
tional problem that can be investigated. Low-variety lan-
guages may be incompatible with the analysis of complex
organization problems (Zadah, 1973). For example, the popu-
lar questionnaire-correlation technique, which uses a lower-
variety language, may limit research projects to relatively
tangible, definable characteristics of organization. Some
causal influence from research models to research language
may also explain the positive slope in Figure 3. The desire to
compare formal characteristics across organizations would
move investigators toward low-variety languages. Whatever
the causal direction, the relationship between language and
model may be important. Problems on Boulding's scale
levels 7 and 8 may require different languages than prob-
lems on scale levels 1 and 2.

63 1/ASQ

This content downloaded from 123.63.98.246 on Thu, 30 May 2019 07:41:55 UTC
All use subject to https://about.jstor.org/terms
 IMPLICATIONS

The purpose of the analysis reported here has been to de-

velop and present a framework that would reveal patterns
about the development of organization studies. This analysis
has not assessed the quality of publications, nor has it sur-
veyed the entire field, nor even all of ASQ. The findings
merely provide a glimpse of changes taking place over time
in organization theory as seen through sample articles from
ASQ.

The pattern taking place from 1959 to 1979 suggests two

related questions: What pattern might we expect in the fu-
ture, say over the next 10 years, and which languages and
models will help us best discover significant new knowl-
edge? Concerning the future pattern of publications, the
data provide several clues. A simple linear projection of the
recent past would suggest increased use of low-variety lan-
guage in the future, that is, continued movement toward
precision and rigor in scientific technique. On the other
hand, the recent trend toward more complex models might
also be expected to continue. If Figure 3 is perceived to be
similar to a table of elements in chemistry, then additional
studies are needed to explore higher-level properties of or-
ganizations on the right-hand side.

But these two trends are contradictory. Quantitative lan-

guage does not fit amorphous, intangible dimensions of or-
ganizations. One trend will have to be reversed, and the
most likely reversal will be the reliance on low-variety lan-
guages. These languages are well suited to quantifiable vari-
ables, linear relationships, and simple, closed-system views
of organization. If investigators continue to be exact, to
quantify, as they examine more complex aspects of organi-
zations, they may tend to oversimplify. A few creative, imag-
inative investigators undoubtedly will find interesting ways to
capture and summarize new aspects of organization with
statistical techniques, but exclusive reliance on these tech-
niques may mean that we interpret the texture of organiza-
tions in a way similar to interpreting Shakespearean plays
exclusively by word counts and ratios. The complex, intangi-
ble, emotional dimensions of organizations probably cannot
be processed through the fine filter of linear statistics. Case
studies and other high-variety techniques may be more ap-
propriate for these dimensions. The December 1979 special
issue of ASQ, devoted to qualitative research, indicates such
a direction. Special issues do not necessarily predict future
trends, because laboratory research did not become popular
after the special issue of 1969, but qualitative techniques
seem to be gaining increasing acceptance among inves-
tigators, and they are compatible with the trend toward
more frequent study of complex organizational phenomena.

As to the best models and languages for future research, no

correct answer emerges from this analysis because organi-
zations contain many realities. Organizations are clockworks,
control systems, growth systems, social systems, all at the
same time; they are simple frameworks, amenable to chart-
ing and mapping, and they are systems of extraordinary
complexity. Organizations are far too vast for any of us to
digest, yet, as in a cafeteria, we can each make our own

632/ASQ

This content downloaded from 123.63.98.246 on Thu, 30 May 2019 07:41:55 UTC
All use subject to https://about.jstor.org/terms
 Evolution of Organization Analysis

selection. Organizations can be many things depending on

what we want to find and the techniques we use for inves-
tigation. Weick (1979), discussing the inquiry into multiple
realities, used a quote from Steinbeck about a fish as illust-
ration (1951: 2-3):
The Mexican sierra has '17-15-9' spines in the dorsal fin. These
can easily be.counted. But if the sierra strikes hard on the line so
that our hands are burned, if the fish sounds and nearly escapes
and finally comes in over the rai, his colors pulsing and his tail
beating the air, a whole new relational externality has come into
being - an entity which is more than the sum of the fish pius the
fisherman. The only way to count the spines of the sierra unaf-
fected by this second relational reality is to sit in a laboratory, open
an evil-smelling jar, remove a stiff colorless fish from formalin solu-
tion, count the spines, and write the truth, '17-15-9.' There you
have recorded a reality which cannot be assailed....
It is good to know what you are doing. The man with his pickled
fish has set down one truth and has recorded in his experience
many lies. The fish is not that color, that texture, that dead, nor
does he smell that way.

The analogy seems perfectly applicable to organization

theory. No single language can represent the many complex
dimensions of organizations. And since research methods
reflect languages used for studying organizations, it is clear
that they are very important. Research methods inform us
only of narrow realities, and provide descriptions that are
inherently incomplete images of organizations. The notions
of multiple realities and imperfect methods suggest that ad-
vocacy by organization scholars of preferred research ap-
proaches is a useless pastime. Knowledge from whatever
source is to be eagerly sought. Organizations are complex
enough for several paths of inquiry. The challenge is not to
decide upon superior methods. The challenge is to embrace
diverse methods that pursue several realities, and to distin-
guish quality in each.

Perhaps organization researchers should consider Stein-

beck's mental preparation (1951: 3-4):
We said, "Let's go wide open. Let's see what we see, record what
we find...."

We determined to go doubly open so that in the end we could, if

we wished, describe the sierra thus: "17-15-9," but also we
could see the fish alive and swimming, feel it plunge against the
lines, drag it threshing over the rail.... And there is no reason
why either approach should be inaccurate. Spine-count description
need not suffer because another approach is also used. Perhaps
out of the two approaches, we thought, there might emerge a
picture more complete and even more accurate than either alone
could provide.

In organization theory, also, we can go wide open. Low-

variety languages enable us to count the spines of organiza-
tion, to chart and map the terrain, and to correlate relevant
ratios. High-variety languages may enable us to hear the
music and poetry of organizations and to capture and com-
municate subtle and intuitive insights about organizations.

In summary, perhaps the most important conclusion to be

drawn from this analysis is an idea suggested by Boulding
(1956) about scientific progress, and expressed about the
time ASQ was beginning: Maybe we can avoid accepting a

633/ASQ

This content downloaded from 123.63.98.246 on Thu, 30 May 2019 07:41:55 UTC
All use subject to https://about.jstor.org/terms
 level of theoretical analysis too far below the complexity of
the empirical world we are investigating. We must not shut
the door on problems which do not fit easily into simple,
mechanical schemes. For all our progress in adopting new
methods and gaining insights in organization theory over the
last twenty years, the notion of system complexity points
out how far we still have to go.

REFERENCES

Alderfer, Clayton P., Robert E.Kap- Carzo, Rocco, Jr., and John N. Fennell, Mary L.
Ian, and Ken K. Smith Yanouzas 1975 "Trends in organizational re-
1974 "The effect of variations in re- 1969 "Effects of flat and tall or- search: An interview with
latedness need satisfaction on ganization structure." Ad- Morris Zelditch, Jr." Seminars
relatedness desires." Ad- ministrative Science Quarterly, on organizations at Stanford
ministrative Science Quarterly, 14: 178-191. University, 2:44-48.
19: 507-532. Coulter, Philip B. Franko, Lawrence G.
Alexander, Ernest R. 1979 "Organizational effectiveness 1974 "The move toward a mul-
1979 "The design of alternatives in in the public sector: The tidivisional structure in Euro-
organizational contexts: A example of municipal fire pro- pean organizations." Adminis-
pilot study."Administrative tection." Administrative Sci- trative Science Quarterly, 1 9:
Science Quarterly, 24: 382- ence Quarterly, 24: 65-81. 493-506.
404. Cressey, Donald R. Fromkin, Howard L.
Argyris, Chris 1 959 "Contradictory directives in 1969 "The behavioral science
1959 "The individual and organiza- complex organizations: The laboratories at Purdue's Kran-
tion: An empirical test." Ad- case of the prison." Adminis- nert School." Administrative
ministrative Science Quarterly, trative Science Quarterly, 3: Science Quarterly, 14: 171-
3: 145-167. 1 -19. 177.

Becker, Selwyn W., and Nicholas Daft, Richard L., and Alexander Glaser, William A.
Baloff MacMillan 1959 "Internship appointments of
1969 "Organization structure and 1977 "Structural control models of medical students." Adminis-
complex problem solving." organizational change: A repli- trative Science Quarterly, 4:
Administrative Science Quar- cation and caveat." American 337-356.
terly, 14: 260-271. Sociological Review, 42:
Grusky, Oscar
667-671.
Beyer, Janice M., and Harrison M. 1959 "Role conflict in organization:
Trice Daft, Richard L., and John C. A study of prison camp of fi-
1979 "A reexamination of the rela- Wiginton cials." Administrative Science
tions between size and vari- 1978 "Language and organization." Quarterly, 3: 452-472.
ous components of organiza- Academy of Management Re-
Guetzkow, Harold
tional complexity." Adminis- view, 4: 179-192.
1959 "Conversion barriers in using
trative Science Quarterly, 24: Dewar, Robert, and James Werbel the social sciences." Adminis-
48-64. trative Science Quarterly, 3:
1979 "Universalistic and con-
Blau, Peter M. tingency predictions of em- 68-81.
1965 "The comparative study of ployee satisfaction and con-
Hage, Jerald, and Michael Aiken
organizations." Industrial and flict." Administrative Science
1969 "Routine technology, social
Labor Relations Review, 28: Quarterly, 19: 426-448.
structure, and organization
323-338. Driver, Michael J., and Siegfried goals." Administrative Science
Bougon, Michel, Karl Weick, and Streufert Quarterly, 14: 366-376.
Din Binkhorst 1969 "Integrative complexity: An
Hall, Richard H.
1977 "Cognition in organizations: approach to individuals and
1962 "Intraorganizational structural
An analysis of the Utrecht groups as information-
variation: Application of the
Jazz Orchestra." Administra- processing systems." Ad-
bureaucratic model." Adminis-
tive Science Quarterly, 22: ministrative Science Quarterly,
trative Science Quarterly, 7:
606-639. 14: 272-285.
295-308.
Boulding, Kenneth E. Evan, William M., and Roberta G.
Hayes, Donald P.
1956 "General systems theory: The Simmons
1969 "The Cornell datalogger."
skeleton of a science." Man- 1969 "Organizational effects of in-
Administrative Science Quar-
agement Sciences, 2: 197- equitable rewards: Two exper-
terly, 14: 222-223.
207. iments in status inconsist-
ency." Administrative Science Hinings, C. R., D. J. Hickson, J. M.
Cafferata, Gail Lee Pennings, and R. E. Schneck
Quarterly, 14: 224-237.
1979 "Member and leader satisfac- 1974 "Structural conditions of in-
tion with a professional asso- traorganizational power." Ad-
ciation: An exchange perspec- ministrative Science Quarterly,
tive." Administrative Science 19: 22-44.
Quarterly, 24: 472-483.

634/ASQ

This content downloaded from 123.63.98.246 on Thu, 30 May 2019 07:41:55 UTC
All use subject to https://about.jstor.org/terms
 Evolution of Organization Analysis

Hummon, Norman P., Patrick Osborn, Richard N., and James G. Rosengren, William R.
Doreian, and Klaus Teuter Hunt 1964 "Communication, organization,
1975 "A structural control model of 1974 "Environment and organiza- and conduct in the 'therapeu-
organizational change." Amer- tional effectiveness." Ad- tic milieu,' " Administrative
ican Sociological Review, 4: ministrative Science Quarterly, Science Quarterly, 8: 70-90.
813-824. 19: 231 -246.
Salancik, Gerald R.
Janowitz, Morris Ouchi, William G., and James D. 1 979 "Field stimulations for organi-
1959 "Changing patterns of organi- Dowling zational behavior research."
zational authority: The military 1974 "Defining the span of con- Administrative Science Quar-
establishment." Administra- trol." Administrative Science terly, 24: 638-649.
tive Science Quarterly, 3: Quarterly, 19: 357-365.
Salancik, Gerald R., and Jeffrey
473-493. Pettigrew, Andrew M. Pfeffer
Keegan, Warren J. 1979 "On studying organizational 1974 "The bases and use of power
1974 "Multinational scanning: A cultures." Administrative Sci- in organizational decision mak-
study of the information ence Quarterly, 24: 570-581. ing: The case of a university."
sources utilized by headquar- Administrative Science Quar-
Pfeffer, Jeffrey, and Gerald R.
ters executives in multina- terly, 19: 453-473.
Salancik
tional companies." Adminis-
1974 "Organizational decision mak- Simpson, R. L.
trative Science Quarterly, 19: ing as a political process: The 1959 "Vertical and horizontal com-
411-421.
case of a university budget." munication in formal organiza-
Lincoln, James R., and Jon Miller Administrative Science Quar- tions." Administrative Science
1979 "Work and friendship ties in terly, 19: 135-151. Quarterly, 4: 188-196.
organizations: A comparative Pondy, Louis R. Smith, Catherine Begnoche
analysis of relational net- 1979 "Influence of internal oppor-
1 969a "Effects of size, complexity,
works." Administrative Sci- tunity structure and sex of
and ownership on adminis-
ence Quarterly, 24: 181-199. trative intensity." Administra- worker on turnover patterns."
Lowin, Aaron, William J. Hrap- tive Science Quarterly, 14: Administrative Science Quar-
chak, and Michael J. Kavanagh 47-60. terly, 24: 362-381.
1969 "Consideration and initiating 1969b "Varieties of organizational Smith, Peter B., David Moscow,
structure: An experimental in- conflict." Administrative Sci- Mel Berger, and Cary Cooper
vestigation of leadership 1969 "Relationships between man-
ence Quarterly, 14: 499-
traits." Administrative Science 505. agers and their work as-
Quarterly, 14: 238-253. sociates." Administrative Sci-
Pondy, Louis R., and Jacob G.
Lynch, Beverly P. ence Quarterly, 14: 338-345.
Birnberg
1974 "An empirical assessment of
1969 "An experimental study of the Steinbeck, John
Perrow's technology con-
allocation of financial re- 1951 The Log from the Sea of Cor-
struct." Administrative Sci- sources within small, hierar- tez. New York: Viking.
ence Quarterly, 1 9: 338-356. chical task groups." Adminis-
Stern, Robert N.
Lynton, Rolf P. trative Science Quarterly, 14:
1979 "The development of an
1969 "Linking an innovative subsys- 192-201.
interorganizational control
tem into the system." Ad- Pondy, Louis R., and Ian 1. Mitroff network: The case of intercol-
ministrative Science Quarterly,
1978 "Beyond open systems mod- legiate athletics." Administra-
14: 398-416.
els of organization." In Barry tive Science Quarterly, 24:
M. Staw (ed.), Research in 242-266.
Mayhew, Bruce H., Jr., and Louis
Organizational Behavior:
N. Gray Sutton, Richard L.
13-40. Greenwood, CT: JAI
1969 "Internal control relations in 1974 "Cultural context and
Press.
administrative hierarchies: A change-agent organizations."
critique." Administrative Sci- Pugh, D. S., D. J. Hickson, and C. Administrative Science Quar-
ence Quarterly, 14: 127-301. R. Hinings terly, 19: 547-562.
1969 "An empirical taxonomy of
Murphy, Thomas P. Tichy, Noel M.
structures of work organiza-
1969 "Political economy and geo- 1974 "Agents of planned social
tions." Administrative Science
graphic distribution of federal change: Congruence of val-
Quarterly, 14: 115-126.
funds for research and devel- ues,cognitions, and actions."
opment: The midwest cast." Pugh, D.S., D. J. Hickson, C. R. Administrative Science Quar-
Administrative Science Quar- Hinings, and C. Turner terly, 19: 164-182.
terly, 14: 426-441. 1969 "The context of organization
Turcotte, William E.
structures." Administrative
Oldham, Greg R., and Daniel J. 1974 "Control systems, perform-
Brass
Science Quarterly, 14: 91-
ance, and satisfaction in two
114.
1979 "Employee reactions to an state agencies." Administra-
open-plan office: A naturally Rosenbaum, James E. tive Science Quarterly, 1 9:
occurring quasi-experiment." 1979 "Tournament mobility: Career 60-73.
Administrative Science Quar- patterns in a corporation."
terly, 24: 267-284. Administrative Science Quar-
terly, 24: 220-241.

635/ASQ

This content downloaded from 123.63.98.246 on Thu, 30 May 2019 07:41:55 UTC
All use subject to https://about.jstor.org/terms
Van de Ven, Andrew H., and Weick, Karl E. White, Harrison C.
Andre L. Delbecq 1979 The Social Psychology of Or- 1969 "Control and evolution of
1974 "A task contingent model of ganizing, 2d ed. Reading, MA: aggregate personnel: Flows
work-unit structure." Adminis- Addison-Wesley. of men and jobs." Administra-
trative Science Quarterly, 19: tive Science Quarterly, 14:
183-197.
1969 "Laboratory organizations and
4-11.
unnoticed causes." Adminis-
Walter, Benjamin trative Science Quarterly, 14: Zadah, Lofti A.
1969 "On sampling, measures, and 294-303. 1973 "Outline of a new approach
aggregations." Administrative to the analysis of complex
Science Quarterly, 14: 131 - systems and decision pro-
133.
cess." IEEE Transactions on
Walton, Richard E., John M. Dut- Systems, Man, and Cyberne-
ton, and Thomas P. Cafferty tics, SMC-3: 28-44.
1969 "Organizational context and
interdepartmental conflict."
Administrative Science Quar-
terly, 14: 522-542.

636/ASQ

This content downloaded from 123.63.98.246 on Thu, 30 May 2019 07:41:55 UTC
All use subject to https://about.jstor.org/terms