BEHAVIORALISM:
Origins
of the Concept
Erkki Berndtson
Department
of Political Science P.O. Box 54 (Unioninkatu 37) 00014 University of Helsinki Finland
Prepared
for Presentation at the XVIIth World Congress of the International Political Science Association August 17-21, 1997 Seoul,
Korea
"By almost
any measure the most important aspect of post-World War II political science has been the rise of 'behavioralism': the controversies
it engendered, its success to coming to dominate much of organized political science, the changes it brought in the matters
to which political scientists attend, and the manner in which they are addressed. A lengthy essay would be necessary to treat
the subject even in moderate detail, and what follows can only be designated a sketch. However, no more (if indeed as much)
is needed by those who have at least a general acquaintance with the 'story.' And, for others, a vast literature - substantive,
polemical, interpretative - is readily available; in fact, it is unavoidable in any moderately extended attention to the written
record of the period." (Waldo 1975: 58)
1. Behavioralism
and the History of American Political Science
In the short history of American academic political science, behavioralism
has played an important role. Even most contemporary historians of the discipline have agreed with Waldo's statement above
(e.g., Seidelman and Harpham 1985: 150; Farr 1995: 198). Behavioralism has, indeed, haunted the discipline for half of its
existence now. And although behavioralism is a product of an American political science and must be understood in that context,
the impact of behavioralism on political science may have been as big in some other countries as it was in the United States.
(1) In that sense the 'story' should have relevance fo all political scientists.
To write
about behavioralism today is important at least for three reasons. First, the question is about the identity. Although the
present state of the discipline is usually labeled as postbehavioralism, "the present postbehavioral era can be understood
only with reference to behavioralism and that, even then, the behavioral revolution is open to different interpretations"
(Farr 1995: 221). But the strong identity is also based on conflict, and although "political scientists have quarreled over
many matters in the contemporary period...the most divisive issue by far has been behavioralism" (Somit and Tanenhaus 1982:
173). Only conflicts reveal important issues.
Secondly,
behavioralism is one of the best research objects to understand the nature of political science and conceptual change within
the discipline. The period following the Second World War was actually one of the two significant moments of academic reform
in the United States, the first one being the rise of the research university after the Civil War (Bender 1997: 4). This academic
reform touched also political science. The 1950s and the early 1960s were the golden days for the support of modern social
science (Warren E. Miller in Baer 1991: 242). In fact, the prestige of social science has clearly declined in the United States
since that time (Lindblom 1997: 227). And behavioralism brought passion for many. It has even be claimed that none of the
contemporary tendencies "has caught the imagination of the discipline as whole and swept it along with the sense of a common
and compelling purpose" (David Easton in Baer 1991: 213; cf. also Lindblom 1997: 230) the way behavioralism did. The reasons
for passion were many, but passion there was.
Thirdly,
from time to time it is useful to look back at history to see, if there is something to learn from history. Social sciences
are eager to adopt new ideas, research topics and approaches. Many times these new ideas may sweep important topics into the
dustbin of history. Maybe it is time to save something important.
Definitions
of Behavioralism
Most histories of political science agree on one thing: the concept of behavioralism has always been
somewhat unclear. As Dwight Waldo stated while writing about the emergence of behavioralism, "what happened was...complicated
- and somewhat obscure" (Waldo 1975: 58). Almost all those trying to define behavioralism have confessed that "every man puts
his own emphasis and thereby becomes his own behavioralist" (Easton 1962: 9) and "attempts at coming to any complete definition
of behavioralism are probably futile given the diversity of those who followed its banner" (Seidelman and Harpham 1985: 151).
And also David Truman has warned that it is a mistake to overstandardize the definition of behavioralism, because "it was
a kind multifaceted expression of dissatisfaction with the constraints and formalities of the conventional political science"
and "that impulse, was the only thing that was in common among really a quite diverse series of efforts" (David Truman in
Baer, et al., eds. 1991). The problem in trying to interpret the meaning of behavioralism simply seems to be that we can always
find "authorities to whom we can turn to press our interpretation" (Easton 1962: 9).
This seems
to indicate that there has been no behavioralism, but many behavioralisms. Accordingly, the first conclusion is that behavioralism
is not a research tradition, but something else. To find out what, a short history of conceptual change is useful.
Behavioralism
as a concept dates from the mid-1950s. As far as I have been able to detect the first one to use it in writing was Dwight
Waldo in his 1956 book "Political Science in the United States" (Waldo 1956: 24) which was an introductory text written mainly
for foreign audiences under UNESCO auspices. The term may have been used earlier, especially at the University of Chicago
(Somit and Tanenhaus 1982: 183), but it became into general use only after 1956 (e.g. Hacker 1959). Furthermore, the concept
seems to have received a wider popularity only after the publication of "The Limits of Behavioralism in Political Science"
(Charlesworth, ed. 1962). The first time the concept appeared in the pages of the American Political Science Review was in
1963 (Mendelson 1963: 593). Mendelson's article "The Neo-behavioral Approach to the Judicial Process: A Critique" was, as
its name indicates, a critique of behavioralism. In the next issue of the Review the concept appeared again in Albert Somit's
and Joseph Tanenhaus's survey of the trends in American political science, now in the form of a neutral research field (Somit
and Tanenhaus 1963).
In the beginning
of the 1950s the concept of behavioralism was not used. Instead, the concept of political behavior was very common from the
beginning of the 1940s having spread since Charles E. Merriam's call for its study in the 1920s. The next step, accordingly,
was to identify those studying political behavior, behavioralists. This happened at the beginning of the 1950s (e.g. Easton
1953: 151), although, instead of behavioralists, many used the term 'behaviorist'. It was above all David Easton, who stressed
the importance of distinguishing behavioralists from behaviorists (Easton 1953: 151).
The conceptual
development seems to coincide with Albert Somit's and Joseph Tanenhaus's periodization of the history of emerging behavioralism
(Somit and Tanenhaus 1982: 185): 1) from the end of the Second World War to 1949 there were only scattered signs of what was
coming, political behavior as a new research area began to be propagated, 2) from 1950 to the mid-1950s was the emerging period,
and, 3) from the mid-1950s to the mid-1960s behavioralism began to claim hegemony within the discipline, at the same time
when its opponents launched a determined counter-attack.
This seems
to indicate that behavioralism was used first either in a neutral introductory meaning (Waldo; Somit and Tanenhaus) or in
a critical fashion (Mendelson). David Easton has also pointed out that:
"behavioralism
was not a clearly defined movement for those who were thought to be behavioralists. It was more clearly definable by those
who were opposed to it, because they were describing it in terms of the things within the newer trends that they found objectionable.
So some would define behavioralism as an attempt to apply the methods of natural sciences to human behavior. Others would
define it as an excessive emphasis upon quantification. Others as individualistic reductionism. From the inside, the practioners
were of different minds as what it was that constituted behavioralism...few of us were in agreement." (David Easton in Baer
et al., eds. 1991: 207)
It was no
wonder that those identifying themselves as behavioralists were more apt to speak about the study of political behavior as
an approach and used the concepts of revolution, mood, movement, persuasion or protest to describe what political behavior
research indicated. From the beginning behavioralism was a political, not a scientific concept. However, when behavioralists
rised to an hegemonic position within American political science at the beginning of the 1960s the concept became more general.
Behavioralists accepted it (Easton 1962) and it became even to designate the stage in the history of American political science
(Easton 1985), while the critics continued to use it in a pejorative sense.
Because
behavioralism cannot be considered as a research tradition, but a political movement to advance something, definitions of
behavioralism give an idea of what the behavioralists wanted. Most introductions to the subject stress methodology and goals
of the discipline (value free research).
One of the
most "influential" definitions of behavioralism has been David Easton's list of its characteristics: 1) search for regularities,
even with explanatory and predictive value, 2) verification with testable propositions, 3) self-conscious examination for
rigorous techniques, 4) quantification for precision when possible and relevant, 5) keeping values and empirical explanations
analytically distinct, 6) systematization as an intertwining of theory and research, 7) pure science preceding the application
of knowledge, and 8) integration of the social sciences (Easton 1962: 7-8; Easton 1965: 7).
Other commentators
have agreed. Somit and Tanenhaus present actually the same eight tenets of behavioralism (Somit and Tanenhaus 1982: 177-179)
and many critics of behavioralism have also focused on methodology as a key element of behavioralism (e.g., Seidelman and
Harpham 1985: 151-152.)
But behavioralism
has been understood also as dealing with something more than only methodology and goals of the discipline. James Farr has
recently analysed behavioralism through three key problem areas: 1) a research focus on political behavior, 2) a methodological
plea for science, and 3) a political message about liberal pluralism (Farr 1995). Although a focus on political behavior is
self-evident in behavioralism, it has been somehow forgotten by those who are more interested in techniques than in actual
problems. Maybe one reason has also been that behavioralists have been eager to demonstrate that the behavioral study of politics
can be applied to all kinds of research areas (Eulau 1962).
Political
message about liberal pluralism refers to a new theory of democracy that was developed by American political scientists after
the Second World War. Especially radical critics of behavioralism have taken a pluralist theory of American politics as their
target, claiming that it has accepted the social and political system uncritically (e.g., McCoy and Playford, eds. 1967).
Locigally behavioralism and pluralism are two distinct phenomena, but, if behavioralism is a political concept or, if it would
like to present itself as a research tradition, pluralism could be understood as a tenet of behavioralism.
4. Explanations
for the Rise of behavioralism
In order to understand behavioralism better as an internally heterogenous movement, it is
necessary to focus on different explanations about its birth. The nature of behavioralism has been a topic of many articles,
but systematic research on its development has been rare. There are general explanatory descriptive studies (Somit and Tannehaus
1982; Waldo 1975; Ricci 1984), historical interpretations of the period through some key scholars (Crick 1959; Seidelman and
Harpham 1985), genealogical studies of the birth of research fields (Gunnell 1993) and countless articles (Dahl 1961; Ball
1993; Farr 1995).
Although
many of these studies are very systematic and thorough in their own ways, they give, however, only a partial picture of the
problem. No full scale attempt to outline factors behind the behavioral movement exists. Using the intellectual matrix of
political science, given explanations thus far may be presented as follows:
Society
Social development.
Deep social crises have made political scientists more keen to realism. Students of the history of the discipline have mainly
referred to three "big things": The Great Depression of the 1930s and the New Deal (Ball 1993), the World War II (Dahl 1961:
Waldo 1975; Ball 1993; Farr 1995) and the Cold War (Ball 1993; Farr 1995). The socializing effect seems to have been most
important in connection with the Great Depression, while it was Second World War that really forged a new relation between
academia and politics. The Cold War marked the whole era of behavioralism and in that way its general effect on behavioralism
was maybe more important than anything else. All three crisis led to a deep dissatisfaction with the "state of the discipline"
(Somit and Tannehaus 1982: 184) and to behavioralism as an answer.
Organizations.
One of the most given explanations for the rise of behavioralism has been a reference to the role of foundations (2), mainly
the emergence of the Ford Foundation as an important provider of funds (Dahl 1961; Somit and Tanenhaus 1982; Seidelman and
Harpham 1985: 154). Many have even argued that the whole concept of behavioralism came into use only because of the policy
of foundations (Geiger 1988: 329). And Bernard Berelson seems to agree:
"What happened
to give rise to the term? The key event was the development of a Ford Foundation program in this field. The program was initially
designated 'individual behavior and human relations' but it soon became known as the behavioral sciences program and, indeed,
was officially called that within the foundation. It was the foundation's administrative action, then, that led directly to
the term and to the concept of this particular field of study." (Berelson 1968: 42) (3)
The foundation
money created also a self-generating process which led to the recruitment of behavioralists. Because behavioralist projects
were funded better than traditional ones, there were a larger supply of behavioralists up for recruitment than others (Hacker
1959: 39-40). It is no wonder that some of the key practioners of behavioralism have been willing to admit, that "it was almost
single-handedly the Ford Foundation that did so much to legitimate empirical social science" (Warren E. Miller in Baer, et
al., eds. 1991: 242).
And David
Easton has argued that "in its orgins it may well be that the concept can be considered an accident" (Easton 1965: 12). Easton
explains this accident with the convergence of the founding of the Behavioral Sciences Division of the Ford Foundation and
the story of some congressmen attacking social sciences as socialist sciences (Easton 1965: 22; cf. Ball 1993).
Foundations
played a key role in financing research at the beginning of the 1950s, but the Ford Foundation terminated its behavioral program
in 1957. One of the reasons was the attack towards foundations. There had already been the House of Representatives Select
Committee (Cox Committee) investigating tax-exempt foundations in 1952, and the foundations were increasingly criticized as
being too liberal (actually it was in the hearings of Cox Committee when social sciences were termed as socialist sciences).
After the
foundations began to diminish their financing of the behavioral sciences, the National Science Foundation stepped in. The
original bill for the establishment of a National Science Foundation, which was given to the Congress in 1945, contained the
establishment of a Social Science Division. Social Sciences came under attack from politicians, however, already then, because
some congressmen saw them representing values foreign to Anmericans. Social scientists counterattacked with rhetorics stressing
the values of science, neutral pursuance of truth, and distancing themselves from social reform. All the values that had helped
to win the Second World War (Ball 1993). But to no avail. The National Science Foundation was established in 1950 without
the Division of Social Sciences.
The Foundation
began to search for opportunies to finance the social sciences, however, already in 1953. In 1958 it founded the Office of
Social Sciences, which was changed into the Division alredy a year later (Lyons 1969: 272). Social sciences were accepted
by politicians at last.
As these
examples show, political science was deeply dependent on the state of democracy in society. A general political climate had
an impact also on individual scholars and universities as organizations.
Political
Science
Scientific
Culture. Scientific culture is tied to many ways to the general social, economic and political culture and development of
a given country. Bernard Crick's thesis that American political science is based on a four-fold relationship between a common
notion of science, the idea of a citizenhip training, the habits of democracy and a common belief in an inevitable progress
(Crick 1959: xv) still merits attention.
On the other
hand, comparing the American and British political science, Andrew Hacker pointed out in the 1950s, how behavioralism was
more natural for an American political scientist than for his/her British counterpart. British political scientists were at
that time mainly library researchers. They could ignore the public, because they were judged by what they were rather than
by what they did. On the contrary, in the United States there has always been a cultural requirement that the professor must
"do" something. Although this is bound to lead to "productive scholarship" or as Benjamin Lippincott once remarked, to work
which reminds "the compilation of a telephone directory" (Lippincott 1993: 157), the seeds for behavioral study of politics
can be found in American culture.
Organizations
and Resources. Different universities played a different role in the rise of behavioralism. Usually one refers to the University
of Chicago. The 'Chicago School of Political Science' led by Charles E. Merriam has been seen by many as a founder of the
behavioralist movement in political science (Dahl 1961). Another kind of impact of the University of Chicago came after the
World War II, when scholars from various disciplines joined together to develop an interdisciplinary behavioral science.
Other forms
of behavioral study of politics were developing elsewhere at the same time. The University of Michigan became a center for
survey research. An important factor was the foundation the Survey Research Center in 1946, which became the Institute for
Social Research in 1948 (Seidelman and Harpham 1985: 153). Later an event of utmost importance was the foundation of the Inter-University
Consortium for Political and Social Research in 1962 Michigan as its homebase. It is important to notice, however, that the
development was very uneven in different places and the spreading of behavioralism as dominant ideology of research was a
long political process.
Other organizations
than universities played a major role in this process. A special importance has been given to the Social Science Research
Council and its Committee on Political Behavior (Dahl 1961; Farr 1995). The Committee was founded in 1949 V.O. Key as its
first chairman. The President of the Social Science Research Council at that time, Pendleton Herring, has simply stated later
that there was a spectrum of changing attitudes about the role of political science and the group "had an opportunity to articulate
ideas that were developing, and it had the means to encourage and reward people who chose to work along lines that the committee
was concerned with" (Pendleton Herring in Baer et al., eds. 1991: 34).
The American
Political Science Association as a peak organization of the discipline did also play an important role. Its impact was strengthened
by the growth of the profession. In 1946 APSA had a membership of 4 000 which had risen to 14 000 in 1966 (Seidelman and Harpham
1985: 154). This growth in itself transformed a learned society to a professional organization. The result was that political
scientists were now able to concentrate more easily on their own problems leaving social problems outside the profession (Seidelman
and Harpham 1985: 151; 185). A good example is the attendance at the annual meetings of APSA. In the 1920s the registrants
ranged between 127 and 292 and it was only in 1939 when there were over 1000 participants.
Academic
freedom. McCarthyism had a severe effect on the moulding of behavioralism (Easton 1985: 139-140; Farr 1995: 211). Some political
scientists referred to this already in the 1950s, claiming that behavioralism provided a convenient escape from pressure from
public appraisal and, because a behavioralist was "a liberal in his politics, but timid in his personality" he was ready to
plunge into the behavioral stream, because he "who has not stuck his neck out cannot be attacked. And if he does question
dominant values, he frequently does it in technical jargon and to a limited audience" (Hacker 1959: 40).
These personal
statements of scholars experiencing the everyday politics of the 1950s may also be corroborated by referring to studies on
the impact of McCarthyism on academia. Paul F. Lazarsfeld's and Wagner Thielens, Jr.'s book "The Academic Mind: Social Scientists
in a Time of Crisis" (1958) was a survey of over 2000 social scientists. Its results showed that "McCarthyism" really had
an impact on academia. Many felt themselves under pressure (see also Diamond 1992).
Individuals.
The role of individuals is more complicated. Behavioralists have presented their own pedigree, which usually starts with Graham
Wallas and Arthur F. Bentley, takes notice of Charles A. Beard and Walter Lippman, and then comes to Charles E. Merriam and
the Chicago School of Political Science.
However,
there have also been conflicting arguments. The first one concerns the role of Charles E. Merriam and his younger colleagues
Harold D. Lasswell and Harold F. Gosnell. For instance, Raymond Seidelman and Edward Harpham (1985) don't count Merriam and
Lasswell as behavioralists, but as representatives of reform political science which was criticized by behavioralists at the
beginning of the behavioral era.
Another
case with conflicting interpretations concerns the role of the émigré scholars during the Second World War. Dahl (1961) and
Somit and Tanenhaus (1982) have claimed that they provided the theoretical background for behavioralism, while John G. Gunnell
(1993) has recently argued for a much more complex relationship between the émigré scholars and American political science.
Many of those who escaped from Europe became leading critics of empirical science of politics.
Other Disciplines
The success
of behavioralism would hardly have been possible without the development of survey methods, interviewing techniques and techniques
of measurement (Truman 1955: 204-209; Dahl 1961; Somit and Tanenhaus 1982; Seidelman and Harpham 1985: 152). On the other
hand, many American universities were organized so that political scientists had an easy access to representatives of other
disciplines, sociologists, social psychologists, etc. Compared to situation in other countries these other social sciences
were well organised and intellectually developed. It was no wonder that interdisciplinary work took root in American universities,
at least to some degree (Hacker 1959).
5. Explaining
the Nature and Rise of Behavioralism
After the brief survey of the meaning of behavioralism and of major explanations
for its birth, the concept still seems elusive. It is also hard to determine the exact weight of different explanations. It
seems reasonable to agree with David Easton, however, that the concept of behavioral sciences was due both to the policy of
the Ford Foundation (3) and to the conservative climate suspicious of anything reminding of socialism. On the other hand,
it must be remembered that the concept of behavioralism was first used either in neutral or critical way, and it was only
later that also behavioralists adopted it.
Political
climate shaped also the discourse on goals of the discipline. The debate on values became important when political scientists
had to defend themselves (being liberal but neutral) against the threat both from the Right and from the Left (Gunnell 1993).
But many
of the tenets of behavioralism would have developed as well under some other label than behavioralism. The thirties and the
New Deal, the Second World War and the Cold War had created the soscial context of American political science, behavioralist
or not. And if we look at research areas, theories, organizing concepts and methodology of political science, all these basic
components of any research tradition, were already pointing towards what came to be called behavioralism.
A focus
on an individual as a basic unit of analysis as well as an interest in empirical study of observable facts are both goals
deeply grounded in American culture. The growing individualism of American society has nurtured methodological individualism
and research areas dealing with the political behavior.
The scientific
study of politics is also a part of American culture. Important forerunners of behavioralism were the national conferences
on the science of politics already in the 1920s. It was only that the development of statistical techniques gave new ammunition
to this endeavor (4).
And, as
was pointed out earlier, behavioralism really was only a loose collection of different research traditions which converged
in a number of points in order to advance their own special interests (5). As Evron M. Kirkpatrick wrote: "the term served
as a sort of umbrella, capacious enough to provide a temporary shelter for heterogenous group united only by dissatisfaction
with traditional political science" (Kirkpatrick 1962: 13). It has even been claimed that it was often difficult to distinguish
true card-carriers in the behavioral movement from fellow-travelers, tolerant sympathizers, occasional supporters, or ambivalent
critics (Easton 1962: 5). And recently Heinz Eulau has argued that:
"It seems
to me that the practioners are less inclined than the bystanders to see this (behavioralism, E.B.) as a great divide in the
discipline's history and tend to be more sensitive to the coexistence of novelties and continuities....I felt at the time,
the historical, legal, or doctrinal approaches to the study of politics 'will persist....I have never believed that behavioralism
occupied the hegemonic position attributed to it by some of its own advocates and by latter-day interpreters. The undue emphasis
on behavioralism as 'revolution' has led to a biased neglect of theory and research that was not behavioral, though done in
the 'behavioral era.'" (Eulau 1997: 585)
It cannot
be denied that in the 1950s and at the beginning of the 1960s the majority of American political scientists were doing something
else than behavioral political science. International relations, American politics and government, comparative government,
public administration and public law were all strong fields, and most of the scholars in those fields were not behavioralists.
A good example
is the Department of Political Science at the University of California (Berkeley), whose chairman from 1948 to 1956 was Peter
Odegard. Odegard was a member of the Ford Foundation Study Committee from 1948 to 1950, but Odegard's own department was very
strong in comparative politics and even in political theory (Watson 1961: 315-343). Although it is good to remember that most
of the money from Ford Foundation went actually to international relations and comparative politics.
How can
all this be accounted for? The obvious paradox may be explained by the fact that most political scientists in the 1950s and
the early 1960s occupied a position somewhere in the middle between behavioralism and traditionalism, but they were intuitively
drawn more towards behavioralism.
Albert Somit's
and Joseph Tanenhaus's survey of the profession at the beginning of the 1960s shows this well. Using factor analysis to study
the attitudes among political scientists, the single most important factor dividing the discipline was "behavioralism". However,
only 10 percent of respondents identified themselves as behavioralists, although 22.6 percent felt that the most significant
work within the discpline was being done by behavioralists. The other side of the coin was that there were 12.3 percent identifying
themselves as political theorists, but only 9.9 percent felt political theory being a field of the most significant work.
Furthermore, when asked which was the field of the least significant work, 7.8 percent said it was behavioral political science,
but 32.4 percent thought that political theory had least significance in the discipline (Somit and Tanenhaus 1963: 939-942).
The American
political Science Review published also in 1956 a report on a Conference on Political Theory and the Study of Politics (Eckstein
1956), which indicated that there were a group of 'behaviorists' and 'anti-behaviorists' in the discipline, but a large majority
were moderates between these two extreme groups. Much of the discussion in the conference dealt with a question of "how and
why had the political 'philosophers' and the political 'scientists' managed to drift so far apart in their work when most
of them were eager to concede that they had much to learn from one another?" and in the final analysis "our villain...was
simply the 'behaviorist' - 'theorist' dichotomy" (Eckstein 1956: 486-487).
Of course,
there were real differences of opinion concerning methods of political science (explanation vs. understanding), its social
role (training vs. cultivating) and its ultimate goal (scientific discipline vs. learned wisdom), but the underlying problem
in many cases seemed to be psychological and practical. People were interested in different problems and there was no real
interest in the study of politics in totality. The problem was a political rather than an intellectual problem: "how to persuade
the practioners in the field to develop a healthy interest and to acquire a healthy preparation in both political science
and political philosophy." The 'behaviorists' were too current-events-conscious and the 'theorists' were not genuinely interested
in political behavior. It seemed to be difficult to reach a compromise "between the study of Plato and the study of local
government in Illinois" (Eckstein 1956: 485).
Anyone familiar
with political scientists knows that this is a real problem. Partly it is an ontological and epistemological problem. There
are political scientists who are convinced that there exists "immortal" questions of politics, such as, freedom, equality
and welfare, which can be studied out of historical context. And there are political scientists who are more interested in
the immediate practical issues of politics which they confront in the everyday life.
But one
must distinguish the elite and the mass. The mass of American political scientists were doing what they had done before. It
is no wonder that Charles E. Lindblom has recently noted that in the 1950s there was an amazing discrepancy between accepted
scientific ideals and actual practice (Lindblom 1997: 233). The elite was defining the discourse. But the elite was not unanimous.
Behavioralism consisted of at least two (still internally divided) research traditions: empirical theory and general theory.
Empirical theory could be divided into those focusing on individuals proper (voting, public opinion) and those adhering to
group approach. Those adhering to general theory, on the other hand, could be divided into various forms of systems theory
- general systems theory, cybernetics, systems analysis, functionalism.
This internal
division in the ranks of behavioralism has not always been understood by all those writing about it (see, however, Farr and
Seidelman, eds. 1993: 203). The division can be personalized using David Truman and David Easton as examples. Easton spoke
mainly for the general theorists who found the established political science wanting. David Truman, on the other hand, spoke
for those more empirically oriented researchers who had been influenced by Charles E. Merriam (Eulau 1969: 5). At the end,
it was Truman's side who took command of the behavioral research.
This was
due to two things. First, Truman was a representative of a true elite within the discipline during behavioral era. For a number
of reasons this elite was able to rise in the ranks of the profession in the 1950s and especially at the beginning of the
1960s (Waldo 1975: 61). Through their strategic positions, this key group of behavioralists was able to raise behavioralism
into a legitimate and hegemonic position within the discipline at the beginning of the 1960s.
Who should
be counted as a member of this core group (elite), is a question which should be studied more carefully, but I would suggest
that at least the following persons should be included: Pendleton Herring, Peter Odegard, V.O.Key, Jr., David Truman, Angus
Cambell, Oliver Garceau, Alexander Heard, Avery Leiserson and Samuel J. Eldersveld. Many of them were members of the Committee
on Political Behavior, while Herring and Odegard acted as key links to foundations. Herring was an executive associate of
the Carnegie Corporation from 1846 to 1948 and Odegard was a member of the Ford Foundation Study Committee (1948-1950), which
planned the Foundation's new activity for the social sciences. Both served also as President of the American Political Science
Association (Odegard 1950-51, Herring 1952-53). Herring was also President of the Social Science Research Council over twenty
years (beginning in 1948). In many respects, however, the key person was V.O. Key, the first chairman of Social Science Research
Council's Committee on Political Behavior. His good friends David B. Truman, Alexander Heard and Oliver Garceau still strengthened
his position.
It is good
to point out two things. Most of these people were what is called empirical theorists and secondly, many of them had studied
pressure groups (or parties). Another interesting thing is that those who wrote the most influencial writings about behavioralism
(Robert A. Dahl, David Easton and Heinz Eulau) were somehow outside the core group (although Dahl was a close friend of Key).
Maybe they were too much general theorists to fit in:
"those who
were particularly interested in quantification and the use of survey research were very critical of the work I was doing,
which was very theoretical. They saw me as a person who was articulating a defense for them, but not as a practioner. And,
although they accepted the importance of empirical theory, for them theory meant a much lower level of generalization than
I was seeking. So those who criticized me as a behavioralist incarnate were driving me into the arms of the behavioralists,
yet the behavioralists didn't have their arms wide open." (David Easton, in Baer et al., eds. 1991: 207)
The division
seems to have been also a question of different generations. Some of the oral histories of American political scientists (Baer,
et. al., eds. 1991) point to this. Those born during the first decade of the 20th century were the first real generation of
empirical students of politics (Charles Hyneman, Pendleton Herring, Belle Zeller, etc.). A generation born in the 1910s became,
however, more theoretically oriented (e.g., Gabriel Almond, Robert A. Dahl and David Easton). A common experience for them
was the depression of the 1930s. The generation born in the 1920s was again more empirical (and conservative) than the previous
generation (e.g., Austin Ranney and Warren E. Miller).
But the
second fact has to be taken into account, too. The reason Truman's side became the core of behavioralism was due to the attitudes
and opinions of the mass. And with the growth of the discipline the mass has become important organizationally (Ricci 1984).
Positive towards behavioralism, but doing very traditional political science, these ordinary political scientists felt more
close to Key's and Truman's concerete studies on American politics than to Easton's theoretical analysis.
If I would
have to make a summary of my argument about the nature of behavioralism and about its rise into an hegemonic position within
American political science, I would use an analogy of American party politics. As American party politics consists of two
major parties which are loose coalitions of different interests, so American political science has tended to create two opposing
parties, whether it is traditionalists vs. behavioralists or behavioralists vs. postbehavioralists does not matter. These
parties have been as undefinable as their counterparts in real political life. But they have caught the imagination of American
political scientists, because American culture is dialectical. They have also formed identities as well as they have been
vehicles in pursuing material interests. To better understand the nature and history of political science, it might be better,
however, to forget behavioralism and postbehavioralism and look at the development of the discipline through different research
traditions competing with each other.
6. The Legacy
of Behavioralism
Terence Ball has recently (1993: 220-221) argued that behavioralism succeeded in the short run, because
of succesful self-promotion, but in the long run it failed, because promises were made that could not be met. But, if the
analysis in the previous chapter is correct, it doesn't matter whether behavioralism succeeded or not. An interesting thing
to look, however, is which aspects of behavioralism have survived in the postbehavioral era and which have been lost.
A few years
ago Theodore Lowi (1992) claimed that American political science was dominated by three subfields of the discipline, public
opinion, public choice and public policy. Of these only public opinion represented the old behavioral study of politics. If
Lowi is correct, the first conclusion is that it is the individualistic perspective inherent in behavioralism that has been
one of the survivors. Although the early behavioralism understood the importance of studying governmental structures and understood
also socialization as an important element in moulding a person's attitudes and behavior, the basic methodological starting
point was methodological individualism, because it was assumed that "collectivities do not exist apart from the conduct of
their individual members" and "political behavior analysis is pre-eminently interested in determining the consequences of
individual political behavior for the functioning of political institutions" (Eulau 1968: 205).
A focus
on individuals has also nurtured quantitative techniques. The early behavioralists were alraedy worried about this trend and,
for example, V.O.Key used to warn that there is a danger that people stop asking good questions when methods are overemphasized
(David Truman in Baer et al, eds., 1991: 148; cf. also David Easton in Baer et al., eds., 1991: 209). Both quantitative and
qualitative analysis had their own place in research:
"the approach
cannot be limited to areas where the possibility of quantification is immediate. The student of political behavior cannot
escape the obligation and must not deny himself the opportunity to ask important questions...simply because his answers must
be more or less qualitative...he is obliged to perform his task in quantitative terms if he can and in qualitative terms if
he must" (Truman 1951: 39).
In regard
to analyzing data with different nethods, American political science has really went into a wrong direction. Partly this may
be due to the untheoretical nature of contemporary research. This untheoretical nature is tied to the major loss in behavioral
legacy, forgetting the need for the integration of social knowledge. The idea has curiously disappeared. Here again the bureaucracy
and special interests have been more powerful. The interdisciplinary idea never went through the ranks of ordinary political
scientists (Geiger 1988: 332).
At the beginning
of the century there were also attempts to maintain close relations with all the social sciences, but already then the drift
went away from this goal. As Albert Lepawsky has noted the new social science associations entered the twentieth century seeking
to counteract their separateness. There were requests to a possible federation of associations, but "it was too late. The
centrifugal forces in this phase of American life were stronger than the centripetal" (Lepawsky 1964: 52).
And this
has been the case ever since (the only success being the founding of the Social Science Research Council in 1924). This has
galvanized also behavioralism. It is logical that the integration of the social sciences was not the major issue for the true
behavioralist elite:
"The student
of political behavior is concerned with the methods and concepts in other disciplines only to the extent that such tools contribute
directly to a more meaningful statement of the phenomena of politics. The ultimate goal of the student of political behavior
is the development of a science of the political process, logically complete within itself. Given the tremendous range and
complexity of political data, such specialization is a continuing need...Many of the concerns of the other behavioral sciences
are not thus relevant" (Truman 1951: 38-39).
This attitude
is completely different than David Easton's argument for the unity of behavioral sciences, which was, according to him, the
real core of behavioralism. Arguing for the behavioral study of politics, Easton wrote that, "if all that behavioralists are
arguing for is the introduction of scientific method and nothing more, why are we not content with calling a spade a spade?"
(Easton 1965 :9)
And Easton
goes on to argue that the real meaning of behavioralism is (besides attention to empirical theory) "in locating stable units
of analysis" (Easton 1965: 13) and "the key idea behind this approach has been the conviction that there are certain fundamental
units of analysis relating to human behavior out of which generalizations can be formed and that these generalizations may
provide a common sense on which the specialized sciences of man in society could be built" (Easton 1965: 14-15).
Referring
to attempts to find the most fruitful unit for a general social theory, Easton (1965: 15-21) refers to concepts such as action,
the decision, functions, systems, power and groups. Although noting that in his own work he has "been exploring the utility
of the system as the major unit, focusing on political life as a system of behavior operating within and responding to its
social environment as it makes binding allocations of values" (Easton 1965: 21), he still leaves a door open for the best
unit of analysis. Easton's real concern is a search for unified social knowledge.
Easton has
not dropped the idea. In his 1969 APSA presidential address he argued again that "social problems do not come neatly packaged
as economic, psychological, political and the like. Our crises arise out of troubles that involve all aspects of human behavior"
(Easton 1969: 1060). This time Easton proposed also the establishment of a Federation of Social Scientists to advance the
integration of the social sciences. But no one seemed to notice this aspect of the new revolution in political science.
The same
has been the case with Easton's article on "The Political Science in the United States" in which Easton wrote:
"in emphasizing
the need to apply whatever knowledge we have to the solution of urgent social issues, we have already run into major difficulties
in trying to reintegrate the various highly specialized disciplines....Application of knowledge to the social problems...requires
the reassembly of the specialized knowledge of the various social sciences". (Easton 1985: 151)
The need
for integration of social sciences is a major legacy of behavioralism, but it has been forgotten as well by the majority of
behavioralists as those criticizing it. The result has been the acceptance of the fragmentation of social knowledge as in
the case of postmodernism. But the integration of social knowledge is as timely as it has always been. And this task doesn't
require of finding any single concept that could unite the social sciences. It only requires that political, economic and
social aspects of human behavior are theoretically linked together.
Individualism,
quantification and fragmentation are the three major problems of social sciences today. Some have even claimed that they are
behind our social crisis today (Dimock 1987; cf. also Bay 1965). To fight these tendencies of modern social science may be
futile, but to be aware of these tendencies is the first requirement of forming ones's identity as a political scientist.
To understand our identity we need a pluralist theory of scientific development.
Notes
1.
Writing about European political science, Ken Newton and Josep Vallès note that "it is, nevertheless, true that the 'behavioral
revolution' transformed large parts of political science in Western Europe, and its effects were so profound that some on
the European side of the Atlantic claim that their traditions of political inquiry were abandoned or forgotten to such an
extent that the American approach colonised the West European profession" (Newton and Vallès 1991: 234-235). Although Newton
and Vallès take the colonization thesis as an exaggeration, they don't deny the American influence (see the articles in EJPR
1991, for instance). In the case of behavioralism it must be remebered that at the beginning of the 1960s political science
was still in many ways an American discipline. Giving his presidential address in the 1965 APSA Meeting on the state of political
science, David Truman did not see anything wrong by saying that "I am willfully going to commit the sin of parochialism by
confining my remarks promarily to the discipline in the United States. In justification I would argue that the problems of
political science are, if only because of the number of practioners involved, chiefly problems of American political science"
(Truman 1965: 865). And Gabriel Almond estimated a year later that "nine out of every ten political scientists in the world
today are American" (Almond 1966: 869). In 1982 William G. Andrews still gave an estimate of 75-80 percent of world's political
scientists being Americans (Andrews 1982: 3). Latest membership figures of national associations belonging to the International
Political Science Association, however, give the percentage of American political scientists in the world today as 41.27 percent
(IPSA 1997). This figure is still consiredably higher than that of any other country (the second largest colony of political
scientists seems to be in India with a percentage of 4.96). It may be, however, that American political scientists still have
an absolute majority, beecause most of the members of the APSA are academics, while in many other countries membership in
local associations consists much more of bureaucrats, politicians, lawyers, students, etc. In any case, a relative decline
of the position of American political science seems to be a reality.
