Southern Political Science Association

Comparative Political Systems

Author(s): Gabriel A. Almond
Source: The Journal of Politics, Vol. 18, No. 3 (Aug., 1956), pp. 391-409
Published by: Cambridge University Press on behalf of the Southern Political Science Association
Stable URL: http://www.jstor.org/stable/2127255 .
Accessed: 26/05/2014 02:40

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
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].

Cambridge University Press and Southern Political Science Association are collaborating with JSTOR to
digitize, preserve and extend access to The Journal of Politics.

http://www.jstor.org

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
 COMPARATIVEPOLITICALSYSTEMS*
GABRIEL A. ALMOND
Princeton University

W Af I PROPOSE to do in this brief paper is to suggest how

HAT
the application of certain sociological and anthropological
concepts may facilitate systematic comparison among the major
types of political systems operative in the world today.
At the risk of saying the obvious, I am not suggesting to my
colleagues in the field of comparative government that social theory
is a conceptual cure-all for the ailments of the discipline. There
are many ways of laboring in the vineyard of the Lord, and I am
quite prepared to concede that there are more musical forms of
psalmody than sociological jargon. I suppose the test of the socio-
logical approach that is discussed here is whether or not it enables
us to solve certain persistent problems in the field more effectively
than we now are able to solve them.
Our expectations of the field of comparative government have
changed in at least two ways in the last decades. In the first place
as American interests have broadened to include literally the whole
world, our course offerings have expanded to include the many
areas outside of Western Europe - Asia, the Middle East, Africa,
and Latin America. Secondly, as our international interests have
expanded and become more urgent, our requirements in knowledge
have become more exacting. We can no longer view political crises
in France with detached curiousity or view countries such as Indo-
China and Indonesia as interesting political pathologies. We are
led to extend our discipline and intensify it simultaneously.
It would simply be untrue to say that the discipline of compara-
tive government has not begun to meet both of these challenges.
As rapidly as it has been possible to train the personnel, new areas
have been opened up to teaching and research; and there has been
substantial encouragement to those who have been tempted to ex-
plore new aspects of the political process both here and abroad and
to employ new methods in such research. It is precisely because
*This paper was preparedfor the Conferenceon The ComparativeMethod
in The Study of Politics, held under the auspices of the Committee on
ComparativePolitics, Social Science Research Council at Princeton, June 2-4,
1955.
[ 391 ]

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
392 THE JOURNAL OF POLITICS [Vol. 18

of the eagerness and energy with which these challenges have been
met that the field is now confronted with the problem of systematic
cumulation and comparison. What appears to be required in view
of the rapid expansion of the field are more comparative efforts in
the tradition of Finer and Friedrich, if we are to gain the maximum
in insight and knowledge from this large-scale research effort.
The problem to which this paper is a tentative and provisional
answer is the following. With the proliferation of courses and
special studies of specific "governments" and groupings of govern-
ments on an area or other bases, is it possible to set up and justify
a preliminary classification into which most of the political systems
which we study today can be assigned? The classifications which
we now employ are particularistic (e.g., American Government,
British Government,the Soviet Union, and the like); regional (e.g.,
Government and Politics of the Far East, Latin America, and the
like); or political (e.g., the British Commonwealth,Colonial Govern-
ment, and the like); or functional (e.g., the comprehensive com-
parative efforts limited to the European-Americanarea, such as
Finer and Friedrich, and the specific institutional comparisons such
as comparative parties, and comparative administration).
Anyone concerned with this general problem of classification
of political systems will find that all of the existing bases of classi-
fication leave something to be desired. Dealing with governments
particularistically is no classification at all. A regional classification
is based not on the properties of the political systems, but on their
contiguity in space. The existing structural classifications, such as
democracy-dictatorship, parliamentary-presidential systems, two-
party and multi-party systems, often turn out to miss the point,
particularly when they are used in the strikingly different political
systems of the pre-industrial areas. There may be a certain use
therefore in exploring the possibilities of other ways of classifying
political systems. What is proposed here is just one of these ways,
and because of the uneven state of our knowledge is necessarily
crude and provisional.
In my own efforts to stand far off, so to speak, and make the
grossest discriminations between types of empirical political systems
operative in the world today, I have found a fourfold classification
to be most useful: the Anglo-American (including some members
of the Commonwealth), the Continental European (exclusive of the
Scandinavian and Low Countries, which combine some of the fea-

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
1956 ] COMPARATIVE POLITICALSYSTEMS 393

tures of the Continental European and the Anglo-American), the

pre-industrial, or partially industrial, political systems outside the
European-Americanarea, and the totalitarian political systems. This
classification will not include all the political systems in existence
today, but it comes close to doing so. It will serve the purpose of
today's discussion, which is not that of testing the inclusiveness
of this classification but rather the usefulness of sociological con-
cepts in bringing out the essential differencesbetween these political
systems.
The terms which I shall use in descriminating the essential
properties of these classes have emerged out of the Weber-Parsons
tradition in social theory.' I shall try to suggest why I find some
of these concepts useful. First, a political system is a system of
action. What this means is that the student of political systems
is concerned with empirically observable behavior. He is concerned
with norms or institutions in so far as they affect behavior. Empha-
sizing "action" merely means that the description of a political
system can never be satisfied by a simple description of its legal or
ethical norms. In other words, political institutions or persons per-
forming political r6les are viewed in terms of what it is that they
do, why they do it, and how what they do is related to and affects
what others do. The term system2 satisfies the need for an inclusive
concept which covers all of the patterned actions relevant to the
making of political decisions. Most political scientists use the term
political process for these purposes. The difficulty with the term
process is that it means any patterning of action through time.
In contrast to process, the concept of system implies a totality of
relevant units, an interdependencebetween the interactions of units,
and a certain stability in the interaction of these units (perhaps
best described as a changing equilibrium).
The unit of the political system is the role. The role, according
to Parsons and Shils, ". . . is that organized sector of an actor's
orientation which constitutes and defines his participation in an in-
teractive processe" It involves a set of complementaryexpectations
'See in particularMax Weber,The Theory of Social And EconomicOrgani.,
nation, trans. by A. M. Hendersonand Talcott Parsons (New York: Oxford
University Press, 1947), pp. 87 ff.
'See David Easton, The Political System: An Inquiry into the State of
Political Science, (New York: Alfred Knopf, 1953), pp. 90 ff.
'Talcott Parsons and Edward A. Shils, eds., Toward a GeneralTheory of
Action (Cambridge:Harvard University Press, 1951), p. 23.

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
394 THE JOURNALOF POLITICS [Vol. 18

concerning his own actions and those of others with whom he inter-
acts. Thus a political system may be defined as a set of interacting
rAles, or as a structure of roles, if we understand by structure a
patterning of interactions. The advantage of the concept of role
as compared with such terms as institutions, organizations, or
groups, is that it is a more inclusive and more open concept. It
can include formal offices, informal offices, families, electorates,
mobs, casual as well as persistent groupings, and the like, in so far
as they enter into and affect the political system. The use of other
concepts such as those indicated above involves ambiguity, forced
definitions, (such as groups) or residual categories. Like the concept
of system it does not prejudice our choice of units but rather en-
ables us to nominate them on the basis of empirical investigation.
While there appear to be certain advantages in these concepts
of political system and role for our purposes, they confront the
political scientist with a serious problem. While he intends the con-
ceplt to have a general application, Parsons appears to have had
before him in elaborating the concept the model of the primary
group - family, friendship, and the like - and not complex social
systems, the units of which are collectivities and not individual
actors. In this sense the sociological concept of system and of role
can only be a beginning of a conceptual model of the political system.
The job of developing additional concepts necessary to handle
macrocosmic social systems such as political systems - national
and international- is still to be done.
My own conception of the distinguishing properties of the poli-
tical system proceeds from Weber's definition- the legitimate
monopoly of physical coercion over a given territory and population.4
The political systems with which most political scientists concern
themselves all are characterized by a specialized apparatus which
possesses this legitimate monopoly, and the political system con-
sists of those interacting roles which affect its employment. There
are, of course, simpler societies in which this function of maintenance
of order through coercion is diffuse and unspecialized; it is com-
bined with other functions in the family and other groupings.
While these systems are also properly the subject matter of poli-
tical science, there are few political scientists indeed with the spe-
cialized equipment necessary to study them.
'From Max Weber: Essays in Sociology, trans. by H. H. Gerth and C.
Wright Mills (New York: Oxford University Press, 1946), p. 78.

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
1956] COMPARATIVE POLITICAL SYSTEMS 395

It may be useful to add a few comments about this definition

of politics and the political in order to avoid misunderstanding. To
define politics as having this distinguishing property of monopoliz-
ing legitimate coercion in a given territory is not the same thing
as saying that this is all that government does. It is the thing that
government does and that other social systems ordinarily may not
do legitimately. Other social systems may employ other forms of
compulsion than physical coercion. Some indeed may legitimately
employ physical coercion on a limited scale. But the employment of
utimate, comprehensive, and legitimate physical coercion is the
monopoly of states, and the political system is uniquely concerned
with the scope, the direction, and the conditions affecting the em-
ployment of this physical coercion. It is, of course, clear that
political systems protect freedoms and provide welfare, as well as
impose order backed up by physical compulsion, but even their pro-
tection of freedom and their provision of welfare is characteristically
backed up by the threat of physical compulsion. Hence it seems
appropriateto define the political system as the patterned interaction
of roles affecting decisions backed up by the threat of physical
compulsion.
The task of describing a political system consists in characteriz-
ing all the patterned interactions which take place within it. It
takes us beyond the legal system into all the roles which occur and
involves our defining these roles in action or behavioral terms. The
concept of system implies that these roles are interdependent and
that a significant change in any one role affects changes in the
others, and thereby changes the system as a whole. Thus the
emergence of pressure groups in the present century produced
certain changes in the party system and in the administrative and
legislative processes. The rapid expansion of executive bureaucracy
was one of the factors that triggered off the development of legis-
lative bureaucracy and pressure group bureaucracy. Changes in
the role of political communication have transformed the electoral
process, the behavior of parties, the legislature, the executive. The
concepts of system and of interdependencelead us to look for these
changes when any specific role changes significantly. It suggests
the usefulness of thinking at the level of the system and its inter-
dependence rather than in terms of discrete phenomena or only

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
396 THE JOURNAL OF POLITICS [Vol. 18

limited bilateral relationships, or relationships occurring only within

the formal-legal role structure.
The fourth concept is orientation to political action. Every
political system is embedded in a set of meanings and purposes.
We speak of "attitudes toward politics," "political values," "ideo-
logies," "national character," "cultural ethos." The difficulty with
all these terms is that their meanings are diffuse and ambiguous.
The concepts of orientation to action and of the pattern variables
are useful since they at least attempt logical distinctness and
comprehensiveness. It is not essential for my purposes to go into
the modes of orientation of action, or into the "pattern variables" in
detail. Parsons and Shils tell us that any orientation to politics
involves three components: the first is perception, or cognition;
the second is preference, involvement, or affect (cathexis); the
third is evaluation or choice through the application of standards
or values to the cognitive and affective components. By cognition
is meant the knowledge and discrimination of the objects, events,
actions, issues, and the like. By cathexis is meant the investment
of objects, issues, etc., with emotional significance, or affect. By
evaluation is meant the manner in which individuals organize and
select their perceptions, preferences, and values in the process of
establishing a position vis-a-vis political action.5
Every political system is embedded in a particular pattern of
orientations to political action. I have found it useful to refer to
this as the political culture. There are two points to be made
regarding the concept of political culture. First, it does not coincide
with a given political system or society. Patterns of orientation
to politics may, and usually do, extend beyond the boundaries of
political systems. The second point is that the political culture is
not the same thing as the general culture, although it is related
to it. Because political orientation involves cognition, intellection,
and adaptation to external situations, as well as the standards and
values of the general culture, it is a differentiatedpart of the culture
and has a certain autonomy. Indeed, it is the failure to give proper
weight to the cognitive and evaluative factors, and to the- conse-
quent autonomy of political culture, that has been responsible for
the exaggerations and over-simplifications of the "national charac-
ter" literature of recent years.
Parsons and Shils, op. cit., pp. 58 ff.

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
1956 ] COMPARATIVE
POLITICALSYSTEMS 397

The usefulness of the concept of political culture and its mean-

ing may perhaps be conveyed more effectively through illustration.
I would argue that the United States, England, and several of the
Commonwealth countries have a common political culture, but
are separate and different kinds of political systems. And I would
argue that the typical countries of continental Western Europe,
while constituting individual political systems, include several
different political cultures which extend beyond their borders. In
other words, they are political systems with fragmented political
cultures.
In an effort to overcome understandable resistances to the
introduction of a new term, I should like to suggest why I find
the concept of political culture more useful than the terms we
now employ, such as ideology or political party. As I understand
the term ideology, it means the systematic and explicit formulation
of a general orientation to politics. We need this term to describe
such political phenomena as these and should not reduce its speci-
ficity by broadening it to include not only the explicit doctrinal
structure characteristically borne by a minority of militants, but
also the vaguer and more implicit orientations which generally
characterize political followings. The term political party also
cannot serve our purpose, for we are here dealing with a formal
organization which may or may not be a manifestation of a political
culture. Indeed, we will be gravely misled if we try to force the
concept of party to mean political culture. Thus the commonly
used distinctions between one-party, two-party, and multi-party
systems simply get nowhere in distinguishing the essential properties
of the totalitarian, the Anglo-American,and the Continental Euro-
pean political systems. For the structure we call party in the to-
talitarian system is not a party at all; the two parties of the Anglo-
American system are organized manifestations of a homogeneous
political culture; and the multi-parties of Continental European
political systems in some cases are and in some cases are not the
organized manifestations of different political cultures.
But the actual test of the usefulness of this conceptual scheme
can only come from a more detailed application of it in developing
the special properties of the classes of political systems to which
we earlier referred.

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
398 THE JOURNAL O0 POLITICS [Vol. 18

THE ANGLO-AMERICAN POLITICAL SYSTEMS

The Anglo-American political systems are characterized by a

homogeneous, secular political culture. By a secular political cul-
ture I mean a multi-valued political culture, a rational-calculat-
ing, bargaining, and experimental political culture. It is a homo-
geneous culture in the sense that there is a sharing of political ends
and means. The great majority of the actors in the political system
accept as the ultimate goals of the political system some combina-
tion of the values of freedom, mass welfare, and security. There
are groups which stress one value at the expense of the others;
there are times when one value is stressed by all groups; but by
and large the tendency is for all these values to be shared, and
for no one of them to be completely repressed. To a Continental
European this kind of political culture often looks sloppy. It has
no logic, no clarity. This is probably correct in an intellectual
sense, since this balancing of competing values occurs below the
surface among most people and is not explicated in any very ele-
gant way. Actually the logic is complex and is constantly referred
to reality in an inductive process. It avoids the kind of logical
simplism which characterizes much of the Continental European
ideological polemic.
A secularized political system involves an individuation of and
a measure of autonomy among the various roles. Each one of the
roles sets itself up autonomously in political business, so to speak.
There tends to be an arms-length bargaining relationship among the
roles. The political system is saturated with the atmosphere of the
market. Groups of electors come to the political market with votes
to sell in exchange for policies. Holders of offices in the formal-legal
role structure tend to be viewed as agents and instrumentalities, or
as brokers occupying points in the bargaining process. The secular-
ized political process has some of the characteristics of a laboratory;
that is, policies offered by candidates are viewed as hypotheses, and
the consequences of legislation are rapidly communicatedwithin the
system and constitute a crude form of testing hypotheses. Finally,
because the political culture tends to be homogeneous and prag-
matic, it takes on some of the atmosphere of a game. A game is a
good game when the outcome is in doubt and when the stakes

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
1956 ] COMPARATIVE POLITICAL SYSTEMS 399

are not too high. When the stakes are too high, the tone changes
from excitement to anxiety. While "fun" is frequently an aspect
of Anglo-American politics, it is rarely a manifestation of Con-
tinental European politics; and, unless one stretches the definition,
it never occurs at all in totalitarian politics.

ROLESTRUCTURE
IN THEANGLo-AMERICAN
POLITICAL
SYSTEMS

The role structure in this group of political systems is (1) highly

differentiated, (2) manifest, organized, and bureaucratized, (3)
characterized by a high degree of stability in the functions of the
roles, and (4) likely to have a diffusion of power and influence
within the political system as a whole.
With regard to the first point, each one of the units- formal
governmental agencies, political parties, pressure groups and other
kinds of voluntary associations, the media of communication, and
"publics" of various kinds - pursues specialized purposes and
performs specialized functions in the system. As was already
pointed out, each one of these entities is more or less autonomous
- interdependent, but autonomous. Certainly there are striking
differences in this respect as between the United States and the
United Kingdom, but their similarity becomes clear in contrast
to the other major types of systems which will be described below.
Secondly, this role structure is manifest and on the surface. Most
of the potential "interests" have been organized and possess bureau-
cracies. Thirdly, there is in contrast to some of the other systems
a relatively high degree of stability of function in the various parts
of the structure. Bureaucracies function as bureaucracies,armies as
armies, parliaments as parliaments. The functions are not ordi-
narily substitutable as among these various institutions and organ-
izations, in contrast to some of the other systems. This is another
way of saying that the political division of labor is more complex,
more explicit, and more stable. There are, of course, striking dif-
ferences between the British and American versions in these re-
spects. For the American system is at the same time more complex
and less stable than the British. There are, for example, many more
pressure groups and types of pressure groups in the United States
for reasons of size, economic complexity, and ethnic and religious
heterogeneity. Furthermore there is more substitutability of func-

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
400 THE JOURNAL0 POLITICS [Vol. 18

tion in the Americansystem, more policy-making by pressure groups

and the media of communication,more intervention in policy-making
through the transient impact of "public moods." But again if we
are comparing the Anglo-American system with, for example, the
pre-industrial or partially industrial systems, the British and
American systems will stand out by virtue of their similarities on
the score of complexity, manifestness, and stability of role structure.
Finally the Anglo-American type of political system is one in
which there is a diffusion of power and influence. This is only
partially expressed in the formal legal phraseology of a democratic
suffrage and representative government. There is an effective as
well as a legal diffusion of power, resulting from a system of mass
communications, mass education, and representation by interest
groups. Here again the British and American versions differ sharply
in terms of formal governmental structure, the relations between
parties and pressure groups, and the system of communication and
education. The net result is a more centralized, predictable role
structure in Britain than in the United States.

THE PRE-INDUSTRIAL
POLITICAL
SYSTEMS

The political systems which fall under this very general cate-
gory are the least well-known of all four of the classes discussed
here. But despite our relative ignorance in this area and our
inability to elaborate the many sub-types which no doubt exist, a
discussion of this kind of political system is analytically useful
since it presents such a striking contrast to the homogeneous,
secular political culture, and the complex and relatively stable role
structure of the Anglo-Americanpolitical system.
The pre-industrial- or partially industrialized and Western-
ized - political systems may be best described as mixed political
cultures and mixed political systems. Nowhere does the need for
additional vocabulary become clearer than in the analysis of these
systems; for here parliaments tend to be something other than
parliaments, parties and pressure groups behave in unusual ways,
bureaucracies and armies often dominate the political system, and
there is an atmosphere of unpredictability and gunpowder surround-
ing the political system as a whole.

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
1956 ] COMPARATIVE POLITICAL SYSTEMS 401

Some clarity is introduced into the understanding of these

systems if one recognizes that they are embedded in mixed political
cultures. What this means is that as a minimum we have two
political cultures, the Western system with its parliament, its elec-
toral system, its bureaucracy and the like, and the pre-Western
system or systems. In countries such as India there are many
traditional political cultures which intermingle with the Western
system. What kind of amalgam emerges from this impingement of
different political cultures will depend on at least five factors: (1)
the type of traditional cultures which are involved; (2) the aus-
pices under which Westernization has been introduced (e.g., West-
ern colonial powers, or native elites); (3) the functions of the
society which have been Westernized; (4) the tempo and tactics
of the Westernization process; (5) the type of Western cultural
products which have been introduced. As a consequence of this
impingement of the Western and traditional political cultures, there
is a third type of political culture which frequently emerges in this
type of system; what in Max Weber's language may be called a
charismatic political culture. It often happens as a consequence of
the erosion of a traditional political culture that powerful forces
are released - anxieties over the violation of sacred customs and
relationships, feelings of rootlessness and directionlessness because
of the rejection of habitual routines. The impact of the Western
rational system on the traditional system or systems often creates
a large potential for violence. One of the typical manifestations
of this conflict of political cultures is the charismatic nationalism
which occurs so frequently in these areas and which may be in
part understood as being a movement toward accepting a new
system of political norms, or a movement toward reaffirming the
older traditional ones, often both in peculiar combinations. To
overcome the resistance of habitual routines backed up by super-
natural sanctions, the new form of legitimacy must represent a pow-
erful affirmationcapable of breaking up deeply ingrained habits and
replacing earlier loyalites. Thus, at the minimum, we must
have in these political systems the old or the traditional political
culture, or cultures, the new or the Western-rational political cul-
ture, and transitional or resultant political phenomena of one kind
or another. Needless to say, this typical mixture of political cul-
tures presents the most serious problems of communication and co-

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
402 THE JOURNAL OF POLITICS [Vol. 18

ordination. We are dealing with a political system in which large

groups have fundamentally different "cognitive maps" of politics
and apply different norms to political action. Instability and un-
predictability are not to be viewed as pathologies but as inescapable
consequences of this type of mixture of political cultures.

ROLE STRUCTURE IN THE PRE-INDUSTRIAL POLITICAL SYSTEMS

These characteristics of the pre-industrial political systems may

be brought out more clearly and systematically in an analysis
of the political role structure which is more or less characteristic.
There is first a relatively low degree of structural differentia-
tion. Political interest often tends to be latent and when it emerges
into politics often takes the form of spontaneous, violent action.
Political parties are unstable; they fragment and consolidate,
appear and disappear. There is ordinarily only a rudimentary
specialized system of communication. Unless there is a bureau-
cracy left by a Western colonial power, the bureaucratic structure
may be only partially developed.
Secondly, because of the absence of a stable and explicit role
structure, there is likely to be a high degree of substitutability of
roles. Thus bureaucraciesmay take over the legislative function, and
armies may and often do the same. A political party may pre-empt
the policy-making function, or a mob may emerge and take the
center of the policy-making stage for a brief interval. In other
words, in contrast to the Anglo-Americanpolitical systems, there
is no stable division of political labor.
A third and most important aspect of these political systems is
the mixing of political role structures. Thus there may be a parlia-
ment formally based on a set of legal norms and regulations; but
operating within it may be a powerful family, a religious sect, a
group of tribal chieftains, or some combination of these. These are
elements of the traditional role structure operating according to
their own traditional norms. The student of these political systems
would be greatly misled if he followed Western norms and ex-
pectations in describing such a decision-makingsystem. What would
be corruption in a Western parliament would be normatively
oriented conduct in a "mixed parliament" of the kind often found
in the regions outside of the Western-EuropeanAmerican area.

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
1956 ] POLITICAL
COMPARATIVE SYSTEMS 403

Thus such concepts as mixed political culture and mixed political

role structures may prepare the field researcher more adequately
than the accepted political science theory and terminology; for in
going to Indonesia or Thailand he will not only have in mind the
Western conception of political process and system and a concep-
tion of the appropriate rOlesof legislatures, bureaucracies, parties,
pressure groups, and public opinion, but will rather look for the
particular pattern of amalgamation of these roles with the tradi-
tional roles. His intellectual apparatuswould enable him to grapple
more quickly and more adequately with political phenomena which
he might otherwise overlook, or treat as pathologies.

TOTALITARIAN POLITICAL SYSTEMS

The totalitarian political culture gives the appearance of being

homogeneous, but the homogeneity is synthetic. Since there are no
voluntary associations, and political communication is controlled
from the center, it is impossible to judge in any accurate way the
extent to which there is a positive acceptance of the totalitarian
order. One can only say that in view of the thorough-going pene-
tration of the society by a centrally controlled system of organiza-
tions and communications, and the special way in which coercion
or its threat is applied, the totalitarian system, in contrast to the
others, tends to be non-consensual. This is not to say that it is
completely non-consensual. A completely coercive political system
is unthinkable. But if one were to place the totalitarian system
on a continuum of consensual-non-consensualit would be located
rather more at the non-consensual end of the continuum than the
others described here. Unlike the other systems where some form of
legitimacy - whether traditional, rational-legal, or charismatic
- underlies the acquiescence of the individual in the political
system, in the totalitarian order the characteristic orientation to
authority tends to be some combination of conformity and apathy.
This type of political system has become possible only in modern
times, since it depends on the modern technology of communication,
on modern types of organization, and on the modern technology
of violence. Historic tyrannies have no doubt sought this kind of
dominion but were limited in the effectiveness of their means.
Totalitarianism is tyranny with a rational bureaucracy, a monopoly

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
404 THE JOURNAL
OF POLITICS [Vol. 18

of the modem technology of communication, and a monopoly of

the modern technology of violence.

ROLESTRUCTURE
IN TOTALITARIAN
POLITICAL
SYSTEMS

I believe Franz Neumann in his Behemoth6 was one of the first

students of totalitarianism who rejected the monocratic model as be-
ing useful in understandingthese systems. He spoke of the peculiar
shapelessness of the Nazi regime, of the fact that there was no stable
delegation of power among the bureaucracy, party, the army, the
organizations of big business, and the like. He concluded, as you
recall, that there was no state under the Nazis. I believe what he
meant to say was that there was no legitimate state. Later students
of totalitarianism such as Hannah Arendt,7 Merle Fainsod,8 Carl
Friedrich,9 Alex Inkeles,1I and Barrington Moore, Jr.,LLhave been
led to similar conclusions about totalitarianism in general, or about
Soviet totalitarianism. Hannah Arendt has painted the most ex-
treme picture, which, while an exaggeration, is useful analytically.
She argues that the "isolation of atomized individuals provides
not only the mass basis for totalitarian rule, but is carried through
at the very top of the whole structure." The aim of this process of
atomization is to destroy solidarity at any point in the system
and to avoid all stable delegations of power which might reduce
the freedom of manoeuver of those at the very center of the system.
"As techniques of government, the totalitarian devices appear simple
and ingeniously effective. They assure not only an absolute power
monopoly, but unparalleledcertainty that all commands will always
be carried out; the multiplicity of the transmission belts, the con-
fusion of the hierarchy, secure the dictator's complete independence
of all his inferiors and make possible the swift and surprising
'Franz Neumann,Behemoth: The Structureand Practice of National Social
ism: (New York: Oxford University Press, 1942), pp. 459ff.
'Hannah Arendt, The Origins of Totalitarianism (New York: Harcourt,
Brace and Company, 1951), p. 388.
8Merle Fainsod, How Russia is Ruled (Cambridge: Harvard University
Press, 1953), pp. 354 ff.
9Carl J. Friedrich, ed. Totalitarianism (Cambridge: Harvard University
Press, 1954), pp. 47ff.
?Alex Inkles in ibid., pp. 88 ff.
"Barrington Moore, Jr., Terror And Progress USSR: Some Sources of
Change and Stability in the Soviet Dictatorship (Cambridge:Harvard Uni-
versity Press, 1954)), pp. 154 ff.

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
1956 ] COMPARATIVE POLITICAL SYSTEMS 405

changes in policy for which totalitarianism has become famous."'2

There are thus at least two distinctive characteristics of the
totalitarian role structure: (1) the predominance of the coercive
roles, and (2) the functional instability of the power roles
bureaucracy, party, army, and secret police. The predominance of
the coercive role structure is reflected in its penetration of all of
the other role structures. Thus all forms of organization and com-
munication become saturated with a coercive flavor. This pre-
dominance of coercion is reflected in the celebrated definition of
the state as "bodies of armed men" in Lenin's State and Revolution.
It is also reflected in the doctrine of the "potential enemy of the
state," a conception under which almost any behavior may be
arbitrarily defined as disloyal behavior. This eliminates the pre-
dictability of the impact of coercion and renders it an omnipresent
force, however limited its application may be in a quantitative
sense.
The functional instability among the power roles has as its
main purpose the prevention of any stable delegation of power,
and the consequent diffusion of power and creation of other power
centers. This pattern was apparently quite marked in the develop-
ment of the Nazi regime and has been observable in the uneasy
balance established in the Soviet Union between party, bureaucracy,
army, and secret police. In the nature of the case there must be a
stabler delegation of power among the economic allocative roles, but
even these roles are penetrated by the coercive role structure and
manipulated within limits. A third class of roles is illustrated by the
electoral process and the representative system, as well as the
practice of "self-criticism" in the party. While there is a set of
norms under which these activities are supposed to influence power
and policy-making, they are rather to be understood as mobilizing
devices, as devices intended to create a facade of consent.

THE CONTINENTALEUROPEAN POLITICALSYSTEMS

We refer here primarily to France, Germany, and Italy. The

Scandinavian and Low Countries stand somewhere in between the
Continental pattern and the Anglo-American. What is most marked
about the Continental European systems is the fragmentation of
"Arendt, op. cit., p. 389.

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
406 THE JOURNAL OF POLITICS [Vol. 18

political culture; but this fragmentation is rather different from

that of the non-Western systems. For in the non-Western systems
we are dealing with mixed political cultures involving the most
striking contrasts. The Western political culture arising out of
a very different development pattern is introduced bodily, so to
speak, from the outside. In the Continental European systems we
are dealing with a pattern of political culture characterized by an
uneven pattern of development. There are significant survivals,
"outcroppings,"of older cultures and their political manifestations.
But all of the cultural variations have common roots and share a
common heritage.
In view of this developmental pattern it may be appropriate to
speak of the Continental European systems as having political sub-
cultures. There is indeed in all the examples of this type of system
a surviving pre-industrial sub-culture (e.g., the Catholic Ancien
Regime areas in France, Southern Italy, and the Islands, and parts
of Bavaria). The historical backgroundof all three of these systems
is characterizedby a failure on the part of the middle classes in the
nineteenth century to carry through a thorough-goingsecularization
of the political culture. Thus another political sub-culture in these
political systems constitutes remnants of the older middle classes
who are still primarily concerned with the secularization of the
political system itself. A third group of political sub-cultures is
associated with the modernized and industrialized parts of these
societies. But because they emerged in an only partially secularized
political culture, their potentialities for "political market" behavior
were thwarted. As major political sub-cultures there are thus these
three: (1) the pre-industrial, primarily Catholic components, (2)
the older middle-class components, and (3) the industrial com-
ponents proper. But the political culture is more complex than
this. Since in the last century the political issues have involved
the very survival of these sub-cultures, and the basic form of
the political system itself, the political actors have not come to
politics with specific bargainable differences but rather with con-
flicting and mutually exclusive designs for the political culture and
political system. This has involved a further fragmentation at the
level of ideology and political organizations. Thus the pre-industrial,
primarily Catholic element has both an adaptive, semi-secular wing
and an anti-secular wing. The middle classes are divided into

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
1956] COMPARATIVE
POLITICAL
SYSTEMS 407

conservative wings in uneasy alliance with clerical pre-republican

elements, and left-wings in uneasy friendship with socialists. Final-
ly, the industrial workers are divided according to the degree of
their alienation from the political system as a whole. The organized
political manifestations of this fragmented political culture take the
form of "movements" or sects, rather than of political parties. This
means that political affiliation is more of an act of faith than of
agency.
Perhaps the most pronounced characteristic of the political role
structure in these areas is what one might call a general alienation
from the political market. The political culture pattern is not
adapted to the political system. For while these countries have
adopted parliaments and popular elections, they are not appro-
priately oriented to these institutions. The political actors come
to the market not to exchange, compromise, and adapt, but to
preach, exhort, convert, and transform the political system into
something other than a bargaining agency. What bargaining and
exchanging does occur tends to take the form of under-the-counter
transactions. Thus demoralization ("transformism") is an almost
inescapable consequence of this combination of political culture and
system. In contrast, the normatively consistent, morally confident
actor in this type of political system is the militant who remains
within the confines of his political sub-culture, continually reaffirms
his special norms, and scolds his parliamentarians.
This suggests another essential characteristic of this type of
role structure, which places it in contrast to the Anglo-American.
There is not an individuation of the political roles, but rather the
roles are embedded in the sub-cultures and tend to constitute
separate sub-systems of roles. Thus the Catholic sub-culture has
the Church itself, the Catholic schools, propaganda organizations
such as Catholic Action, Catholic trade unions, or worker organiza-
tions, a Catholic party or parties, and a Catholic press. The Com-
munist sub-culture - the sub-culture of the political "alienates"
- similarly has a complete and separate system of roles. The
socialist and "liberal" sub-cultures tend in the same direction but
are less fully organized and less exclusive. Thus one would have
to say that the center of gravity in these political systems is not
in the formal legal role structure but in the political sub-cultures.
Thus "immobilism"would appear to be a normal property of this

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
 408 THE JOURNAL OF POLITICS [Vol. 18

kind of political system, and it is not so much an "immobilism" of

formal-legal institutions as a consequence of the condition of the
political culture. Needless to say, this portrayal of the Continental
European political system has been exaggerated for purposes of
contrast and comparison.
Two other general aspects of the role structure of these coun-
tries call for comment. First, there is a higher degree of substitu-
tability of roles than in the Anglo-American political systems and
a lessor degree than in the non-Western systems. Thus parties
may manipulate pressure groups in the sense of making their
decisions for them (the Communist case); interest groups such as
the Church and Catholic Action may manipulate parties and trade
unions; and interest groups may operate directly in the legislative
process, although this last pattern occurs in the Anglo-American
system as well. The "immobilism" of the formally political organs
often leads to a predominance of the bureaucracy in policy-making.
A second general characteristic, which is a consequence of the
immobilism of the political system as a whole, is the ever-present
threat of what is often called the "Caesaristic" breakthrough. As
in the non-Western area, although the situations and causes are
different, these systems tend always to be threatened by, and
sometimes to be swept away by, movements of charismatic national-
ism which break through the boundaries of the political sub-
cultures and overcome immobilism through coercive action and
organization. In other words, these systems have a totalitarian
potentiality in them. The fragmented political culture may be trans-
formed into a synthetically homogeneous one and the stalemated
role structure mobilized by the introduction of the coercive pattern
already described.

* * *

In conclusion perhaps the point might be made that conceptual

and terminological growth in the sciences is as inevitable as the
growth of language itself. But just as all the slang and neologisms
of the moment do not find a permanent place in the language, so
also all of the conceptual jargon which the restless minds of scholars
invent - sometimes to facilitate communication with their col-
leagues and sometimes to confound them - will not find its

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions
1956 ] COMPARATIVE
POLITICALSYSTEMS 409

permanent place in the vocabulary of the disciplines. The ultimate

criterion of admission or rejection is the facilitation of under-
standing, and this, fortunately enough, is not in the hands of the
restless and inventive scholar, but in the hands of the future schol-
arly generations who will try them out for "fit." If I may be
permitted to conclude with a minor note of blasphemy, it may be
said of new concepts as it was said of the salvation of souls . . .
"there shall be weeping and gnashing of teeth, for many are called
but few are chosen."

This content downloaded from 68.44.150.245 on Mon, 26 May 2014 02:40:57 AM

All use subject to JSTOR Terms and Conditions