Political Witch Hunts: The Sacred and the Subversive in Cross-National Perspective

Author(s): Albert James Bergesen

Source: American Sociological Review, Vol. 42, No. 2 (Apr., 1977), pp. 220-233
Published by: American Sociological Association
Stable URL: http://www.jstor.org/stable/2094602
Accessed: 03-01-2016 06:35 UTC

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 support@jstor.org.

Sage Publications, Inc. and American Sociological Association are collaborating with JSTOR to digitize, preserve and extend
access to American Sociological Review.

http://www.jstor.org

This content downloaded from 134.129.120.3 on Sun, 03 Jan 2016 06:35:33 UTC
All use subject to JSTOR Terms and Conditions

POLITICAL WITCH HUNTS:

THE SACRED AND THE SUBVERSIVE IN
CROSS-NATIONAL PERSPECTIVE*
ALBERT JAMES BERGESEN
University of A rizona
American Sociological Review 1977, Vol. 42 (April):220-233
This paper proposes a general theory of political witch hunts, viewing them as ritual mechanisms for the periodic rejuvenation of collective sentiments in national societies. Ultimate
national purposes require not only their worshipers, but also their enemies. When these sacred
forces penetrate daily reality, then their opposites-subversives-will also appear in daily institutional life. The corporateness of societies, as expressed in their political system, is theoretically
linked to the penetration of transcendent reality into daily life, and a witch-hunting dispersion
index is proposed to measure the extent to which subversion is ritually discovered throughout a
society's social institutions. The overall rate of witch hunting is also measured. Data on rates
of political it-itch hunting between 1950 and 1970 for 39 countries is presented to evaluate
the general theoretical argument. The data suggest that as societies politically express more
of their corporate national interest they ritually cleanse more institutional areas, as measured by the dispersion index. A long wit/ the representation of the corporate national interest, the overall rate of witch hunting is significantly affected by country size, lev el of economic development and the relative power of the state. The dispersion of uw'itchhunting, on
the other hand, is unaffected by these control variables and seems to be a more purely
Durkheimian phenomenon.

Modem national societies, like primitive societies, must periodicallyrenew the

meaning of corporate existence. Where
rites were once performedto symbolizations of the tribe, they are now performed
to symbolizations of the nation-state.
Durkheim (1965) observed that sacredness required profanity, and sacred national purposes seem to requirethose who
would undermineand subvertthem. Similarly, as Durkheimargued that crime is a
normalaspect of social life, so are political witch hunts. The nation-state'screation of political subversives, no less than
the community's manufacture of deviance, is a mechanism for renewing
common moral sentiments and redefining
the contours of social reality.
lIhistheoreticalperspective on political
witch huntingderivesfrom and links Durk* This is a revised version of a paper read at the
American Sociological Association meetings,
Montreal, 1974. Janine Blair assisted with the computerwork andcommentsfrom MorrisZelditch,Jr.,
John W. Meyer and Ronald Schoenberghave been
most helpful. I especially want to thank Beverly
Duncan and Otis Dudley Duncan for their suggestions and assistance on an earlierdraftof this paper.

heim's observations on religious ritual

and myth and the social functions of
crime. His work in religionand crime has
a common concern with the function of
ritual activity in creating and maintaining
collective reality, whether symbolic representations in the analysis of primitive
religion or moral boundariesin the analysis of crime. However, there has been little intellectual contact between later students of the sociology of religion (for
example, Swanson, 1964; 1967; Douglas,
1966; 1970;Goffman, 1956;Berger, 1969;
Luckmann, 1967)and the sociology of deviance (Erikson, 1966). Moveover, this
theoretical formulationis congruous with
two empirical findings which recur in
studies of political trials (Cohen, 1971;
Kirchheimer, 1961), purges (Brzezinski,
1956; Conquest, 1968; Dallin and Breslauer, 1970), rectification campaigns
(MacFarquhar, 1960; Goldman, 1965;
Solomon, 1971; Baum and Teiwes, 1968)
and loyalty controversies (Bell, 1964;Lipset, 1955; Parsons, 1955). Efforts at exposing, unmasking and discovering subversives seem oriented as much toward
dramatizingtheir presence as toward ac-

220

This content downloaded from 134.129.120.3 on Sun, 03 Jan 2016 06:35:33 UTC
All use subject to JSTOR Terms and Conditions

POLITICAL WITCH HUNTS

tually apprehendingthem; and, accordingly, the arrests, trials, confessions and

other acts which comprisea modernpolitical witch hunt seem largely ritualisticin
character.
The ritual mobilizationof a community
searching for imaginary enemies traditionally has been viewed as largely
epiphenomenal-the real causes of witch
huntingare thought to lie elsewhere. The
trials, purges and political terrorof Communist regimes, for instance, have been
seen as an outcome of the fear and
paranoia generated by totalitarianstates
and as an "instrumentof the state" used
by elites to thwart potential opposition
and maintainpolitical power (Brzezinski,
1956). Similarly, the hysteria of the
McCarthy period has been dismissed as
the paranoidprojections of social groups
experiencing status stress (Bell, 1964). In
contrast, I want to propose a theory of
political witch hunts explicitly based on
their distinctly mythical and ritualistic
character.

221

Durkheim (1965) called positive rites

where the Nation and the People are worshipped: coronations, inaugurations,nationalholidays and memorialdays. Others
are negative rites, like witch hunts.
Throughsuch activities as the purge, trial,
investigation, accusation, arrest and imprisonment, society creates its own internal enemies to ritually reaffirmthe very
sacred national purposes which subversives are supposedly undermining. The
corporate nation needs not only its worshipers, but also its enemies. This process
of moralrevitalizationis complex, centering on the interrelationshipof corporate
society, its collective representationsand
the ritualwhich rejuvenatesand redefines
them.
The Penetration of Ordinary Life by
Sacred Forces

Social realityis a matterof definition.It

is neither naturally mundane, with ordinary people and propanemotives, nor sacred, with transcendentpolitical purposes
individuals and social events.
animating
THEORY
Depending upon how closely a nation's
collective myths are mergedwith the proThe Sacred and the Subversive
fane and ordinary,more or less ultimate
For Durkheim,the moral reaffirmation political significance can be infused into
of collective life involved two causally in- daily existence. The penetration of daily
terrelatednotions-one, that society pre- life by sacred forces is, in effect, a varisents itself througha varietyof symboliza- able, and not the singular property of
tions or collective representationsranging primitive religious systems as we have
from materialobjects to systems of ideas; traditionallybelieved.'
the other, that society mobilizes itself
In pluralisticWestern societies, by and
throughrites performedto these symbolizations as a means of periodicallyrenewI Bellah (1970:27), writing on religious evolution,
ing their presence and simultaneouslyre- argues
that one of the distinguishing characteristics
newing the larger social order they sym- of primitive religion is "the very high degree to
bolically represent. Like primitive which the mythical world is related to the detailed
societies, modem national societies pre- features of the actual world. Not only is every clan
and local group defined in terms of the ancestral
sent themselves through numerous sym- progenitors
and the mythical events of settlement,
bolizations. Some are materialobjects like but virtually every mountain, rock, and tree is exflags, emblems and political leaders. plained in terms of the actions of mythical beings."
Others are symbolic representations of The same could be said for the so-called modern
the nationalcollectivity itself - images of religious situation. In certain highly ideological
societies, for example, virtually every individual acthe Nation of the People as organic en- tion
is linked to the transcendent world of History or
or
such
as
tities,
political ideologies
Nature and every event is seen as an instance of the
Communism, Socialism, Fascism or De- mythical forces of "imperialism," "socialism" or
mocracy. As in primitive societies, rites "capitalism." Because our religious representations
are at the same time the theory of our own universe,
are performedto these modern collective we
have had a very difficult time assigning some
representationsof the corporate national beliefs to the category "myths" and others to what
community. Some of these rites are what we take as constituting our "real" reality.

This content downloaded from 134.129.120.3 on Sun, 03 Jan 2016 06:35:33 UTC
All use subject to JSTOR Terms and Conditions

222

AMERICAN SOCIOLOGICAL REVIEW

large, life is understoodas the interaction

of groups and personages governed by
secular and profane motivations of the
here and now. Life is not filled with the
rumblingand movement of Marx's historically preordainedstrugglebetween social
forces in "constant opposition to one
another, carrying on an uninterrupted,
now hidden, now open fight." Conversely, the two most pronounced examples of sacred political forces mingling
with people in daily life are Fascist and
Communistsocieties where there is a very
close connection between religious representations and daily existence.

daily activity becomes the realization of

these transcendent realities. The ritualistic creation of enemies, as a means of
renewing the presence of these sacred
forces, centers on discovering "enemies
of the people," the nation, and even
enemies of Nature and History itself.
Terror... its chief aim is to make it possible
for the force of nature or of history to race
freely through man . . . [it] singles out the
foes of mankindagainst whom terror is let
loose, and no free action of eitheropposition
or sympathy can be permitted to interfere
with the elimination of the "objective
enemy" of History or Nature, of the class or
the race. Guiltand innocence become senseless notions; "guilty" is he who standsin the
Underlyingthe Nazis' belief in race laws as
way of the natural or historical process
the expressionof the law of naturein man, is
which has passed judgement over "inferior
Darwin's idea of man as the product of a
races," over individuals "unfit to live,"
naturaldevelopment which does not necesover "dying classes and decadent peoples."
sarilystop with the presentspecies of human
Terror executes these judgements, and bebeings, just as under the Bolsheviks' belief
fore its court, all concerned are subjectively
in class-struggleas the expression of the law
of historylies Marx'snotionof society as the
innocent: the murdered because they did
product of a gigantic historical movement
nothing againstthe system, and the murderwhich races accordingto its own law of moers because they do not really murder but
tion to the end of historicaltime when it will
execute a death sentence pronounced by
abolish itself. (Arendt, 1973:463)
some higher tribunal.The rulers themselves
do not claim to be just or wise, but only to
These ideologies can be consideredDurkexecute historical or natural laws; they do
heimian representationsof the corporate
not apply laws, but execute a movement in
social order. Individualscome and go but
accordancewith its inherent law. Terror is
groups persist: the idea of an unending
lawfulness, if law is the law of the moveuniverse of evolving Nature or History is
ment of some suprahumanforce, Nature or
the perfect symbolization for the corpoHistory. (Arendt, 1973:465)

rate continuity of society over the temporary lives of mundane individuals. From
the point of view of highly corporate
societies, those countries in which groups
and classes do not realize their participation in the transcendentreality of historical development are said to have "false
consciousness." From the other side,
these corporate societies are described as
highly "ideological" and their people as
"brainwashed." Each side claims its experience of reality to be correct and the
other's false. What is experienced,
though, is mediatedby each society's definition of reality, and there is no absolute
reality independent of society and its imposed system of classificationsand definitions.
Political ideologies, like the sacred
forces of History and Nature, or the ideas
of the People or the Nation can penetrate
and merge with ordinary reality so that

Ritual Transformations: Trial, Purge,

Accusation, Investigation
There is no clear line separating ordinary mundane reality and the larger purposes and personages of a transcendent
reality such as History or Nature. One can
see himself, his work and others as the
embodiment of the laws of history working themselves out here on earth, as when
factory work is experienced as building
socialism and realizing the historical role
of the proletariat. One can also find oneself engaged in mortal combat with these
cosmic forces and being charged with
being an enemy of socialism and conspiring against the laws of social development. In the first situation, one is marching with the sacred forces that have penetrated one's existence and, in the other,
one is marching against them. In both

This content downloaded from 134.129.120.3 on Sun, 03 Jan 2016 06:35:33 UTC
All use subject to JSTOR Terms and Conditions

POLITICAL WITCH HUNTS

situations, one will probably be doing

identical activities-only the definitions
will have changed.
Political witch hunts are the ritual
mechanisms that transform individuals,
groups, organizationsor culturalartifacts
from things of this world into actors
within a mythical universe. These rituals
are the social "hooks" that keep sacred
transcendentforces present in the lives of
ordinarypeople and relevantfor everyday
institutional transactions. In effect,
Berger's (1969) "Sacred Canopy" extends to different lengths, more or less
penetrating everyday reality, with ritual
being the "lock" which keeps sacred reality tied to daily reality. The discovery of
"wreckers" in factories, "hostile elements" in the Party bureaucracy,
"bourgeois thoughts" in literature, and
"anti-state" activity in governmentis the
ritual activity which functions to reaffirm
the presence of the sacred struggle between Capitalismand Socialismwithinthe
fabric of everyday life.
The crimes one is chargedwith and the
motives imputed are not of this world.
Through some ritual transformational
logic, they become part of the mythical
reality of political ideology where, for
example, "proletarianvirtue" is battling
"bourgeois selfishness" and the
"socialist and capitalist lines" are struggling for supremacy.Trivialactivityis ritually transformedinto actions of large historical forces. Using one kind of fertilizer
or another on a communalfarm is transformed into the mythicalworld of "taking
the capitalist road" or "following the
socialist road." Readinga book or seeing
a play is transformed into being
"poisoned by bourgeois thoughts" or
being a "communist sympathizer." Correspondence with friends abroad can be
"restoring
transformed into acts
capitalism," making one an "imperialist
agent" and "counterrevolutionary."
When these ritual convulsions occur,
ordinary reality loses its usual meaning
and human beings mingle with mythical
beings, playing roles in a cosmic drama,
As HannahArendtobserved, questions of
guilt and innocence are irrelevant; they
are causal relationsrooted in the structure
of meaning of mundanereality and make

223

no sense when the cletinition of reality has

been ritually shifted to the mythical world
of political ideology. What one "really"
did or did not do is irrelevant, for the
"doing" has meaning in a world that has
been supplanted by a mythical universe
with its own characters, motives and
crimes. Being an "ultra-leftist," "ultrarightist," or "left in appearance but right
in essence" makes no sense within the
meaning structures of this world; they are
acts, motives and types of people from the
world of political ideology.
One does not enter this mythical world
through any act one performs. The sacred
and profane are two different worlds,
there is nothing one can do in one world
to enter into the other. Hence, guilt
and innocence have never mattered during political witch hunts. The most trivial
infraction, or no action at all, has the same
weight as the most serious crime. It is the
activity of the trial, the purge, the accusation or the self-confession which transforms individuals from one reality to the
other.
It is no wonder then that witch hunts
appear irrational, terrifying and unreal. In
some sense, they are truly unreal and irrational, for their logic derives from the
symbolic significance of the ritual encounters between mythical beings and forces
and not from the actualities of human
conduct. There is no intrinsic quality to
that which is treated as sacred and there is
no intrinsic quality to individuals considered subversive and dangerous. Durkheim
wondered how such insignificant things as
lizards, frogs, turkeys, ants and caterpillars could, in and of themselves, engender
the sense of the sacred; and we have also
wondered how the wording of a Girl Scout
Handbook, a play exploring the thoughts
of a conscientious objector, or interest in
the United Nations, race relations and
civil liberties could be considered dangerous, subversive and un-American. Anything can serve as a vehicle for the designation "sacred" or "subversive," and
anyone can become "an enemy of the
people" and "think dangerous thoughts."
As Arendt (1973:462) observed:
Totalitarianlawfulness,defying legality and
pretending to establish the direct reign of
justice on earth, executes the law of History

This content downloaded from 134.129.120.3 on Sun, 03 Jan 2016 06:35:33 UTC
All use subject to JSTOR Terms and Conditions

224

AMERICAN SOCIOLOGICAL REVIEW

or of Nature without translating it into

standardsof right and wrong for individual
behavior. It applies the law,directly to mankind without botheringwith the behavior of
men.

are they embodied in specific political institutions to the extent to which the
CommunistParty is considered the literal
embodimentof the "proletarianwill." Finally, the American People or Public
Opiniondo not have a mythicallink to the
specifics of historical evolution as does
Societal Differences in Witch-Hunting
the
Communist's Proletariat. The ManActivity
ifesto stated, "The history of all hitherto
Collective representations mirror the society is the history of class struggles.
social order, and Durkheim'sobservation Freemanand slave, patricianand plebian,
that the attitude of respect toward our lord and serf, guildmaster and jourgods is similar to the attitude of respect neyman . . ." and now, in our times, the
toward social authorityprovides the now bourgeoise and proletariat. In compariwell-known (Swanson, 1964; 1967) theo- son, the "When in the course of human
retical linkage between corporate social events . . ." beginning to the Declaration
groups and religious experienceswith sa- of Independence is a much more casual
cred forces and spirits.
and almost offhand reference to the
As mentioned earlier, tMe extent to primordialorigins of the American nation
which transcendent reality merges with compared with the lockstep progression
daily reality appearsto be a variableand, of History in the Manifesto that has
as these sacred forces are symbolic repre- broughtforth present-day Socialist counsentations of the corporate social order, tries. There is also no single body of literavariationin the extent to which that cor- ture for the liberal democraticWest comporateness is expressed should be re- parable to that of Marxismdefining, say,
flected in variationsin the extent to which America's "Historical Role." There are
the sacred cosmos penetrates mundane no "sacred texts" and no common set of
reality.2 The more the corporate interest intellectual heroes like Marx, Lenin and
of the nation as a whole is politically ex- Mao. Marxism provides one version of
pressed, the more sacred forces should ultimatereality, which is modifiedto each
intervene in daily life, and the more particularsocial system, but is still linked
witch huntingthere should be to reaffirm to the transcendentprocess of social dethese collective representations and se- velopment and the laws of History.
cure their presence in daily affairs. In efExpressing corporateness: political
fect, as the corporateinterest of the soci- party systems. Swanson (1967; 1971) arety is weakened, so is the strengthof the gues that social collectivities have a corgods. They become less powerful, more porate existence; they can make collecelusive and more tenuously tied to the tive decisions and take collective action.
For national societies, the structural
specifics of daily life.
For example, representations of the mechanismthroughwhich collective decicorporatenessof the United States are not sions are formulatedand collective action
as well developed as those of Communist taken is the institution of government.
countries. The idea of the American Collectivities vary in the degree to which
People, or Public Opinion, as a force or they allow the expression of the corporate
spirit in our lives does not penetrate our interest of the collectivity as a whole, as
existence to the extent to which repre- opposed to the interests of the constituent
sentationslike "the thoughtsof Chairman groups within the society. Political party
Mao" do in China (Schwartz, 1968). Nor systems are the most common structural
mechanismfor expressinggroupinterests,
2 This theoretical discussion leans heavily on the
whether corporate or constituent.3
work of Swanson (1964; 1967) on corporate groups
and experiences with the sacred in daily life. His
ideas on immanence and constitutional systems are
the backbone of my thinking about these matters,
and those familiar with his work will see the strong
effect he has had on my thinking.

I The expression of corporate versus constituent

interests need not always be mutually exclusive, although it most often appears that way. One exception would be a country like France which has a

This content downloaded from 134.129.120.3 on Sun, 03 Jan 2016 06:35:33 UTC
All use subject to JSTOR Terms and Conditions

225

POLITICAL WITCH HUNTS

Multi-party systems, such as the European parliamentarystates, allow the most

penetrationof constituent group interests
into the structureof government. Specific
groups-agricultural, religious, working
class-each have their own party and,
with the proportionalrepresentationelectoral arrangements multi-party systems
usually have, these parties are virtually
guaranteed some seats in the legislature.
Two-partyarrangements,like the AngloAmerican countries, collapse many
specific group interests into two broadly
based parties. The emphasis is upon what
the different groups have in common,
rather than what separates them, and
more of the common corporateinterest of
the society is manifested. Finally, there
are one-partystates, whetherCommunist,
Fascist or one-party nationalist. Here,
only the interest of the collectivity as a
whole is represented by the single party.
Partialinterests are not provideda formal
role throughpartiesin the politicalorganization of these nations as corporate entities.
Some Hypotheses

The penetration of sacred forces into

ordinaryreality is not the sole propertyof
primitive belief systems, and the ritualistic search for enemies as a means of reaffirmingthese transcendentforces is not a
propertyof any one kind of society; both
vary accordingto the extent to which corporate, as opposed to constituent, group
interests are politically expressed. We
have been referring to daily affairs and
everyday reality in quite general terms.
We can refer to them more precisely in
terms of the differentinstitutionalspheres
of which daily reality is composed. We
can speak of the extent to which religious,
political, economic or educationalinstitutions are found infested and pollutedwith
subversion.
From the preceding theoretical discussion, the following hypotheses can be
multi-partysystem allowing for the expression of
constituent group interests and a strong and centralized national bureaucracywhich expresses the
corporateinterests.

made: (1) Other things being equal, there

should be a positive relationship between
the expression of corporate interests and
the distribution of witch hunting across a
society's social institutions, reflecting the
more extensive penetration of daily institutional reality by sacred forces. (2) We
should also expect the expression of society's corporate interest to be positively
related to increased frequency of witch
hunting in general. One-party states
should experience subversion in more institutional areas and have a higher overall
rate of witch hunting than two-party
states, and they, in turn, should experience subversion in more institutional
areas and have a higher overall rate than
multi-party states.
DATA AND

METHOD

Independent Variables
Countries were chosen for analysis if
they had a stable party system from 1950
through 1970. A stable system is defined
as one in which the party system did not
change and the country was not involved
in major political conflicts. The attempt
was to isolate, as much as possible, the
effect of party system upon a country's
propensity tor witch hunting. The countries are listed in Appendix 1. A country's
party system was coded from Blondel's
(1973) classification of national legislatures. Countries were coded one, two and
three, representing one-, two- and multiparty systems. The 39 countries differ in
population size, level of economic development and the degree to which political power is concentrated within the state,
as well as with respect to political party
system. The former factors may be seen
as rivals to party system in accounting for
cross-national differences in witch hunting
and, as I am arguing that other things
being equal, corporateness should be related to the dispersion and rate of witch
hunting, I have compiled measures of
each country's 1960 population (IBRD,
1973), 1960 per capita GNP (IBRD, 1973)
and per capita internal security forces
(Taylor and Hudson, 1971) to control for
these factors.

This content downloaded from 134.129.120.3 on Sun, 03 Jan 2016 06:35:33 UTC
All use subject to JSTOR Terms and Conditions

226

AMERICAN SOCIOLOGICAL REVIEW

Dependent Variables
Frequency of witch hunting. Counts of

witch huntingactivityare obtainedby coding government activities (for example,

trials, purges, arrests) from news sources
whenever these activities charge someone
with threateningor standingin opposition
to the nationalinterest. Since government
represents the social structure through
which the nation itself takes collective action, only acts by a country's national
government will be coded (this includes
the military). Charges of subversion by
private citizens, for instance, will not be
coded. The New

York Times Index,

1950-1970, was chosen as the basic

source because of its broad international
coverage.4 The scheme used for coding
provides for 19 different kinds of government acts, rangingfrom mere charges of
subversive activity to large-scale purge
trials. Each category is mutually exclusive. The specific code categories are: (1)
warnings of danger to the nation, (2)
specific charges, (3) discovery of plots, (4)
calls of public action to thwart subversion, (5) proposed actions, (6) new laws,
ordinances or executive decrees, (7) restrictions of personal activity, (8) expulsion of agents of foreign governments, (9)
deportations, (10) arrests, (11) imprisonment, (12) resignations, (13) purges, (14)
court actions, (15) mass mobilizations,
(16) censorship campaigns, (17) political
trials, (18) sentences following trials and
(19) government investigations. This
scheme is intended to cover the great
variety of government activity a society
can employ in creatingthe idea of subversion in its midst.
The coding was done by four graduate
students. Activities were assigned codes
independently by two coders and the
value of Robinson's (1957) coefficient of
4 There are numerous sources of bias involved in
using press reports. There are actions taken by political authorities in identifying and prosecuting subversives that never reach the pages of the Times.
This would seem particularly true of Communist
states and countries with little contact with the
United States. There is also undoubtedly a bias in
terms of the number of events reported for large and
prestigious nations over those smaller and less
known. For a discussion of some of the problems in
coding news sources, see Danzger (1976).

agreement was found to be .85. This compares favorably with the reliability
achieved by other researchers who have
coded news sources (see Gurr, 1968;
Feierabend and Feierabend, 1966; Banks,
1971).
The coding scheme also provides for 36
different institutional areas in which subversion might be discovered. The information available about the activities made
it impossible to determine the institutional
areas for 57 percent of the government
activities coded. For the present analysis,
these 36 areas are grouped into seven gen(1)
categories:
institutional
eral
government-national and local, (2) military personnel and facilities, (3) educational institutions and students, (4)
economy, (5) intellectuals, (6) religious
groups and institutions and (7) a category
for foreigners.
The coding procedure works as follows.
The Times Index is read for each country.
When one of the code activities (a trial,
purge, arrest) is encountered, it is coded
along with the institutional area where the
subversion was located. This procedure
generates the basic data on both the overall volume of witch hunting (the number of
government activities) and its institutional
location. Some countries could not be
coded for the full 21 years (such as the
African countries, which did not become
independent until around 1960), so each
country's total number of activities was
divided by the number of years coded,
creating a witch-hunting per year variable.
The Witch-Hunting Dispersion Index.
To measure the dispersion of witch-hunting activity throughout a society's institutional space, a population diversity index
is employed. This 'dispersion index,
adapted from Simpson's (1949) index of
population diversity, is defined as
Nj (Nj- 1)
D= 1 -E
N(N-1)
where Nj equals the number of witchhunting activities in the jth institutional
category, N equals the total number of
witch-hunting activities and D equals the
degree of dispersion of witch hunting
across institutional categories. The higher
the score on the index, the more dispersed

This content downloaded from 134.129.120.3 on Sun, 03 Jan 2016 06:35:33 UTC
All use subject to JSTOR Terms and Conditions

POLITICAL WITCH HUNTS

227

for country size (population), level of

economic development (GNP) and the
relative power of the state (internalsecurity forces) are entered into the following
regressionanalysis along with the variable
political party system. It was logged to
help correct for the skewed distributionof
the witch-huntingper year variable. The
correlationmatrixfor the regression analysis is presented in Appendices 2 and 3.
The regression analysis in Table 1
strongly supports our first hypothesis
that, other things being equal, there is a
positive relationship between party system and the dispersion of witch hunting
FINDINGS
throughouta society's institutionalspace.
The political party variable has a strong
The Dispersion of Subversion through Ineffect, with a beta of .575. It is also the
stitutional Space and the Overall Rate of
only
unstandardizedcoefficient twice its
Witch Hunting
standard error. The effect of population
We can measure the extent to which and GNP are negligiblewith betas of .047
different institutional areas are polluted and .042, respectively, and the indicator
with subversion using the dispersion of state power, internal security forces,
index mentioned earlier. The higher the has a small negative effect with a beta
score the more dispersion; the lower the of -.242.
score the more witch hunting is concenThe second hypothesis stated that,
trated within a few institutional areas. other things being equal, there is a posiAppendix 1 presents the total number of tive relationship between party system
witch-hunting activities by institutional and the overall rate of witch hunting.
area for each country. Control variables Politicalparty system was significantlyre-

the witch huntingacross institutions.This

index was computed across the
categories: government, military, education, economy, intellectuals, religion,
foreigners and agents of foreign governments. At least ten government activities
were arbitrarilychosen as a minimumto
compute an index score. No index was
computed for the following countries, as
they had less than ten acts within their
institutional areas: Chad, Guinea, Ivory
Coast, Senegal, Tanzania,Australia,New
Zealand, Colombia, Belgium, Iceland and
Ireland.

Table 1. Regression Coefficients of Witch-HuntingDisperson Index and Log Witch Hunting per Year
on Political Party System, Internal Security Forces, Population and per Capita Gross National Product
IndependentVariables
Political
Party
System

Dependent
Variables
Witch-Hunting
DispersionIndex
Log Witch-Hunting
Activities per Year

46.162
(22.917)
.698*
(.307)

Internal
Security
Forces

Population

Per
Capita
GNP

UnstandardizedCoefficientsa
-.762
.0027
.00473
(.688)
(.0118)
(.0298)
.193*
.000538*
.0009*
(.0088)
(.000174)
(.00041)

Constant

R2

774.20
(69.89)
2.675
(.950)

.271
.446

StandardizedCoefficients
Witch-Hunting
Dispersion Index
Log Witch-Hunting
Activities per Year

.575
.455

-.242
.313

.047

.042

.438

.429

Sources: Political Party System-One-, Two- and Multi-Party. Internal Security Forces-Internal
Security Forces per thousand working age population (Taylor and Hudson, 1971). Population-1960 population in millions (I.B.R.D., 1973). GNP=1960 Gross National Product, per capita (I.B.R.D., 1973).
Standard errors in parentheses.
* p<.05.

This content downloaded from 134.129.120.3 on Sun, 03 Jan 2016 06:35:33 UTC
All use subject to JSTOR Terms and Conditions

228

AMERICAN SOCIOLOGICAL REVIEW

lated (p<.05) to the overall rate of witch

hunting, but so were the control variables.
All of the independent variables were significant at the .05 level and all the unstandardized coefficients were twice their
standard errors. The overall rate does not
seem to be a simple function of the representation of corporate or constituent
interests.
One-party countries have the highest
rate of witch-hunting activity. They have a
median 6.1 log witch-hunting activities per
year compared with only 1.6 for two-party
and 1.2 for multi-party. The large score
(Appendix 1) for the United States should
be interpreted cautiously, as it is undoubtedly inflated by the very extensive coverage the Times gives the U.S. Except for
Ghana, the African one-party states have
very little activity. Most one-party witch
hunting, therefore, was conducted by
We
have
Communist
countries.
categorized countries by party system,
but we have no way of further differentiating among Communist states as to the representation of partial or corporate interests within these one-party regimes. Some
support for the applicability of the general
theoretical argument that witch hunting is
related to representing solely the collective interest at the expense of competing
or partial interests is suggested by Hannah
Arendt. She observes that, even for oneparty Communist and Fascist states, the
further elimination of competing interests
is associated with an increase in witch
hunting and political terror.
terror increased both in Soviet Russia and
Nazi Germanyin inverse ratio to the existence of internal political opposition, so
that it looked as though political opposition
had not been the pretext of terror . . . but
the last impedimentto its full fury. [In Soviet
Russia] . .. full terrordid not breakloose in
the twenties but in the thirties, when the
opposition of the peasant classes was no
longer an active factor in the situation.Khrus-chev,too . . . notes that "extreme
repressivemeasureswere not used" against
the opposition during the fight against the
Trotskyites and the Bukharinites,but that
"the repressionagainst them began" much
later after they had long been defeated.
Terror by the Nazi regime reached its
peak duringthe war, when the Germannation was actually "united." Its preparation

goes back to 1936when all organizedinterior

resistance had vanished and Himmler proposed an expansion of the concentration
camps. (Arendt, 1973:393)
To assess more systematically the importance of Communist countries per se, as
opposed to the more general Durkheimian
idea of corporateness, one would want to
consider examining other historical forms
of one-party states, developing an indicator of corporateness other than party system, or searching for a means of differentiating, among one-party states, their
relative expression of corporate and constituent interests.
The Distribution of Subversion across
Specific Institutional Areas
We can more closely examine the distribution of witch-hunting activity by looking at the percentage distribution of ritually discovered subversion in different institutional areas (see Table 2). The Govand
ernment,
Military,
Foreigners
categories generally have the highest percentages of witch-hunting activity. If we
exclude the Religion category, because of
confounding factors for many of the
newer one-party states which will be discussed later, then these three categories
account for 64.4 percent of the total in
one-party states, 74.4 percent in twoparty states and 73.9 percent in multi-party
countries. This suggests that the creation
of subversion surrounds structures and
persons who are related in some fashion
to the distinctly corporate aspect of a
country's existence. Government and military provide the organizational structure
through which nations attain their corporate existence and seem imbued with
larger political significance and, accordingly, those who would subvert those sacred national purposes. Foreigners, on the
other hand, relate to the corporate nation
in a different manner. The very definition
of foreigners, as outsiders and nonmembers of the national collectivity, is
derived from their relationship to the corporate nation as a whole.
Multi-party states discover a very large
proportion (42 percent) of their subversion among foreigners. This could reflect

This content downloaded from 134.129.120.3 on Sun, 03 Jan 2016 06:35:33 UTC
All use subject to JSTOR Terms and Conditions

229

POLITICAL WITCH HUNTS

Table 2. Percent Distribution of Witch-Hunting Activities by Institutional Area and Political Party
System
Multi-Party

Two-Party

One-Party
Institutional
Area

Percent

Percent

Government
Military
Education
Economy
Intellectuals
Religion
Foreigners

26.4
4.5
11.2
5.7
10.8
21.9
19.4

549
93
233
119
225
455
402

33.1
19.3
12.2
6.6
6.4
1.9
20.6

212
124
78
42
41
12
132

15.0
15.0
9.2
3.6
12.8
1.9
42.5

54
54
33
13
46
7
153

100.0

2076

100.0

641

100.0

360

Total

the low degree to which their institutional

structuresare imbuedwith largerpolitical
significance. These countries, as nationstates, have a corporate existence, although they do not formally express as
much of their distinctly corporate interests as two- and one-partystates. Accordingly, sacred nationalpurposes are not as
extensively infused into their institutional
structures and, consequently, they conduct much of their witch hunting outside
their institutional infrastructure, e.g.,
among foreigners.
The percentages among the general
institutional categories,
Education,
Economy and Intellectuals,are all quite
similar. There is, though, a large discrepancy between the proportion of witch
hunting centering on persons from the
general category of religious groups and
institutions for one-party countries (22
percent) and two- and multi-partycountries (2 percent). This high percentagefor
one-party states could reflect the churchstate struggles involved in the nationbuilding process of those new states
which emerged following the Second
World War. Partial evidence for this is
found in comparingthe amount of witch
hunting which occurred during the early
1950-1955period for new and older oneparty states, with the latterhavingalready
passed through many of the pangs of
nation-building.The new states (China,
East Germany, Albania, Bulgaria,
Czechoslovakia, Hungary, Poland,
Rumania and Yugoslavia) averaged 73
their within-religiouspercent of
institutions witch hunting during the
1950-1955period. One-partystates estab-

Percent

wishedbefore 1945 (the USSR, Spain and

Portugal) averaged only seven percent of
their within-religious-institutions activity
during this early period. The new states
also accounted for some 86.6 percent of
all one-party witch hunting around religious institutions which further suggests
that most of this activity was tied to the
problems of newly emerging states.
Subversion within Government
We also can examine the proportion of
witch hunting within the four subcategories composing the general area of
government. Table 3 shows that the vast
proportion of witch hunting in government
is concentrated within national bureaucracies (one-party, 63.4 percent; twoparty, 78.8 percent; multi-party, 71.1 percent). There are, though, some interesting
differences. As the political representation of the corporate national interest is
compromised with the increasing representation of constituent group intereststhat is, moving from one- to multiparty-the proportion of subversion discovered around the national executive officer, who represents the interests of the
collectivity as a whole, gradually decreases (one-party, 20 percent; two-party,
10.4 percent; multi-party, 1.9 percent).
The inverse relationship holds for the
proportion of subversion discovered
within national legislatures. As the representation of constituent interests increases, the proportion in legislatures increases (one-party, 4.3 percent; twoparty, 6.1 percent; multi-party, 11.3 percent). It seems that where the expression

This content downloaded from 134.129.120.3 on Sun, 03 Jan 2016 06:35:33 UTC
All use subject to JSTOR Terms and Conditions

AMERICAN SOCIOLOGICAL REVIEW

230

Table 3. Percent Distribution of Witch-Hunting

System

National Executive
Officer
National Bureaucracy
National Legislature
Local Government
Total

by Political Party
Multi-Party

Two-Party

One-Party
Government

Activities within Government

Percent

Percent

Percent

20.0
63.4
4.3
12.3

70
222
15
43

10.4
78.8
6.1
4.7

22
167
13
10

1.9
71.7
11.3
15.1

1
38
6
8

100.0

350

100.0

212

100.0

53

of the nation's corporate interest is

paramount (one-party) the office which
stands for the collectivity as a whole, the
nationalexecutive officer, is the source of
more subversiveactivity (20 percent)than
the legislature(4.3 percent), which is the
structure representing constituent group
interests. Conversely, where the collectivity is structuredso that constituent interests are expressed at the expense of the
corporate national interest (multi-party),
then legislatureshave a higherproportion
(11.3 percent) than the nationalexecutive
officer (1.9 percent), who represents the
now-compromisedcorporate interest. Finally, two-party countries express less
corporate interests than one-party but
more than multi-party.Consequently,they
have more witch hunting around their
executive officer than multi-party, but
less than one-partycountries, and more in
their legislatures than one-party, but less
than multi-partycountries.
SUMMARY AND CONCLUSION

Althoughthe most extreme instances of

political terror and witch hunting are associated with totalitarian regimes, the
ritualistic search for imaginary enemies
should be conceptualized as a variable,
not the singular property of totalitarian
states. The substance of the charges and

accusations and the kinds of ritual may

vary from country to country, but the
sociological process is identical. Nations,
as corporate entities, are all searching for
the same thing: the mythical enemy which
stands in symbolic opposition to the collectivity as a corporate whole.
The perpetuation of social reality
through the complex interaction of ritual
and a mythical universe populated with all
sorts of extraordinary spirits and forces is
similarly not the sole characteristic of
primitive religious systems. The penetration of the sacred into daily reality is also
a variable. Modern men also mingle and
walk among their gods and find themselves in mortal combat with the mythical
forces of Nature and History or The
People and The Nation. These are Durkheimian representations of the corporate
reality of modern societies. The more
corporate reality that is present, the
stronger, more clearly defined and more
closely merged with everyday reality are
those symbolic representations which
mirror that corporate reality. Daily life
becomes filled with transcendent political
significance and, simultaneously, the
enemies of the sacred purposes. The ritual
creation of oppositions to representations
of corporate social reality is one of the
fundamental forms of the modern religious life.

This content downloaded from 134.129.120.3 on Sun, 03 Jan 2016 06:35:33 UTC
All use subject to JSTOR Terms and Conditions

POLITICAL WITCH HUNTS

231

Appendix 1. Total Number of Witch-HuntingActivities by Institutional Area

Activities by InstitutionalArea
Party System
and Country

Total
Activities

Totale

3347
532
475
406
405
324
316
242
197
109
94

2076
318
300
307
288
169
195
159
108
66
51

549
88
95
57
65
55
43
52
3
27
27

93
20
7
9
15
3
13
5
1
4
1

Ghanab

90

45

15

Portugal
Albania
Senegalc

62
52
15

27
20
5

0
11
3

Guinea b
Tanzania'

10
9

6
8

6
3

One-Party
Total
China
Czechoslovakiaf
Poland
U.S.S.R.
East Germany
Hungaryg
Yugoslavia
Spain
Bulgaria
Rumania

Chadc
Ivory Coastc
Two-Party
Total
United States
Great Britain
Philippines
Australia
Austriaa
Colombiab
NewZealand
Multi-Party
Total
France
West Germanya
Italy
Sweden
Canada
Switzerland
Netherlands
Finland
Norway
Denmark
Belgium
Ireland
Iceland
Luxembourg
a

Govern- Mili- Edu- Econ- Intel- Reliment

tary cation omy lectuals gion

Foreigners'

233
23
16
43
15
42
13
16
34
9
1

119
18
22
6
27
13
8
3
14
2
0

225
19
27
49
21
3
23
36
31
3
2

455
107
57
95
39
45
49
27
19
6
8

402
43
76
48
106
8
46
20
6
15
12

5
1
0

14
1
0

2
0
0

5
0
0

1
1
0

0
6
2

1
3

0
1

0
2

0
0

0
0

0
1

5
1

4
1

3
1

0
0

0
0

0
0

0
0

0
0

1
0

1267
1040
123
34
29
25
14
2

641
507
87
20
6
15
5
1

212
166
37
6
0
1
2
0

124
99
21
2
0
0
2
0

78
70
3
3
1
1
0
0

42
38
2
2
0
0
0
0

41
35
2
0
1
3
0
0

12
11
0
1
0
0
0
0

132
88
22
6
4
10
1
1

826
275
157
103
64
51
49
25
24
22
20
20
15
1
0

360
112
53
43
29
12
30
24
18
12
15
7
5
0
0

54
23
13
4
1
2
1
2
2
0
1
1
4
0
0

54
20
1
6
7
0
1
3
2
6
7
1
0
0
0

33
8
9
4
1
3
2
3
1
0
2
0
0
0
0

13
3
5
1
4
0
0
0
0
0
0
0
0
0
0

46
27
6
6
0
2
0
1
0
0
2
1
1
0
0

7
2
1
0
0
0
0
0
4
0
0
0
0
0
0

153
29
18
22
16
5
26
15
9
6
3
4
0
0
0

1956 and later.

1958 and later

c 1950 and later.
1964 and later.
Not all activities were described completely enough to permit assignment to an institutional area.
f Because of foreign occupation, not coded during 1968.
g Because of foreign occupation, not coded during 1956.
h Combinedwith category agents of
foreign governments.

This content downloaded from 134.129.120.3 on Sun, 03 Jan 2016 06:35:33 UTC
All use subject to JSTOR Terms and Conditions

232

AMERICAN SOCIOLOGICAL REVIEW

Appendix 2. Correlation Coefficients among the Variables Party System, Internal Security Forces,
Population, per Capital GNP and Log Witch-Hunting Activities per Year (N=34)
Variables
Variables

PARTY SECFOR

Political Party System

Internal Security Forces
Population
Per Capita GNP
Log Witch-Hunting
Activities per Year

PARTY
SECFOR
POP
GNP
LWHA

1.00
.202
.149
-.701
.283

1.00
.446
.055
.314

POP

1.00
-.056
.446

GNP

Mean

S.D.

1.00
.055

2.21
21.62
44.69
829.27
5.81

.91
22.74
114.16
672.16
1.40

Appendix 3. Correlation Coefficients among the Variables Party System, Internal Security Forces,
Population, per Capital GNP and Witch-Hunting Dispersion Index (N=25)
Variables
PARTY SECFOR

Variables
Political Party System
Internal Security Forces
Population
Per Capita GNP
Log Witch-Hunting
Dispersion Index

PARTY
SECFOR
POP
GNP
WHDI

1.00
.393
.195
-.694
.460

1.00
.212
-.283
-.038

POP

1.00
-.119
.205

GNP

1.00
.294

Mean

S.D.

2.16
.94
36.44
24.10
58.88 130.85
907.16 673.58
.852
.075

REFERENCES

Danzger, M. Heroert
1976 "Validating conflict data." American
Arendt, Hannah
Sociological Review 40:570-84.
1973 The Originsof Totalitarianism.New York: Douglas, Mary
HarcourtBrace Jovanovich.
1966 Purity and Danger. Harmondsworth:PenBanks, ArthurS.
guin.
1971 Cross-PolityTime Series Data. Cambridge:
1970 Natural Symbols. New York: Pantheon.
Massachusetts Institute of Technology Durkheim,Emile
Press.
1965 The Elementary Forms of the Religious
Baum, Richardand FrederickC. Teiwes
Life. Tr. J. W. Swain. New York: Free
1968 Ssu-ch'ing: The Socialist Education
Press.
Movementof 1962-1966.Berkeley:Center Erikson, Kai T.
for Chinese Studies, University of Califor1966 WaywardPuritans:A Study in the Sociolnia.
ogy of Deviance. New York: Wiley.
Bell, Daniel
Feierabend, Ivo K. and Rosalin Feierabend
1964 "Interpretationsof Americanpolitics." Pp.
1966 "Aggressive behaviors within polities,
47-73 in Daniel Bell (ed.), The Radical
1948-1962: a cross-nationalstudy." JourNew
Right.
York: Anchor.
nal of Conflict Resolution 10:249-71.
Bellah, Robert N.
Goffman,Erving
1970 Beyond Belief. New York: Harper and
1956 "The natureof deference and demeanor."
Row.
AmericanAnthropologist58:473-502.
Berger, Peter L.
Goldman,Merle
1969 The Sacred Canopy. New York: Anchor
1965 Literary Dissent in Communist China.
Books.
Cambridge:HarvardUniversity Press.
Blondel, Jean
Gurr,Ted
1973 Comparative Legislatures. Englewood
1968 "A causal model of civil strife: a comparaCliffs, N.J.: Prentice-Hall.
tive analysisusingnew indices." American
Brzezinski, ZbigniewK.
Political Science Review 62:1104-24.
1956 The PermanentPurge.Cambridge:Harvard IBRD (InternationalBank for Reconstructionand
University Press.
Development)
Cohen, Stephen F.
1973 WorldEconomic Atlas. Washington,D.C.
1971 Bukharin and the Bolshevik Revolution.. Kirchheimer,Otto
New York: RandomHouse.
1961 PoliticalJustice. Princeton:PrincetonUniConquest, Robert
versity Press.
1968 The Great Terror.New York: Macmillan.
S. M.
Lipset,
Dallin, Alexanderand George W. Breslauer
1955 "The radicalright:a problemfor American
1970 Political Terror in Communist Systems.
democracy." British Journalof Sociology
Ca.:
Stanford
Stanford,
University Press.
6:176-209.

This content downloaded from 134.129.120.3 on Sun, 03 Jan 2016 06:35:33 UTC
All use subject to JSTOR Terms and Conditions

REFORM ORGANIZATIONSAND CAREERS

233

Simpson, E. H.
1949 "Measurement of diversity." Nature
163:688.
Solomon, RichardH.
1971 Mao's Revolutionand the Chinese Political
Culture.Berkeley:University of California
Press.
Swanson, Guy E.
1964 The Birth of the Gods. Ann Arbor: The
University of MichiganPress.
1967 Religionand Regime. Ann Arbor:The University of MichiganPress.
1971 "An organizational analysis of collectivities." American Sociological Review
22: 17-25.
36:607-23.
Schwartz, BenjaminI.
1968 "The reign of virtue: some broad perspec- Taylor, Charles L. and Michael C. Hudson
1971 Handbook of Political and Social Indicatives on leader and party in the Cultural
tors. Vol. II. Ann Arbor: Inter-University
Revolution." The China Quarterly 35
Consortiumfor Political Research.
(July-September):1- 17.

Luckmann,Thomas
1967 The Invisible Religion. New York: Macmillan.
MacFarquar,Roderick
1960 The Hundred Flowers Campaignand the
Chinese Intellectuals.New York: Praeger.
Parsons, Talcott
1955 "Social strainsin America." Pp. 209-38 in
Daniel Bell (ed.), The Radical Right. Garden City: Anchor.
Robinson, W. S.
1957 "The statistical measurement of agreement." American Sociological Review

SOCIAL REFORM ORGANIZATIONS AND SUBSEQUENT CAREERS

OF PARTICIPANTS: A FOLLOW-UP STUDY OF EARLY
PARTICIPANTS IN THE OEO LEGAL SERVICES PROGRAM*
HOWARD

S. ERLANGER

University of Wisconsin, Madison

American Sociological Review 1977, Vol. 42 (April):233-248
This paper considers the extent to which participation as a salaried professional in a reformoriented organization affects the participant's subsequent career. This issue is studied in the
context of one such organization, the OEO sponsored Legal Services Program, which was
probably the largest and best known organization oriented to the redistribution of professional
services in the late 1960s. Because of the paucity of literature on the consequences of participation in reform organizations, a related literature, that of the consequences of participation in
the-student movement of the sixties, is drawn upon for insight, yet also critically examined.
Comparison of the subsequent careers of 228 lawyers in Legal Services in 1967 to those of 981
other lawyers who were practicing law in 1967 indicates that participation in the program has
an important effect on both the distribution of professional services and the rendering of
reform-oriented pro bono (free or reduced fee) work. In contrast to previous studies, the
explanation offered here differentiates between various components of socialization. In addition, the importance of job market factors is stressed. A further di/Jerence from previous
wi'orkis the consideration, albeit brief, of the effects of variation in experience in the organization.

In the study of social reform

movements and organizations, a good
deal of attention has been paid to the
* This paper is part of a broader study of legal
rightsactivities in which I am collaboratingwith Joel
F. Handler and Ellen Jane Hollingsworth. I am
gratefulto them and to our earliercollaborator,Jack
Ladinsky,for their extensive work on all phases of
the project;to James Fendrich,Thomas McDonald,
Gerald Marwell and, especially, Felice Levine for
their comments on an earlier draft; to Hal
Winsborough, Arthur Goldberger and Robert
Hauserfor methodologicaladvice; andto IreneRodgers, Pramod Suratkar, Anna Wells and Nancy
Williamson for research and programmingassistance. This work was supportedby funds grantedto
the Institutefor Researchon Povertyat the University of Wisconsinby the Office of EconomicOppor-

characteristicsof participantsat the time

of entry, but relatively little to the effects
of participationon the subsequentcareers
of participants. There are many reasons
for this; most obviously, current participants are relatively easy to locate, while
former participantsare not. In addition,
much of the literatureon participationrelates to the social reform activities of the
1960s, for which it is only now practicalto
collect follow-up data.
The activism of the 1960swas most evident among college youth; hence, there is
tunity pursuantto the EconomicOpportunityAct of
1964. Responsibilityfor what follows remains with
the author.

This content downloaded from 134.129.120.3 on Sun, 03 Jan 2016 06:35:33 UTC
All use subject to JSTOR Terms and Conditions