Theories of Revolution in Contemporary Historiography

Author(s): Perez Zagorin

Source: Political Science Quarterly, Vol. 88, No. 1 (Mar., 1973), pp. 23-52
Published by: The Academy of Political Science
Stable URL: http://www.jstor.org/stable/2148647
Accessed: 19-03-2015 14:42 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.

The Academy of Political Science is collaborating with JSTOR to digitize, preserve and extend access to Political Science
Quarterly.

http://www.jstor.org

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
 Theoriesof Revolutionin
Contemporary Historiography,

PEREZZAGORIN
The Universityof Rochester

Revolutionhas been the subjectof investigationby

philosophersand historians since almost the beginning of the
Westernintellectualtradition.Plato and Aristotlewere the first to
deal with revolutionas a theoreticalproblemand Thucydidesthe
first to give it reflectivehistoricaltreatment.2No one who readsit
is likely to forget Thucydides'penetratingaccountof revolution-
ary psychologyand the effectsof class war upon the Greekstates.
It is highly criticalof revolution,yet is, on the whole, the fruit of
profoundinsight and detachedobservation.Fromthe time of Thu-
cydidesto the eighteenthcentury,a successionof historians,most
of them men of affairswho wrote on history and politics as an
avocationor as the occupationof exile or retirement,have thrown
fitful light upon revolutionand civil war. To this companybelong
Sallust, Machiavelli, Commines, Davila, de Retz, Harrington,
Clarendon,and Montesquieu,together with some other well or
'Other discussions of this subject which will be found useful are Law-
rence Stone, "Theories of Revolution," World Politics, XVIII (1966); Henry
Bienen, Violence and Social Change (Chicago, 1968), especially chap. 3;
and Charles Tilly, "Revolutionsand Collective Violence," in Fred I. Green-
stein and Nelson Polsby, eds., Handbook of Political Science (forthcoming).
I am obliged to ProfessorTilly for allowing me to see a copy of his paper.
2 Plato, Republic, Bk. VIII; Aristotle, Politics, Bk. V; Thucydides, The

PeloponnesianWar, III, 81-84.

Volume LXXXVIIINumberl March1973 23

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
24 I POLITICAL SCIENCE QUARTERLY

little known authors.It was not until the nineteenthand twenti-

eth centuries,however,that revolutionbecameone of the supreme
and central preoccupationsof historiography.Since the French
Revolution,the volumeof work devotedto the variousrevolutions
of the Western countrieshas increasedenormously,while, more
recently,interesthas extendedfartherto includethe revolutionsof
Asia andAfricaas well.
The prominencethat revolutionhas assumedin historicalstudy
is generallythought to be mainly due to the influenceof revolu-
tions themselvesin shapingthe modernworld. This influencebe-
came powerful with the French Revolution, which both by its
actualcharacterand by the mythologiesit inspired,openedan era
of profoundchangein human affairs.In 1850, Alexis de Tocque-
ville, commentedon the extentof the change:
Forsixty years we have been deceivingourselvesby imaginingthat we
saw the end of the Revolution.It was supposedto be finished on the
i8th Brumaire,and again in 1814; I myself thought in 1830, that it
mightbe overwhenI sawthatdemocracy, havingin its marchpassed
overand destroyedeveryotherprivilege,hadstoppedbeforethe an-
cientand necessaryprivilegeof property.I thoughtthat, like the ocean,
it hadat last foundits shore.I was wrong.It is now evidentthatthe
tide is rising,andthatthe sea is still enlargingits bed;thatnot only
havewe not seenthe end of the stupendous revolutionwhichbegan
beforeourday,butthattheinfantjustbornwill scarcelyseeit. Society
is not in processof modification,but of transformation.3
Thesewordswerewrittenmorethan a centuryago, but who would
say that they are no longer applicable?Althoughthe revolutionof
which Tocquevillespoke has constantlyassumedfresh forms, al-
though it has metamorphosedinto communist,fascist, racial,and
colonialrevolutions,the processthat he recognizedhas neverthe-
less retainedall of its dynamism.
During the last few years the word "revolution"has also been
heard a great deal in the United States. The Americanpostwar
consensus, which some had imagined to be permanent,broke
down in the 1960s. The so called"end of ideology,"happilycele-
bratedin the 1950S by many academicintellectuals,turnedout to
be an illusion. Variousrevolutionaries,both real and merely self-
styled, cameupon the scene, and in the angry babbleof argument
I Alexis
de Tocqueville, Memoirs, Letters, and Remains of Alexis de
Tocqueville(London,i86i), I, 423.

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
 THEORIESOF REVOLUTION| 25

that followed, the possibilityof a revolutionin the United States

alsobeganto be discussedfor the firsttimesincethe 1930s.
A numberof revolutionshave recently proclaimedthemselves
in America:the blackrevolution,the studentrevolution,the wom-
en's revolution,and that miraculousoccurrenceprophesiedby Pro-
fessor Reichand known as ConsciousnessIII or the youth revolu-
tion.4Therehave been manifestationsof a similarkind in Europe
and Japan.The belief in the necessity of a redemptiverevolution
that will cleansethe advancedindustrialworld of its accumulated
evils has taken root in variousquarters.At the same time, signif-
icant developmentshave occurredalso in revolutionarystrategy.
The teachingsof Mao Tse-tungand GeneralGiaphave gaineddis-
ciples in the West. Regis Debray's importantbook about Latin
America,RevolutionIn the Revolution?,contains a rationalefor
the ideas of Castroand Guevara,which are intendedto shake up
the Communistparties, launch guerillainsurgencies,and thereby
createrevolutionarysituationscapableof bringingCommuniststo
power.5
Admittedly,much of what passes for revolutionnowadays is
only light-mindedtalk and frivolousposturing.Some of it, how-
ever, deservesto be taken seriously.Moreover,the increasingfre-
quency of violent episodes of protest and repressionhas aroused
questions about the stability of Americaninstitutions and their
adaptivenessto peacefulchange. Connectedwith this questioning
has been a sudden outburstof writings and studies dealing with
the causesof violencein America.Partlyfor the same reasonsand
partlybecauseof the revolutionarychallengesoperatingin the un-
derdevelopedworld, therehas also been a markedrevival of intel-
lectual interest in the problem of revolution. Thus, revolutions,
which the revolutionistLeonTrotskycalled "the mad inspirations
of history,"have acquireda freshpertinence.6It is thereforeworth
seeing the place of revolutionat presentin the intellectualland-
scapeof historyandthe socialsciences.

I
To proceed,we must raise the importantand controversialques-
'Charles Reich,The Greeningof America(New York,1970).
6Regis Debray, Revolution in the Revoltution?(New York, 1967).
' Leon Trotsky, My Life (New York, 1930), 334.

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
26 ! POLITICAL SCIENCE QUARTERLY

tion of the definitionof revolution, the sourceof considerabledis-

agreement.A glance at the history of the word itself may provide
us with some help in this matter.7It still seems not to be general-
ly realizedthat beforethe seventeenthcentury,revolution possess-
ed quitedifferentconnotationsfrom those associatedwith it today.
The word was then scarcelyused at all in a politicalsense, but re-
ferredmainly to the circularmotion of the planets. During the
seventeenthcenturyit graduallyacquireda politicalmeaningalso,
but retainedthe idea of circularity.Thus, even as a politicaloccur-
rence, revolution was understoodmerely as a synonym for the
cycle of changein states, a cycle of turbulentups and downs. The
first rebellionof modern times widely called a revolutionby its
contemporarieswas the EnglishRevolutionof i688. Nevertheless,
these same contemporaries, as well as many laterwriters,did their
best to depictthe revolutionagainstJamesII as conformingto the
circularmodel-as a "restoration,"a "return"to a systemof legal-
ity that had been violatedby a tyrannicalking. Accordingly,what
is noticeablyabsent from the meaning of revolution, even at the
end of the seventeenthcentury,is any connectionwith innovation
and the inaugurationof a new order.For the purposesof this dis-
cussion,let it sufficeto say that after 1789, revolution vastly en-
largedits reference.The upheavalin Franceinfusedthe termwith
a new potency and made it a call to action, a shibboleth,a mys-
tique. Marxismin due course reinforcedthis significance.It be-
came linked with ideas of progressand the consciousshaping of
history. It began to signify the willed, deliberateeffort to createa
new society, a new humanity, and a new world. Nineteenth-
century thinkers,whether conserva-tive,liberal, or socialist, were
largelyin accordin viewing revolutionas a phenomenonof epoch-
al changeandinnovation.
Thus, by an indefiniteprocessof extension,revolution has come
to containa vague, assortedmedleyof meaningsand to be applied
to developmentsof the most varied kinds. Historians,in conse-
quence,have given us an almost endlessnumberof differentrev-
olutions,such as, for instance,the industrial,commercial,scientif-
Cf. Perez Zagorin, The Couirt and the Country (New York, 1970),
chap. i, and the literature cited there for a fuller discussion of this
subject.

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
 THEORIES OF REVOLUTION | 27

ic, and Protestantrevolutions;the intellectual,educational,and

military revolutions;and the urban,neolithic,population,second
industrial,and sexual revolutions.This loose and confusinguse of
the term merely illustrateshow closely the idea of revolutionhas
becomeequatedwith change.It would seem that all macroprocess-
es of change,whateverthey are, must be describedas revolutions.
Nevertheless,few of these revolutionshave anything in common
with one anotheror with the conceptof revolutionitself defined
in any precisesense. Hence, althoughI have little hope that they
will do so, historians'would add greatly to clarity if they ceased
the proliferationof revolutionsand reservedthe term for a single,
reasonablywell markedclassof events.
What should its definitionbe? A recentsuggestion,adoptedby
some political scientists, is to substitutefor revolution the term
internal war, definedas "any resortto violence within a political
order to change its constitution,rulers, or policies."8While the
use of internal war has provokedsome illuminatingdiscussion,it
is hard to see the superiormerit claimedfor it. It also has the de-
fect that certainkinds of revolutions,such as some colonialrevolts
or military coups d'etat, would not normally be called internal
wars. In addition, internal war is analogous to civil war and,
therefore,ratherthan being sui generis, is best seen as one possible
phase or stage in the developmentof a numberof differenttypes
of revolution.Hence it seems preferableto retain the well estab-
lished word revolution in a clearly delimited context describing
changewhich is characterized by violenceas a means and a speci-
fiablerangeof goalsas ends.
A well known line of thoughtproposesto restrictthe term rev-
olution solely to movements with goals involving far-reaching
changesin social structure,class domination,institutions,and ide-
ology. This view has been characteristicof the Marxistapproach
although it is held by non-Marxistsas well. In effect, it accepts
only the greatest revolutionsas revolutions.A recent version of
this opinion is advancedby S. P. Huntington:"A revolutionis
a rapid,fundamentaland violent domesticchangein the dominant
8 Harry Eckstein, "On the Etiology of Internal Wars," History and
Theory, IV (1965), 133; cf. also, H. Eckstein, ed., Internal War: Problems
and Approaches(New York,1964).

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
28 | POLITICAL SCIENCE QUARTERLY

values and myths of society, in its political institutions, social

structure,leadership,government activity, and policies."9This
definition,however,is arbitraryin what it excludesand is, there-
fore, of little use to the historian.It has no room for most peasant
revolts, urban insurrections,and provincialor national separatist
rebellions,to mention only a few. To adopt it would only result
in a pointlessnarrowingof the field for both the comparativeand
theoreticalstudyof revolution.
Perhapsit is impossibleto establish a completelysatisfactory
definitionof the term, so complex are the phenomenaand vari-
ables to be included.Of the definitionsavailable,the most useful
appearsto be the one adoptedby ChalmersJohnson in his two
books.10It conceivesa revolutionas violence directedtowardone
or more of the following goals: a changeof government(person-
nel and leadership),of regime (form of governmentand distribu-
tion of politicalpower), or of society (social structure,system of
propertycontrol and class domination,dominantvalues, and the
like). It would be desirableto add to this list a changeof govern-
mental policy as well. With that addition,Johnson'sformulation
seems reasonablyexact and appropriateto all or most varietiesof
revolution, whatever their differencesin aim, scale, or social
character.
Acceptingthis working definition,we are in a position to ask
in what ways the problemof revolutionhas been attacked.Three
possiblelines of inquiryexist, which even though they may over-
lap, can nonetheless be clearly distinguishedfrom one another.
The first, which is also the best worked and most familiar,is his-
toricalin the strictestsense. It is directedto the investigationof a
specificindividualrevolutionlike the Frenchor the English Rev-
olution or the Fronde. Historians taking this approach have
achievedoutstandingsuccesses,with the result that almost every-
thing we know about revolutionscomes from their detailed re-
searchesand generalaccounts.The secondkind of inquiryis com-
parative.Its procedureis to deal with revolutionsas a class, select-
ing from this class two or more instancesin order to investigate
'Samuel P. Huntington, Political Order in Changing Societies (New
Haven, 1968), 264.
10Chalmers Johnson, Revolution and the Social System (Stanford, 1964)
and C. Johnson,RevolutionaryChange(Boston,1966).

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
 THEORIES OF REVOLUTION | 29

any significantrelationshipsbetween them. In this field, like that

of comparativehistory in general,historiansand social scientists
have accomplishedsomething,but not a great deal so far. Finally,
the third kind of inquiryis theoretical.Its purposeis to establish
a theory of revolutioncapableof explainingits causes, processes,
and effectsas a type of change.Varioushypotheseshave been ad-
vanced within the social sciences,but despite the suggestiveness
of some of these, nothing has appearedthat qualifiesas a general
theory of revolution.Furthermore, among theoriststherehas been
little progressiveaccumulationof ideas. The generaltheory of rev-
olution remains subject to confusion, doubt, and disagreement.
Even elementaryquestionsof definition,terminology,and delimi-
tationof the fieldto be explainedarestill not settled.

II
Before looking at recent work on the theory of revolution,it is
useful to mention some earlierattemptsto deal with the subject.
Previousefforts to establisha theory of revolutionhave concen-
tratedprimarilyon causation.Other problems,such as the classi-
ficationof revolutions,the investigationof the dynamicprocesses
involved in revolutions,and the study of the long-range conse-
quences of revolutions,have been neglected by comparison.To
accountfor revolution,many factorsand conditionshave been ad-
duced.In Europeanhistory, the FrenchRevolutionaffordsthe clas-
sic case in which causal primacyhas been imputedto factors of
the most diversekinds-social, economic,intellectual,psycholog-
ical, and so on. Tocquevillethought that revolutionwas likeliest
to happenwhen oppressionwas being lightened.Marx,on the oth-
er hand, connectedthe developmentof revolutionwith intensified
oppressionor exploitation.In 1944, the AmericanhistorianLouis
Gottschalkpublishedan article proposingfive general causes of
revolutions,such as provocationby regimes, solidifiedpublic op-
inion, hopefulnessof change." A list of this kind, however,is too
nebulous to contributemuch to the developmentof a theory of
revolutionarycausation.
Betweenthe two world wars, priorto the recentrenewalof in-
'" Louis Gottschalk, "The Causes of Revolution," American Journal of
Sociology, L (1944), i-8.

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
30 | POLITICAL SCIENCE QUARTERLY

terest,some writingsof considerablemeritdealingwith the theory

of revolutionappearedin the United States.Among these were the
works of two sociologists,PitirimSorokinand LyfordP. Edwards,
and a political scientist, George S. Pettee.'2The best known of
earlierwritings, however, is historianCraneBrinton'sThe Anat-
omy of Revolution, still probablythe most influentialas well as
the most widely read book on revolutionto have been written in
this country.13 Prior to the appearanceof his book, in a mono-
graphon the Jacobins,Brintontested severalhypothesesabout the
role of psychologicaland economicfactorsin attractingpeople to
the revolutionarycause. He decided from an analysis of their
membershipthat the Jacobinswere neithermisfits nor failuresand
that they included a large cross section of French society.
-His skepticalconclusionwas that the Jacobinswere actuatednot
by materialor any other interests,but by a quasi-religiousfaith.14
In The Anatomy of Revolution,Brintonattemptedto discovercer-
tain uniforrnitiesby systematicallycomparingthe English,Amer-
ican, French,and RussianRevolutions.In selectingthese four, he
followedmost scholarship,which has tendedto concentrateon in-
stances drawnfrom the small class of so called great revolutions.
Limitingthe comparisonto this category,however,affordsa dubi-
ous basis for establishinguniformities.Even among the great rev-
olutions, moreover, a number, like the Dutch Rebellion or the
Mexican and TurkishRevolutions,for instance, are hardly ever
considered.For Brintonsome of the uniformitiesin the inception
of revolution are an economicallyadvancing society, an ineffi-
cient, financiallyhard-pressedgovernment,class antagonisms,de-
sertion of the existing orderby the intellectuals,and loss of self-
confidenceby the ruling elite. He also perceivesadditionalunifor-
mities in the successivestages of revolution:the controlof moder-
ates gives way to the rule of extremists,which, in turn, gives way
to a reign of terrorfollowed by a Thermidoreanreaction.Further
uniformitiesappearin the effects of revolution, including great
' Pitirim Sorokin, The Sociology of Revolution (Philadelphia, 1925);
Lyford P. Edwards, The Natural History of Revolution (Chicago, 1927);
GeorgeS. Pettee,The Processof Revolution(New York,1938).
'3CraneBrinton,The Anatomy of Revolution(New York,1938).
14 Crane Brinton, The Jacobins: An Essay in the New History (New

York, 1930).

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
 THEORIES OF REVOLUTION | 31

propertytransfers,the replacementof one ruling class by another,

and the achievementof more efficient, centralizedgovernment.
It is doubtfulwhetheranyone would acceptthese as uniformi-
ties since so many exceptionsto them can be found. The English
Revolution,for instance,was not causedby class antagonism,was
not precededby the desertionof the intellectuals,and did not pro-
duce a reign of terror.Similarly,a politicalscientisthas writtenof
the AmericanRevolutionthat "we find neithervictory of the ex-
tremists, nor the terror,nor the Thermidor,nor yet the 'tyrant'
dictatorwho reestablishedorder. .*..15 Furthermore,neither in
Englandnor in Americadid one ruling class replaceanother. In
my opinion, Brinton's generalizationsresult from an uncritical
tendencyto treat the FrenchRevolutionas a model, a predisposi-
tion he sharedwith many others.Hencethey apply muchbetterto
the Frenchthan to the otherrevolutions.Althoughbrilliantlysug-
gestive, The Anatomy of Revolutionseems to me to offer in the
main a numberof partial resemblancesrather than uniformities
in any truesense.
Another early attempt to deal with the theory of revolution
derives from Marxism,which has remainedextremelyinfluential
in the historicaltreatmentof revolution.It has dominatedthe his-
toriographyof the FrenchRevolutionto such an extent that its in-
terpretationof the Revolution has become virtually canonical.16
Marxist theories have also been frequently advancedto explain
the EnglishRevolution,and they figure as well in currentcontro-
versies about the so called "generalcrisis of the seventeenthcen-
tury."'7Marxismhas been, as well, one of the main influences
upon the historical conceptualizationof the nineteenth-century
revolutionsin Franceand otherEuropeancountries.The much dis-
cussedquestionof the rivoluzionemancata,or failed revolutionof
Carl J. Friedrich,Revolution: Nomos, VIII (New York, 1966), 6.
16Cf. Alfred Cobban, The Myth of the French Revolution (London, 1955)
and A. Cobban, The Social Interpretationof the French Revolution (Cam-
bridge, 1964).
17 From the large and continually growing literature on this subject, I

need cite here only the papers contained in Trevor H. Aston, ed., Crisis
in Europe, 1560-1660 (New York, 1965); Lawrence Stone, ed., Social
Change and Revolution in England, 1540-1640 (London, 1965); and Rob-
ert Forster and Jack P. Greene, eds., Preconditions of Revolution in Early
Modern Europe (Baltimore,1970).

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
32 | POLITICAL SCIENCE QUARTERLY

the Italian Risorgimentohas its origin in Marxist conceptionsof

the historicalprocess.18And althoughit containsno statementof
a formaltheory of revolution,Marxistideas have provideda con-
siderablepart of the explanatoryframeworkof E. H. Carr'sgreat
Historyof SovietRussia.9
Marxismhas been a source of insight to historians of widely
differentviews. In speakingof the Marxist theory of revolution,
however, we are referring to Marxism in its integral charac-
ter, not to Marxist ideas diluted in eclectic mixtures nor to
those of Marx'sideas that are acceptedby nearly all studentsof
society. The outstandingmerit of 'theMarxistmodel is its system-
atic integrationof the factorsof the economy,the social structure,
the state, and ideology in order to explain revolution.Its defects
include the attributionof ultimate causal determination,largely
or wholly, to economicfactors;its mistakenassumptionsthat eco-
nomic class is always the dominantcollectivityin a social struc-
ture and that class conflict is the sole source of revolutionary
change; finally, its grossly simplifiednotion of an evolutionary
successionof societies,with primitivecommunismgiving way, in
turn, to slavery, feudalism,capitalism,and so on, to which there
corrrespondsan equally simplifiedclassificationof revolutions.
Historically,Marxist scholarshiphas been most interested in
bourgeoisrevolutionand the transitionfrom feudalismto capital-
ism. Its conceptualand terminologicalconfusions on the subject
are notorious.Feudalismhas come to include anything from the
eleventh to the nineteenthcenturiesand to designate conditions
as far apartas a full-blownseigneurialregime, absolutemonarchy,
and an orderin which landed aristocraciesmerely enjoy superior
social prestigeand authority.The bourgeoisieand bourgeoisrev-
olution are almost equally elastic categories.When Marxisthisto-
rians speak of seventeenth-and eighteenth-centuryFranceas a
feudal society or describethe imperialismof HapsburgSpain as
"the highest stage of feudalism,"the result can only be confu-
sion.20Admittedly,the analysis of class and of social structures
18 Cf. A. W. Salomone, "The Risorgimento and the Political Myth of

'The Revolution that Failed,'" in A. W. Salomone, ed., Italy from the

Risorgimento to Fascism (New York,1970).
19 E. H. Carr,History of Soviet Russia (London,1950-71).
' Cf. Cobban, Social -Interpretation,for some illustrations. The feudal
conception of seventeenth-centuryFrance is advanced in B. Porshnev, Les

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
 THEORIES OF REVOLUTION | 33

presentscomplicatedproblemsand is a difficultundertaking.Marx-
ist conceptsused for this purpose,however, have not been rigor-
ously or even clearlydeveloped.
With the progressof research,the inadequaciesof the Marxist
theory of revolution have become increasinglyevident. Marxist
explanationsof the EnglishRevolutionhave fallen into abeyance
becauseof the damagingcriticismsto which they have been sub-
jected. In the case of the FrenchRevolution,attemptsto preserve
the Marxistschemeresemblethe additionof epicyclesto the Ptol-
emaicsystemin orderto save the phenomena.The accumulationof
knowledge about the complexitiesof Frenchsociety has made it
clear that the Marxist analysis is far too abstractand too crude.
Marxiststracethe revolutionto a conflictbetween the aristocracy
and the bourgeoisie,but concreteevidence for so reductionista
dichotomyis lacking. A leading Frenchhistorian,PierreGoubert,
has recentlypointed out in a penetratingstudy of the ancien re'-
gime the absurdityof any longer trying to explain the revolution
as "the triumphof an unidentifiablecapitalistbourgeoisieover an
unidentifiablefeudal aristocracy."'"Severecriticismsof the Marx-
ist class analysis have also come from ProfessorsMousnierand
Cobban.22 Even E. H. Carr,who is more sympatheticto the Marx-
ist view, has pointedout that the FrenchRevolutioncannotbe de-
scribedas a bourgeoisrevolutionin the sense that it was begun or
led by an identifiableclass.23When Marxists resort to explana-
tions of a still more generalcharacter,the difficultiesbecomeeven
worse. Forexample,ProfessorSoboul,who has done importantre-
search on the popularmovement in revolutionaryParis, locates
the ultimatecause of the FrenchRevolutionin the contradiction
between the productiveforces and the relations of production.24
soulevements populaires en France de 1623 a 1648 (Paris, 1963). Cf. also,
D. Parker, "The Social Foundationof French Absolutism, 1610-1630," Past
and Present, LIII (1971), 67-89. For the "feudalism"of seventeenth-century
Spain, cf. P. Vilar, "Le temps du 'Quichotte,'" Europe,XXIV (1956), trans-
lated in New Left Review, LXVIII (1971), 59-71.
n PierreGoubert,L'AncienRegime(2nd ed.; Paris, 1969), 257.
22 Cf. the papers collected in Roland Mousnier, La plume, la faucille, et
le marteau (Paris, 1970); Cobban, Social Interpretation.
' E. H. Carr,The October Revolution (New York,1969), 16-17.
24 Albert Soboul, The Parisian Sans Culottes and the French Revolu-

tion, 1793-94 (Oxford, 1964). The doctrinarism which overlays the acute
perception and rich concreteness of Soboul's work has been noted by

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
34 | POLITICAL SCIENCE QUARTERLY

This propositionis so murkyand abstractthat it eludes any possi-

bility of verification.Perhapsit should be receivedas a profession
of faithratherthanas an historicalstatement.
If one asks aboutMarxismthe questionthat Crocein a famous
essay asked about the philosophy of Hegel-what is living and
what is dead in the Marxist theory of revolution?-the answer
must be that, in largepart,it has been superseded.Its inadequacies
might have been thought to have been conclusivelyrevealedwhen
Marx'sown predictionthat revolutionswould occur in the most
advancedcountriesof the West did not come true, and, instead,
socialismtriumphedin the backwardpeasant societies of Russia
and China.With the growth of researchon many revolutions,so
muchof the Marxistmodelhas been disprovedor shown to be un-
tenable because of its oversimplificationsthat it has become in
many ways an obstacleto furtherunderstanding.What remainsof
value in this model are some of the problemsit poses for investi-
gation and its emphasis upon the importanceof economic and
social-structuralfactorsin the analysis of change and revolution.
These will undoubtedlyset challengesto historians, sociologists,
andpoliticalscientistsfor a long timeto come.

III
Since about 1960 there has been a noticeablerevival of interest
among social scientists other than historians in revolution as a
theoreticalproblem.This interesthas manifesteditself in a num-
ber of writingsby politicalscientistsand sociologistsdealing with
revolutionin various theoreticalcontexts. At the same time, the
present position of historiographywith respect to the investiga-
tion of revolutionalso presentscertainfeaturesthat call for men-
tion. AlthoughI have been referringto history as a social science,
many historians would either reject or wish to qualify that de-
scription.In additionto the continuousfragmentationof histori-
ographyinto innumerablespecialties,therehas been an increasing
separationbetween the humanisticand social-scienceapproaches
another historian of the popular movement in the French Revolution; cf.
Richard Cobb, The Police and the People: French Popular Protest, 1789-
1820 (Oxford,1970), 120-21, 186.

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
 THEORIES OF REVOLUTION i 35

to historicalinquiry.This seems to have reachedthe point where

it may be asked if history remainsany longer a single discipline.
The same questionhas been raised in a recent volume edited by
David Landesand CharlesTilly which is part of the survey of the
behavioraland social sciencesconductedunderthe auspicesof the
SocialScienceResearchCounciland the NationalAcademyof Sci-
ences.25There the humanisticand social-sciencemodes are con-
trastedas quite differenttypes of inquiry.Probablymost histori-
ans would place themselvessomewherebetweenthe two extremes,
but nearer the humanistic than the social-scienceapproach.In
view of the distinction,however, it is significantthat the same
survey should indicatethat revolutionis one of the focal subjects
of social-sciencehistory.This is certainlytrue of muchof the more
interestingwork being done. Social-sciencehistory is not charac-
terizedprimarilyby the use of quantitativetechniques;it is mainly
characterized by a certainmannerof formulatingproblems,by its
effort at empiricalrigor, by its use of theories,models, and ideal
types, by its interestin comparativeand interdisciplinarystudies,
and by the orientationof its researchtowardthe understandingof
whole societies.In history, as in other fields, most of the current
effortsto gain a clearerunderstandingof revolutionproceedfrom
or areinfluencedby this approach.
An example of current scholarshipis the research on what
might be called revolutionarypopulations.The actors in revolu-
tions belong either to elite groups or to the masses. Systematic
analysis of the characteristicsof each group using comparative
data on the individualmemberscan lead to enlighteningresults.
Researchof this kind can center on anything from the member-
ship of formalinstitutionsand rulingbodies to the participantsin
protestincidentsas shortlivedas a riot. Fromsuch work it is pos-
sible to gain a much fuller and more preciseknowledgethan we
have ever had before of recruitment,of oppositionto revolution,
of leadership,of popularactivism,and even of individualmental-
ities. The collectivehistory of the Parisiansans culottesby Albert
Soboulis an outstandingexample.He reconstructedas far as pos-
sible in all its lineamentsthe mass movementin Paris during the
5David Landes and Charles Tilly, eds., History as Social Science
(EnglewoodCliffs,N. J., 1971).

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
36 | POLITICAL SCIENCE QUARTERLY

Jacobinascendancyand the Terrorby close investigation of the

people who composedit.26 George Rude has performeda similar
task in his dissection of the crowds and crowd behavior in the
FrenchRevolution and other social disturbances.27Another in-
stanceof a differentkind is the detailedelite studies,like the analy-
sis of the Soviet Politburoand of Nazi officialdom,done by politi-
cal scientists.28Still another example is the investigation of the
membershipof the Long Parliamentby Douglas Brunton and
D. H. Pennington,from which it appearsthat the Royalist and
Parliamentarian membersas a body were socially and economical-
ly more or less indistinguishablefrom one another.29This re-
searchis not only valuablefor the understandingof particularrev-
olutions,but is also relevantto the theory of revolutionbecauseit
-canprovidedata for modelsand for the testing of explanationsde-
rivedfromgeneraltheories.
Anotherkind of work is the comparativehistory of revolution,
of which there are various recent examples.A prominentone is
R. R. Palmer'sThe Age of the DemocraticRevolution,which gives
a valuable account of political developmentsbetween 11760 and
18oo and also conceives of the FrenchRevolutionas the central
event in an interrelatedupsurge of revolutions extending over
Americaand WesternEurope.30Most of the discussionprovoked
by Palmer'sbook has centeredon the question of whether there
reallywas an eighteenth-centuryrevolutionof the West. Although
Palmerhas put much materialtogetherwith great skill, his work
containslittle on the wider problemof revolutionitself. It touches
passingly on the causes of revolution,on revolutionarypsychol-
ogy, on the basis of political allegiances,and similar problems,
but these reflectionsare too vague to be very illuminating.Palmer
has also written an essay on generalizationsabout revolution
which is equally cursory and vague in regard to theoretical
' Soboul. Similarly valuable is the work of Cobb,despite his "total rejection
of sociology and quantification,"p. xv.
7George Rude, The Crowd in the French Revolution (Oxford, 1959),
and G. Rud6, The Crowd in History (New York, 1964).
' HaroldLasswelland Daniel Lerner,eds., World RevolutionaryElites (Cam-
bridge,Mass., 1965).
9Douglas Brunton and D. H. Pennington, Members of the Long Par-
liament (London,1954).
80 R. R. PalmeT, The Age of the Democratic Revolution (Princeton,

1959-64).

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
 THEORIES OF REVOLUTION j 37

issues.3'
Work of more directtheoreticalinteresthas appearedin recent
writing dealing with sixteenth- and seventeenth-centuryrevolu-
tions and "the crisis of the seventeenthcentury."The differences
arisingfromthese studies,which addressthemselvesto basic ques-
tions concerningthe societies, governments,and revolutions of
early modernEurope,have led to some of the liveliest controver-
sies in contemporaryhistoriography.Contributionsto the subject
have come fromH. R. Trevor-Roper, EricHobsbawm,J. H. Elliott,
LawrenceStone, J. H. Hexter, Perez Zagorin, Roland Mousnier,
and others,who have providedsearchingdiscussionsof the deter-
minants of social structureand political differencesin connection
with the revolutionsof the period.32Models for these revolutions
have also been suggested,such as the court-countryoppositionor
the conflictbetweenthe warlike,top-heavyabsolutiststate and so-
ciety. One of the most active scholarsin all these problemsis Pro-
fessor Mousnierof the Sorbonne.He has compiledan extremely
detailedquestionnaireto guide researchon rebellionsin Francebe-
tween the end of the MiddleAges and the FrenchRevolution.In a
numberof historicaland comparativestudies,he has also stressed
the importanceof aristocraticleadershipfor the type of revolt that
occurredin the societiesof the ancienregime.33
Millenarianism,anotherlargetopic,has been the subjectof com-
parativestudies,in which the centralproblemis its relationshipto
modernrevolutionarymovements.On this there has been a con-
vergence of researchby historians,anthropologists,and sociolo-
gists.34 While by no means uniform in character,millenarian
movementsin Europehave usually displayedtwo features:first,
Si
R. R. Palmer, "Generalizationsabout Revolution: A Case Study," in
Louis Gottschalk, ed., Generalizationin the Writing of History (Chicago,
1963).
3 Cf. the writings cited above in n. 17 and the literature referred to
in these works.
9Roland Mousnier, Recherches sur les soulbvements populaires en
France de 1485 I 1787: Questionnaire (Paris, Centre de Recherchessur la
Civilisation de l'Europe Moderne, n.d.); R. Mousnier, Fureurs paysannes:
Les paysannes dans les revoltes du xvii' siecle (France, Russe, Chine)
(Paris,1967); R. Mousnier,La plume.
s Cf. Sylvia Thrupp, ed., Millennial Dreams in Action, in Comparative
Studies in Society and History, Supplement II (The Hague, 1962), and
Yonina Talmon, "Millenarism," in the International Encyclopedia of the
Social Sciences (New York, 1968), X, and Y. Talmon, "Pursuit of the

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
38 | POLITICAL SCIENCE QUARTERLY

the completerejectionof the existing world orderand the expecta-

tion of a futurenew world purgedof the evils of the old; second,
an ideology of total transformrLationderivedfrom Jewish-Christian
apocalypticbeliefs. A further feature frequentlyassociatedwith
millenarianismis the presenceof a charismaiicleaderor a prophet.
Many writers have noted the typological resemblancebetween
such religious movementsand secularrevolutionarymovements,
and some have also arguedthe historicaldependenceor intercon-
nectionof the two. One approachto the problemis exemplifiedin
Hobsbawm's Primitive Rebels, which treats millenarianismas
a prepolitical phenomenon arising in traditional societies and
among the backwardstrata of more advancedsocieties.Millenar-
ian movementsin this view are a revolutionaryexpressionof de-
-privedand oppressedgroups who have not yet attainedpolitical
consciousnessand who lack the means of politicalorganization.3"
An alternativeapproachis taken by NormanCohn in The Pursuit
of the Millennium,who stressesnot only socioeconomicfactorsbut
also the psychologicalelementsin millenarianismwhich tend to-
ward producingoutbreaksof collectivehysteria and irrationality.
In the case of rebellionssuch as the GermanPeasantWar of 1525,
Cohn is inclined to see millenarianismas a factor injectedby a
small numberof fanaticalenthusiastsinto a much largerpopular
movementwith realisticaims. He also points out the importance
of intellectualsin framing the ideology of millenarianrevolts.36
Hobsbawm'sanalysis regards millenarianismboth as the ex-
pressionin mysticalterms of social and economicinterestsand as
the precursorof realisticpoliticalstruggle.It also assumesan evo-
lutionary scheme, distinguishingarchaicfrom modern forms of
rebellion, the highest stage of which is revolutionaryMarxism.
His assumption,however, is questionable.It seems more likely
that ratherthan being a phenomenonpeculiarto primitivesocial
forms, futurist fantasies of transformationand perfectionare a
possibilityin any society and are capableof combiningwith high-
ly realisticbeliefs and a sophisticatedpoliticalorganization.They

Millennium: The Relation Between Religious and Social Change," in Barry

McLaughlin,ed., Studiesin SocialMovements(New York,1969).
3 EricHobsbawm,PrimitiveRebels(Manchester,1959).
36Norman Cohn, The Pursuit of the Millennium (Fairlawn, N. J., 1957),
and N. Cohn,"MedievalMillenarism,"in Thrupp,37-39.

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
 THEORIES OF REVOLUTION | 39

are an important element, for instance, both in Leninist and Mao-

ist thought and in the utopian expectations of contemporary youth
revolutionaries. Millenarianism can therefore be considered not
only as defining one particular species of rebellion, but also as a
possible component in some secular types of revolution as well.
Perhaps the most noteworthy recent comparative study of rev-
olution is Barrington Moore's Social Origins of Dictatorship and
Democracy.37 Moore's book exemplifies the current intellectual
trend in the study of revolution by being the workoLa-sociologist
concentrating entirely on a set of historical problems. What it
seeks to explain is the process of transition from traditional agrar-
ian to modern industrial society as experienced in a number of
countries in the East and West; England, France, the United States,
Japan, China, and India are examined, with frequent reference
made also to Germany and Russia. The balance between the analy-
sis of historical events and the drawing of theoretical general-
izations is admirably maintained. Whatever criticisms historians
may make, they are not likely to accuse it of the jejune and mech-
anical treatment often apparent in the writings of nonhistorian
social scientists when they use historical materials.
Moore identifies three basic patterns in the modernizing process.
The first leads via bourgeois revolution to capitalism and democ-
racy of the Western type. Its instances include England, France,
and also the United States whose civil war Moore explains as a
revolution caused by the political incompatibility between slavery
and a democratic capitalist order. The second, in which the bour-
geois revolution fails, leads to capitalism with much weaker dem-
ocratic features. This was the pattern for Germany and Japan
which resulted in fascism. The third leads through great peasant
revolutions to communist regimes that force the society into mod-
ernization. The examples are Russia and China. To these three
basic patterns, Moore adds a fourth, exemplified by the case of
India, in which, despite the achievement of parliamentary democ-
racy, the impulse to modernization is weak because there has been
neither a bourgeois nor a peasant revolution.
One variable is critical in the patterns described by Moore-
namely, the way in which the traditional agrarian sector in each
3 Barrington Moore, Social Origins of Dictatorship and Democracy:

Lord and Peasant in the Making of the Modern World (Boston, 1966).

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
40 | POLITICAL SCIENCE QUARTERLY

case has been subjectedto capitalistrelationsand modernization,

which, he believes, was, in turn, determinedby the respective
roles of peasants,landedelites, urbanbourgeoisies,and the state.
Wherethe processhas been carriedout by the bourgeoisieand part
of the peasants,as in France,or by capitalisticlandlordstogether
with the bourgeoisie,as in England,the result is democraticcapi-
talist regimes. Where it has been accomplishedby landlordsand
the state bureaucracywithout the presenceof a strongbourgeoisie,
as in Germanyand Japan,the result has been authoritarianand
fascist regimes. Finally, where landlordsand state bureaucracies
inhibitedthe processand the bourgeoisiewas too small and weak
to effect it, as in China and Russia, the impetus has come from
peasantmasses mobilizedby communistparties,which then made
thepeasantstheirfirstvictims.
Moore does not proposea general theory of revolution;rather
he presentssome generalizationsabout a certainrange of experi-
ence in which revolutionoccupiesthe centralplace.The occurrence
or failureof revolution,as well as its specificcharacter,are viewed
as crucialfor the ensuing type of development.Revolutionis thus
placedwithin the historicalprocessas a decisive point of conflict
having significant systemic consequences.Moore dismisses the
idea that revolutionis necessaryto remove the obstaclesto mod-
ernization.He points out, however, that modernizationachieved
underconservativeauspicesretainsmany old structuresand hence
leads to quite differentresultsfrom those producedby a great rev-
olution. He accepts the conception of a bourgeois revolution,
though he does not necessarilymean the seizure of power by a
bourgeoisie.Revolutions,Moore holds, should be comparedand
classifiedwith referenceto their broadinstitutionalconsequences.
Consequently,becauseof the kind of politicaland social orderthat
resultedfrom them, he links the English and FrenchRevolutions
with the AmericanCivil War as stages in the developmentof the
bourgeois-democratic revolution.
Thereare a numberof seriousweaknessesin Moore'swork. For
one thing, in laying such heavy stress on economicrelations,he
underestimatesor disregardsthe role of values and even of politics
in social change.Thus in the case of Germany,conservativemod-
ernizationmay be a necessarycause of fascism, but it can hardly
be a sufficientcause since it does not include other political cir-

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
 THEORIES OF REVOLUTION | 41

cumstancesand value orientationsthat were fundamentalto the

outcome.In the case of Englandand France,to call both products
of a bourgeoisrevolutiondoes nothing to explainwhy the subse-
quentFrenchpoliticalorder,in contrastto the English,was so un-
stable and divided against itself. Moreover,Germanfascism, like
Italian, was itself a revolution with broad systemic features.
Moore'smodel, however,cannotincludefascism as a type of rev-
olution. Strongreservationsalso arise in connectionwith Moore's
analysisof the AmericanCivil War. Since he agreeswith the con-
clusion of recentscholarshipthat Northerncapitalismand South-
ern plantationslaverywere not antagonistic,but existedin a prof-
itablerelationwith one anotheras economicsystems, it is difficult
to see what groundthere can be for explainingthe Civil War as a
bourgeois-democratic, capitalistrevolutionof the same type as the
EnglishandFrench.
Moore's work, nonetheless,must be regardedas a significant
synthesis in the comparativetreatmentof revolution. It is also
notablefor its avoidanceof excessiveabstractionand its firmsense
of historicalreality. An approachof this kind, linking the com-
parativeand theoreticaldimensionsof inquirywith the historical,
is one of the most likely to lead to a betterunderstandingof the
nature and effects of revolutionin large-scaleprocessesof social
change.
In addition to comparativestudies, social scientists have also
tried to develop some general theories and explanationsof rev-
olution. For instance,the questionof how revolutionis relatedto
poverty and to economicand social oppressionhas always been
puzzling.AlthoughMarxexplainedrevolutionsas the effectof in-
creasingmisery and exploitation,Tocqueville,Brinton,and others,
while not denying, of course,the importanceof poverty, have ar-
gued that some of the greatest revolutionsoccurredin societies
where the level of life was improving.Tocquevilleis most explicit
on this point, declaringthat an unprecedentedadvancein the pros-
perity of the nation took place before the FrenchRevolutionand
that this promoteda spirit of unrest and discontentwhich was
highest in those parts of Francewhich had experiencedmost im-
provement.38Today most students hold that revolution is least
"Cf.' Alexis de Tocqueville, The Old Regime and the Revolution, trans.
S. Gilbert (New York, 1955), Pt. 3, chap. 4. For Marx the intensification

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
42 | POLITICAL SCIENCE QUARTERLY

likely among populationsexperiencingstatic conditions of pov-

erty, low living standards,and economicbackwardness.Compara-
tive data assembledby political scientists also indicate a positive
correlation between political instability and rapid economic
growth.39What then is the connectionbetween poverty and op-
pression,economicdevelopment,andrevolution?
To this questiona politicalscientist,J. C. Davies, has proposed
an answer in the form of a model that combinesfeaturesof both
Marx and Tocqueville.40It is based on the view that a situation
most favorableto revolution exists when a prolongedperiod of
economicgrowth is followed by a short period of sharp reversal.
New expectations,needs, and standardshave then been created
that are suddenly checked.With the J-curve,as Davies calls the
downwardswing, an acutediscrepancyappearsbetweenmounting
wants and expectationsand the possibility of their satisfaction,
which, in turn, establishesthe potentialfor revolution.Revolution
is accordinglyseen as the effectneitherof poverty and miserynor
of improvementbut of a certainsequenceof the two. Davies has
triedto fit his modelto severalcases,and althoughit cannotbe ac-
ceptedas a generalexplanationof revolution,it is clearlypertinent
to certainrevolutionslike the French,which began when contrac-
tion, bad harvests, and famine succeeded a secular phase of
growth. LawrenceStone has pointedout that the J-curvecan also
apply to other sectorsbeside the economic.Thus troublemay arise
if a phase of liberal governmentalconcessionsis followed by a
phase of politicalrepressionor if a phase of easier entry into the
elite is followed by a phase of aristocraticreactionand closure.4'
A muchbroaderapproachfoundedon similarpremiseshas been
of oppression is only one of the conditions of revolution. There are also
other conditions, including the development of class consciousness in the
oppressed. As regards capitalism, Marx gives a concise analysis of the
"law" of capitalist production leading to the deteriorationof the worker's
position in Wage-Labourand Capital and in Value, Price and Profit both
in SelectedWorks (New York,n.d.), I.
3 Cf. M. Olson Jr., "Rapid Growth as a Destabilizing Force," and Ivo

K. and Rosalind L. Feierabend, "Aggressive Behaviors Within Polities,

[948-1962," both in J. C. Davies, ed., When Men Revolt and Why (New
York, 1971).
40J C. Davies,"Toward a Theory of Revolution," in Davies, ed., When
Men Revolt.
4 Stone, "Theories,"172.

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
 THEORIES OF REVOLUTION j 43

recently developed in a book by Ted Gurr, Why Men Rebel.42

Gurr'spoint of departureis the theory of relative deprivation,
which may be definedas the discrepancythat individualsperceive
betweenwhat they expectand what they are likely to obtain.Rela-
tive deprivationmay exist either because value expectationsin-
creasefaster than value capabilitiesor becausevalue expectations
remain the same while value capabilitiesdecline. In both cases
men will feel anger, frustration,and discontent.If relativedepri-
vation is strong enough and becomesfocused on politicalobjects,
a situation arises that may lead to violence and revolution.
Whetherrevolutiondoes occuror not dependson a numberof vari-
ables, includingthe scope and intensity of the relativedeprivation
and the actionstakenby the regime.Gurrhas elaboratedthe logic
of this analysis with great thoroughness.Thus the expectations
subjectto relative deprivationare not necessarilylimited to the
economicand the material.The expectationsmay be about power
and politicalparticipationor status,personaldevelopment,and be-
lief. Gurrattemptsto specifythe diversecausesof relativedepriva-
tion and to explicatethe conditionsaffectingboth regimesand dis-
sidents that may culminatein revolution.The analysis gives close
considerationto the role of legitimacy, to the effects of violence
under various circumstances,and to the determinantsof support
for rebels and regimes. It also suggests methods such as opinion
surveysto measuresomeof the variablesinvolved.
Gurr'swork is of exceptionalinterestboth in its range and pre-
cision. The main questionthat it raises concernsits fundamentally
psychologicalformulationand treatmentof the problemof revolu-
tion. The theoryof relativedeprivationis derivedfrom the studies
done by social psychologistson frustrationand aggression.It thus
begins by locating the cause of dissatisfactionand discontentin
the individual.Startingthere,it generalizesthe processto account
for many people'sdiscontentand then by a furtherprogressionto
the likelihoodthat these same peoplewill becomeviolent, will di-
rect their violence against politicaltargets, and eventuallyengage
in revolution.The focus is accordinglyon the side of personalex-
perienceleadingto the formationof discontent.
Thereare severalimportantdrawbacksto this view. Apartfrom
the practicaldifficultiesin determiningthe scope and intensity of
42Ted Robert Gurr, Why Men Rebel (Princeton, 1970).

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
44 | POLITICAL SCIENCE QUARTERLY

relative deprivationboth in the present and even more in earlier

periodswhere the data are scanty, a very long distanceseparates
discontentfrom revolution.In orderfor the one to lead to the oth-
er, many factorsmust intervene,such as developmentsin the so-
cial and economicstructure,conflictsand changes in the political
system, the emergenceof ideologies,and the like. Such factorsare
crucialto the dynamiccausal processof revolution.Furthermore,
they are not attributesof states of mind, but propertiesof political
and social situations or systems. Gurr'stheory, however, cannot
take accountof structuraldevelopmentsor politicalconflictsexcept
as they are relatedto the psychologicalphenomenonof relativede-
privation.This inability considerablydiminishesboth the signifi-
canceand the possibilityof analyzingadequatelynonpsychological
factors.For that reason,althoughthe relevanceof a psychological
approachis undeniable,it seems to be in itself a limited and in-
completeone.
In explaining revolutions,it is obviously necessary to distin-
guish long-rangepreconditionsand processesfrom immediatepre-
cipitatorswhich, by chance,act to releaserevolutionarysituations.
Only the preconditionscan be incorporatedinto causal theories.
Any attempt,however,to elevate a particularset of preconditions
into generalcauses runs into difficulties.HarryEckstein,a poli,ti-
cal scientist, has classified the commonly advanced hypotheses
dealing with the causes of revolution,or, as he prefers to call it,
internalwar.43Some emphasizeconditionsof the social structure,
like too much or too little social mobility, the emergenceof new
social classes, and the inadequatecirculationof elites. Others em-
phasize ideologicalconditions,like the conflict of belief systems
and social myths or the existence of corrosivephilosophiesand
unrealizablevalues. Others stress economic conditions, such as
poverty and exploitationor rapid growth. Still others emphasize
politicalconditionssuch as governmentoppression,division with-
in governingclasses,or the excessivetoleranceof alienatedgroups.
With such an abundanceof partiallyinconsistenthypotheses,any
of which might yet be valid in a particularcase, the obstaclesto a
causaltheoryareapparent.
Ecksteinoffers some suggestionsto improvethe explanationof
the preconditionsfor revolution.First,he points out that histori-
'Eckstein, "Etiology of Internal Wars."

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
 THEORIES OF REVOLUTION | 45

ans and theoristshave tendedto pay more attentionto rebelsthan

to incumbents,but that both should receive equal consideration.
This can hardlybe gainsaidin view of the importanceof elite be-
havior in creating the preconditions.Some writers, like Pareto
for instance, have even held that revolutionsare largely due to
changesin elite characteristics,leading to estrangementfrom non-
elites, incompetence,and the failure of nerve." Second, any ex-
planationshould take into accountbehavioralas well as structural
factors.This is contraryto the commonapproach,which has tend-
ed to concentrateon.the structuralor so calledobjectiveconditions
of revolutions.The same conditions,however,may and do produce
differentresults becauseorientations,values, and attitudesmedi-
ate between the social environmentand action. Hence, culture,
ideology, and the role of intellectualsin maintainingthe values of
legitimacymust also be stressedin the formationof preconditions.
Third, explanationsshould strive for broad propositionsdescrib-
ing generalsocial processeswhich can apply to a large variety of
specificpreconditions.This last suggestionis no doubt valid, but
at the same time there is also the danger that such propositions
maybecomevacuousor trivialthroughtheirvery generality.
Based on these ideas, Eckstein has proposed an explanatory
model containingeight variableswhich in combinationestablish
the presenceand extent of the preconditionsfor revolution.The
negative variables working against revolution are the facilities
(especiallyfor coercion)availableto incumbents,effectiverepres-
sion, adjustive concessions, and diversionarymechanisms.The
positive variablesworking in favor of revolutionare the ineffi-
ciency of elites, disorientingsocial processes,subversion,and the
facilitiesfor violence availableto insurgents.With this paradigm,
Ecksteinholds that it is possible to assess the potential for rev-
olution and the balancebetween the favoring and countervailing
factors.
Eckstein'smodel seems a useful one in its ability to incorporate
differentsets of particularpreconditions.At the least it provides
an inventoryof factorsto be checkedoff in explainingrevolutions.
It would thereforebe of considerableinterestto see it appliedto an
historicalcase in order to test its fit to the complexityand un-
tidinessof the events.
"Cf. V. Pareto,SociologicalWritings, ed. S. E. Finer (London,1966), 55-58.

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
46 | POLITICAL SCIENCE QUARTERLY

There are a number of other angles from which recent social-

science scholarship has attacked the problem of revolution. For
instance, in an historical monograph on the Vendee, Charles Tilly,
a sociologist, has based an analysis of a counterrevolution on the
effects of urbanism and market penetration in a region of tradi-
tional subsistence agriculture.45A noticeable trend has also been
to subsume revolution under theories of violence.46 The assump-
tion implied in such studies is reflected in Gurr's work as well as
in the attempts of political scientists to establish precise indices
and measures of conflicts' within states.47 The identification of
violence and revolution is a mistake, however, for while the use
of violence is certainly one of the defining characteristics of rev-
olution, the two are not identical phenomena. A people may be
prone to violence, like Americans, without necessarily being rev-
olutionary. Many acts of collective or political violence-such as
banditry, labor strife, lynchings, food, race, and antiforeign
riots-may bear no relation to revolution or to the formation of
the preconditions for revolution. A great deal of violence, more-
over, proceeds from agencies of the state both in their "normal"
activities and in repressive actions, such as those carried out
against peaceful demonstrations with no revolutionary import. Al-
so, the major kind of collective violence is carried out by govern-
ments when they make external war. Hence, though revolution
and violence have some common features, they are different phe-
nomena requiring different theories for their elucidation.
Revolution is often discussed in connection with theories of
modernization. The study of modernization has become one of the
preoccupations of current political science and focuses on the prob-
lem of stability in societies undergoing rapid social and economic
change, its interest being for obvious reasons concentrated mainly
on the contemporary scene. Modernization itself is sometimes re-
garded as a revolution rather than as the possible outcome or char-
'4 Charles Tilly, The Vendee (Cambridge,1964); cf. Charles Tilly's "The
Analysis of a Counter-Revolution,"History and Theory, III (1963), 30-58.
4 Cf. the discussion in Bienen, and in H. L. Nieburg, Political Violence
(New York, 1969).
4 Cf. Ivo K. and Rosalind L. Feierabend and Ted Robert Gurr with
Charles Ruttenberg, The Conditions of Civil Violence: First Tests of a
Causal Model (Princeton, 1967); see also some of the papers in Hugh
Davis Graham and Ted Robert Gurr, eds., Violence in America: Historical
and ComparativePerspectives(New York,1969).

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
 THEORIES OF REVOLUTION | 47

acteristic of a certain type of revolution. Thus one writer speaks of

"the revolution of modernization," which he calls "the first rev-
olution in history to set a new price upon stability in any system
of society; namely, an intrinsic capacity to generate and absorb
continuing transformation."48 This paradoxical observation re-
peats the familiar mistake of making revolution synonymous
with any rapid, large-scale change. It also assumes that a society
that has achieved modernization will be free of violent change
and can remain stable even if transformation is continuous and in-
cessant. Both assumptions are doubtful and pertain more to the
utopia of modernization theorists than to the real world.
Probably the outstanding work devoted to revolution in the con-
text of modernization is Samuel P. Huntington's Political Order
in Changing Societies, an impressive synthesis of the conditions
of stability and instability. As I have previously remarked, Hunt-
ington restricts the definition of revolution exclusively to trans-
formations of the most sweeping character which are effected by
violence. His argument is that social and economic changes such as
urbanization, industrialization, the spread of literacy, of educa-
tion, and of communication facilities, and so on lead to a rise in
political consciousness, the mobilization of new groups into poli-
tics, and the multiplication of political demands. Traditional soci-
eties launched on the process of modernization usually lack the
political institutions and organizations capable of bearing the
heavy new strains put on them. There is consequently an imbal-
ance between socioeconomic growth and the political capacity to
assimilate the new forces mobilized in the process. The result is
instability, disorder, and, in certain cases, revolution. According-
ly, in Huntington's view, revolution is an aspect of moderniza-
tion. It is very unlikely to occur either in highly traditional soci-
eties or in highly modern ones and is least probable in both dem-
ocratic and communist political systems because of the capacity of
each to absorb new groups. Huntington also distinguishes a West-
ern and an Eastern pattern of revolution. In the Western pattern-
of which the French, Russian, and Mexican Revolutions are in-
stances-the decay of political institutions and the collapse of the
old regime precedes the process of political mobilization and the
8 Manfred Halpern, "The Revolution of Modernization in National and
International Society," Revolution: Nomos, VIII (New York, 1966), 179.

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
48 ! POLITICAL SCIENCE QUARTERLY

creation of a new order. In the Eastern pattern, which includes the

Chinese and Vietnamese Revolutions, political mobilization and
the shaping of new institutions come first and the overthrow of
the existing orderlast.49
I will not attempt to assess this challenging thesis in detail, but
instead will merely express some critical reservations. Hunting-
ton's theory is based on a three-stage classification of societies as
traditional, transitional or modernizing, and modern. While these
categories are useful for some purposes, they are much too large
and undifferentiated to describe the course of development as the
historian sees it. To place French, Russian, and Mexican societies,
for example, within this classification and then to treat the revolu-
tions as the effects of modernization adds little to our understand-
ing of them. The contrasts between the societies, economies, cul-
tures, and states concerned are in many respects far too wide to
justify such a procedure. It is not clear, moreover, why there
should be two patterns of revolution, since Huntington's model
appears to require that the political mobilization of new groups
must always precede the breakdown of the traditional regime.
With only one pattern, however, difficulties arise because certain
revolutions cannot be made to fit the historical sequence. Further-
more, throughout, the analysis is continually open to the danger
of advancing by a circular argument. Instead of consistently treat-
ing revolution and modernization as variables between which, in
every case, an empirical connection must be established, Hunting-
ton sometimes joins them by definition. Thus, in explaining why
China underwent a revolution and Japan did not, he tends to judge
whether the political stresses resulting from modernization were
severe enough to cause a revolution merely by the fact that a
revolution did, indeed, occur.
This problem poses the larger issue of whether modernization,
as such, can be accepted as the cause of revolution. There is, to be
sure, a good deal of evidence that rapid social and economic change
is apt to produce instability.50 Modernization itself, nevertheless,
is merely a name for an ensemble of complex processes.5' Unless
' Huntington,chaps.5-6.
SOHuntington,chap.i, and the referencescited in n.39 above.
5 The difficulties in defining modernization are well known; cf. Rich-

ard Rose, "Modern Nations and the Study of Political Modernization,"

in Stein Rokkan, ed., ComparativeResearch Across Cultures and Nations

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
 THEORIES OF REVOLUTION | 49

these are disentangledand their effects separatelytested and veri-

fied, it is difficultto see how reliableconclusionscan be drawnas
to their connection with revolution. If we follow Huntington,
modernizationleads to revolutionvia the politicalmobilizationit
engenders."Revolution,"he writes, "is the extremecase of the ex-
plosion of politicalparticipation."52 But while it is true that great
revolutionsare accompaniedby unprecedentedmass involvement
in politics,how can the fact of modernizationaccountfor the sup-
port of a revolutionby groupsalreadyinvolved-by intellectuals,
bourgeoisies,and dissidentaristocracies,for example,as has often
been the case?The alienationof participantsin politicallife, and
the withdrawalof their support from existing regimes, is fre-
quentlyof profoundimportancein the inceptionof revolution,yet
it is a phenomenonon which Huntington'stheory throws little
light.
A commonfeaturein explanationsof revolutionsis to attribute
them to some incompatibilityor stress within the ongoing insti-
tutions and processesof society; the most general theory of this
kind has been developedby ChalmersJohnson.53Johnson'sstart-
ing point is the model of a functioningsocial system in a state of
equilibrium.Equilibriumin this case does not mean the absence
of changebut rathersynchronizationbetweenthe differentsectors
of the system as changeproceeds.Here Johnsonis largely depen-
dent for his ideas on the sociologicaltheories of Talcott Parsons
concerningthe requirementsof a functioningsocial system. The
process of revolutionaccordinglybegins with disequilibriumand
dysfunction.These may ensue from changes in both values and
the environmentand such changes may originate either from
within the system or be introducedfrom without. Dysfunction
signifies a lack of integrationbetween the sectors or subsystems
of the social system as a whole. It occurswhen the changetaking
place is too much for the adaptiveand adjustiveprocessesof the
system. If dysfunctionis not dealt with by purposefulmeasuresto
reestablishequilibrium,then it will multiply with widening ef-
(Paris, 1968). On the diverse components of modernization, cf. Daniel
Lerner, "Comparative Analysis of the Processes of Modernization," in
Rokkan. A general treatment is Cyril E. Black, The Dynamics of Modern-
ization (New York,1968).
2 Huntington,266.
" See Revolution and the Social System and RevolutionaryChange.

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
50 1 POLITICAL SCIENCE QUARTERLY

fects of disorientation and protest within the society. At this point

the action of elites is crucial. If the elite relies on force to cope
with the pressures of dysfunction, the result will be what Johnson
calls a power deflation. On the other hand, if the elite fails to in-
stitute policies capable of reintegrating the system, it will suffer a
loss of authority. Sheer intransigence on its part may also be fatal.
Multiple dysfunction and elite inefficiency create a revolutionary
situation. What then makes revolution highly probable or certain,
according to Johnson, is the intervention of some unforeseen trig-
gering factor or accelerator.This could be an event such as a defeat
in war or any other incident that undermines an army's loyalty to
a regime and persuades revolutionaries that they have the chance
to win power. Johnson's formula can therefore be summarized as
follows: change plus disequilibrium plus multiple dysfunction
plus elite failure plus an acceleratorlead to revolution.
To account for the varieties of revolution, Johnson has also de-
vised a classification based on four criteria: the targets of revolu-
tionary activity, whether government, regime, or society; the
identity of the revolutionaries, whether elites, masses, or both;
the goals of the revolutionary ideology, whether reformist, nation-
building, eschatological, etc.; finally, whether the revolution is a
spontaneous or a calculated movement. With the help of these cri-
teria, Johnson then distinguishes six types of revolution. The first
is the jacquerie, a spontaneous peasant outbreak that does not chal-
lenge the legitimacy of the regime and appeals to traditional auth-
orities for redress of grievances. The second type is the millenarian
rebellion, which may be actuated by the same grievances as the
jacquerie, but aims at a totally new order with the expected aid of
supernatural forces. The third type is the anarchistic rebellion,
which is a reaction against change or modernization and seeks to
reinstitute an idealized older order. The fourth type, which is also
the rarest, is the Jacobin-communist or great revolution and is the
one that leads to the most far-reaching changes. The fifth type is
the conspiratorial coup d'etat, an attempt by a small elite to intro-
duce social change by violence. Finally, the sixth type is the mili-
tarized mass insurrection, a creation of the twentieth century,
which is a people's war of guerilla insurgency led by a dedicated
elite and inspired by nationalist and communist ideologies.
This typology of revolution is the best that has yet been sug-
gested and can be useful in discriminating particular revolutions,

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
 THEORIES OF REVOLUTION | 51

which may, of course,-possess characteristicsof more than one

type. It cannot, however, be regardedas complete.There is no
place in it, for example, for the religious-civilwars of the six-
teenth and seventeenthcenturiesled by aristocraciesnor for the
separatistrebellionsor urbanrevolts of the same period.Another
flaw is that the Jacobin-communist type of revolutionis mislead-
ingly named, since some great revolutionsin this category, like
the MexicanRevolutionor the Nazi conquestof power, were nei-
ther Jacobinnor communist.One might also consider whether
what Johnsoncalls the militarizedmass insurrectiondoes not real-
ly belong to the type of great revolutions and differs only in
makinguse of new politicalandmilitarystrategies.
Turningfrom this typology to Johnson'smodel as a whole, it
representsthe most ambitious attempt made thus far to apply
social-systemstheory to revolution.Its principaldefectsare due to
its too abstractand overly general character.The assumption,for
instance, that a nonrevolutionarysociety is a system in equilib-
rium is one that arouses considerableskepticism.Has any such
system ever existed?Surely it assumesthat societies are far more
integratedthan they possibly can be. A more plausibleview is
that imperfectintegrationand some degree of disequilibriumare
characteristicof every social system. But such talk is, in any case,
unilluminatingin the absence of definite criteria for specifying
equilibriumand disequilibrium.The criteria that Johnson pro-
poses for this purposeare unfortunatelyof little help. Thus, equi-
librium is describedas a condition of synchronizationbetween
values and the social division of labor.54Such a formula,however,
suffers from extremevagueness unless clear indicatorsare given
by which to identify and distinguishthis conditionfrom its op-
posite. In the case of disequlibrium,Johnsonmentions somewhat
more precisecriteria,such as a rising suiciderate or increasedide'
ological activity.55 The drawback here, however, is that we have
no reason to accept these factors as necessarily reflecting disequi-
librium unless we first accept the theory they are supposed to con-
firm. Problems arise as well with regard to the conception of dys-
function. Every social system contains many instances of poor
governmental performance, elite failure, and out-of-date institu-
tions. Since these are all presumably dysfunctional, how to recog-
5 Johnson,RevolutfonaryChange,56, and chap.3.
rz
Ibid., chap.6.

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions
52 | POLITICAL SCIENCE QUARTERLY

nize dysfunctionas a special-conditionresulting from disequilib-

rium becomes a puzzling question. Considerationsof this kind
suggest that equilibrium,understoodeven as an ideal conditionto
which actualsystems are merely in some approximaterelation,is
not a good foundationon which to erecta theory of revolution.A
final point is that Johnson'swork concentratesso heavily on the
social system that it is relativelyuninformativeabout the specific
conflictspresentin revolution,the genesis of revolutionaryopposi-
tion, or the actions undertakenby governmentsand rebels. In
placinghimself on the heights of generaltheory,he has not avoid-
ed the dangerof making his fundamentalpropositionsso expan-
sive that they verge towardemptinessand thus fail to entail the
concretecorrelativesby whichtheiradequacycanbe tested.

IV
After this review of theoriesof revolution,the main conclusionto
be drawnis that the subjectis in a lively but disorderlystate. A
numberof promisingideas have been propounded,but they do not
fit into any coherentscheme.Whethera generaltheory of revolu-
tion is attainableis a questionone might well wish to ask, in view
of the disagreementsand lack of success of past and present ef-
forts. Of course, no answer to this question can be given on a
priori grounds, and every theoreticalcontributionshould be as-
sessed on its merits. Nevertheless,I admit to considerableskepti-
cism in the matter.As an historian,I am inclined to believe that
above a certainlevel of generalitysocial-sciencetheory is too re-
mote from reality to be eithervery interestingor useful. A global
approachtoward the problemof revolutionseems to me, there-
fore, less likely to be profitablethan one that strives, for instance,
to clarifythe differenttypes of revolutionand to developa model
of the causalprocessof each type. Yet despitethe reservationshis-
toriansare apt to feel about the utility of grandtheory, they need
to take accountof whatevertheoriesare availableand to try to use
and improvethem. In the presentposition of history, the union of
theoreticaland comparativeinvestigationwith the more tradition-
al methods of inquiry is one of the most importantlines of ad-
vancetowarda deeperunderstandingof thepast.

This content downloaded from 169.230.243.252 on Thu, 19 Mar 2015 14:42:46 UTC
All use subject to JSTOR Terms and Conditions