The Element of Surprise in the East European Revolution of 1989 Author(S): Timur Kuran Source: World Politics, Vol

Trustees of Princeton University

Now Out of Never: The Element of Surprise in the East European Revolution of 1989 Author(s): Timur Kuran Source: World Politics, Vol. 44, No. 1 (Oct., 1991), pp. 7-48 Published by: Cambridge University Press Stable URL: http://www.jstor.org/stable/2010422 . Accessed: 05/06/2011 12:52

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=cup. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

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

Cambridge University Press and Trustees of Princeton University are collaborating with JSTOR to digitize, preserve and extend access to World Politics.

http://www.jstor.org NOW OUT OF NEVER The Elementof Surprisein theEast EuropeanRevolution of 1989

By TIMUR KURAN*

I. UNITED IN AMAZEMENT UR jaws cannot drop any lower," exclaimed Radio Free Eu- rope one day in late 1989. It was commentingon the electrify- ing collapse of Eastern Europe's communistregimes.' The politicallandscape of the entire region changed suddenly,astonishing even the most seasoned political observers.In a matter of weeks entrenchedleaders were overthrown,the communistmonopoly on power was abrogated in one countryafter another, and persecutedcritics of the communistsystem were catapulted into high office. In the West the ranks of the stunnedincluded champions of the view thatcommunist totalitarianism is substantiallymore stablethan ordinary authoritarianism.2"It has to be conceded," wrote a leading proponentof this view in early 1990, "that those of us who distinguishbetween the two non-democratictypes of governmentunderestimated the decay of Communist countriesand expectedthe collapse of totalitarianismto take longer than has actually turned out to be the case."3 Another acknowl- edged her bewildermentthrough the titleof a new book: The Withering Away of the TotalitarianState. . . And OtherSurprises.4

* This researchwas supportedby the National Science Foundation under grantno. SES- 8808031. A segmentof the paper was draftedduring a sabbatical,financed partly by a fellow- ship fromthe National Endowmentfor the Humanities,at the Institutefor Advanced Study in Princeton.I am indebtedto WolfgangFach, Helena Flam, JackGoldstone, Kenneth Ko- ford,Pavel Pelikan, Jean-PhilippePlatteau, Wolfgang Seibel, Ulrich Witt,and threeanon- ymous readersfor helpful comments. 1 BernardGwertzman and Michael T. Kaufman,eds., The Collapseof Communism,by the Correspondentsof "The New YorkTimes" (New York: Times Books, 1990),vii. 2 For an earlystatement of thisthesis, see Hannah Arendt,The Originsof Totalitarianism, 2d ed. (1951; reprint,New York: World Publishing,1958), pt. 3. Arendt suggested that communismweakens interpersonalbonds rooted in family,community, religion, and pro- fession,a situationthat makes individualsterribly dependent on the goodwill of the stateand thus blocks the mobilizationof an anticommunistrevolt. ' Richard Pipes, "Gorbachev's Russia: Breakdown or Crackdown?" Commentary,March 1990,p. 16. 4Jeane J.Kirkpatrick, The WitheringAway of the TotalitarianState ... And OtherSurprises (Washington,D.C.: AEI Press, 1990). A decade earlierKirkpatrick had articulateda variant of Arendt's thesis,insisting that the communistsystem is incapable of self-propelledevolu-

WorldPolitics 44 (October 1991),7-48 8 WORLD POLITICS Even scholarswho had rejectedthe conceptof a frozenand immobile region were amazed by the eventsof 1989. In 1987 the American Acad- emy of Arts and Sciences invited a dozen specialists,including several living in Eastern Europe, to prepare interpretiveessays on East Euro- pean developments.As the Daedalus issue featuringthese essays went to press,the uprisingstook off,prompting many authorsto change "whole sentences and paragraphs in what were once thoughtto be completed essays." Daedalus editor Stephen Graubard remarksin his prefaceto the issue: "A quarterly journal has been obliged to adapt, inconveniently, but in some measure necessarily,the techniques of a weekly or even a daily newspaper."5 Graubard proudly points out that even before the last-minuterevisions the essays offeredremarkable insights into the in- tellectual,social, and political stirringsthat were transformingthe region. But he concedes thatneither he nor his essayistsforesaw what was to happen. Recalling that in a planning session he had asked whether anything could be done to avoid publishing "an issue that will seem 'dated' three years afterpublication," he continues: "Was this passage a premonitionof all that was to follow? One wishes that one could claim such extraordinaryprescience. Regrettably, it did not reallyexist."6 Wise statesmen,discerning diplomats, and giftedjournalists were also caught off guard. So too were futurologists.John Naisbitt's celebrated Megatrends,which sold eight million copies in the early 1980s,does not predict the fall of communism.7As the Economistobserved even before the East European Revolution had run its course, 1989 turned out to be a year when "the most quixotic optimists"were repeatedly"proved too cautious."8 Within Eastern Europe itselfthe revolutioncame as a surpriseeven to leading "dissidents."In a 1979 essay, "The Power of the Powerless," Vaiclav Havel recognized that the regimesof Eastern Europe were anythingbut invincible.They mightbe toppled,he wrote,by a "social movement, an "explosion of civil unrest,"or a "sharp conflictinside an ap- tion. See Kirkpatrick,"Dictatorships and Double Standards,"Commentary, November 1979, pp. 34-45. 5 Graubard, "Preface to the Issue 'Eastern Europe ... Central Europe ... Europe,' Daedalus 119 (Winter 1990),vi. 6 Ibid., ii. 7 Naisbitt,Megatrends: Ten New DirectionsTransforming Our Lives (New York: Warner Books, 1982). The monthsfollowing the East European Revolutionsaw the appearance of John Naisbitt and Patricia Aburdene,Megatrends 2000: Ten New Directionsfor the 1990's (New York: William Morrow, 1990). This sequel characterizesthe East European developments of the late 1980s as an unforeseen"political earthquake" and thenpredicts that the 1990swill witnessthe furthererosion of communism(chap. 3). 8 Economist,November 18, 1989,p. 13. SURPRISE IN THE EAST EUROPEAN REVOLUTION 9 parentlymonolithic power structure,"among other possibilities.9This essay is at once a brilliantprobe into the communist system'sstability and a penetratingprognosis of its ultimatedemise. Yet it steersclear of speculation on the timing of the collapse. It is replete with statements such as "we must see the hopelessnessof tryingto make long-rangepre- dictions" and "far-reachingpolitical change is utterlyunforeseeable," although it ends on a cautiouslyoptimistic note: "What if [the 'brighter future']has been here for a long time already,and only our own blindness and weakness has preventedus fromseeing it around us and within us, and kept us fromdeveloping it?."10 Eight years later Havel himselfwould exhibit "blindness" to events that were ushering in a "brighterfuture." Less than three years before the revolutionhe commented as follows on the rousing welcome given by a Prague crowd to visitingSoviet leader Mikhail Gorbachev:

I feel sad; this nation of ours never learns. How many times has it put all its faith in some external force which, it believed, would solve its problems? . . . And yet here we are again, making exactly the same mistake. They seem to think that Gorbachev has come to liberate them from Hu- sa~k!

In late 1988,with less than a year to go, Havel was stillunsure about the directionof events:

Maybe [the Movement forCivil Liberties]will quickly become an integral featureof our country'slife, albeit one not particularlybeloved of the regime. . . . Perhaps it will remain for the time being merely the seed of something that will bear fruitin the dim and distant future.It is equally possible that the entire"matter" will be stamped on hard.'2

Other Czechoslovak dissidentswere just as unprepared for the revolution. In November 1989 JanUrban suggestedthat the opposition con- testthe national electionsscheduled forJune 1991-only to be ridiculed by his friendsfor making a hopelesslyutopian proposal."3Within a matter of days, they were all celebratingthe fall of Czechoslovakia's com- munistdictatorship.

9 Havel, "The Power of the Powerless" (1979), in Havel et al., The Powerof thePowerless: Citizensagainst the State in Central-EasternEurope, ed. JohnKeane and trans.Paul Wilson (Armonk,N.Y.: M. E. Sharpe, 1985),42. 10 Ibid., 87, 89, 96. 11Havel, "Meeting Gorbachev" (1987), in William M. Brintonand Alan Rinzler, eds., WithoutForce or Lies: Voicesfrom the Revolution of CentralEurope in 1989-90 (San Francisco: MercuryHouse, 1990),266. 12 Havel, "Cards on the Table" (1988), in Brintonand Rinzler (fn. 11), 270-71. 13 Sidney Tarrow, " 'Aiming at a Moving Target': Social Science and the Recent Rebel- lions in Eastern Europe," PS: PoliticalScience and Politics24 (March 1991), 12. 10 WORLD POLITICS A few months beforethe revolution,in neighboringPoland negotia- tions were under way between the communistregime and Solidarity,the trade union that for years had been demanding political pluralism. To the surpriseof almost everyone,the regimeagreed in April 1989 to hold open elections for a pluralistic parliament. In elections scheduled for June all 100 Senate seats and 161 of the 460 Assembly seats would be contestable.Exceeding the wildest expectations,Solidarity won all but one of the Senate seats in addition to all of the Assembly seats it was allowed to contest.Stunned by the enormityof this success, Solidarity officialsworried that the electoratehad gone too far,that victorywould forceSolidarity into making bold politicalmoves simplyto satisfyraised hopes. They feared thatsuch moves would provoke a communistcrack- down. The significantpoint is thatneither the governmentnor Solidarity was prepared for such a lopsided result.The April accord was designed to give Solidaritya voice in Parliament,not to substantiateand legitimate its claim to being the voice of the Polish people.14 We will never know how many East Europeans foresawthe eventsof 1989-or at least the impending changes in theirown countries.But at each step, journalisticaccounts invariablypainted a pictureof a stunned public. For example, two days afterthe breachingof the Berlin Wall, the New York Times carried an article in which an East German remarks: "It's unfathomable.If you had told me that one week ago, I wouldn't have believed it. Mentally,I stillcan't. It will take a few days beforewhat this means sinks in.""5 I know of only one systematicstudy of relevance.Four months after the fall of communismin East Germany,the Allensbach Instituteasked a broad sample of East Germans: "A year ago did you expect such a peaceful revolution?" Only 5 percent answered in the affirmative,although 18 percent answered "yes, but not that fast." Fully 76 percent indicated that the revolutionhad totallysurprised them.16 These figures are all the more remarkable given the "I knew it would happen" fal-

14 On the electionsand the reactionsthey generated, see the reportsof JohnTaglibue, New YorkTimes, June 3-6, 1989. The eventsleading up to the April accord have been chronicled and interpretedby Timothy Garton Ash, "Refolution:The Springtimeof Two Nations," New YorkReview of Books,June 15, 1989,pp. 3-10. He observed: "Almost no one imagined that the great gulf between 'the power' and 'the society,'between Jaruzelskiand Walesa, could be so swiftlybridged" (p. 6). For another informativeaccount of Poland's political transformation,see Elie Abel, The ShatteredBloc: Behind the Upheaval in EasternEurope (Boston: Houghton Mifflin,1990), chap. 4. 15 New YorkTimes, November 12, 1989,p. 1. 16 Question 36 on the East German Surveyof the Institutfur Demoskopie Allensbach, February 17-March 15, 1990, Archive no. 4195 GEW. I am indebtedto Elisabeth Noelle- Neumann, directorof the institute,for agreeing to insertthis question into a broader survey on East German politicalopinions. SURPRISE IN THE EAST EUROPEAN REVOLUTION 11 lacy-the human tendencyto exaggerateforeknowledge."7 Even trained historianssuccumb to thisfallacy, portraying unanticipated events as inevitable, foreseeable,and actually foreseen."8In view of this fallacy,if East Germans had been asked a year before the revolution,"Do you expect a revolutionin a year's time?" the percentageof unqualified negative answers would undoubtedlyhave been even higher. The events that sealed the fateof East Germany's communistregime took offin the final days of summer,when thousands of East German vacationers in Hungary took advantage of relaxed border controls to turntheir trips into permanentdepartures for West Germany. The East German governmentresponded by restrictingits citizens'access to Hun- gary, only to see thousands show up at the West German embassy in Prague. In the ensuing days it acceded to a seriesof face-savingarrange- mentsby which the vacationerscould depart forthe West, but only after firstreturning home. Each new concession prompted furtherwaves of emigrants,however, confuting the government'sexpectation that the exodus would taper offquickly.19 The governmentwas not alone in failing to anticipatewhere eventswere headed. Thousands of East German citizens rushed to join the exodus preciselybecause theyfelt their chances of reaching the West would never again be so good. Had they known that the Berlin Wall was about to come down, few would have left in such haste, leaving behind almost all their possessions,including their cars. It might be said that some veryknowledgeable observersof the communist bloc had predictedits disintegrationbefore the centurywas out. As earlyas 1969,for instance, the Soviet dissidentAndrei Amalrik wrote that the Russian Empire would break up within a decade and a half. Although it is temptingto credit Amalrik with exemplaryforesight, a rereadingof his famous essay shows thathe expected the Soviet Empire to meet its end followinga protractedand devastatingwar with China, not through a stringof popular upheavals. In fact,he explicitlystated that the Soviet systemof governmenthad left people too demoralized

17 Baruch Fischhoff,"Hindsight * Foresight: The Effectof Outcome Knowledge on Judgmentunder Uncertainty,"Journal of ExperimentalPsychology: Human Perceptionand Performance1 (August 1975), 288-99; and Baruch Fischhoffand Ruth Beyth," 'I Knew It Would Happen'-Remembered Probabilitiesof Once-Future Things," OrganizationalBe- haviorand Human Performance13 (February1975), 1-16. 18 David Hackett Fischer,Historians' Fallacies: Towarda Logic ofHistorical Thought (Lon- don: Routledgeand Kegan Paul, 1971),chaps. 6-8. 19For a compilationof pertinentreports from the New YorkTimes, see Gwertzman and Kaufman (fn. 1), 153-84. Superb eyewitnessaccounts include Timothy Garton Ash, "The German Revolution,"New YorkReview of Books, December 21, 1989,pp. 14-19; and George Paul Csicsery,"The Siege of Nogradi Street,Budapest, 1989," in Brintonand Rinzler (fn. 11), 289-302. 12 WORLD POLITICS and too dependent on authorityto participatein a spontaneous uprising.20 So Amalrik did not reallyforesee the eventsof 1989. Like a broken watch thattells the correcttime everytwelve hours,he got the timingof the firstcrack in the empire essentiallyright, but on the basis of a spu- rious forecastof events. This is not to suggestthat the East European explosion came as total surpriseto everyone.Though most were astonishedwhen it happened, and though few who saw it coming expectedit to be so peaceful,a small number of commentatorshad prophesied that the revolutionwould be swift and remarkably bloodless. Havel, despite his above-quoted remarks, is one of these. And Vladimir Tismaneanu, a Romanian emigre livingin the United States,came close to predictingmajor change. About a year before the collapse of the Romanian regime, he depicted it as "probably the most vulnerable" in Eastern Europe. Sensing an "all-pervasive discontent,"he observedthat "the Brasov riotsin November 1987, when thousands of citizens took to the streets,chanted anti-Ceausescu slogans and burned the dictator'sportraits, represent an unmistakable signal for Moscow that uncontrollableviolence may flare up in Roma- nia."21Tismaneanu failed to place the Romanian uprisingin the context of an upheaval spanning all of the Soviet Union's Warsaw Pact allies. Nor did he predict that Romania would be the last Soviet satellite to overthrow its government.It is remarkable nonethelessthat he diag- nosed the Romanian regime's vulnerability.Like Havel, he succeeded where many Western observersfailed, because he understoodthe weaknesses thatunderlay the apparentstability of the communistsystem. This understandingprepared him for the type of explosion that eventually occurred,although, as discussed furtheron, it did not endow him with the abilityto predictwhen the revolutionwould break out. While the collapse of the post-World War II politicalorder of Eastern Europe stunned the world, in retrospect it appears as the inevitable con- sequence of a multitudeof factors.In each of the six countriesthe leadership was generallydespised, loftyeconomic promisesremained unful- filled,and freedomstaken forgranted elsewhere existed only on paper. But if the revolutionwas indeed inevitable,why was it not foreseen? Why did people overlook signs thatare clearlyvisible after the fact?One of the central argumentsof this essay is preciselythat interactingsocial and psychologicalfactors make it inherentlydifficult to predictthe out- 20 Amalrik,Will the Soviet Union Survive until 1984? (1969) (New York: Harper and Row, 1970),esp. 36-44. 21 Tismaneanu, "Personal Power and Political Crisis in Romania," Governmentand Op- position24 (Spring 1989), 193-94. SURPRISE IN THE EAST EUROPEAN REVOLUTION 13 come of politicalcompetition. I shall argue that the East European Rev- olution was by no means inevitable. What was inevitable is that we would be astounded if and when it arrived. "The victim of today is the victorof tomorrow,/ And out of Never grows Now!"22 Brecht's couplet capturesperfectly our central paradox: seeminglyunshakable regimes saw public sentimentturn against them with astonishingrapidity, as tinyoppositions mushroomed into crushing majorities. Currentlypopular theoriesof revolutionoffer little insight into this stunningpace; nor for that matterdo they shed light on the element of surprise in previous revolutions.All lay claim to predictive power, yet none has a track record at veritableprediction. The next section brieflycritiques the pertinentscholarly literature. Without denying the usefulnessof some received theoriesat explaining revolutionsof the past, I go on to present a theorythat illuminates both the process of revolutionarymobilization and the limitsof our abilityto predictwhere and when mobilizations will occur. Subsequent sections apply this argument to the case at hand. The termrevolution is used here in a narrow sense to denote a mass- supported seizure of political power that aims to transformthe social order. By thisdefinition it is immaterialwhether the accomplishedtrans- fer of power brings about significantsocial change. With regard to the East European Revolution,it is too early to tell whetherthe postrevolu- tionaryregimes will succeed in reshapingthe economy,the legal system, internationalrelations, and individual rights-to mention just some of the domains on the reformistagenda. But even if the ongoing reforms all end in failure,the upheavals of 1989 can continueto be characterized as a regionwide revolution.

II. RECEIVED THEORIES OF REVOLUTION AND THEIR PREDICTIVE WEAKNESSES

In her acclaimed book Statesand Social Revolutions,Theda Skocpol treats social revolutions as the product of structuraland situational conditions.23Specifically, she argues thata revolutionoccurs when two conditions coalesce: (1) a state's evolving relationswith other statesand local classes weaken its abilityto maintain law and order, and (2) the elites harmed by this situationare powerless to restorethe statusquo ante yet

22 BertoltBrecht, "Lob der Dialectic" (In praiseof dialectics,1933), in Gedichte(Frankfurt: Suhrkamp Verlag, 1961),3:73; poem translatedby Edith Anderson. 23 Skocpol,States and Social Revolutions:A ComparativeAnalysis of France, Russia, and China (Cambridge: Cambridge UniversityPress, 1979). 14 WORLD POLITICS strongenough to paralyze the government.Through theirobstruction- ism the elitesgenerate a burstof antielitesentiment, which setsin motion an uprisingaimed at transformingthe social order. The appeal of Skoc- pol's theorylies in its invocationof structuralcauses to explain shiftsin the structureof political power. It does not depend on such "subjective" factors as beliefs, expectations,attitudes, preferences, intentions, and goals, although these do creep into structuralistcase studies, including those of Skocpol herself. Tracking emotionsand mental statesis a treacherousbusiness, which is why the structuralistschool considers it a virtue to refrainfrom ap- pealing to them. Social structuresare ostensiblyeasier to identify,which would seem to endow the structuralisttheory with predictivesuperiority over "voluntarist"theories based on "rational choice." Theories that fall under the rubric of rational choice have certainlybeen unsuccessfulat predictingmass upheavals. What theyexplain well is the rarityof popular uprisings.24The crucial insightof the rational-choiceschool is that an individual opposed to the incumbentregime is unlikelyto participate in effortsto remove it, since the personal risk of joining a revolutionary movement could outweigh the personal benefitthat would accrue were the movement a success. It is generallyin a person's self-interestto let others make the sacrificesrequired to secure the regime's downfall, for a revolution constitutesa "collective good" -a good he can enjoy whether or not he has contributedto its realization. With most of the regime's opponents choosing to freeride, an upheaval may fail to materialize even if the potential revolutionariesconstitute a substantialma- jority.Yet fromtime to time revolutiondoes break out, and thispresents a puzzle that the standard theoryof rational choice cannot solve. The standard theorysimply fails to make sense of why the firstpeople to challenge the regime choose selflesslyto gamble with theirlives.25 With respectto the East European Revolutionin particular,the standard theoryilluminates why, for all theirgrievances, the nations of the region were remarkablyquiescent forso many years.It does not explain why in 1989 theirdocility suddenly gave way to an explosivedemand for change. For its part, the structuralisttheory elucidates why the revolution broke out at a timewhen the Soviet Union was emittingincreasingly convincingsignals that it would not use forceto tryto preservethe East

24 The seminal contributionis Mancur Olson, The Logic of CollectiveAction: Public Goods and the Theoryof Groups(1965; rev. ed., Cambridge: Harvard UniversityPress, 1971). 25 This point is developed by Michael Taylor, "Rationalityand RevolutionaryAction," in Taylor, ed., Rationalityand Rcvolution(Cambridge: Cambridge UniversityPress, 1988),63- 97. Taylor also offersan illuminatingcritique of structuralism. SURPRISE IN THE EAST EUROPEAN REVOLUTION 15 European statusquo. But it explains neitherwhy the old order collapsed so suddenly in several countriesat once nor why the eventsof 1989 out- distanced all expectations. Neither school has come to termswith its predictiveweaknesses. That granted,can the deficienciesin question be overcome by incorporating additional relationshipsinto these theories?It would seem, on the basis of reasons developed below, thatperfect predictability is an unachievable objective. The theorydeveloped here accommodates some of the major featuresand implicationsof these two theories,with the added virtue, however,of illuminatingwhy major revolutionscome as a surpriseand why,even so, theyare quite easily explained afterthe fact. Like all unanticipatedrevolutions, the East European Revolution is generatingmultitudes of retrospectiveexplanations that draw attention to its diversecauses and warningsigns. To cite just one example, an essay writtenshortly after the fall of the East German regime begins with a flashbackto April 1989: two passengerson an East German train,mutual strangers,share with each othertheir negative feelings about the regime, within earshot of others-a highly uncommon event, because of the ubiquityof informants.This opening gives the impressionthat East Ger- many was obviouslyreaching its boiling point, although the rest of the essay makes clear that the East German uprising was in fact scarcely anticipated.26Like so much else now rolling off the presses,this essay leaves unexplained why events seen in retrospectas harbingersof an imminentupheaval were not seen as such beforethe actual revolution. Not that signs noticed in retrospectare necessarilyfabrications. The availabilityheuristic, a mental shortcutwe use to compensatefor our cog- nitivelimitations, highlights information consistent with actual eventsat the expense of informationinconsistent with them.27Accordingly, events considered insignificantwhile the regime looked stable may suddenly gain enormous significanceafter it falls. Among all the events that are consistentwith a particularoutcome, those thatfit into the models at our disposal will be the ones that attractattention. Thus, a structuralistwill be predisposed to treatas significantthe structuralsigns of the coming revolution.These signs need not be imaginary,but there is nothing in

26 Edith Anderson, "Town Mice and CountryMice: The East German Revolution,"in Brintonand Rinzler (fn. 11), 170-92. 27 On the availabilityheuristic, see Amos Tverskyand Daniel Kahneman, "Availability: A Heuristic for JudgingFrequency and Probability,"Cognitive Psychology 5 (September 1973), 207-32. The biases that this heuristicimparts to the use of historicalknowledge are discussedby ShelleyE. Taylor, "The AvailabilityBias in Social Perceptionand Interaction," in Daniel Kahneman, Paul Slovic, and Amos Tversky,eds., Judgmentunder Uncertainty: Heuristicsand Biases (Cambridge: Cambridge UniversityPress, 1982), 190-200. 16 WORLD POLITICS the structuralisttheory-or, for that matter,in the standard theoryof rationalchoice-that explains why it is betterat explanationthan at prediction. This paradox is seldom appreciated,partly because the authors of retrospectiveaccounts do not always concede theirown bafflement. They generallywrite as though theirfavored theoryshows the revolution to have been inevitable,seldom pausing to explain why,if this is so, theythemselves had not offeredunambiguous, unequivocal forecasts. If one bete noire of the structuralistschool is the rational-choiceap- proach to the study of revolutions,another is the relative-deprivation approach. According to this thirdapproach revolutionsare propelled by economic disappointments,that is, by outcomes that fall shortof expectations.If the consequentdiscontent becomes sufficientlywidespread, the resultis a revolt.28With respectto the major revolutionsshe investigates, Skocpol correctlyobserves that theybegan at times when levels of dis- contentwere by historicalstandards not unusual. More evidence against the relative-deprivationtheory comes fromCharles Tilly and his associ- ates, who find that in France the level of collective violence has been uncorrelated with the degree of mass discontent.29Thus, the relativedeprivation theoryneither predicts nor explains. The reason is simple. While relative deprivation is doubtless a factorin every revolution in history,it is too common in politicallystable societiesto provide a com- plete explanation for everyobserved instability.By implication,to treat relativedeprivation as an unmistakablesign of impending revolutionis to subject oneselfto a continuousstring of alarms,mostly false.

III. PREFERENCE FALSIFICATION AND REVOLUTIONARY BANDWAGONS

So mass discontent does not necessarilygenerate a popular uprising against the political statusquo. To understandwhen it does, we need to identifythe conditionsunder which individuals will display antagonism toward the regime under which theylive. Afterall, a mass uprising results frommultitudes of individual choices to participatein a movement for change; there is no actor named "the crowd" or "the opposition."

28 For two of the major contributionsto this approach, see JamesC. Davies, "Toward a Theory of Revolution,"American Sociological Review 27 (February1962), 5-19; and Ted R. Gurr, WhyMen Rebel (Princeton:Princeton University Press, 1970). 29 David Snyderand Charles Tilly, "Hardship and CollectiveViolence in France, 1830 to 1960," AmericanSociological Review 37 (October 1972), 520-32; and Charles Tilly, Louise Tilly,and RichardTilly, The RebelliousCentury: 1830-1930 (Cambridge: Harvard University Press, 1975). For much additional evidence against the theoryof relativedeprivation, see Steven E. Finkel and JamesB. Rule, "Relative Deprivationand Related PsychologicalThe- ories of Civil Violence: A CriticalReview," in Louis Kriesberg,ed., Researchin Social Move- ments,Conflicts and Change(Greenwich, Conn.: JAI Press, 1986),9:47-69. SURPRISE IN THE EAST EUROPEAN REVOLUTION 17 The model presented here is in agreement with the rational-choice school on this basic methodologicalpoint, although it departs in impor- tant ways fromthe standard farein rational-choicemodeling. Consider a societywhose membersare indexed by i. Each individual member must choose whether to support the governmentin public or oppose it; depending on his public acts and statements,each person is perceived as eithera friendof the governmentor an enemy,for the political statusquo or against. In private,of course,a person may feel torn between the governmentand the opposition,seeing both advantages and disadvantages to the existingregime. I am thus distinguishingbetween an individual'sprivate preference and public preference.The formeris ef- fectivelyfixed at any given instant,the lattera variable under his control. Insofaras his two preferencesdiffer-that is, the preferencehe expresses in public diverges from that he holds in private-the individual is en- gaged in preferencefalsification. Let S representthe size of the public opposition,expressed as a per- centage of the population. Initiallyit is near 0, implyingthat the government commands almost unanimous public support. A revolution,as a mass-supportedseizure of political power, may be treated as an enormous jump in S. Now take a citizen who wants the governmentoverthrown. The likelyimpact of his own public preferenceon the fateof the government is negligible: it is unlikelyto be a decisive factorin whetherthe government stands or falls.But it may bringhim personal rewards and impose on him personal punishments.If he chooses to oppose the government, for instance, he is likely to face persecution,though in the event the governmentfalls his outspokennessmay be rewarded handsomely.Does this mean that our individual will base his public preferencesolely on the potential rewards and punishmentsflowing from the two rival camps? Will his privateantipathy to the regime play no role whatsoever in his decision? This does not seem reasonable,for history offers count- less examples of brave individuals who stood up for a cause in the face of the severestpressures, including torture. On what, then,will our disaffectedindividual's choice depend? I sub- mit that it will depend on a trade-offbetween two payoffs,one external and the other internal.30 The externalpayoff to siding with the oppositionconsists of the just- discussed personal rewardsand punishments.In net terms,this payoffis apt to become increasinglyfavorable (or increasinglyless unfavorable) 30 For a detailed analysisof this trade-off,see Timur Kuran, "Privateand Public Prefer- ences,"Economics and Philosophy6 (April 1990), 1-26. 18 WORLD POLITICS with S. The larger S, the smaller the individual dissenter'schances of being persecutedfor his identificationwith the oppositionand the fewer hostile supportersof the governmenthe has to face. The latterrelation- ship reflectsthe fact that governmentsupporters, even ones privately sympatheticto the opposition,participate in the persecutionof the government'sopponents, as part of their personal effortsto establish con- vincing progovernmentcredentials. This relationshipimplies that a rise in S leaves fewer people seeking to penalize membersof the public opposition. The internalpayoff is rooted in the psychologicalcost of preference falsification.The suppression of one's wants entails a loss of personal autonomy,a sacrificeof personal integrity.It thus generateslasting dis- comfort,the more so the greaterthe lie. This relationshipmay be cap- tured by postulatingthat person i's internalpayoff for supportingthe oppositionvaries positivelywith his privatepreference, xA. The higherxi, the more costlyhe findsit to suppresshis antigovernmentfeelings. So i's public preferencedepends on S and xi. As the public opposition grows, with his private preferenceconstant, there comes a point where his externalcost of joining the oppositionfalls below his internalcost of preferencefalsification. This switchingpoint may be called his revolu- tionarythreshold, P. Since a thresholdrepresents a value of S, it is a number between 0 and 100. If xi should rise,7 will fall. In otherwords, if the individual becomes more sympatheticto the opposition,it will take a smaller public opposition to make him take a stand against the government.The same will be true if the governmentbecomes less efficient,or the oppositionbecomes more efficient,at rewarding its supportersand punishing its rivals. In fact,anything that affects the relationshipbetween S and the individual's externalpayoff for supporting the oppositionwill change his revolutionary threshold.Finally, P will fall ifi develops a greaterneed to stand up and be counted, for the internalcost of preferencefalsification will then come to dominate the externalbenefit at a lower S.31 This simple frameworkoffers a reason why a person may choose to voice a demand for change even when the price of dissent is very high and the chances of a successfuluprising very low. If his privateopposi-

31 The theoryoutlined in this sectionis developed more fullyin Timur Kuran, "Sparks and Prairie Fires: A Theory of UnanticipatedPolitical Revolution," Public Choice 61 (April 1989),41-74. A summaryof the presentformulation was deliveredat the annual convention of the AmericanEconomic Association,Washington, D.C., December 28-30, 1990.This pre- sentationappeared under the title"The East European Revolutionof 1989: Is It Surprising That We Were Surprised?"in theAmerican Economic Review, Papers and Proceedings81 (May 1991), 121-25. SURPRISE IN THE EAST EUROPEAN REVOLUTION 19 tion to the existingorder is intenseand/or his need for integrityis quite strong,the sufferinghe incurs for dissent may be outweighed by the satisfactionhe derives from being true to himself.In every society,of course,there are people who go againstthe social orderof the day. Joseph Schumpeteronce observed that in capitalistsocieties this group is domi- nated by intellectuals.Their position as "onlookers" and "outsiders" with much time for deep reflectioncauses them to develop a "critical attitude" toward the statusquo. And because of the high value theyat- tach to self-expression,they are relativelyunsusceptible to social pressures.32The same argumentapplies to noncapitalistsocieties. As a case in point,a disproportionatelylarge share of the East European dissidents were intellectuals. Returningto the general model, we can observe that individuals with differentprivate preferencesand psychologicalconstitutions will have differentrevolutionary thresholds. Imagine a ten-personsociety featur- ing the thresholdsequence A = {0, 20, 20, 30, 40, 50, 60, 70, 80, 100}.

Person 1 (T1 = 0) supportsthe opposition regardlessof its size, just as person 10 (T10 = 100) always supportsthe government.The remaining eight people's preferencesare sensitiveto S: depending on its level, they opt forone camp or the other.For instance,person 5 (P = 40) supports the governmentif 0 < S < 40 but joins the oppositionif 40 ' S ' 100. Let us assume that the oppositionconsists initially of a single person,or 10 percentof the population,so S = 10. Because the nine other individuals have thresholdsabove 10, thisS is self-sustaining;that is, it consti- tutesan equilibrium. This equilibrium happens to be vulnerable to a minor change in A. Suppose that person 2 has an unpleasantencounter at some government ministry.Her alienation from the regime rises, pushing her threshold down from20 to 10. The new thresholdsequence is A' = {0, 10, 20, 30, 40, 50, 60, 70, 80, 100}.

Person 2's new thresholdhappens to equal the existingS of 10, so she switchessides, and S becomes 20. Her move into the oppositiontakes the form of tossing an egg at the country'slong-standing leader during a government-organizedrally. The new S of 20 is not self-sustainingbut self-augmenting,as it drives person 3 into the opposition.The higherS of 30 then triggersa fourthdefection, raising S to 40, and this process

32 Schumpeter,Capitalism, Socialism and Democracy,3d ed. (1950; reprint,New York: Harper Torchbooks, 1962),chap. 13. 20 WORLD POLITICS continues until S reaches 90-a new equilibrium. Now the firstnine individuals are in opposition,with only the tenthsupporting the government. A slight shiftin one individual's thresholdhas thus generated a revolutionarybandwagon, an explosivegrowth in public opposition.33 Now consider the sequence

B = {0, 20, 30, 30, 40, 50, 60, 70, 80, 100}, which differsfrom A only in its thirdelement: 30 as opposed to 20. As in the previous illustration,let 1 fall from20 to 10. The resultingse- quence is

B' = {0, 10, 30, 30, 40, 50, 60, 70, 80, 100}.

Once again, the incumbent equilibrium of 10 becomes unsustainable, and S rises to 20. But the opposition'sgrowth stops there,for the new S is self-sustaining.Some governmentsupporters privately enjoy the sight of the leader's egg-splatteredface, but none followsthe egg throwerinto public opposition.We see thata minorvariation in thresholdsmay dras- ticallyalter the effectof a given perturbation.And in particular,an event that causes a revolutionin one settingmay in a slightlydifferent setting produce only a minor decline in the government'spopularity. Neither private preferencesnor the corresponding thresholds are common knowledge. So a societycan come to the brink of a revolution withoutanyone knowing this,not even those with the power to unleash it. In sequence A, for instance,person 2 need not recognize that she has the abilityto set offa revolutionarybandwagon. Even if she senses the commonnessof preferencefalsification, she simplycannot know whether the actual thresholdsequence is A or B. Social psychologistsuse the term pluralisticignorance to describe misperceptionsconcerning distributions of individual characteristics.34In principle,pluralistic ignorance can be mitigatedthrough polls thataccord individualsanonymity. But it is eas- ier to offerpeople anonymitythan to convincethem thatthe preferences they reveal will remain anonymousand never be used against them. In any case, an outwardlypopular governmentthat knows preferencefal- sificationto be pervasive has no interestin publicizing the implied fra-

33Lucid analysesof bandwagon processesinclude Mark Granovetter,"Threshold Models of Collective Behavior,"American Journal of Sociology83 (May 1978), 1420-43; and Thomas C. Schelling,Micromotives and Macrobehavior(New York: W. W. Norton, 1978). 34 Under the termimpression of universality,the conceptwas introducedby Floyd H. All- port,Social Psychology(Boston: Houghton, Mifflin,1924), 305-9. The termpluralistic igno- rancewas firstused by Richard L. Schanck, "A Study of a Communityand Its Groups and InstitutionsConceived of as Behaviorof Individuals,"Psychological Monographs 43-2 (1932), 101. SURPRISE IN THE EAST EUROPEAN REVOLUTION 21 gilityof its support,because this might inspire the disaffectedto bring their antigovernmentfeelings into the open. It has an incentiveto discourage independentpolling and discreditsurveys that reveal unflatter- ing information. We have already seen that the thresholdsequence is not fixed. Any- thing that affectsthe distributionof privatepreferences may alter it, for instance,an economic recession,contacts with other societies,or inter- generational replacement.But whatever the underlyingreason, private preferencesand, hence, the thresholdsequence can move dramatically against the governmentwithout triggering a revolution.In the sequence C = {0, 20, 20, 20, 20, 20, 20, 20, 60, 100} the average threshold is 30, possibly because most people sympathize with the opposition. Yet S = 10 remains an equilibrium. It is true, of course,that a revolutionis more likelyunder C than underA. C features seven individuals with thresholdsof 20, A only one. A ten-unitfall in any one of the seven thresholdswould triggera revolution. The point remains that widespread disapproval of the governmentis not sufficientto mobilize large numbersfor revolutionaryaction. Anti- governmentfeelings can certainlybring a revolutionwithin the realm of possibility,but other conditionsmust come togetherto set it off.By the same token,a revolutionmay break out in a societywhere privatepreferences,and thereforeindividual thresholds,tend to be relativelyunfa- vorable to the opposition.Reconsider the sequence A', where the average thresholdis 46, as opposed to 30 in C. Under A' public opposition darts from10 to 90, whereas under C it remainsstuck at 10. This simple com- parison shows why the relative-deprivationtheory of revolutionhas not held up under empiricaltesting. By treatingthe likelihood of revolution as the sum of the individual levels of discontent,the relative-deprivation theoryoverlooks the significanceof the distributionof discontent.As our comparisonbetween A' and C indicates,one sufficientlydisaffected person with a thresholdof 10 may do more for a revolution than seven individuals with thresholdsof 20. Imagine now that a superpowerlong committedto keeping the local governmentin power suddenlyrescinds this commitment, declaring that it will cease meddling in the internalaffairs of other countries.This is preciselythe type of change to which the structuralisttheory accords revolutionarysignificance. In the presentframework, such a change will not necessarilyignite a revolution.The outcome depends on both the preexistingdistribution of thresholdsand the consequentshifts. Since the postulated change in internationalrelations is likely to lower the ex- 22 WORLD POLITICS pected cost of joining the opposition,people's thresholdsare likelyto fall. Let us say that everythreshold between 10 and 90 drops by 10 units. If the preexistingthreshold sequence were A, B, or C, the resultwould be an explosion in S from10 to 90. But suppose that it were D = {0, 30, 30, 30, 30, 30, 30, 30, 30, 100}.

The structuralshock turnsthis sequence into DI = {0, 20, 20, 20, 20, 20, 20, 20, 20, 100}.

Fully four-fifthsof the population is now willing to switch over to the oppositionbut onlyif someone else goesfirst. No one does, leaving S at 10. Structuralfactors are thus partof the story,yet by no means the whole story.While theycertainly affect the likelihood of revolution,they cannot possibly deliver infalliblepredictions. A single person's reaction to an event of global importancemay make all the differencebetween a massive uprisingand a latentbandwagon that never takes off.So to suggest, as the structuralistsdo, that revolutionsare broughtabout by deep historical forces with individuals simply the passive bearers of these forces is to overlook the potentiallycrucial importance of individual characteristicsof little significancein and of themselves.It is always a conjunctionof factors,many of them intrinsicallyunimportant and thus unobserved,if not unobservable,that determinesthe flow of events. A major global event can produce drasticallydifferent outcomes in two settingsthat differ trivially. Structuralism and individualismare not rival and mutually incompatible approaches to the study of revolution,as Skocpol would have it. They are essentialcomponents of a single story. We can now turn to the question of why with hindsightan unanticipated revolutionmay appear as the inevitableconsequence of monumen- tal forcesfor change. A successfulrevolution brings into the open long- repressedgrievances. Moreover, people who were relativelycontent with the old regime embrace the new regime,and they are apt to attribute theirformer public preferencesto fearsof persecution. Reconsider the thresholdsequence A' = {0, 10, 20, 30, 40, 50, 60, 70, 80, 100}.

The relativelyhigh thresholdsin A' are likely to be associated with private preferencesmore favorableto the governmentthan to the opposition.35Person 9 (T9 = 80) is much more satisfiedwith the government than,say, person 3 (T3 = 20). As such she has littledesire to join a move-

35 Relativelyhigh thresholdsmay also be associatedwith relativelygreat vulnerabilityto social pressure. SURPRISE IN THE EAST EUROPEAN REVOLUTION 23 ment aimed at topplingit. Rememberthat public oppositionsettles at 90, she being the last to jump on the revolutionarybandwagon. The impor- tant point is this: person 9 changes her public preferenceonly afterthe opposition snowballs into a crushingmajority, making it imprudentto remain a governmentsupporter. Having made the switch,she has everyreason to feigna long-standing antipathyto the toppled government.She will not admit that she yearns for the status quo ante, because this would contradicther new public preference.Nor will she say that her change of heart followed the government'scollapse, because thismight render her declared sympathyfor the revolutionunconvincing. She will claim thatshe has long had serious misgivingsabout the old order and has sympathizedwith the objectives of the opposition. An unintended effectof this distortionis to make it seem as though the toppled governmentenjoyed even less genuine support than it actuallydid. This illusion is rootedin the veryphenomenon responsible for making the revolutiona surprise:preference falsification. Having misled everyone into seeing a revolutionas highlyunlikely, preference falsification now conceals the forcesthat were working against it. One of the consequences of postrevolutionarypreference falsification is thus to make even less comprehensiblewhy the revolutionwas unforeseen. The historiansof a revolutionmay appreciate the biases that afflict people's postrevolutionaryaccounts of their prerevolutionarydisposi- tionswithout being able to measure the significanceof thesebiases. Con- sider the sequence C' = {0, 10, 20, 20, 20, 20, 20, 20, 60, 100}. Like A', this sequence drivesS from 10 to 90, implyingthat nine out of ten individuals have an incentiveto say that theydespised the prerevo- lutionaryregime. If thresholdsbelow 50 reflectprivate support for a revolution,and those above 50 privatesatisfaction with the statusquo, eight of the nine would be telling the truth,the one liar being person 9 (T9 = 60). It follows from the same assumption that four of the nine would be lying if the thresholdsequence were A'. But once again, because thresholdsare not public knowledge,historians may have difficulty determiningwhether the prerevolutionarysequence was A or C-or for thatmatter, whether the postrevolutionarysequence is A' or C'. Before moving to the East European Revolution,it may be useful to commenton how the foregoingargument relates to threesources of con- troversyin the literatureon revolutions:the continuityof social change, the power of the individual,and the significanceof unorganized crowds. 24 WORLD POLITICS The proposed theorytreats continuous and discontinuouschange as a single,unified process. Private preferences and the correspondingthresh- olds may change gradually over a long period during which public opposition is more or less stable. If the cumulativemovement establishes a latent bandwagon, a minor event may then precipitatean abrupt and sharp break in the size of the public opposition. This is not to say that privatepreferences change onlyin small increments.A major blunder on the part of the government may suddenly turn private preferences against it. Such a shiftcould also occur in responseto an initial,possibly modest, increase in public opposition.The underlyinglogic was expressedbeau- tifullyby Alexis de Tocqueville: "Patientlyendured so long as it seemed beyond redress,a grievance comes to appear intolerableonce the possibility of removing it crosses men's minds."36In terms of our model, Tocqueville suggeststhat the thresholdsequence is itselfdependent on the size of the public opposition.If so, a revolutionarybandwagon may come about as the joint outcome of two mutuallyreinforcing trends: a fall in thresholdsand a rise in public opposition. Imagine that public oppositionrises sufficiently to convincethose privately sympathetic to the governmentthat a revolutionmight be in the making. This realization induces many of them to think about possible alternativesto the status quo. Their thinkingstarts a chain reactionthrough which privatepreferences shiftswiftly and dramaticallyagainst the government.The consequent changes in the thresholdsequence cause the revolutionarybandwagon to accelerate. The theorydepicts the individual as both powerless and potentially verypowerful. The individual is powerlessbecause a revolutionrequires the mobilization of large numbers,but he is also potentiallyvery pow- erfulbecause under the rightcircumstances he may set offa chain reaction that generates the necessarymobilization. Not that the individual can know preciselywhen his own choice can make a difference.Al- thoughhe may sense thathis chances of sparkinga wildfireare unusually great, he can never be certainabout the consequences of his own opposition.What is certainis thatthe incumbentregime will remain in place unless someone takes the lead in moving into the opposition. As we saw in the previous section,the standard theoryof rational choice depicts the potentialrevolutionary as paralyzed by the realization of his powerlessness.Many social thinkerswho, like the presentauthor, accept the logic of collectiveaction have struggledwith the task of explaining how mass mobilizationsget started.One of the proposed expla-

36 Tocqueville, The Old Regime and the FrenchRevolution (1856), trans. Stuart Gilbert (Garden City,N.Y.: Doubleday, 1955), 177. SURPRISE IN THE EAST EUROPEAN REVOLUTION 25 nationsrests on a cognitiveillusion: the individual overestimateshis personal political influence.Another invokes an ethical commitment:the individual feels compelled to do his fair share for the attainmentof a jointlydesired outcome.37The approach used here, which is not incompatible with these explanations,places the burden of sparkingthe mobilization process on the individual's need to be true to himself.This approach is consistentwith the fact that revolutionaryleaders tend to be surprisedwhen theirgoals materialize. The cognitive-illusionexplana- tion is not: people who challenge the governmentout of an overestima- tion of their personal abilityto direct the course of historywill not be surprisedwhen theirwishes come true.The approach of thisessay is also consistentwith the factthat some people risk theirlives fora revolution even as the vast majorityof the potentialbeneficiaries refrain from doing theirown fairshare. Finally, the outlined theoryaccords organized pressure groups and unorganized crowds complementaryroles in the overthrowof the government.Organized oppositionsenhance the externalpayoff to dissent, both by providing the individual dissenterwith a supportnetwork and by raisingthe likelihoodof a successfulrevolution. They also help shatter the appearance of the invulnerabilityof the status quo, and through propaganda, they shiftpeople's private preferencesin favor of change. Charles Tilly is thereforeright to draw attentionto the structuraland situational factorsthat govern a society'spattern of political organization.38But as Pamela Oliver warns, we must guard against overempha- sizing the role of organization at the expense of the role of the unorganized crowd. A small differencein the resourcesat the disposal of an organized oppositionmay have a tremendousimpact on the outcome of its efforts.39This observationmakes perfectsense in the contextof the theory developed here. Where a small pressure group fails to push a bandwagon into motion a slightlybetter organized or slightlylarger one might.

IV. EAST EUROPEANCOMMUNISM ANDTHE WELLSPRINGOF ITS STABILITY

Communist partiescame to power in Russia, and thenin Eastern Europe and elsewhere, with the promise that "scientificsocialism" would pio-

37 Each of these is developed by Steven E. Finkel, Edward N. Muller, and Karl-Dieter Opp, "Personal Influence,Collective Rationality, and Mass PoliticalAction," American Polit- ical ScienceReview 83 (September1989), 885-903. 38 Tilly, FromMobilization to Revolution(Reading, Mass.: Addison-Wesley,1978). 39 Oliver, "Bringing the Crowd Back In: The Nonorganizational Elements of Social Movements," in Louis Kriesberg,ed., Researchin Social Movements,Conflict and Change (Greenwich,Conn.: JAI Press, 1989): 11:1-30. 26 WORLD POLITICS neer new dimensions of freedom,eliminate exploitation,vest political power in the masses, eradicate nationalism,and raise standardsof living to unprecedentedheights-all this,while the state was witheringaway. They did not deliveron any of these promises.Under theirstewardship, communism came to symbolize repression,censorship, ethnic chauvin- ism, militarism,red tape, and economic backwardness. The failuresof communism prompted a tinynumber of Soviet and East European citizens to criticizeofficial policies and establishedinsti- tutions.Such dissidentsexpressed theirfrustrations through clandestine self-publications(samizdat) and writingspublished in the West (tamiz- dat). Given the chasm between the rhetoric of communism and its achievements,the existenceof an opposition is easily understood. Less comprehensibleis the rarityof public opposition-prior, that is, to 1989. The few uprisings that were crushed-notably, East Berlin in 1953, Hungary in 1956, and Czechoslovakia in 1968-are the exceptionsthat prove the rule. For most of several decades, most East Europeans dis- played a remarkable tolerance for tyrannyand inefficiency.They remained docile, submissive,and even outwardlysupportive of the status quo. This subservienceis attributablepartly to punishmentsmeted out by the communist establishmentto its actual and imagined opponents. In the heydayof communism a person speaking out against the leadership or in favor of some reformcould expect to sufferharassment, lose his job, and face imprisonment-in short,he could expect to be denied the opportunityto lead a decent life. Even worse horrorsbefell millions of suspected opponents. Justthink of the forced-laborcamps of the Gulag Archipelago and of the liquidationscarried out under the pretextof historical necessity."We can only be rightwith and by the Party,"wrote a leading theoreticianof communism,"for historyhas provided no other way of being in the right."40Such thinkingcould, and did, serveto justify horriblecrimes against nonconformists. Yet officialrepression is only one factorin the endurance of communism. The systemwas sustainedby a general willingnessto supportit in public: people routinelyapplauded speakers whose message they disliked, joined organizationswhose mission theyopposed, and signed de- famatoryletters against people they admired, among other manifesta- tions of consent and accommodation. "The lie," wrote the Russian novelist Alexander Solzhenitsynin the early 1970s, "has been incorpo- rated into the state systemas the vital link holding everythingtogether, 40 The words of Leon Trotsky,cited by Arendt(fn. 2), 307. SURPRISE IN THE EAST EUROPEAN REVOLUTION 27 with billions of tiny fasteners,several dozen to each man."'41 If people stopped lying,he asserted,communist rule would break down instantly. He then asked rhetorically,"What does it mean, not to lie?" It means "not sayingwhat you don't think,and that includes not whispering,not opening your mouth, not raising your hand, not casting your vote, not feigninga smile, not lending your presence,not standing up, and not cheering."42 In "The Power of the Powerless," Havel speaks of a greengrocerwho places in his window, among the onions and carrots,the slogan "Work- ers of the World, Unite!" Why does the greengrocerdo this,Havel won- ders. Is he genuinelyenthusiastic about the idea of unityamong the workers of theworld? Is hisenthusiasm so greatthat he feelsan irrepressibleimpulse to acquaintthe publicwith his ideals? Has he reallygiven more thana moment'sthought to how such a unificationmight occur and what it would mean? Havel's answer is worthquoting at length: The overwhelmingmajority of shopkeepersnever think about the slogans theyput in theirwindows, nor do theyuse themto expresstheir real opin- ions. That posterwas deliveredto our greengrocerfrom the enterprise headquartersalong withthe onions and carrots.He put themall intothe windowsimply because it has beendone that way for years, because everyone does it,and becausethat is theway it has to be. If he wereto refuse, therecould be trouble.He could be reproachedfor not having the proper "decoration"in hiswindow; someone might even accuse him of disloyalty. He does it becausethese things must be done ifone is to get alongin life. It is one ofthe thousands of details that guarantee him a relativelytranquil lifein "harmonywith society," as theysay.43 So our greengrocerputs up the assigned slogan to communicatenot a social ideal but his preparednessto conform.And the reason the display conveysa message of submissionis thatevery submissive greengrocer has exhibitedthe same slogan foryears. By removingthe poster-or worse, replacingit with one thatreads "Workers of the World, Eat Onions and Carrots!"-our greengrocerwould expose himselfto the charge of sub- version. He thereforedisplays the required slogan faithfullyand fends off trouble. In the process, he reinforcesthe perceptionthat society is solidlybehind the Party.His own prudence thus becomes a factorin the willingness of other greengrocersto promote the unity of the world's

41 Solzhenitsyn,"The Smatterers"(1974), in Solzhenitsynet al.,From undertheRubble, trans. A. M. Brock et al. (Boston: Little,Brown, 1975),275. 42 Ibid., 276; emphasisin original. 43 Havel (fn.9), 27-28. 28 WORLD POLITICS workers. Moreover, it pressures farmers,miners, bus drivers, artists, journalists,and bureaucratsto continuedoing and saying the thingsex- pected of them. Effortsto prove one's loyaltyto the political status quo often took more tragic formsthan a greengrocer'sdisplay of a well-worn Marxist slogan. People tattled on each other. And they ostracized and vilified nonconformistswho were sayingor doing thingsthat they admired. The Romanian dissidentNorman Manea writesof authors who "persecuted theircolleagues on the 'blacklist'with tireless, diabolical energy."44In the same vein, the Polish dissidentPiotr Wierzbicki writesabout a famous composer who went out of his way to alert the governmentto an anti- Soviet insinuationon the sleeve of a recordby a Pole living abroad. The squealing composer knew that this informationwas likely to block the local performanceof his fellow Pole's music. He did it to prove his loyalty to the regime-to earn, as it were, a certificateof normalcy.45 In 1977 a group of Czechoslovak intellectualsestablished a loose association,Charter 77, dedicated to the basic human rightsthat Czecho- slovakia agreed to respectby signingthe Helsinki accords of 1975.46The governmentresponded by detaining the spokesmen of Charter 77 and launching a nationwide campaign against the association.47In the course of this campaign millionsof ordinarycitizens expressed their opposition to Charter 77 by signingstatements of condemnation,sending hate let- tersto newspapers,and ostracizingits signatories.Many an opponent of Charter 77 did so in betrayalof his conscience. It is true of course that some who participatedin this campaign saw Charter 77 as a menacing organizationbent on tarnishingCzechoslova- kia's image abroad. And the tale-bearingPolish composermay well have had motives other than a desire to please the regime,for instance,jeal- ousy or professionalcompetition. But East Europeans turned against each other routinelyeven in the absence of such motives. Let us returnto the storyof the greengrocer.Havel asks us to "imagine thatone day somethingin our greengrocersnaps and he stops putting up the slogans." The greengroceralso "stops voting in elections he knows are a farce"; he "begins to say what he really thinksat political meetings"; and he "even findsthe strengthin himselfto expresssolidar-

44 Manea, "Romania: Three Lines with Commentary,"in Brintonand Rinzler (fn. 11), 327. 45 Wierzbicki,"A Treatise on Ticks" (1979), in Abraham Brumberg,ed., Poland: Genesis ofa Revolution(New York: Random House, 1983),205. 46 The Charter77 declarationis reproducedin Havel et al. (fn.9), 217-21. 47 See Timothy Garton Ash, The Uses of Adversity:Essays on the Fate of CentralEurope (1983-89) (New York: Random House, 1989),esp. 61-70. SURPRISE IN THE EAST EUROPEAN REVOLUTION 29 itywith thosewhom his consciencecommands him to support."In short, he makes "an attemptto live withinthe truth."48 Here are the likely consequences of this revolt: [The greengrocer]will be relievedof his postas managerof theshop and transferredto the warehouse.His pay will be reduced.His hopes fora holidayin Bulgariawill evaporate.His children'saccess to highereduca- tionwill be threatened.His superiorswill harass him and hisfellow work- erswill wonderabout him. Most of thosewho applythese sanctions, however,will notdo so fromany authentic inner conviction but simply under pressurefrom conditions, the same conditionsthat once pressuredthe greengrocerto displaythe official slogans. They will persecutethe green- grocereither because it is expectedof them, or to demonstratetheir loyalty, or simplyas partof thegeneral panorama, to whichbelongs an awareness thatthis is how situationsof thissort are dealt with,that this, in fact,is how thingsare alwaysdone, particularly if one is not to becomesuspect oneself.49 The brillianceof thisvignette lies in its insightsinto the pressuresthat kept East Europeans outwardly loyal to theirinefficient, tyrannical regimes. Officialrepression met with the approval of ordinarycitizens and indeed was predicatedon theircomplicity. By falsifyingtheir preferences and helping to discipline dissenters,citizens jointlysustained a system thatmany consideredabominable. According to Havel, the crucial "line of conflict"ran not between the Partyand the people but "througheach person," for in one way or another everyonewas "both a victim and a supporterof the system."50 The same idea found vivid expression in a banner hung above the altar in an East German church: "I am Cain and Abel.""5 The implied intrapersonalconflict is rooted of course in the clash between the individual's drive to exercise autonomy and his need for social acceptance. Until 1989 most East Europeans tended to resolve this chronic clash in favor of social acceptance. By thus avoiding an open battle with communism, theyacquiesced to battle silentlywith themselves.In the process, most achieved a measure of outer security,though at the expense of inner peace. Not that communist rule managed to do away altogetherwith the human propensityto protest.As Wierzbicki points out, newspapers received lettersof complaintin abundance-about shabbyhousing, the ne- glected grave of some poet or other,and the sloppilypainted fence of a

48 Havel (fn.9), 39; emphasisin original. 49 Ibid., 39. 50 Ibid., 37. 51 Timothy Garton Ash, "Eastern Europe: The Year of Truth," New York Review of Books,February 15, 1990,p. 18; emphasisin original. 30 WORLD POLITICS children's playground.Yet protesterstended to stay within a Party-de- fined zone of acceptability:they refrained from probing too deeply into issues and avoided challengingcommunism itself. A schoolteacherwrit- ing furiousletters about a defectiveappliance would not bringherself to blame the systemthat produces useless appliances. Nor would she sign a letterexpressing solidarity with dissidentsor join a demonstrationfor freedomof speech.52 The typical East European feignedopposition to the few dissidents, though in private he applauded theirmission. Havel suggeststhat this admirationwas coupled with a resentment:people who lacked the courage to be true to themselvesfelt threatened by displays of integrityon the part of others.They thus treatedopen defiance "as an abnormality, as arrogance,as an attack on themselves,as a formof dropping out of society."53If it is true that the "iron in the soul" of another reminded a conformistof the lack of iron in his own, this would have served as an additional obstacle to overtopposition.54 Another such obstacle was pluralistic ignorance: people alienated from the communistregime did not know how widely theiralienation was shared. They could sense the represseddiscontent of theirconform- ist relatives and close friends;they could observe the hardships in the lives of their fellow citizens; and they could intuit that past uprisings would not have occurred in the absence of substantialdiscontent. Still, they lacked reliable, currentinformation on how many of their fellow citizens favored a change in regime. The government-controlledpress exploited this ignorance by stressingthe "unityof socialist society"and its "solidarityin supportingthe Party." Insofar as such propaganda led potential revolutionariesto underestimatethe prevalence of discontent, it weakened theirincentives to join the minusculeopposition. Governmentsthroughout history have recognized the significanceof preferencefalsification and out of self-interesthave tried to keep themselves informedabout the privatepreferences of theirconstituents. Louis XIV told his heir that "the art of governing"consists in "knowing the real thoughtsof all the princesin Europe, knowing everythingthat people tryto conceal from us, their secrets,and keeping close watch over them."55So it is that the communistgovernments of Eastern Europe conducted numerous surveysto find out the true thoughtsand feelings

52 Wierzbicki (fn.45), 206-7. 53 Havel (fn.9), 37. 54 The metaphorbelongs to BarringtonMoore, Jr.,Injustice: The Social Bases of Obedience and Revolt(White Plains, N.Y.: M. E. Sharpe, 1978). 55 Cited by NorbertElias, Powerand Civility(1939), trans.Edmund Jephcott(New York: Pantheon,1982), 197. SURPRISE IN THE EAST EUROPEAN REVOLUTION 31 of their subjects. If the fact that theykept the resultssecret is any indi- cation, these were not entirelyflattering to them or theirpolicies. Infor- mation for publication "was checked beforehandand given the appro- priateinterpretation," to keep it fromemboldening the regime'sdeclared and potentialopponents.56 It would be an exaggerationto suggest that all East European sup- porters of communist rule were privatelyopposed to the status quo. Some benefitedhandsomely from the system,and othersfelt threatened by major reform.Nor did those who became conscious of the failuresof communism necessarilylose faith in officialideals. Even leading dissidents remained sympatheticto centralplanning and collectiveownership and ever suspiciousof the free-enterprisesystem.57 By and large,they felt that communism was betrayed by self-servingleaders, not that it was inherentlyunworkable. These observationsare consistentwith opinion polls of East Europe- ans travelingabroad conducted by Western organizations in the 1970s and early 1980s. With remarkable consistencyand for each nation, the data showed that in free elections offeringa full spectrumof choices, including a Democratic Socialist Party and a Christian Democratic Party,the Communist Party would receive at most a tenthof the vote, and the socialistswould invariablybe the winners.58 Further systematicevidence is contained in surveysconducted from 1970 onward forthe benefitof the leadershipby the Central Institutefor Youth Development in Leipzig. Now being declassified,these surveys suggestthat until the mid-1980smost East Germans accepted the official goals of socialism.In 1983,46 percentof a sample of trade school students endorsed the statement"I am a devoted citizen of the German Demo- cratic Republic," whereas 45 percentendorsed it with reservationsand only 9 percentrejected it. And in 1984,50 percentagreed that "socialism will triumph throughoutthe world," whereas 42 percent agreed with reservationsand 8 percentdisagreed. Between 1970 and 1985,the results showed little variation.59They may, of course, have been based on a

56 Jifi Otava, "Public Opinion Research in Czechoslovakia," Social Research55 (Spring- Summer 1988),249. Every issue of the Czechoslovak government'sofficial bulletin on public opinion stated: "We remind all researchersthat this bulletinis not meant for the public, which means not even foryour friendsand acquaintances,but servesexclusively as internal materialfor poll-takers and thosewho collaboratewith us" (p. 251 n. 2). 57 See Vladimir Tismaneanu, The Crisisof Marxist Ideology in EasternEurope: The Poverty of Utopia (London: Routledge,1988), esp. chap. 4. 58 Henry 0. Hart, "The Tables Turned: If East Europeans Could Vote," Public Opinion 6 (October-November 1983), 53-57. The surveysreported by Hart cover Czechoslovakia, Hungary,Poland, Romania, and Bulgaria. 59 "Daten des Zentralinstitutsfur Jugendforshung Leipzig" (Mimeograph),Tables 1 and 2. These tableswere compiled by Walter Friedrich,the directorof the institute,and distrib- 32 WORLD POLITICS flawed methodology,as was much public opinion researchdone in East- ern Europe. But, as we shall see later, it is highlysignificant that after 1985 this same methodologyregistered a sustained deteriorationboth in the citizenry'sattachment to the regimeand in its faithin socialism. It thus appears that while the East Europeans overwhelminglydis- liked the regimes under which they were living, they were much less troubledby the principlesof socialism-at least until the mid-1980s. To make sense of this finding,we need to touch on the cognitiveimplica- tionsof preferencefalsification. Disaffected citizens choosing to conform to the regime's demands typicallypaid lip service to officialgoals, used Marxist jargon, and made excuses for communism's shortcomingsby pointing to the ostensiblyworse failuresof capitalism. In the process, they unavoidably kept their fellow citizens uninformedabout those of their private beliefs that were inimical to the status quo. Worse, they knowinglyexposed one anotherto falsefacts and misleadingarguments. In short, they distortedpublic discourse. Since public discourse influ- ences what is noticedand how eventsare interpreted,this distortion un- doubtedly affectedthe evolution of East European private preferences. East Europeans subjected fromearly childhood to predictionsof the im- minent demise of capitalism and to theoriesof the incontrovertiblesu- periorityof communismmust have become more or less conditioned to think in Marxist terms,developing some mental resistanceto the fundamental flaws of theirsocial order.60 If this reasoningis correct,Marxist discourse would also have blunted the abilityof East Europeans to articulatean alternativeeconomic order. Vladimir Shlapentokh points to a paradox here. The socialist worker mistruststhe market order, even though he obtains his treasured blue jeans through the only free market to which he has access-the black market. Likewise, the enterprisemanager who turns regularlyto the underground economy for vital spare parts dreads economic liberalization. Shlapentokh ascribes such inconsistenciesto a disjunctionbetween the "pragmatic" and "theoretical" layers of the individual mind.61 uted to the participantsat a conferenceheld in Ladenburg in February 1991, under the auspices of the Gottlieb Daimler and Karl Benz Foundation. Elisabeth Noelle-Neumann broughtthe documentto my attention;John Ahouse translatedit into English. 60 For a fullerargument on how preferencefalsification distorts public discourseand how, in turn,this distortion warps the evolutionof people's privatepreferences, see Timur Kuran, "The Role of Deception in PoliticalCompetition," in AlbertBreton et al., eds., The Compet- itiveState (Boston: Kluwer-Nijhoff,1990), 71-95. 61 Though Shlapentokhdevelops theargument with respect to the Soviet Union, it applies also to Eastern Europe. See Shlapentokh,Soviet Public Opinionand Ideology:Mythology and Pragmatismin Interaction(New York: Praeger,1986); and idem,Public and PrivateLife ofthe Soviet People: ChangingValues in Post-StalinRussia (New York: Oxford UniversityPress, 1989). SURPRISE IN THE EAST EUROPEAN REVOLUTION 33 Known in cognitivepsychology as mentalpartitioning, this phenomenon is an inevitableconsequence of the mind's limitationsin receiving,stor- ing, retrieving,and processinginformation. People are simplyunable to incorporateinto a single, comprehensivemodel the multitudesof variables and relationshipsthat bear on their happiness; they thus ignore many interconnectionsand treat closely related phenomena as unre- lated.62 For our purposes,the importantimplication is this: an East European confronteddaily with communism'sshortcomings would not necessarily have taken them as a sign of the unworkabilityof the system.He could easily have turned against individual functionarieswithout losing faith in the systemin which theyoperated. Some East Europeans did of course recognize that specificshortcomings were part of a general patternof failure. Many were intellectualswith much time to think and thus to make the mental connectionsnecessary for identifying the system'sfun- damental flaws. But many othersdid not make these connections,partly because the prevailingpublic discourseprovided no help. So processes rooted in preferencefalsification kept privateopposition to communism far fromunanimous. This does not negate the fact that vast numbersremained outwardly loyal to communistrule primarilyout of fear. But for widespread preferencefalsification, the communist regimes of Eastern Europe would have faced severepublic opposition,very possiblycollapsing before 1989. In view of its profoundimpact on both private and public sentiment,preference falsification may be characterized as the wellspringof the communistsystem's stability.

V. THE REVOLUTION

The foregoingargument has two immediate implications.First, the regimes of Eastern Europe were substantiallymore vulnerable than the subservienceand quiescence of theirpopulations made them seem. Mil- lions were prepared to stand up in defianceif ever theysensed that this was sufficientlysafe. The people's solidaritywith their leaders would then have been exposed as illusory,stripping the veneer of legitimacy from the communistmonopoly on power. Second, even the support of those genuinelysympathetic to the statusquo was ratherthin. Though many saw no alternativeto socialism,their many grievancespredisposed them to the promise of fundamental change. Were public discourse

62 See John H. Holland et al., Induction:Processes of Inference,Learning, and Discovery (Cambridge: MIT Press, 1986); and Amos Tverskyand Daniel Kahneman, "The Framing of Decisions and the Rationalityof Choice," Science211 (January1981), 453-58. 34 WORLD POLITICS somehow to turn against socialism,they would probablyawaken to the possibilitythat theirlives could be improved. But what would catalyze the process of revolutionarymobilization? With hindsightit appears thatthe push came fromthe Soviet Union. In the mid-1980s festeringeconomic problems,until then officiallydenied, convinced the top Soviet leadership to call forperestroika (restructuring) and glasnost(public openness). Repressedgrievances burst into the open, including dissatisfactionwith communistrule itself.And with Mikhail Gorbachev's rise to the helm in 1985, the Soviet Union abandoned its long-standingpolicy of confrontationwith the West, to seek accommodation and cooperation.63In Eastern Europe thesechanges kindled hopes of greaterindependence and meaningfulsocial reform. Lest it appear that these developmentsprovided a clear signal of the coming revolution, remember that Havel dismissed a Czechoslovak crowd's jubilation over Gorbachev as a sign of naivete. He was hardly alone in his pessimism. Even if Gorbachev wanted to liberate Eastern Europe, a popular argumentwent, it was anythingbut obvious that he could. Surely, the militaryand hard-line conservativeswould insist on retainingthe Soviet Union's strategicbuffer against an attack fromthe West. Nor was thisthe only obstacle to liberation.Economic and ethnicten- sions within the Soviet Union could provide the pretextfor a conservative coup. There was always the precedentof Khrushchev,toppled in 1964. About the time that Havel was exuding pessimism,a joke was making the rounds in Prague: "What is the differencebetween Gor- bachev and Dubcek [the deposed leader of the 1968 Prague Spring]?" The answer: "None--except Gorbachev doesn't know it yet."64Signifi- cantly,in the fall of 1989 Moscow was rifewith rumorsof an impending coup.65Some observersexpected Gorbachev to survive but only by re- versing course and becoming increasinglyrepressive.66 An old Soviet joke expresses the underlyingthinking. Stalin leaves his heirs in the Party two envelopes. One is labeled, "In case of trouble, open this." Trouble arises and the envelope is opened ceremoniously:"Blame me."

63 For details,see RobertC. Tucker, PoliticalCulture and Leadershipin SovietRussia: From Lenin to Gorbachev(New York: W. W. Norton, 1987),chap. 7. 64Economist, July 18, 1987,p. 45. 65 Z [anonymous],"To the Stalin Mausoleum,"Daedalus 119 (Winter 1990),332. 66 With the revolution,the notionthat Gorbachev would turnto the armyand the KGB in a bid to stayin power lostplausibility. It regainedplausibility in late 1990 withthe resignation of his foreignminister, Eduard Shevardnadze,who publiclyaccused Gorbachev of plotting with hard-linersto createa repressivedictatorship. SURPRISE IN THE EAST EUROPEAN REVOLUTION 35 The otherenvelope is labeled, "In case of more trouble,open this." More troublecomes and the second envelope is opened: "Do as I did."67 In support of their predictionthat the conservativeelements in the leadership would prevail sooner or later,pessimists frequently invoked the conservatismof the Soviet people. In a widely discussed 1988 article, for instance,a Russian social scientistargued that seven decades of bu- reaucraticregimentation had suppressed individual creativity,reorient- ing the "Soviet value systemaway fromrevolutionary transformation to conservativeimmobility." Communism had quashed the very personal qualities on which the reformistswere counting.68In June 1989 another Soviet observerwould confess:"For threeyears I have tried to findout whetheror not thereis mass supportfor perestroika, and now I feel I can conclude that it does not exist." He blamed not only the individual citizen's fear of change but also the Soviet ethic that identifiessocial justice with economic equality.69The upshotof such comments,to which scores more from diverse sources could be added, was that Soviet citizens tended to be deeply suspicious of Gorbachev's intentions.Many com- mentatorsinferred that Gorbachev's reformswere doomed, reasoning that he could not relyon the masses forprotection against a conservative challenger. As Gorbachev was tryingto restructurethe Soviet Union, Poland was testingthe limitsof its freedomfrom Moscow. The struggleto legalize Solidarityhad already given the countrya taste of pluralism,and gov- ernmentcensorship was being relaxed in fitsand starts.Everyone recognized that this softeningenjoyed Gorbachev's approval. Yet few in- formed people put much faith in Gorbachev's ability to push the liberationof Eastern Europe substantiallyforward, and once again it was not clear thathe intendedto try."Dissidents throughoutEurope," wrote the Economistin mid-1987, sound "sceptical" when talking about Gor- bachev. "This is not because theyquestion [his] reformingzeal. It is simply that many thinkingpeople in Eastern Europe have come to believe that real change in Communist countriescannot be imposed from the top-or fromoutside-but must emerge frombelow."70 Plenty of events lent credence to this reasoning.For instance,Gorbachev did not prevent

67 Recorded by Daniel Bell, "As We Go into the Nineties: Some Outlines of the Twenty- firstCentury," Dissent 37 (Spring 1990), 173. 68 Igor Kon, "The Psychologyof Social Inertia"(1988), Social Sciences20, no. 1 (1989), 60- 74. 69 Gennadii Batyagin,TASS, June28, 1989,quoted byElizabeth Teague, "Perestroikaand the Soviet Worker,"Government and Opposition25 (Spring 1990), 192. 70 Economist,July 18, 1987,p. 45. 36 WORLD POLITICS the East German regimefrom falsifying the resultsof local electionsheld in the springof 1989 or fromendorsing China's massacre at Tiananmen Square that summer. Nor did he keep the East German regime from using forceto disperse small demonstrationsagainst these two acts.7" In sum, prior to the actual revolutionit was not at all clear that the Soviet Union would sit back if its six Warsaw Pact allies tried to overthrow their communistregimes. Statements,events, and trends that in retrospectappear as unmistakable signs of an explosion in the making coexisted with many signs that pointed in the directionof inertia and continued stability.Some of Gorbachev's actions did indeed suggestthat he wanted to institutefundamental reforms in many areas, includingthe Soviet Union's relationshipwith its East European satellites.But there were many reasons to expect his effortsto end in failure. Yet since the revolutionit has seemed as thoughGorbachev engineered the liberationof Eastern Europe. In fact,he was a masterat puttingthe best face on events that had pushed past him. In the fall of 1989 there were many reportsthat events were going much furtherand/or faster than Gorbachev wanted. He was reportedlywilling to permitmoves toward democracy, provided the communistswere not humiliated and Eastern Europe's militaryties to the Soviet Union were preserved.And like leaders in Washington,Paris, Bonn, and elsewhere,he was reluctant to support anythingthat might disturb Europe's hard-won peace. But when the peoples of Eastern Europe grabbed politicalpower, pushed the communistsaside, and proclaimed theirintention to leave the Warsaw Pact, Gorbachev just accepted realityand gave his blessingto eventsgen- erated by forcesbeyond his control.One is reminded of the horseman who, thrown from his horse, explains with a smile that he has "dis- mounted." The point remains that the Soviet reformmovement fueled expectations of a freerEastern Europe, reducingfor growing numbersthe perceived riskof challengingthe statusquo. In termsof the model described in Section III, the movement lowered the revolutionarythresholds of East Europeans, making it increasinglyeasy to set in motion a revolutionarybandwagon. But no one could see that a revolutionwas in the making, not even the Soviet leader whose moves were helping to establish the still-latentbandwagon. Recall that revolutionarythresholds are influencedalso by people's private preferences.Since privatepreferences are governed to a consid- erable extentby public discourse,the dissentgenerated by Soviet glasnost 71 TimothyGarton Ash, "GermanyUnbound," New YorkReview of Books, November 22, 1990,p. 12. SURPRISE IN THE EAST EUROPEAN REVOLUTION 37 probablypushed the privatepreferences of East Europeans against communism and communistrule. The East German surveysdiscussed above provide dramatic evidence to thiseffect. They show thatafter 1985 East German attachmentto socialism steadilydeteriorated. By October 1989 only 15 percentof the surveyedtrade school studentsendorsed the statement "I am a devoted citizen of the German Democratic Republic," down from46 percentin 1983. Fully 60 percentendorsed it with reservationsand 25 percentrejected it. In the same monthas few as 3 percent continuedto believe that"socialism will triumphthroughout the world," down from50 percentin 1984. Just27 percentagreed with reservations and a whopping 70 percentdisagreed.72 The contrastbetween the figures for 1989 and those for 1983-84 is striking.It points to a massive rise in discontentin the second halfof the decade, a rise thatmust have lowered the revolutionarythresholds of millionsof individual East Germans. What specificevents set the revolutionarybandwagon in motion? One must recognize thatattempting to answer this question is akin to trying to identifythe spark that ignited a forestfire or the cough responsible fora fluepidemic. There were many turningpoints in the East European Revolution,any one of which mighthave derailed it. One turningpoint came in earlyOctober, when East German officials refusedto carryout Partyleader Honecker's order to open fireon street demonstrators.On October 7 Gorbachev was in Berlin for celebrations marking the fortiethanniversary of the German Democratic Republic. With scores of foreignreporters looking on, crowds took to the streets, chanting,"Gorby! Gorby!" And the police clubs went into action. West German televisionimmediately played these events back to the rest of East Germany.The scenes alerteddisgruntled citizens in everycorner of the countryto the pervasivenessof discontent,while the government's weak responserevealed its vulnerability.A peaceful protestbroke out in Leipzig on October 9. Honecker ordered the regional Partysecretary to block the demonstration,by force if necessary. But bloodshed was averted when Egon Krenz, a Politburo member in charge of security, flew to Leipzig and encouraged the securityforces to show restraint. Local leaders-some of whom had already appealed for restraint-accepted this contraventionof Honecker's order, and tens of thousands marched withoutinterference. Sensing the shiftingpolitical winds, more and more East Germans throughoutthe countrytook to the streets.The East German uprisingwas now in fullswing. As the regimetried to stem the tide througha stringof concessions,the swelling crowds began to

72 "Daten des Zentralinstitutsfuir Jugendforschung Leipzig" (fn.59), Tables 1 and 2. 38 WORLD POLITICS make increasinglybold demands. Within a month the Berlin Wall would be breached,and in less than a year the German Democratic Re- public would become part of a unified,democratic Germany.73 Another turningpoint came on October 25, during Gorbachev's state visitto Finland. Two monthsearlier a Solidarityoffical had formedPo- land's firstnoncommunist government since the 1940s, following the Communist Party's stunningdefeat at the polls. A legislativedeputy to Gorbachev had declined detailed commenton the grounds that the developments were a domestic matter for the Poles.74The communists were in retreatin Hungary, too. In meetingswith dissidentgroups the Hungarian Communist Partyhad endorsed freeparliamentary elections. Then, in the belief that its candidates would do poorly running under the banner of communism,it had transformeditself into the Hungarian Socialist Party.75This was the firsttime that a ruling communistparty had formally abandoned communism. With the world wondering whether the Soviet Union had reached the limitsof its tolerance,Gor- bachev declared in Finland that his countryhad no moral or political rightto interferein the affairsof its East European neighbors.Defining this position as "the Sinatra doctrine,"his spokesman jokinglyasked re- porterswhether they knew the Frank Sinatra song "I Did It My Way." He went on to say that "Hungary and Poland are doing it their way." Using the Western term for the previous Soviet policy of armed inter- vention to keep the governmentsof the Warsaw Pact in communist hands, he added, "I think the Brezhnev doctrineis dead."76Coming on the heels of major communist retreatsin Poland and Hungary, these commentsoffered yet another indicationthat Gorbachev would not try to silence East European dissent. If one effectof thissignal was to embolden the oppositionmovements of Eastern Europe, another must have been to discourage the govern- ments of Eastern Europe fromresorting to violence unilaterally.This is not to say thatGorbachev enunciatedhis Sinatra doctrinewith the inten- tion of encouraging East European oppositionsto grab for power. Nor is it to say that the revolutionwould have peteredout in the absence of this move. By the time Gorbachev renouncedthe Soviet Union's rightto intervene,opposition movementsin Poland, East Germany, and Hun- gary already commanded mass support,and it is unlikelythat anything

73 This account draws on Ash (fn. 19); Anderson(fn. 26); and theNew YorkTimes reports compiled in Gwertzmanand Kaufman (fn. 1), 158-60, 166-84,216-22. 74 New YorkTimes, August 18, 1989,p. 1. 75 Ibid., October 8, 1989,p. 1. For a fulleraccount of the transformation,see Abel (fn. 14), chap. 2. 76 New YorkTimes, October 26, 1989,p. 1. SURPRISE IN THE EAST EUROPEAN REVOLUTION 39 shortof massive brutalitywould have broken theirmomentum and re- stored the status quo ante. Nonetheless, some incumbent communist leaders were seriouslyconsidering a militarysolution, and the procla- mation of the Sinatra doctrinemay well have tipped the balance against the use of force. Had even one East European governmentresorted to forceat this stage, the resultmay well have been a series of bloody and protractedcivil wars. Justas we cannot be certainthat a delay in announcingthe new Soviet doctrine would have altered the course of history,we will never know whetherthe contraventionof Honecker's order to shoot had a significant impact on the subsequent flow of events.What can be said is this: had Honecker's subordinatesenforced his order, the growth of the opposition would have slowed, and later demonstrationswould probably not have stayed peaceful. The same historicalsignificance can be attributed to the restraintshown by the individual soldierson dutyduring the dem- onstrationand by the individual demonstrators.In the tense atmosphere of the demonstrationa shot firedin panic or a stone thrown in excite- ment could have sparked a violentconfrontation. It was an extraordinary conjunction of individual decisions that kept the uprisingpeaceful and preventedthe revolutionfrom being sidetracked. The success of antigovernmentdemonstrations in one country in- spired demonstrationselsewhere. In early November, Sofia was shaken by its firstdemonstration in fourdecades as several thousand Bulgarians marched on the National Assembly. Within a week, on the very day throngsbroke throughthe BerlinWall, Todor Zhivkov's thirty-five-year leadership came to an end, and his successor began talking of radical reforms. Up to thattime Czechoslovakia's communistgovernment had yielded littleto its own opposition.Conscious of developmentselsewhere, it had simply promised economic reformsand made minor concessions on travel and religion.77These retreatsencouraged the swelling crowds to ask for more. On November 24, just hours afterAlexander Dubvek ad- dressed a crowd of 350,000 in his firstpublic speech since 1968,the Com- munist Partydeclared a shake-up in the leadership,only to face a much largerrally of people shouting,"Shame! Shame! Shame!" The new gov- ernmenttried to placate the demonstratorsby vowing to punish the com- mandant of the paramilitaryforces that had roughed up protestorsa week earlier.Unimpressed, the oppositionleaders labeled the announced changes "cosmetic" and promisedto redouble theirpressure. The success

77 Ibid., November 16, 1989,p. 1. 40 WORLD POLITICS of the general strike they called for November 27 led the Communist Party to capitulate within a matter of hours to their major demands, including an end to its monopoly on political power.78"Not since the Paris crowd discoveredthat the dreaded Bastille contained only a hand- ful of prisonersand a few terrifiedsoldiers has a citadel fallenwith such ease," wrote theEconomist a few days later."They just had to say boo."79 This bringsus back, forone last time,to Havel's brilliant1979 essay. He predicted there that when the greengrocersdecided they had had enough, communism would fall like a house of cards. So it turned out: when the masses took to the streets,the support for the Czechoslovak governmentjust vanished. The mobilization process followed the pat- terns of East Germany and Bulgaria. Emboldened by signals from the Soviet Union and the successesof oppositionmovements in neighboring countries,a few thousand people stood up in defiance,joining the tiny core of long-persecutedactivists. In so doing theyencouraged additional citizens to drop their masks, which then impelled more onlookers to jump in. Before long fear changed sides: where people had been afraid to oppose the regime,they came to fearbeing caught defendingit. Party members rushed to burn theircards, assertingthey had always been re- formistsat heart. Top officials,sensing that theymight be made to pay for standing in the way of change and for any violence,hastened to accept the opposition's demands, only to be confrontedwith bolder ones yet. Had the civilian leadership or the top brass attemptedto resist the opposition,the transferof power would not have been so swift,and cer- tainly not so peaceful. One of the most remarkable aspects of the East European Revolution is that,with the partialexception of Romania, the securityforces and the bureaucracyjust meltedaway in the face of growing public opposition.Not only did stateofficials shy away fromputting up a fight,but many crossedover to the oppositionas a transferof power appeared increasinglylikely. This is highlysignificant, for a defection fromthe inner establishmentis an unusually good indicatorof the pre- vailing politicalwinds. A Politburomember distancing himself from the Partyleader does more to expose the regime'svulnerability than a greengrocer who stops displaying the obligatoryMarxist slogan. In turn, a defiantgreengrocer does more harm to the regime'simage than does an obstreperousprisoner in solitaryconfinement.

78 For an eyewitnessaccount of these events,see Timothy Garton Ash, "The Revolution of the Magic Lantern,"New YorkReview of Books, January 18, 1990,42-51. See also Abel (fn. 14), chap. 3. 79 Economist,December 2, 1989,p. 55. SURPRISE IN THE EAST EUROPEAN REVOLUTION 41 In the simple model of Section III the perceivedstrength of the public opposition is measured by S, the share of societypublicly in opposition. This variable treatsall individuals equally: with ten individuals, each individual carriesa weightof 10 percent.But in reality,as I have argued, membersof societydiffer in theircontributions to the perceivedstrength of the opposition. So a more realistic measure of perceived strength would be some unequallyweighted indicator of public opposition,where the weights correlatewith levels of relativeinfluence. Such a weighted measure would assign a Politburo member more weight than a greengrocer,and the lattermore weightthan a nameless prisoner.Were we to introduce this refinementinto our model, the central argument would remain unaffected:with public preferencesstill interdependent,there would remain the possibilityof a latent,unobserved bandwagon.80My reason for abstractingfrom this refinementin Section III was to keep the presentationsimple. Some of the officialswho distancedthemselves from the Partyor even moved into the opposition as the uprisingstook offmay at heart have disliked the communistsocial order. Many othersundoubtedly acted for opportunisticreasons ratherthan out of conviction.Sensing the immi- nent collapse of the old order, theyabandoned it in hopes of findinga place in the order about to be born. A few chose to resist,but the speed of the anticommunistmobilization leftmost of them with insufficient time to plan and execute a coordinated response.Had the mobilization been slower, theymight well have managed to mount a credible,effec- tive response.8" Timothy Garton Ash, an eyewitnessto the mobilizationsin Hungary, Poland, East Germany,and Czechoslovakia, characterizes1989 as East- ern Europe's "year of truth."82This designationis accurate insofaras it captures the end of feignedsupport for communism. But it conceals the push the revolutiongot frompreference falsification on the part of those who sympathizedwith the statusquo. As noncommuniststhrew off their masks in joy and relief,many genuine communistsslipped on masks of theirown-masks depictingthem as the helpless functionariesof a re- pressive system,as formerpreference falsifiers thrilled to be speaking theirminds afteryears of silentresentment. Yet Ash's label is meaningful in another sense as well. The floweringof anticommunistdiscourse has

80 For a demonstration,see Kuran (fn.31). 81 The pace of events was undoubtedlya key factoralso in the failureof conservative groups in the Soviet Union to block EasternEurope's liberation.Had eventsproceeded more slowly,they might have had timeto oust Gorbachevand order the Red Armyinto action. 82 Ash (fn.51). 42 WORLD POLITICS exposed the officialideology more clearlythan ever beforeas a heap of sophistry,distortion, and myth. It has awakened millions of dormant minds, confrontingcitizens resignedto the statusquo with the conflicts between the pragmatic and theoreticallayers of theirbeliefs. This is to say neitherthat the thoughtsof everyEast European are now internally consistentnor that Marxist thinkinghas ceased. Rather, it is to suggest thatthe transformationof public discoursehas opened many to new possibilities. In the days followingthe fall of Czechoslovakia's communistregime, a banner in Prague read: "Poland-10 years,Hungary-10 months,East Germany-10 weeks, Czechoslovakia-10 days."83The implied accel- eration reflectsthe fact that each successfulchallenge to communism lowered the perceived risk of dissent in the countriesstill under communistrule. In termsof our model, as revolutionarythresholds in neighboring countriesfell, the revolutionbecame increasinglycontagious. Had thisbanner been prepareda few weeks later,it mighthave added "Romania-10 hours." As the Czechoslovak uprisingneared its climax, the executive committeeof the Romanian Communist Party was busy reelectingNicolae Ceauaescu as presidentand interruptinghis acceptance speech with standing ovations. Three weeks later protestsbroke out in the western provinces,but they were brutallyput down by the securityforces. Confident of his abilityto preventa replayof the events that had brought down other communistregimes, Ceauaescu leftfor a state visit to Iran, but the protestsintensified. Upon his returnhe organized a rally to denounce the "counterrevolutionaries,"but when he startedto speak he was booed. Television broadcastthe look of shock on his face, and the Romanian revolt was on. The consequent change of regime turned out to be bloodier than the previous five,because the securityforces responsible for the earlier massacre resistedthe revolution. They caused hundreds of deaths beforethey were beaten by the army. Ceauaescu tried to escape but he was caught and summarilyexecuted.84 Yet again, the world watched a nation jump with littlewarning from quiescence and subservience to turbulence and defiance. As the year went out, commentatorswere still marvelingat the speed with which the politicallandscape of Eastern Europe had changed. Long-persecuted dissidentsnow occupied high governmentpositions. In Czechoslovakia, forinstance, Havel was president,Dubcek, chairmanof the Federal As-

83 "Czechoslovakia: The Velvet Revolution,"Uncaptive Minds 3 (January-February1990),

84 For the New YorkTimes reportsof these events,see Gwertzman and Kaufman (fn. 1), 332-39. SURPRISE IN THE EAST EUROPEAN REVOLUTION 43 sembly,and JirvDienstbier (a Charter77 signatoryserving time as a coal stoker),foreign minister. All six countriesbegan planning freeelections and committedthemselves to economic liberalization.Some even moved to withdraw fromthe Warsaw Pact.

VI. THE PREDICTABILITY OF UNPREDICTABILITY

Unexpected as theywere, these developmentsnow seem as though they could easily have been predicted.Was it not obvious that the economic failuresof communism had sown the seeds of a massive revolt? Was it not self-evidentthat the East Europeans were just waiting foran opportunityto topple their despised dictators?Did not the severe domestic problems of the Soviet Union necessitateits withdrawal from Eastern Europe, to concentrateits resourceson economic reforms?Retrospective accounts of 1989 offera panoply of such reasons why the East European Revolution was inevitable. "It is no accident that Mikhail Gorbachev declined to intervene,"writes one commentator85-this,in a volume peppered with commentson how 1989 surprisedone and all. This essay has shown that the warning signs of the revolution remained cloudy until it was all over. Moreover, the unobservabilityof private preferencesand revolutionarythresholds concealed the latent bandwagons in formationand also made it difficultto appreciate the significanceof events that were pushing these into motion. The explanation forthis predictivefailure transcends the particularitiesof Eastern Europe: this is afterall hardlythe firsttime a major social uprising has come as a big surprise. The French Revolution of 1789 shocked not only Louis XVI and his courtiersbut also outside observersand the rioterswho helped end his reign. Yet it had many deep causes-all expounded at great length in literallythousands of volumes. This paradox is one of the centralthemes of Tocqueville's Old Rigime and the FrenchRevolution. "Chance played no part whatever in the outbreak of the revolution," he observes. "Though it took the world by surprise,it was the inevitableoutcome of a long period of gestation,the abruptand violentconclusion of a process in which six generationsplayed an intermittentpart."86 In this centurythe Nazi takeoverof Germany took place with astonishing speed. Within a few monthsentrenched political institutions were turned upside down, all democraticopposition was destroyed,and a la-

85 William M. Brinton,"Gorbachev and the Revolutionof 1989-90,"in Brintonand Rinz- ler (fn. 11), 373. 86 Tocqueville (fn.36), 20. 44 WORLD POLITICS bor movement with millions of members was driven underground.87 Though it was not foreseen,there is no shortageof explanationsfor the rise of Nazism. The Iranian Revolution of 1979-80 offersyet another example of an unanticipated uprising. There now exists a panoply of competingexplanations, including ones that invoke class conflicts,fail- ures of governance,foreign exploitation, economic reversals,the disaf- fectionsof bazaar merchants,and Islamic ideology.88Yet for all their differences,students of this revolutionagree that it stunned almost everyone-the Shah and the Ayatollah Khomeini, the CIA and the KGB, statesmen,diplomats, academics, and journalists. The veryrevolution that prepared the ground forthe firstcommunist regime in historywas an unforeseenevent. Weeks before the Russian Revolutionof February 1917 Lenin told an audience in Switzerland that Russia's great explosion lay in the distantfuture and thatolder men like himselfwould not live to see it.89And with just days to go, foreignob- serversin Petrogradwere advising theircapitals that the monarchywas stable and secure.90But the tsar fell,and before the year was over the communistshad gained full controlof the government.It has since been recognized that Marxist scholarshipdid not prepare us for the world's firstsuccessful communist revolutionoccurring in, of all places, back- ward, semifeudalRussia.91 Nor did Marxist scholarship-or for that matter,non-Marxist scholarship-anticipate the midcenturyuprisings in the communiststates of Eastern Europe. "The Hungarian uprisingof October 1956 was a dramatic, sudden explosion,apparently not organized beforehandby a rev- olutionarycenter; neither outsiders nor the participantshad anticipated anythinglike the irresistiblerevolutionary dynamism that would sweep the country."Thus begins The UnexpectedRevolution, a monograph on this failed attemptto overthrowcommunism that is replete with evi-

87 Detlev J.K. Peukert,Inside Nazi Germany:Conformity, Opposition, and Racismin Every- day Life, trans.Richard Deveson (New Haven: Yale UniversityPress, 1987),27-28 and pas- sim. 88 For a sample of explanations,see Hamid Algar, The IslamicRevolution in Iran (London: Muslim Institute,1980); Said Amir Arjomand, "Iran's Islamic Revolutionin Comparative Perspective,"World Politics 38 (April 1986),383-414; Shaul Bakhash, The Stateand Revolu- tionin Iran (London: Croom Helm, 1984); Nikki R. Keddie, Rootsof Revolution(New Ha- ven: Yale UniversityPress, 1981); and RobertLooney, Economic Origins of the Iranian Revo- lution(Boulder, Colo.: WestviewPress, 1982). 89 Leonard Schapiro,The RussianRevolutions of 1917: The Originsof ModernCommunism (New York: Basic Books, 1984), 19. 90William H. Chamberlin,The Russian Revolution,1917-1921 (New York: Macmillan, 1935), 1:73-76. 91 Furtherevidence concerning the elementof surprisein theFrench, Russian, and Iranian revolutionsmay be foundin Kuran (fn.31), secs. 2, 6-7. SURPRISE IN THE EAST EUROPEAN REVOLUTION 45 dence of widespread preferencefalsification right up to the uprising.92 Prior to October 1956 writerswho were to play leading roles gave not the slightestsign of opposition to the political status quo. For another example, clerical employees remained docile and submissive until the uprising in which they participated,often hiding theirgrievances even fromfamily members.93 The Prague Spring of 1968 offersanother example of an unforeseen attempt to crack the wall of communism. In a retrospectiveaccount, Havel writesthat in 1967 the entirenation was behaving like the Good Soldier Svejk, accommodating itself to the regime's demands. "Who would have believed . . . that a year later this recently apathetic, skepti- cal, and demoralized societywould stand up with such courage and in- telligenceto a foreignpower!" "And," he continues,"who would have suspectedthat, after scarcely a year had gone by,this same societywould, as swiftlyas the wind blows, lapse back into a stateof deep demoraliza- tion far worse than its original one!"94 This tallyof unanticipateduprisings could be expanded, but the point has been made: the revolutionof 1989 was not the firstto surprise us. Time and again entrenchedauthority has vanished suddenly,leaving the victorsastonished at theirtriumph and the vanquished, at theirdefeat. Should we conclude, along with JohnDunn, that revolutionsare in- eluctable "factsof nature,"events that fail "to suggestthe dominance of human reason in any form"?95In other words, is the culprithuman irrationality?The argumentdeveloped in thispaper does not point in this direction.It suggests,on the contrary,that predictivefailure is entirely consistentwith calculated, purposefulhuman action. Underlyingan ex- plosive shiftin public sentimentare multitudesof individual decisions to switch political allegiance, each undertakenin response to changing in- centives.So just as a failure to predicta rainstormdoes not imply that the clouds obey no physical laws, a failure to predict some revolution does not implyindividual irrationality. Dunn also suggeststhat revolutionshave too many determinantsto make them amenable to a grand, comprehensivetheory. Shunning the futileexercise of constructinga theorywith universal applicability,we ought to focus,he says,on the particularitiesof each situation.Although

92 Paul Kecskemeti,The UnexpectedRevolution: Social Forces in the Hungarian Uprising (Stanford,Calif.: StanfordUniversity Press, 1961), 1. 93 Ibid., 60, 84-85. 9Havel, Disturbingthe Peace: A Conversationwith Karel Hvz~dala (1986), trans.Paul Wil- son (New York: AlfredA. Knopf, 1990), 109. 95 Dunn, ModernRevolutions: An Introductionto theAnalysis of a PoliticalPhenomenon, 2d ed. (New York: Cambridge UniversityPress, 1989),2-3. 46 WORLD POLITICS I agree that revolutionsare complex events broughton by a symphony of interactingvariables, I depart fromDunn on the usefulnessof general theorizing:obstacles to forecastingparticular revolutions do not preclude useful insightsinto theprocess of revolution.Even if we cannot predict the time and place of the next big uprising,we may prepare ourselves mentallyfor the mass mobilization thatwill bring it about. Equally im- portant,we can understand why it may surprise us. There are other spheresof knowledge where highlyuseful theories preclude reliable pre- dictions of specificoutcomes. The Darwinian theoryof biological evolution illuminates the process whereby species evolve but without enabling us to predict the futureevolution of the gazelle. Sophisticated theoriesof the weather elucidate why it is in perpetualflux but without making it possible to say with much confidencewhether it will rain in Rome a week fromnext Tuesday. Such general theorieshave a common virtue:they reveal the source of their predictivelimitations. The reason theycannot predictinfallibly is not simply that theycontain large numbersof variables. In each theory variables are related to one anothernonlinearly; that is, a small perturbation in one variable, which normallyproduces small changes in other variables, may under the right set of circumstanceshave large consequences. Consider the theoryof climatic turbulencedeveloped by Ed- ward Lorenz. It shows thata sparrowflapping its wings in Istanbul-an intrinsicallyinsignificant event-can generatea hurricanein the Gulf of Mexico. This is because the weatherat any given location is related to its determinantsnonlinearly. In other words, its sensitivityto other variables, and theirsensitivities to one another,are themselvesvariable. Ac- cordingly,variable x may be imperviousto a jump in y from20 to 200, yet exhibithypersensitivity ify risesa bit higher,say, to 202. It may then start to grow explosively,effectively feeding on itself.The notion that small events may unleash huge forcesgoes against much of twentieth- centurysocial thought,with its emphasis on linearity,continuity, and gradualism. But in contextsas differentas technologicaldiffusion and cognitivedevelopment it is the key to understandinga host of otherwise inexplicablephenomena. What endows intrinsicallyinsignificant events with potentiallyexplo- sive power in the contextof political change is that public preferences are interdependent.Because of this interdependence,the equilibrium levels of the public opposition are related to the underlyingindividual characteristicsnonlinearly. A massive change in privatepreferences may leave the incumbentequilibrium undisturbed,only to be followed by a tiny change that destroysthe status quo, settingoff a bandwagon that will culminate in a verydifferent equilibrium. Partlybecause of prefer- SURPRISE IN THE EAST EUROPEAN REVOLUTION 47 ence falsification,the nature of the interdependenceis imperfectlyobserv- able. This is why a massive rise in public oppositionmay catch everyone by surprise. Because preferencefalsification afflicts politics in everysociety, major revolutionsare likely to come again and again as a surprise.This is not to assert the impossibilityof accurate prediction.If we possessed a reliable technique for measuring people's revolutionarythresholds, we would see what it would take to get a revolutionstarted. And if we understoodthe determinantsof these thresholds,we would know when the required conditionswere about to be met. For all practicalpurposes, however, such informationis available only in highlyincomplete form. In any case, there is an irremovablepolitical obstacle to becoming suffi- cientlyknowledgeable: vulnerableregimes can block the productionand disseminationof informationpotentially harmful to theirown survival. Censorship and the regulation of opinion surveys-both widely prac- ticed in prerevolutionaryEastern Europe-are two of the policies that serve these objectives. I have deliberatelycharacterized the source of unpredictabilityas im- perfectobservability, as opposed to unobservability.The degree of imper- fectionobviously constitutesa continuum.Societies with strong democratic traditionsexhibit less imperfectionthan ones with nonexistentor fragiledemocratic freedoms. This is because thereis less preferencefal- sificationin the formergroup, at least with respectto the politicalsystem itself.Accordingly, one can track the course of antigovernmentor anti- regime sentimentmore confidentlyfor Norway, Switzerland,or France than for Pakistan, Brazil, or Ghana. This is why developmentsin Paki- stan are more likelyto catch the world offguard than are developments in Norway; by implication,Norway's political futurecan be predicted with greaterconfidence than can thatof Pakistan. Most countriesof the world lie closer to Pakistan than to Norway as regards the significance of preferencefalsification in sustainingtheir political regimes. This emphasis on unpredictabilityshould not be considered offensive to the scientificspirit: accepting the limitsof what we can expect from science is not an admission of defeat.On the contrary,establishing these limitsof knowledge is itselfa contributionto the pool of useful knowledge. It is also a necessary step toward charting a realistic scientific agenda. "To act as if we possessed scientificknowledge enabling us to transcend [the absolute obstacles to the predictionof specificevents]," wrote Friedrich Hayek in his Nobel Memorial Lecture, "may itselfbe- come a serious obstacle to the advance of the human intellect."96

96 Hayek, "The Pretenceof Knowledge" (1974),American Economic Review 79 (December 1989),6. 48 WORLD POLITICS The predictionof unpredictabilityis not to be confusedwith the un- falsifiabilityof the underlyingtheory. The theorydeveloped in thisessay is fully falsifiable.It implies that political revolutionswill continue to surpriseus, so a stringof successfulpredictions would renderit suspect. Simply put, it can be falsifiedby developing some theoryof revolution that forecastsaccurately. In principle,if not in practice,the presented theorycan also be falsifiedby showing that preferencefalsification was not a factorin unanticipatedrevolutions of the past.