You are on page 1of 35

Trustees of Princeton University

Clouds, Clocks, and the Study of Politics Author(s): Gabriel A. Almond and Stephen J. Genco Source: World Politics, Vol. 29, No. 4 (Jul., 1977), pp. 489-522 Published by: Cambridge University Press Stable URL: http://www.jstor.org/stable/2010037 . Accessed: 04/11/2013 07:17
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

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

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

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND THE STUDY OF POLITICS


By GABRIEL A. ALMOND and STEPHEN J.GENCO*

politicalsciencehas in recent to becomescientific, its eagerness withitsontological base.It has tended to lose contact decadestended as natural events lendingthemand phenomena events to treat political and theother logicas is foundin physics selves to thesameexplanatory in as a phase in be understood part This tendency may sciences. hard of ontological in two steps, and as a diffusion, revolution, thescientific hard scisuccessful fromthe strikingly assumptions methodological and then fromthesebelland economics, to psychology ences: first politicalscience, anthropology, wetherhuman sciencesto sociology, In adoptingthe agenda of hard science,the social and even history. were encouragedby the and politicalsciencein particular, sciences, of science whichlegitimated this schoolof thephilosophy neopositivist homogeneity. assumptionof ontological and meta-methodological and and somepsychologists of science somephilosophers Morerecently, to human abouttheapplicability have had secondthoughts economists of strategy used in hard science.It may be usefulto subjectmatters of politicalscientists. to the attention bringthesearguments
POPPER S METAPHORS

IN

Carl Hempel,and Karl Popper,who along withR. B. Braithwaite, homogeErnestNagel has argued the thesisof meta-methodological of reality, and its has stressed the heterogeneity more recently neity, uses the of scientific He model explanation. a single unamenability to notions the commonsense of clouds and clocksto represent metaphor in physical He asks us to systems. of determinacy and indeterminacy fromthe mostirregular, disorderly, stretching imaginea continuum most to the and orderly, "clouds"on theleft regular, and unpredictable As thebestexampleof a deterministic "clocks"on theright. predictable Toward Poppercitesthe solar system. near the clock-extreme, system we would findsuch phenomenaas penduthisend of the continuum cars.As an exampleof a system near and motor clocks, lums,precision he citesa cluster of end of the continuum, the other, indeterminate, exceptthat gnatsor small fliesin which each insectmovesrandomly
* An earlierversion at the Edinburgh IPSA Congress, of thispaper was delivered AugustI976.

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

490

WORLD POLITICS

backtoward toofarfrom theswarm. itturns thecenter whenitstrays of Nearthisextreme we wouldfind gas clouds, theweather, schools fish, human societies and,perhaps a bitcloser toward thecenter, indihuman vidual beings and animals. The Newtonian revolution inphysics popularized thenotion-which thiscommonsense was to persist forapproximately 250 years-that inexplaining arrangement wasinerror. The success ofNewton's theory and predicting a multitude of celestial and earthbound events by his notNewton himself-to lawsof motion led most thinkers-although werebynature theuniverse andall itsparts embrace theposition that had Phenomena clocklike andin principle completely predictable. that theappearance ofindeterminacy wereviewed as being merely poorly and to be found understood; in time, they alsowereexpected regular ofscience after affirmed predictable. Thus, thereigning model Newton thatall nature laws or, to put it in was governed by deterministic of Popper's metaphor, "all cloudsare clocks-even themostcloudy In theI920'S, thedevelopment of quantum theory challenged this clocklike modelofnature and supported theviewthat indeterminacy andchance were fundamental toall natural processes. With this discovnow thedominant viewheld ery, Popper's metaphor was inverted; orin other that all clocks areclouds; that "tosome degree words, only of cloudiness."' different cloudsexist, though cloudsof very degrees of modelwith thischange Manyscientists and philosophers greeted ofdeterminism from thenightmare itseemed tofree them relief, since that ofhuman choices andgoals. denied theefficacy ButPopper hiscentral that"indeterminism goeson to argue point, ofhuman is notenough" to account fortheapparent ideas autonomy in thephysical "If determinism is true, thenthewholeworld world. isa perfectly flawless all clouds, all organisms, running clock, including all animals, all men.If,on theother Pierce's or Heisenberg's hand, or someother form of indeterminism is true, thensheer chance playsa rolein ourphysical world. Butis chance major really more satisfactory inthenegative. Popper answers Although physicists andphilosophers ofhuman choice basedupontheunpredicthavetried tobuildmodels ofquantum these toocircumscribed. ability jumps,4 he rejects as being
1 Karl R. Popper,"Of Clouds and Clocks: An Approach to the Problemof Rationalityand the Freedomof Man," in Popper,Objective Knowledge:An Evolutionary Approach(Oxford:ClarendonPress I972), 2io; emphasis in original. 3 Ibid.,226; emphasis 2 ibid.,2I3; emphasis in original. in original. 4ArthurH. Compton, The Freedomof Man (New Haven: Yale University Press
I935) .

clouds."'

?"3 thandeterminism

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

491

modelmaybe a model He acknowledges that"thequantum-jump for. . . snapdecisions.... Butaresnapdecisions so very interreally ofhuman esting? Arethey characteristic behavior-of rational human "I do notthink behavior?" so.... Whatwe needfor He concludes: rational human indeed animal understanding behavior behavior-and -is something in character, intermediate between chance perfect and perfect determinism-something intermediate between clouds perfect and perfect clocks.... For obviously whatwe wantis to understand how suchnon-physical as purposes, things deliberations, plans,decisions, theories, intentions, andvalues, canplaya part in bringing about physical in thephysical changes world"5 of arriving Popper's method at a solution to thisproblem seems, liketheproblem to politics to be relevant and political science. itself, His conjecture is that is essentially oneofcontrol; theproblem i.e.,the ofbehavior control and other ofthephysical aspects world byhuman ideasor mental abstractions. Thus,he states that"thesolution must explain freedom; and it mustalso explain how freedom is notjust chance but, theresult of a subtle rather, between interplay something almost random orhaphazard, andsomething likea restrictive orselective control-such as an aimor standard-though nota castcertainly iron control." Accordingly, herestricts thescope ofacceptable solutions to those that "conform totheidea ofcombining freedom and control, andalsototheideaof'plastic control,' as I shallcallitin contradistinction toa 'cast-iron' control."' an evolutionary reaches Popper solution to thisproblem-one that trial stresses anderror orvariation elimination, andselective retention.7 Onlysucha theory canaccommodate plastic control, andthushuman freedom. Once thisis seen,theproblem of therelationship between ideasandbehavior becomes solvable: "Forthecontrol ofourselves and ofouractions byourtheories and purposes is plastic control. We are notforced tosubmit ourselves tothecontrol ofourtheories, for we can discuss them critically, and we canreject them freely ifwe think that fallshort of ourregulative they standards. Not onlydo our theories control us,butwe can control ourtheories (and evenourstandards): isa kind there offeedback here."8 Popperconcludes: "We haveseenthatit is unsatisfactory to look upontheworldas a closed physical system-whether a strictly deter229; emphasis in original. ibid.,23I-32; emphasis in original. 7 See Donald T. Campbell,"Variationand SelectiveRetention in Socio-cultural Evolution," GeneralSystems XIV (i969). Yearbook, 8 Popper(fn. I), 240-4I; emphasis in original. 6

5Popper (fn. I), 228,

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

492

WORLD POLITICS

ministic ora system system inwhich whatever isnotstrictly determined is simply dueto chance; on sucha viewoftheworld human creativenessand human freedom can onlybe illusions.... I havetherefore offered a different view oftheworld-one in which thephysical world is an opensystem. This is compatible withtheviewof theevolution of lifeas a process of trial-and-error elimination; and it allowsus to understand rationally, though farfrom fully, theemergence of biologicalnovelty and the growth of humanknowledge and human freedom."9 ThusPopper tells us that themodels of explanation appropriate to thephysical sciences willnotenable us to cometo grips withhuman and cultural phenomena, and that whilewe can increase our underofthem, standing we cannot explain them fully because oftheir creative andemergent properties.
THE ONTOLOGICAL PROPERTIES OF POLITICS

social Popper's essay presents us with three waysofconceptualizing ofplastic Politandas a system controls. reality-as a clock, as a cloud, ofpolitical science to explain, is clearly icalreality, which itis thetask ofideas-human Itconsists third best captured bythe conceptualization. with constant and intense interaction decisions, goals,purposes-in human andthephysical world. At thecenter of other ideas, behavior, to command, and decisions-decisions this arechoices complex system The political universe has organization; makedemands. obey, vote, elites to command or notto command, whatto commakedecisions howtoimplement andsubjects makedecicommands. Citizens mand, or notto comply; sions to comply, howto comply to makedemands, how to makedemands, or notto makedemands. That is theheart ofpolitics, ourdiscipline matter is committed to exploring thesubject andunderstanding. events arenotsimply these as arethe The relations among reactive, arenotreadily to causeamenable ofphysical encounters objects; they or metaphors. models thisis because "clocklike" and-effect Basically, ofelites andcitizens arenotfixed thebehavioral repertories repertories. from learn inpolitics havememories; The actors they experience. They calculative havegoals,aspirations, strategies. Memory, learning, goal intervene between and and problem "cause" seeking, solving "effect," variable. anddependent between independent in a vacuum; arenotmadeandimplemented Political decisions they and opportunities. ofconstraints aresubject to a complex These array
9lbid.,254-55.

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

493

ofpolitics-range from constraints-the necessities therelatively hard variety or ecological to thequite represented byenvironmental limits soft fashions andfads. Constraints variety illustrated define bypassing the"operational andexhibit milieu" ofpolitical actors'0 varying degrees ofmanipulability. oftechnology, Some, likegeography orthelevel are difficult to alter evenin thelongrun;in theshort are run, they practically nonmanipulable. likecultural andpublic Others, values opinion, arerelatively easyto manipulate in somecircumstances, moreintracin others. table Butmanipulation is very in principle. rarely impossible Even relatively hardenvironmental constraints-such as therelation between material resource needsand population-can sometimes be altered as a consequence of man'screative, The adaptive capacities. agricultural revolution someio,ooo years agomultiplied times bymany thenumber ofpeople in a given capable ofbeing sustained and space, theindustrial revolution ofthelasttwocenturies itbymany multiplied Theseontological of political affairs properties are plainforall to see;they arenotmatters on which reasonable candiffer. persons Social scientists who-for whatever ormethodological reasonsphilosophical denythemand viewhumanbehavior as simply and consereactive quently susceptible tothesameexplanatory logicas "clocklike" natural phenomena aretrying a science tofashion on empirically based falsified presuppositions. That becomes clearwhentheir explanatory schemes arethought ofin terms of their ownbehavior as scientists. Insofar as theyacknowledge of scientific the importance scientific memory, creativity, calculative strategies, goal seeking, and problem solving in their ownwork, they in some must degree acknowledge these qualities in thehuman and social material they investigate and seekto explain. The implication ofthese ofhuman complexities and socialreality is thattheexplanatory of thehardsciences strategy has onlya limited to thesocialsciences. application Models, procedures, and methodoloto explore a world giescreated in which clocklike and cloudlike characteristics predominate will capture onlya partof themuchricher worldof socialand political interaction. Thus,a simplesearch for regularities and lawful relationships among variables-astrategy that has led to tremendous in thephysical successes sciences-will notexplainsocialoutcomes, butonlysomeof theconditions affecting those outcomes. Because the ofpolitical properties differ from reality those ofphysical
10 Harold Sprout and Margaret The EcologicalPerspective Sprout, on Human Affairs (Princeton: Princeton University PressI965).

times again.

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

494

WORLD POLITICS

alsodiffer from those of ofpolitical regularities reality, theproperties are soft. we discover Theyare The regularities physical regularities. rather exhibit ofprocesses that plastic soft because arethe outcomes they in history and involve reTheyare imbedded thancast-iron control. of humanmemories, of largenumbers current "passings-through" and choices human among goal-seeking impulses, learning processes, halfto havea short we discover appear alternatives. The regularities creative searching, life.Theydecayquickly because of thememory, itself mayconandlearning them. Indeed, socialscience that underlie includes not only increasingly tribute to thisdecay,sincelearning itself. from scientific research butfrom learning experience, can be ofpolitical theories boundedness The softness and historical Political scientists arejustifiably proud illustrated bya fewexamples. It is theclosest to a scientific thing oftheir ofvoting behavior. theory a setofwhat tobe "coverIt hasgenerated appear theory that we have. of thevoting correlates deciand attitudinal inglaws"-demographic of Downsian model the The deductive arrived at. sion,inductively of different distributions of voter attiforparty systems consequences Butevena casual basiclaw ofpolitics. tudes lookslikean evenmore in thelastthirty years shows ofvoting research review ofthefindings ofhardscience howunstable regularities are,and howfarshort these ourefforts fall.Modern research on them must inevitably to stabilize in studies ofAmerican elecbehavior progress voting madeitsgreatest ofrapideconomic in theI950's andearly tions growth i960's, a period of American voting behavior in Students and low-intensity politics. that andpredict American voting maintained they could explain period behavior and "candidate of"party identification" on thebasis image"; ofthis effort to a secondary role."The result issues seemed toplay only theory ofvoting explanation wasa psychological produce a hardcausal identification and candidate behavior basedon party image.But this donein theearlyI970's was soonto be challenged by studies theory which datafrom theI930's and latei96o's. Theseearlier include and voters their on the as making choices laterperiods showAmerican basis ofcandidates' issue extent than wastrue positions toa fargreater of oftheI950's andearly i960's. Recent writers the "decomposispeak oftheindividuation oftheparty ofvoting tion" behavior, and system, ofthe"ideologization" ofAmerican Andoneof theleading politics.'2
The VoterDecides (Evanston,Ill.: Row Peterson others, The American Campbell and others, Voter(New York: Wileyi960). 12 Norman R. Petrocik, The Changing Nie, Sidney Verba,and John American Voter Press I976), 345ff; WalterDean Burnham, (Cambridge:HarvardUniversity Critical Elections and theMainsprings of American Politics(New York: Nortoni97o).
11 Angus Campbelland
I954);

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

495

oftheMichigan which collaborators group produced theoriginal partyand identification theory now acknowledges thatthe demographic attitudinal correlates of voting behavior are onlyloosely related, and that theonly we canaspire tois "some kindoftheory orderly specificationoftheconditions under which they vary."'3 Political socialization theory is still engaged in a futile effort to imputerelatively fixed valuesand weights to agents of socializationadult family, school, workplace, media ofcommunication, experiences, andthelike.'4 Likevoting research, socialization research in itsthrust toward scientific has overlooked thelarger parsimonious explanation ofvariables. historical context andtheinherent and instability Jennings Niemi,'5 in oneof themost sophisticated studies of political socializaofparents andteachers tion that theimpact on ever undertaken, report of highschoolseniors was surprisingly weak. thepolitical attitudes to register thefact that thehighschool seniors were Theyfailed they cohort the first of the were the class of sampling i965, post-World War II baby boom. which to a considerable extent It was a generation downin the socialized itself, and it turned socialization theory upside latei960's byproviding innovators oftheyouth rebellion. thecultural is now slowly Like voting behavior socialization actheory, theory of variables. The impact of the knowledging theinherent instability agents of socialization varies withchanges in demographic and social and issues. All that we can and political events structure, technology, tois a collection aspire ofpropositions specifying theconditions under which these impacts tendto vary. ofthese thrusts Perhaps themost vulnerable intohardscience were theefforts in theearlyi960's to disof students of American politics cover therelationships between politics andpublic The problem policy. had beensetbyearlier workwhich argued thatcharacteristics of the voter competition, system-party participation, apportionment, political forpublic as measand thelike-had important consequences policy and particularly uredbythelevelof public expenditures, by welfare A series of statistical studies the political, expenditures. comparing of theAmerican characteristics states in and public economic, policy todemonstrate that these i960's proceeded theI950's andearly political
"PublicOpinionand VotingBehavior," in Fred I. Greenstein 13 PhilipE. Converse, and Nelson W. Polsby,eds., Handbook of PoliticalScience,IV (Boston: AddisonWesley1975), 126. 14 For a recent reviewof theliterature, see David 0. Sears,"Political Socialization," in Greenstein and Polsby(fn. 13), 93ff. 15M. Kent Jennings and RichardG. Niemi,The PoliticalCharacter of Adolescence Press1974). (Princeton: Princeton University

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

496

WORLD POLITICS

on thepolicy variables had little variables. independent impact When forlevelofeconomic the effect of controlled these development, political differences This finding was washed led to theremarkable away. conclusion thateconomic and other environmental variables explain public policy much better thanpolitical variables.'6 Therearetwoaspects ofthis in public research that arenotepolicy forourpurposes. worthy The first is theextraordinary of constriction thetime andspace in this perspectives effort totest a global proposition concerning therelationship between economics, politics, and public policy. The factthatthese weretheAmerican states in the 1950's-a periodof political thanin the 1930's, did not regstability-rather ister as limiting thekindsof inferences thatcouldbe drawn. Politicalscientists no studying these problems brought historical perspective tobearon their research-no memories ofwar, and depresrevolution, sion, andoftheir well-known topolitics andpublic relationships policy. was no recognition there ofthefact Second, that varienvironmental ablescannot directly produce public policy, that choice must political in thenature ofthecaseintervene between andthat them, historically this intervention has beenvery largeindeed. Socialmobilization has sought to explain theory and predict trends towardpoliticization, and de-ideologization democratization, from trends toward urbanization, industrialization, communication, andeducation-only to discover thatwhenthese relationships are examined historically, humanintractability and inventiveness, as well as sheer chance, complicates these patterns enormously.'7 The prophet of the endofideology'8 hasbecome theprophet ofthepostindustrial society' and,currently, theprophet of socialdisjunctions and cultural exhausSocialscientists arefinding that do a better tion."0 they jobofexplaining whenthey follow thecourse ofhistory, using sophisticated methodologiesto isolate necessary sequences and constraints, butalways aware oftheroleof chance and human inventiveness in producing theoutcomes they areseeking to explain. In their fascination with powerful regularities anduniformities that havetheproperties ofcausalnecessity or highprobability, socialscien16See ThomasR. Dye,Understanding PublicPolicy(EnglewoodCliffs, N.J.:PrenticeHall I972), 243-48, for a reviewof thisliterature and a fullerformulation of these findings and inferences. 17For a reviewof thisliterature, see GabrielA. Almond,ScottC. Flanagan,and Robert J.Mundt,eds., Crisis,Choiceand Change (Boston:LittleBrownI973), 8ff. 18 Daniel Bell, The End of Ideology(New York: Free Pressi96o). '9 Daniel Bell, The Comingof Post-Industrial Society(New York: Free PressI973). 20 Daniel Bell, The Cultural Contradictions of Capitalism (New York: Basic Books I976).

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

497

tists much ofsocialandpolitical haveoverlooked thefact that change hastobe explained norbyweakregularineither bystrong regularities butby accidental had a low probaties, that conjunctions-by events of occurring. withparof particular bility The concatenation leaders ticular historical contexts is a matter of chance-offortune-rather than necessity. Scholars canexplain wasripefor revolution why Russia in I9I7; and they can explain ofLenin's and someaspects personality operational code; butthey cannot explain whythetwoconjoined to produce theBolshevik Revolution, only that they conjoined bychance. The problem is similar to that of thebiologist seeking to explain the emergence of a new species. nichein He can describe an ecological terms ofconstraints andopportunities; butfor theniche tobe occupied, thechance occurrence of an appropriate mutation or setof mutations in somerespects Although the problem is similar to thatof the in fundamental the biologist, between it differs ways.The interplay constraints oftheecological niche oftheprocess of andtherandomness tobe sure, is a matter mutation, oftrial and error. The search process is a random one,and largely genetic. In humanaffairs, the search in addition process hasimportant conscious, planful It involves aspects. notonlythechanceconcatenation of a revolutionary political niche witha Lenin,butwitha scheming, contriving, willing, improvising abouttheconstraints Lenin, constantly probing, and learning testing, and opportunities within thenichehe is striving to occupy. Once he doesoccupy theniche and thepopulation it,he transforms occupying itinways that willconstrain (butagainnotdetermine) future adaptive efforts. If we are to understand political reality, we haveto cometo grips notonly with itsdeterminate aspects but, most particularly, with For itis this itscreative, lastcharacadaptive, problem-solving aspects. human isthe teristic which andwhich isthe essentially property, unique ofthesocialsciences. and explanatory mechanism challenge
THE CLOCK MODEL OF POLITICAL SCIENCE

is required.

"behavioral" in political The nowdominant, 'tradition science tends on three andmethodological torest which epistemological assumptions from thehardsciences: ithastaken thepurpose of science is (I) that of regularities the discovery laws of, socialand in, and ultimately scientific political means processes; (2) that explanation thedeductive of individual events under"covering subsumption laws"; and (3) that theonlyscientifically relevant between in the events relationships to a physicalistic worldare thosewhichcorrespond of conception

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

498

WORLD POLITICS

arehighly interrelated, and each causalconnection. Theseassumptions of politics. forthestudy carries important substantive implications mustfirst in political science (i) The emphasis on generalizations WhenDavid Eastonarguedin be understood in historical context. and reliable whenit increases I953 that "knowledge becomes critical organization, when,in short, in generality and internally consistent to generalized statements applicable itis castin theform ofsystematic a tradition ofparticular hewasspeaking numbers against large cases,"'" andinstitutional casestudies ofideographic, descriptive, noncumulative, (witha fewnotable excepthat haddominated much ofthediscipline A similar concern animated thebehavioral tions)forseveral decades. of Truman in theearlyI95O's.22The long-term polemics and others from result of this attempt to shift emphasis description praiseworthy ofthenotion toexplanation, ofgenhowever, hasbeentheenshrining of the eralization aspirations as the sine qua non of the scientific apparent in the recently profession. This is perhaps mostreadily For example, in literature. Scarrow, burgeoning "scope and methods" announces that"Generalizations his Comparative Political Analysis, are the hallmark of all scientific while Conwayand endeavor,"23 An Introduction, that in Political declare "thefuncFeigert, Analysis: as beingthe establishment of tionof science is generally perceived whichexplain withwhichthe laws or theories thebehavior general such is concerned."24 Evena sophisticated study, particular discipline Social Inquiry, as Przeworski's and Teune'sLogic of Comparative that:"The pivotal of this somewhat assumption states dogmatically is that socialscience research, including comparative inquiry, analysis aboutsocialphenomena. shouldand can lead to general statements thathumanand socialbehavior can be exThis assumption implies ofgeneral in terms Introduced lawsestablished byobservation. plained ofpreference, thisassumption will notbe logihereas an expression cally justified."25 on generalizations of thisemphasis is to The substantive impact on regularities, ofresearch focus and stable theattention uniformities,
(New York: KnopfI953), 55. Easton,The Political System in the David B. Truman,"The Impact on PoliticalScienceof the Revolution in Heinz Eulau, ed., Behavioralism in PoliticalScience Behavioral Sciences," reprinted
21 22

(New York: Athertoni969).


23

Harper & Row i969), 33.

An Introduction (New York: Political Analysis: HowardA. Scarrow, Comparative

24 MargaretConway and Frank B. Feigert,Political Analysis:An Introduction (Boston:Allynand Bacon 1972), 17. Social Inquiry 25Adam Przeworski and HenryTeune, The Logic of Comparative (New York: Wiley1970), 4.

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

499

patterns of association in political of unique processes at theexpense or low-probability events or political As Frohock outcomes. expresses itin TheNature ofPolitical with Inquiry, "Science is concerned establishing causal andgeneral relations laws.To do this thesocial scientist must on systematic concentrate ofhuman conduct. patterns Onlyas an event is a recurring ofa general class canitbe treated instance scientifically."" We arenotarguing herefortheviewthat do notoccur regularities in political processes or that validgeneralizations cannot be made.As we noted above, political regularities-albeit exist and are soft-clearly crucial to political inquiry. Rather, ourcriticism is aimedat positions thatsee regularities and generalizations as theonlyproper of objects scientific political inquiry. This seems to us an unnecessary delimitationof thescopeof thediscipline's If political subject matter. reality isbest viewed as a conjunction ofchoice andconstraint, andas a source ofboth regularity andinnovation, then political should notbe science limited to a consideration ofonly partofthis A purefocus on reality. as "the hallmark ofall scientific generalizations wouldseem endeavor" tocondemn ittojustsucha limitation.
(2) The concern with generalizations andregularities-and theconcomitant willingness tolimit thescope ofpolitical science toonly those aspects of political reality thatare generalizable-is closely associated witha particular of explanation in political conception This inquiry. position in the"scopeand methods" is also reflected literature. Alan in hisScopeandMethod Isaak, ofPolitical Science, that declares politicalscientists must accept the"scientific fact oflife" that "every sound explanation andprediction contains atleast onegeneralization; without generalizations couldbe no explanations there or predictions."27 SimiandFeigert larly, that Conway argue "Explanations in science require ... lawsortheories which arewellestablished.... Explanation occurs when tobe explained thefacts canbe deduced as a logical consequence ofthelawsortheory and ... other known facts."28 of The model explanation alluded to hereis theso-called "covering law"ordeductive-nomological (D-N) model in thephilosodeveloped phyofscience byR. B. Braithwaite,29 CarlHempel,30 and others. The
I967),
27

26Fred M. Frohock,The Nature of Political Inquiry (Homewood, Ill.: Dorsey

Conwayand Feigert (fn.24), 27. Braithwaite, Scientific Explanation (Cambridge: Cambridge University PressI953). 30 Hempel,Aspects Explanation (New York: Free PressI965); see also of Scientific Ernest Nagel,The Structure Braceand World i96i). ofScience(New York: Harcourt,
28 29

i969), 8o.

Isaak, The Scope and Method of Political Science (Homewood, Ill.: Dorsey

14I.

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

500

WORLD POLITICS

when is explained something modelis that this basicideaunderlying of a moregeneral classof things. to be a member it has beenshown isknown caseofwhat itas a special something istoexhibit "To explain whentheparto themodel, according in general.""3 Thisis achieved, general law (or setof laws) that from a more ticular caseis deduced "covers" cases. Thatis whygeneralirelevantly similar itandall other explanations. rolein deductive zations playsucha fundamental that from thefact derives oftheD-N model power The explanatory that whichis denecessitates lawslogically from covering deduction ofwhat on thebasis us that, "explains" bytelling The deduction duced. was to be thecasein question know(thegeneralization), we already Thisnotion of"itwasto be thewayitdid.32 it hadtooccur expected: ofexplanaconception ofthedeductive stands at thecenter expected" and explanation between forthecloseassociation tion, and accounts an explanaoftheD-N model, Foradherents inthemodel.33 prediction a prediction tionthatwouldnot be equallycapableof supporting therefore, It is notsurprising, explanation.34 as a true wouldnotqualify termiin Popper's models" systems-"clock thatcloseddeterministic As Hempelputsit: to D-N explanation. mostamenable nology-are to theD-N modelare conforming examples ofexplanations "The best of deterministic character.... [T]he laws theories basedon physical in forthechanges ofstate aredeterministic bysucha theory specified ofthat atanyonetime, deterthey state system the that, given thesense or later, earlier time."35 at anyother, mineitsstate to thedegree that that theD-N modellosesitsusefulness It is clear in theexplanation to thelaw or lawswarranting there areexceptions maintain that "all A's areB's" and If we cannot legitimately question. "someA's areB's,"thenthe fora law asserting settle onlythat must and ourexplanation oftheoccurrence ofB linkis dissolved deductive Thisstate ofaffairs, is justwhat however, continues tobe problematic. that ofplastic control. we can means Plasticity is implied bythenotion
33 in theSocial Sciences(Chicago: AldineAtherPaul Diesing,Patterns of Discovery ton I97I), 164. while its obverseis maintained See Hempel (fn. 30), 367, wherethisposition as an explanation-is mustalso qualify put aside.This modificathata validprediction and prediction" thesisof explanation has not always tionof the so-called"symmetry scientists. See, e.g.,Oran Young,"The Perilsof Odysseus: been appreciated bypolitical in Raymond Tanterand Richard Theories in International Relations," On Constructing Relations (Princeton:Princeton Ullman, eds., Theoryand Policy in International Press1972), 183. University
35Hempel (fn. 30),
351;

31AbrahamKaplan, The Conductof Inquiry(San Francisco: Chandler1964), 339. 32Ibid

see also Nagel (fn. 30),

323.

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

501

in principle, thatthere will be exceptions to anygeneralizations expect, to us. Thus, thatare of interest we mightformaboutthephenomena the moreour subject matter exhibits plasticcontrol, the less it will be amenableto simpleD-N explanations. is closelyassociatedwith the idea of (3) The notionof causality covering-law explanation by bothpoliticalscientists and philosophers in of science. R. B. Braithwaite, forexample, describes causality strictly termsof covering laws: "[T]he statement thatsome particular event of a general is theeffect of a setof circumstances involves theassertion law; to ask forthecauseof an eventis alwaysto ask fora generallaw whichappliesto theparticular event."36 This formulation is echoedby politicalscientists. Dahl Thus, Robert arguesthat"If we wishto explainan event, E, in a strictly causal manner,we consider E as an effect and bringit undersomegeneralization of the form:'EveryeventC is accompanied laterby an eventE.' . . . The C is called the cause,E the effect."37 Similarly, Isaak maintains "If saying that, that'A causesB' is tantamount to 'B alwaysfollows A,' thentheyare bothreducible to 'If A, thenB.' In otherwords,we can known as a causal relationship expresswhat is traditionally without All of thesecharacterizations reston the notionof causality as an statusacquired? As explanatory concept.But how is thisexplanatory to theliterature on causation can be seenfrom evena cursory exposure the concepts are broad and amand conditions,39 "cause" and "effect" biguous. One elementof theirmeaning seems to stand out in any As account,however: the principleof "same cause, same effect."40 Hempel puts it, "as is suggested by the principle'same cause, same that [a givensetof] circumstances effect,' the assertion jointlycaused and wherever of a given eventimpliesthatwhenever circumstances
36Braithwaite (fn.29), 2; see also Hempel(fn.30), 348-49. in the Studyof Politics," in Daniel Lerner,ed., Cause 37Dahl, "Cause and Effect and Eflect(New York:FreePressi965), 87. 38 Isaak (fn. 27), 95. 39See, e.g., ErnestSosa, Causationand Conditionals(Oxford: OxfordUniversity PressI975); MylesBrand, ed., The Natureof Causation(Urbana: University of Illinois
40 There are manydisputes the philosophical statusof causality concerning thatgo well beyondthis consensual elementof its meaning-forexample,the problemof a constant whether the causal connection represents or conjunction, logical necessity, of thetemporal "natural" necessity; and theproblem and contiguity ordering of causes of thesein terms relevant For a discussion to political and effects. scienceresearch, see Georg Henrik von Wright, Explanationand Understanding (Ithaca, N.Y.: Cornell PressI971). University

using the termcause."38

Press i976).

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

502

WORLD POLITICS

ofthekindto be explained an event takes thekindin question occur, cautious formulation: more place."'" Or,in Abraham Kaplan's slightly in terms of somerelation of is usually "Causalconnection analyzed of is at a implication: thegrammar the'if-then' conjunction least startIfthecauseoccurs, itseffects occur."42 It is this ingpoint. then element on causal of"samecause, that sameeffect" confers explanatory power in theworld. it,"causality" relations Without becomes simply another problematic and essentially unexplained relationship between twoor more things, events, or processes. oftherelationship between cause Thisphilosophical characterization of to notion cast-iron control. The andeffect is closely related Popper's ofthecauseis theexplanacause produces theeffect, andtheexistence as narrowly ofpure tion oftheeffect. A world causeandeffect, defined with ofcausality would identification bythis covering-law explanation, a worldthat couldnotbe other than be a worldwithout exceptions, is completely aliento theworldof whatit is. Sucha world, we feel, forsurprise and innovation is inherent politics, in which thepotential ifnotmost, in many, situations. oftheexplanatory oftheinflexibility of In spite and aridity concept haveattempted to couch scientists causality, however, many political in terms ofthenotions ofcause their analyses ofpolitical phenomena an of often mixture formalized andeffect. The result is odd definitions ofsuch a mixture, substance. As an example we andunrelated empirical ofpolitical that hasmade a brief lookatonebranch might take analysis on the formulations-the literature of useofcausal considerable concept isexplicitly ofcause andeffect as a invoked power. Herethe relationship for a necessary, connection Forexbetween events. metaphor dependent "for 'C haspower Simon that theassertion has stated ample, Herbert 'C'sbehavior over the assertion causes R'sbehavior.' R,'wecansubstitute Ifwe candefine we can define thecausalrelation, or influence, power Andrew and viceversa."" Similarly, McFarland asserts that authority, of poweror influence basedon suchconcepts "definitions as force, incentives orutilities, andminimum are... reducible coalitions winning to causalterms."44 Morerecently, JackNagel has defined poweras follows: "A power actual orpotential, is an actual relation, orpotential between causalrelation thepreferences of an actor regarding an out43Simon, ModelsofMan (New York: Wiley1957), 5. in Pluralist Power and Leadership 44McFarland, Stanford UniSystems (Stanford:
29.

Hempel (fn. 30), 348-49. 42Kaplan,"Noncausal in Lerner(fn.37), 146. Explanation,"


41

versityPress i969),

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

503

itself."45 edition AndRobert comeandtheoutcome Dahl,in thelatest maintain of his Modern seems to with Political Analysis, (although hislong-standing viewthat causation is fundamental somecaveats) to "whenwe singleout understanding powerand influence relations: in order to give influence from ofhuman interaction all other aspects itspecial what we focus attention on is attention, interests us andwhat that oneormore ofthepersons in this interaction getwhat they want, oratleast getcloser toactin towhat they want, bycausing other people some particular way.We wanttocallattention toa causalrelationship whatA wants between andwhat B does."46 it How is theword"cause" usedin these definitions? being Clearly in thesense is notbeing usedas an explanatory concept, described by For an explanation to be truly causalin that philosophers of science. in question as we haveseen, wouldhaveto be sense, therelationship (i) cast-iron, (2) generalizable, to covering-law and (3) amenable wouldseemto apply explanation. None of these properties to power relationships. Thereis no "necessity" in theoutcome inherent of an attempt to assert poweroveranother as there is in a causal person, connection between two physical The target of the power objects. attempt may, for anynumber ofreasons, actdifferently than thepower wielder wouldhavehimact.Thisis because a power relationship does not involve cast-iron oftwochoosing control; instead, itisan interaction and mutually constraining individuals, each withhis own resources, goals, purposes, interests, and strategies. The intentions and resources of thefirst certainly constrain thechoices and actions of thesecond, butthey do notdetermine in anysort those choices andactions ofcastiron sense. This"looseness offit" between thebehavior and intentions ofactors involved in an attempt to exercise power means that their relationship is notreadily is itparticularly generalizable; neither to strict amenable As HartandHonore haveputit: "Thestatecovering-law explanation. ment thatone person did something because . . . another threatened no implication orcovert carries assertion that ifthecircumstances him, action thesame would a statement were nordoessuch repeated, follow; foritsdefense, as ordinary causalstatements require do, a generalization.. .." Theseconsiderations lead us to conclude thatthepower
46Dahl, ModernPoliticalAnalysis(3rd ed.; EnglewoodCliffs, N.J.: Prentice-Hall in original. 1976), 30; emphasis 47H. L. A. Hart and A. M. Honore, Causationin the Law (Oxford: Clarendon
52. 1975),

45Nagel, The Descriptive Analysisof Power (New Haven: Yale University Press
29.

Press1959),

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

504

WORLD POLITICS

at leastnotin theexplanatory senseofthe relationship is notcausal, Thisconclusion wouldseem, in one sense, tobe shared byDahl and whousecausallanguage in their many of theother political scientists definition their of power of power. If we examine empirical analyses in real-world relations situations rather political thantheir definitions, we find of thecomplex careful and precise examinations interactions thatcontribute on simplistic without reliance notions to outcomes, of "samecause, sameeffect." In suchsubstantive to analyses-as opposed definition is recognized and indeterminateness making-plasticity is often handledwithsophistication and insight. toobserve in this Whatwe seem areaofpolitical particular research, is a rhetorical or metaphorical-rather thanexplanatory-usage then, ofcausallanguage in formalizations anddefinitions. Thisaccounts for the lack of a subsequent commitment to actualcausal analysis in research. substantive The somewhat incongruous gap can perhaps best be explained as an attempt ofpolitical on thepart tocreate scientists a "halo effect" around their theoretical formulations. Our longing for fullscientific status hasledus tocreate a kindof"cargo cult," fashioningcardboard imitations ofthetools andproducts ofthehardsciences in thehopethat ourincantations wouldmakethem real. Thesethree elements oftheimplicit informs logicthat much ofpolitical science research today appear to imply a substantive modelofthe political world which closely resembles thedeterministic "clock model" outlined byPopper. Thatis notto saythat anypolitical scientists actuallyseethepolitical world this way;no doubt we wouldall agree that itoften appears tobe quite porous, andunpredictable. irregular, Rather, it is to say thatthe arsenal of meta-methodological principles and procedures we have borrowed from thephysical sciences-or, more froma certain correctly, philosophical perspective on the physical sciences-has cometous with an array ofsubstantive assumptions that all proclaim theprinciple "all clouds areclocks." Ifwe search only for in political generalizations and regularities ifwe couchour processes, ofthecovering-law explanations onlyin terms and ifwe view model, political relationships as ultimately causalin nature, we are committingourselves-whether we recognize it or not-to a disciplinary research to strip program designed awaythecloudlike and purposive of political in orderto exposeits "true"clocklike aspects reality structure. If politics is notclocklike in itsfundamental structure, then
48For further arguments along similarlines,see Terence Ball, "Power,Causation and Explanation," viii (Winter1975), I89-214. Polity,

term.48

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

505

thewhole program is inappropriate. this We believe tobe thecase:the current quandary in political canto a large science extent be explained bythefact that, bythemselves, "clock-model" areinapproassumptions priate fordealing withthesubstance of political phenomena.
THE ADOPTION OF THE CLOCK MODEL AND ITS EFFECTS ON POLITICAL RESEARCH AND PEDAGOGY

The movement toward hardscience in thestudy is a phase ofpolitics in thescientific revolution ofthelastseveral decades. The great breakthroughs in physics and biology, and theextraordinary increases in research funding as science became a national asset, created a mood of sanguine expectations. It is not surprising thatpolitical scientists sought to share in thisexciting and remunerative adventure. Political science was invited to imitate thehardsciences by some of themoreinfluential philosophers of science on thegrounds that political reality lent itself tothesame powerful methods that hadproven so effective in physics and biology. That is one of thebasictenets of thelogicalpositivist tradition in thephilosophy of science,49 and has beena starting point for many books andarticles designed to showthe socialsciences and history a "truly" howto achieve scientific status.50 In addition, there was immediate evidence ofthesuccess of thehardscience strategy within thesocialsciences themselves. and Psychology economics had beenthefirst disciplines in thesocialsciences to move in this direction, demonstrating thepossibilities of experimental methods, sophisticated quantitative methods, computer and simulation, mathematical modelling. The combination of philosophical legitimationandthedemonstrated progress ofpsychology and economics was toresist. impossible As a consequence of these and demonstration legitimations effects, the incentive structure ofpolitical science toencourage anorientabegan tionmodelled on thephysical sciences. The pressures forconformity canbe measured in terms ofprestige, journal publications, fellowships, and grants. Majorsources of research funding and graduate fellowsuchas theNational ships, Science havebeendominated Foundation, by the hard sciences; the socialscience divisions have been junior
Prediction, and 'Imperfect' Knowledge," in Herbert Feigl and GroverMaxwell,eds., Minnesota Studiesin the Philosophy of of Science: Vol. 3 (Minneapolis:University Minnesota Press I962); RichardS. Rudner, Philosophy of Social Science (Englewood Cliffs, N.J.:Prentice-Hall On theEvolvingStandard I966); Rudner, "Comment: View in Philosophy ofScience," American Political ScienceRetview, Vol. 66 (September I972).
50See, e.g., Nagel (fn. 30);

49See von Wright (fn.40), chap. i.

Hempel (fn. 30), chap. 9; May Brodbeck, "Explanation,

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

506

WORLD POLITICS

hasbeenthe andthepolitical science section partners in these agencies, of hardscience thathavetheappearance most junior of all. Projects forgaining substantial research support. havehad theinside track of hard consequence of thisimitation Perhaps themost important criterion on method as theprimary for science has beenan emphasis theleading Today, in political science. judging thequality ofresearch rather than bytheir methodologies research traditions tend tobedefined of organizationof thisprinciple their substantive foci.One result the consequence ofit-has beenthat although certainly nota necessary by its technical to be measured primarily valueof thisworkseems of the problems by the importance virtuosity, and onlysecondarily treated orilluminated. drive toward In thelasttwodecades there has beena tremendous in a thistrend in political science. Rikercelebrated quantification Review when Political Science recent communication totheAmerican of thearticles in recent issues of he commented that sometwo-thirds employing sophistianalysis thatjournalwerebasedon quantitative contributed to major catedstatistics.5' has undoubtedly Quantification Butit has also in political and other socialsciences. advances science exercises led to a significant number of pseudo-scientific thatexhibit in thephysical of research sciences. theform butnot thesubstance whentheuseof quantification is Suchstudies become more prevalent toward rather as a means understanding than treated as an endin itself hasrecently been Irrelevant quantification concrete political problems. in international relations,52 comparaofsearching critiques thesubject and elsewhere. tive studies,54 politics,5" policy in political science has movedincreasingly Quantitative analysis methods. Butthestructure of the statistical toward moresophisticated oftencomesinto conflict with the data in social science research statistical The problems theory. confirmatory assumptions underlying to nonrandom, statistical methods nonin applying complex involved notbe minimized.55 Muchof the or nonadditive data should linear,
AmericanPoliticalScience 51William H. Riker,quoted in "EditorialComment," Vol. 68 (June1974), 733-34. Review, in PoliticalScience,"WorldPolitics, Data Analysis 52EdwardR. Tufte,"Improving
xxi (July i969). 53Andrew vii (JulyI975). Are Not Enough,"Comparative Politics, Mack,"Numbers
54 ii

"A Critical Look at Quantitative Methodology," PolicyScience, Ralph E. Strauch, (Winter1976). 55See, e.g., HaywardR. Alker,"The Long Road to International Relations Theory: WorldPolitics, xviII (July1966); HubertM. Nonadditivity," Problems of Statistical in Variables: The Problemof Multicollinearity," Independent Blalock, "Correlated Analysis of Social Problems(Reading,Mass.: Edward R. Tufte,ed., The Quantitative Addison-Wesley 1970).

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

507

of the is lostwhenthestructure inferential methods powerof these of thetheory. These requirements to therigid datadoesnotconform enough to lead somestatisticians, formidable haveproven difficulties data-analytic to devise alternative Tukey, at Princeton, suchas John advanced as powerful as themost notnearly that, although techniques characwith idiosyncratic compatible the methods, aremore statistical into tohavefallen Herewe seem data.56 andpolitical teristics ofsocial developphases ofThird-World oftheearly tothat a trap comparable pooragricultural ment introduced into were when "high technologies" We are disruptive consequences. fortheir without regard countries which technology, levelof statistical discovering that an intermediate of socialdata,is more thespecialcharacteristics takesintoaccount than are the verysophisticated to the social sciences appropriate methods. in political science on statistics Running parallel to thisemphasis of simple, logically and theconstruction is an interest in mathematics in comparative has been advocated models. This approach rigorous scientists who arguethat"political by Holt and Richardson, politics ifthediscipline is toprogress scientifically. turn tomathematics" must one: "In pathfrom thestatistical this to distinguish Theyarecareful about we are nottalking an appealformoremathematics, making witha basisforrigorous a science . . . [S]tatistics provides statistics. thatthecrying needin comparative induction. Our critique suggests this is where mathematics, for deduction and more rigorous politics is is echoedby A. James This statement is relevant."57 not statistics, others. OranYoung,"andmany Gregor, is that measure with models mathematical they usually The difficulty For ofthephenomena beingmodelled. to thecomplexities up poorly advocates theuse of modelling OranYoung,who strongly example, has candidly observed that"The in international relations, methods that its little is may display hazard of this products procedure inherent fortheindefinite ofinternational relations relevance to therealworld a matheon theother that hand, argue future."59 HoltandRichardson, mustnecessarily science takea radically oriented political matically
56Tukey, David Data Analysis (Reading,Mass.: Addison-Wesley Exploratory i977); forth(Reading,Mass.: Addison-Wesley, C. Hoaglin,A FirstCoursein Data Analysis coming). 57Robert in ComparaParadigms M. Richardson, Jr., "Competing T. Holt and John

Science Review, Vol. 62 (June i968), 425-39; Young (fn. 34). 59Ibid.) i96.

of Comparative in Holt and JohnE. Turner,eds., The Methodology tive Politics," Research (New York:Free Press1970), 70. 58 Gregor, American "Political Scienceand theUses of Functional Political Analysis,"

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

508

WORLD POLITICS

circumscribed viewofpolitical reality, cutting itself free from problem "A science solving: that is heavily committed to dealing withsocially andmorally relevant problems finds little usefor this kind ofparadigm or forthecommitment to mathematics thatit requires. For political science to advance, itmust shedthis professional commitment to solvingsocial andmoral problems."6" One aspect ofthemathematical topolitics approach deserves special mention: theuseofrational-choice models toexplain behavior. political Thesemodels are particularly interesting because takethemost they intractable elements ofpolitical processes-the individual andcollective ofpolitical choices totreat them actors-and try Some deterministically. analysts ifpolitical haveargued that is ever tobe a true science science, notion the ofrationality must Forexample, be itscentral Riker concept. and Ordeshook drawan explicit on the analogy between rationality onehandandthenotion ofmechanism on theother: ... itis clear that theassumption ofrationality andtheassumption of in theexplanation mechanism play comparable roles ofthesocial and world. physical The mechanical assert that there assumptions is something about things that assures us they will(usually) move regularly, and therationality assumption asserts that there is something about people that makes in a regular them behave (usually) way.In each case, thefunction is togeneralize about theregularity.61 The kindof regularities and Ordeshook are concerned Riker with here areofa special type-"postulated" as opposed to "observed" regularities. thatchoices in empirical situations Granting fail to usually exhibit thedegree of regularity forwarranting necessary deductive explanations and theories, Rikerand Ordeshook chooseto build a ofpolitics on thefoundations ofhowpeoplewouldactif they theory maximizers. ofcourse, rational leadstoa theory were that utility This, tomodel well.Butthesubstantive lossisconsidered fails political reality ofthemethodological in light ofpostuacceptable gain:"The method is positively more because it permits latedregularity theeasy efficient, a of and and offers generation hypotheses single parsimonious explanationofbehavior."62
61WilliamH. Rikerand PeterC. Ordeshook, An Introduction to PositivePolitical
60 Holt and Richardson (fn.57), 70-71.

A sympathetic Theory (Englewood Cliffs, N.J.: Prentice-Hall ii. 1973), yet sober choicemodelsforexplaining of theutility of rational and predicting evaluation coalition and CabinetFormations behavior is offered by AbramDe Swann,CoalitionTheories (San Francisco: Jossey-Bass 1973). 62Riker and Ordeshook we can onlyassumethat (fn. 6i), 11-12. By "explanation," mean "definition," Rikerand Ordeshook since the postulation of rationality defines a but does not explainit in any way. (hypothetical) typeof behavior,

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

509

The popularity of rational-choice in political would models science bepuzzling toanyone whowasnotfamiliar thecurrent with hierarchy ofmethodological and substantive priorities in thefield. Butwiththis hierarchy in mind, someparticularly perplexing exercises become understandable. For example, in the recently published Handbookof Political Science, J.Donald Mooncontributes a pieceon "The Logic ofPolitical Inquiry."6" This article begins very promisingly byarticutheD-N model lating ofexplanation as wellas an important alternative to it, the interpretive model,whichexplains behavior in terms of in motives, intentions, rulesand norms, etc.Notingserious defects both models, Moonturns to thetaskofsynthesizing thetwoin order to create a morecomprehensive framework forpolitical explanation. out notto be a synthesis at all; instead it But the"synthesis" turns of a substitution of a rational consists actor "modelof man"forthe ofexplanation. the"looseness" and interpretive model Thiseliminates lackofregularity of empirical choice is captured that bytheinterpreforit "presuppositions the tivemodeland substitutes [that]specify withdescriptions of decisional of theactors premises which, together theactions which their for about situations, provide therationale bring theoverall to pattern of socialbehavior . . . that. . . theorists desire explain."64 to Riker Liketheregularities ofinterest and Ordeshook, these "presuppositions" arepostulated (specified) a prioi. Theyreplace theconchoice and action with causaland lawlike ofempirical tingent aspects to an algorithm a are reduced assumptions. Thus,choices specifying calculation. The netresult necessary outcome from a necessary utility ofthis ofchoice in terms ofcausesubstantive reduction is a definition whichis to say,a definition of choicethat and-effect relationships; this conclusion denies theexistence ofchoice! wouldappear Certainly if we werenotfamiliar withthecurrent of method strange priority As it is, we can see thatMoon is in political science. oversubstance with thetask offitting hisrecalcitrant matter to the struggling subject ofnecessity notion that strict ofa methodological bears little exigencies ofpolitical choice. resemblance totherealities andformalizaThe stress on reductionist explanation, quantification, ofgraduate If a political curricula. tion hasalsoled to an overloading and sociologist, must be a statistician, then scientist some psychologist, in be set has to aside order to ofthetraditional curriculum makeroom
63Moon, "The Logic of Political of OpposedPerspectives," Inquiry:A Synthesis in and Polsby(fn. IA I. Greenstein 14Ibid., I94.

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

510

WORLD POLITICS

who has taught and techniques. Anyone forthese newer disciplines in thelasttwenty science ofpolitical graduate department in a major and technicizing ofnarrowing process this inexorable willrecall years have beenrerequirements theforeign-language of thecurriculum; from five to havedropped requirements examination duced, thefield it had become evento two.By themid-ig6o's, perhaps four to three, withlittle science a Ph.D. in political to become forsomeone possible political foreign history, political theory, ofpolitical ifanyknowledge American politics much about andeven relations, international systems, "Training remarked: hasrecently Alker As Hayward andgovernment. such methods quantitative in multivariate intensively students graduate a sophistifordeveloping makes lesstimeavailable analysis as factor and said about beenthought of whathas classically catedawareness for is inappropriate particularly training life.... Thusmodern political about systems in which many questions politics modern understanding raised."65 are continually restructuring of thegraduate and technicization thisnarrowing Accompanying tradiof older intellectual has beena demoralization the curriculum and Political science. theory andin political in thesocial sciences tions inand descriptive law and publicadministration, public philosophy, andsecondary peripheral, defensive, haveall become stitutional analysis science tradition a large ofthe political As a result, part matters. subject generations. effectively to younger transmitted being is no longer exis nota setof methods hereis that"science" Whatwe suggest philosophers as the neopositivist physics, from mathematical tracted a commitment and to explore it is ultimately haveus believe; might The means ofempirical reality. a given segment tounderstand attempt in "good"science, be secondary: inpursuing this goalshould employed matter being matter rather thansubject methods arefit tothesubject of notion in orderto fitit to a preordained truncated or distorted have should that socialscientists Thisis thelesson "scientific method." it and, haveignored they from sciences. Instead, learned thephysical Kaplanhascalledthe whatAbraham in theprocess, haveundermined is to redeem itself, "Social If socialscience of inquiry."" "autonomy own notions of 'goodscience,' their their needto construct scientists to their subject particular appropriate own methodological approach that there is some up thenotion giving matter.... Thisviewimplies
65 I3), VII, I57. (fn. 66

and Poisby Alker, "Polimetrics: Its DescriptiveFoundations,"in Greenstein Kaplan (fn. 31), 3.

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

511

in thephysical to basicresearch in thesocialsciences closeanalogy sciences."67


SECOND THOUGHTS IN PSYCHOLOGY AND ECONOMICS

of "scientific has acquired our discipline Much of theknowledge in disciplines thetwo"pace-setting" through method" hasbeenfiltered at the If we look closely and economics. socialscience-psychology in theuse of whichhavepioneered disciplines of these state present we and experimentation, models, mathematical methods, statistical and disillusionment. ofsomedoubt evidence find has overthelastcoupleof science, muchlikepolitical Psychology, the concerning debate" "great constant a nearly decadesentertained the discipline. underlying principles and methodological conceptual be conceptualofpsychology, matter man,as thesubject How should and acquire, to hope psychology should ? knowledge kindof ized What in someparticipants Lately, be pursued? best knowledge howcanthis orthoofthe established critical andmore more this havebecome debate assumptions. sacrosanct previously doxyand havebegunto question inanydiscipline, minority dissenting arenottheinevitable Thesecritics in theprofession-leaders leaders someof therecognized butinclude conceptions thevery in creating havebeeninstrumental who,in fact, nowquestion. they has beentaken of man"in psychology ofthe"image The problem was and lucid discussion trenchant A particularly up manytimes. to theSociety address Cheinin his i962 presidential byIsidor offered "among Cheinargued that ofSocialIssues. Study for thePsychological of an impotent imageof Man is-that . . . theprevailing psychologists and bytwodistinct determined completely withitsresponses reactor, on impinging offactors: (i) theforces sets interacting, albeit separate, ... momentary in thelatter term (including itand (2) itsconstitution that this is obviously He heldthat false, image states)."68 psychological ourcardinal obligation can clingto itonly"byviolating psychologists tosupport faith in oursubject scrupumaintain matter, as scientists-to and to observe without willful fully thatwhichwe observe, lously bias."69 attests and whatobservation to,is clearly Whatthisimagedenies,
67

of Social Issues,xviii (Octoberi962), 3. 69Ibid. Similararguments are made in Rom Harre and P. F. Secord,The ExplanaI972). tion of Social Behavior(Totowa, N.J.: Rowmanand Littlefield

6i, 62. I974), I03 (Summer 68 Chein, "The Image of Man," Journal

of Social Science," Daedalus,Vol. "On theNatureand Condition Marc J.Roberts,

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

512

WORLD POLITICS

a helpless, manis "an active, that responsible agent, notsimply powerChein "I am saying that we should notpermit lessreagent." continues: ourselves to be seduced, as so many of us havebeen, bythose pretenMan tious high order conceptualizations ofPsychology that would deny in the thequality offreedom-and, that is inalienably his,thequality makeMan,as a psychological inaccessible."70 denial, agent, to Popper's. The deterThis argument bearsa strong resemblance isrejected of"cast-iron over choice andaction minist control" assumption fora conception thatallowsforthe autonomy of humanactionin as wellas in responding creating, to,theworld. Interestingly enough, ofviewing event Cheinclaims to be a determinist-in thesense every as having thatmotives necessary and sufficient conditions-but argues ofhuman thus andpurposes share in thedetermination actions, bringLikePopper, Chein is human ingthem under direct control. therefore, suchas purposes, with ofhow"mental events" concerned thequestion in can playa partin bringing about deliberations, plans, etc., change thephysical world. In hispresidential Assoaddress before theAmerican Psychological in i975,DonaldCampbell ciation calledon psychologists to showa bit "all scientific ofepistemic andtorecognize that is humility, knowledge andincompletely corroborated atbest." indirect, presumptive, obliquely in psychology He went on to argue that reductionism must be seenas a first in a long-term notas an endin itself: step research strategy, of ourfield Considering thecomplexities and ourmodels from the of thesuccessful a strategy of deliberate initial overhistory sciences, topsychology. hastoberecommended But this simplification guarantees in theearly that ofdevelopment stages thetheoretical orthodoxy will be misleadingly willportray reductionistic, humans as more simple Ifpsychologists machines than they actually are. atsuch a stage were to that of their losetheperspective thisviewwas a product long-term strategy, were instead to exaggerate thedegree of perfection of their topropagate andwere these current theories, immature theories as final ofpopular could be destructive values.... Here thenetresult truth, initial the ofdeliberate a science again, requiring strategy oversimplificatoadopt a demeaning, scholars tion recruit may overeager mechanistic, view ofhuman nature.7' reductionistic some havemanaged tomove atleast the Today, psychologists beyond research basedupona mechanistic imageof man,and are pursuing thenewer in anduseful realistic more conception. Among approaches
in original;i8. 2; emphasis and Between and Social Evolution Between theConflict Biological xxx (DecemberI975), II20, American Psychologist, and MoralTradition," Psychology
70Chein (fn. 68),

"On 71 Campbell,

II2I.

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

513

which examines theory," is "attribution for example, social psychology, the "naive thatconstitute hypotheses and working the assumptions own behaviors interpret their peopleas they of ordinary psychology" has in this field, Lee Ross, oneoftheleaders ofothers. andtheactions of thisapproach: summed up thesignificance
cultheory in socialpsychology of attribution ascendancy The current of man. discipline's conception to upgrade that minates a longstruggle
of radical behavior(S-R) automaton No longerthe stimulus-response and cognitive processor beyondthe rank of information ism,promoted man has at last been awarded a status seeker, psychological consistency him. For man, in the who investigates equal to that of the scientist is an intuitive who seeks of attribution psychologist theory, perspective about actors and their to explain behavior and to draw inferences environments.72

Lee Cronbach. Reflecting psychologist consideration by theeducational in experimental overthe last two social psychology on his experience "Should social scienceaspireto decades,Cronbachasks the question, to laws?" He observes that"Social scientists generally, reducebehavior in particular, have modelledtheirworkon physical and psychologists them to restructure to amassempirical generalizations, aspiring science, laws intocoherent theory. intomoregeneral laws,and to weld scattered That lofty is farfromrealization."" aspiration Cronbachargues,is The essential withthismethodology, difficulty mutathatsocialsciencelaws,unlikephysical laws,seem to be highly "At one time a ble. As he puts it, "Generalizations decay."Further, situation the existing describes conclusion well, at a latertimeit acit is valid only as countsfor ratherlittlevariance,and ultimately of an empiricalproposition may be great or The half-life history. the half-life of relations the shorter small. The more open a system, in thetaskof building theories to be." He compares within it are likely if a "It is as we needed a mechanical with problem: assembly thisway grossof drycells to power an engineand could make one a month. cells beforewe had half the would leak out of the first The energy ofourgeneralizations."" So itis withthepotency completed. battery
72 Ross,"The Intuitive in theAttribuDistortions and His Shortcomings: Psychologist X Social Psychology, in L. Berkowitz, ed., Advancesin Experimental tion Process," Pressi977), i74. (New York: Academic American Psychology," 73Cronbach,"Beyond the Two Disciplinesof Scientific xxx (February ii6, I25. I975), Psychologist, 74Ibid.,I22-23.

it can thekindofknowledge second problem, Whatofpsychology's hasrecently beengiven careful man?Thatissue about toattain expect

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

514

WORLD POLITICS

At theendofthisarticle, of aspiration which recounts twodecades toward a nomological psychology, Cronbach writes:
Socialscientists arerightly from proud ofthediscipline wedraw the naturalscience sideofourancestry. Scientific discipline is whatwe uniquely addtothetime-honored man. an identificaofstudying Too narrow ways tion with science, however, hasfixed oureyes uponan inappropriate goal. The goalofourwork, I haveargued is nottoamass here, generalizations ineachgeneration is topindownthe task ofthesocial scientist contempoand the rary he shares withthehumanistic scholar facts. Beyond that, into andto intheeffort togaininsight artist contemporary relationships, realign theculture's viewof man withpresent realities.75
atop whicha theoretical towercan some day be erected.... The special

like psychology and socialpsychology, has also been Economics, in recent itstroubles havebeensurhaving The critical themes years. is seenas isolated and inbred, withits thefield prisingly consistent; formal little to theempirical models world bearing very resemblance with which economists tobe concerned. Thesecriticisms aresupposed haveforquitesometimebeenthestock-in-trade of suchestablished and John gadflies of theprofession as Gunnar Kenneth GalMyrdal braith. that for hasargued economists havefailed Myrdal, to example, ofan inappropriate produce relevant knowledge because commitment to themethods ofthesimpler natural sciences:
to emulate amongmy economic whattheyconceive of as colleagues themethods of thenatural sciences by constructing utterly simplified models, often givenmathematical dressing.... It shouldbe clear, of a form, however, thatthisadoption whichthe in more natural scientists, simple, pointed questions, can usefor analysis andpresentation, doesnotreally makethesocial sciences more scientific, ifthat form is notadequate to social and therefore, reality notadequate for ofit.76 theanalysis Galbraith usedtheoccasion ofhis i972 presidential Similarly, address to the AmericanEconomicAssociation to chide the profession forits failureto come to grips with practicaleconomicproblems:"Neoclassicalor neo-Keynesian economics, thoughproviding unlimited ophas a decisive flaw.It offers fordemanding portunities no refinement, thatnow besetthe the economic usefulhandleforgrasping problems
modern society..
75Ibid.,
I26.

In recentdecades . . . therehas been a strenuous, even strained, effort

. No arrangementfor the perpetuationof thought

Vintage I972),

76GunnarMyrdal, Againstthe Stream:CriticalEssayson Economics(New York:


I43.

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

515

is secure ifthat thought doesnotmakecontact with theproblems that itispresumed tosolve."77 Thesedoubts a bitmore andconcerns havelately become widespread. "a significant MarcRoberts, a younger economist, asserts that proportion ofrecent theoretical work in economics hasbeenoflittle scientific value. Many papers explore questions notbytheworld posed itself, but bysomeone else's model."78 Theseviews seem tobe shared bysomeof oftheeconomic themost respected leaders OskarMorestablishment. an in in thatecogenstern, important paperpublished 1972, argues nomicsis in a crisisbecauseit lacks the concepts, and methods, todeal adequately with socialandpolitical itneeds philosophy reality. of current Following a discussion equilibrium theory, Morgenstern observes: is striking; thetime for The contrast with hascome economic reality theory toturn around to"face themusic." the andthe ofproving There is,ofcourse, always possibility temptation noempirical allsorts oftheorems which have relevance whatsoever ... is whether what Yettheultimate criterion thetheorem asserts is what inreality. isfound Onecannot but bereminded ofHansChristian help oftheEmperor's Andersen's clothes.79 story who won theNobelMemorial Wassily Leontief, Prizefortheinvention of input-output an evenmorepessimistic has struck analysis, note.In hispresidential address to theAEA, giventwoyears before Leontief that"The uneasiness Galbraith's, argued is [in economics] ofthe caused towhich notbytheirrelevance practical problems presenttheir butrather address inadedayeconomists efforts, bythepalpable withwhich to solvethem..... quacyofthescientific means they try formathematical Uncritical enthusiasm formulation tendsoften to oftheargument conceal content theephemeral substantive behind the "In no other of algebraic He concluded that formidable front signs." has so massive and sophisticated a statistical ofempirical field inquiry results."" beenusedwithsuchindifferent machinery as in psychology, wouldseemto be in economics, The problems much likePopper, substantive. primarily Morgenstern, sounding points
"Powerand the UsefulEconomist," American Economic Re77John K. Galbraith, view, Vol. 63 (MarchI973), 2. 78 Roberts (fn. 67), 6o. 79Morgenstern, "ThirteenCritical Points in Contemporary Economic Theory," Journal of Economic Literature,x (DecemberI972), ii64-65. 80 Leontief, "Theoretical and Nonobserved Assumptions Facts,"American Economic Vol. 6i (March I971), Review,
I, 2,

in original. 3; emphasis

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

516

WORLD POLITICS

tothe failure ofeconomics todealseriously with the nonphysical aspects ofeconomic processes:
preferences, states ofinformation, expectations, etc., etc., thatdetermine themovement andsignificance ofthephysical components ofthewhole economic phenomenon. We arefar from having more than broad notions of howto describe and measure their share in a concrete situation. Do we evenhavea goodmethodology we couldapply?81
... theoverwhelming on the physical emphasis aspectsof the economic process. . . seemsone-sidedwhen we realize thatit is plans,decisions,

in commenting on theinability ofeconomic Robert theory. Heilbroner, remarks that to predict thecourseof a national of economics economy, "it may be thatthisis less possiblethan it was, becausethe economy and so much now is so muchmorea creature of decision itself making, of sheerinterplay of impersonal thatprediction lesstheoutcome forces, becomesinherently moredifficult."83 This major problemin economicswould seem to have important forpoliticalscience.For what the economists are saying implications is becoming matter it is thatto theextent theirsubject morepolitical, to scientific and formalistic is becoming lesssusceptible methodologies. of shifting in of thepossibility the economy The impactof decisions, of the impersonal undermines the regularity forces new directions, and modellingexerallowed forsuccessful thatpreviously predictive does not augur well forthosewho envision an cises.This conclusion formalizedpoliticalscience.Indeed, the tendency seems eventually economics morelike to be in theopposite direction; maybe becoming science! political have had to deal with A second and relatedproblemeconomists mention:the problemof decaying Like psydeserves generalizations.
81Morgenstern (fn.79), 1187-88.
82 83

It usedto be,and apparently if still is in muchofeconomic theory notpractice, that these decisions and expectations couldbe discounted because they tended to cancel one another outin theclassical market situation. Today,however, many economists attribute a largepartof howextenthediscipline's empirical dilemma toa failure toappreciate nowoverride of themarket. sively political decisions themechanisms "in placeofthemarket now Galbraith that observes system, we must assume that forapproximately halfof all economic output there is a power or planning system."82 The effect of this injection of planning intotheeconomic process has beento upset thepredictive capabilities

May

(fn. 77), 4. Galbraith New York Times Magazine "Economists in Recession," Quoted in Wade Greene,
12, 1974,

p. 64.

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

517

in itsattempt economics hasbeenunsuccessful tobuildlasting chology, As Leontief models matter. empirical ofitssubject putsit: In contrast tomost physical sciences, westudy a system that is not only butalsoina state I have in mind exceedingly complex ofconstant flux. notthe obvious change in the variables . . . that ourequations aresupposed toexplain, butthe basic structural relationships described bythe form andthe parameters ofthese equations. In order toknow the what ofthese structural shapes relationships areat anygiven actually time, wehave tokeep them under continuous surveillance.84 Thesesecond thoughts in economics the and psychology illustrate degree towhich thetwobellwether disciplines arenowreassessing their earlier explanatory strategies and meta-methodological commitments. Clearly, their attempts todealwith thecomplexities ofsocial reality in terms of a modelof scientific method from the physical borrowed thanthey sciences has run intomoredifficulties had expected. The of thiseffort to bring the humanenterprise underthe ambivalence has beencaptured categories and logicof thehardsciences by the Albert who points out economist and socialphilosopher Hirschman, "A Passion in a recent book-in a section entitled forthePossible"itas their todiscover that "Most social scientists conceive exclusive task anduniform andstress stable . . ." regularities, relationships, sequences rather thanrecognizing "themultiplicity and creative disorder of the humanadventure." thatthesocialscientists wouldbe He maintains if their search forgeneral surprised and even"distraught laws were crowned with success," andconcludes, . . . all the total "Quite possibly in thesocialsciences, successive theories and models and theimmense that aremotivated ifunconscious, efforts go intothem, bythenoble, desire of thesocialworld to demonstrate theirreducibility to general to affirm laws! In no other so conwaywouldit havebeenpossible "85 offreedom and creativity. as therealm clusively thesocialworld is experiencing itself a process ofre-evaluaofscience Thephilosophy similar to that tionand reorientation and taking placein psychology which we have usedas a metaphorical Thearticle economics. byPopper is butoneexample ofa more for ourownthinking trend guide general ofPolanyi,87 andthat in thefield exemplified byhiswork86 Hanson,88
A BiasforHope (New Haven:Yale University 0. Hirschman, Albert Press1971), 27. and Refutations Conjectures (New York: Basic Books I963); The Logic of Scientific Discovery (New York: Basic Books 1959). 87 Michael Polanyi, Personal of ChicagoPress1958). Knowledge(Chicago:University 88 Norwood R. Hanson,Patterns of Discovery(Cambridge:Cambridge University and Explanation: A Guide to Philosophy Press 1958); Observation of Science (New York: Harperand Row i97i).
85

84Leontief (fn. 80), 3.

8"Popper (fn. i);

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

518

WORLD POLITICS

others. and many Kuhn,89 Quine,90 Lakatos,91 Toulmin,92 Today,the pre-eminent position heldby logicalpositivism in thephilosophy of science seems to be weakening. Philosophers of science no longer see their roleas one of legislating the"rules" of science; they are more ofresearch. likely topursue descriptive andexplanatory modes Science as a logical is viewed as an activity or a process, notsimply product. an appreciation Accordingly, is beginning to develop forthedegree to whichscience-humankind's loftiest intellectual achievement-is grounded and dependent uponbasiccommon senseand informal as wellas formalized substantive knowledge.93 Philosophers are learning about howscience more grows andhowitprospers. The newer literain the ture philosophy ofscience isrich in insights andimplications for theenterprise ofsocialscience.
IMPLICATIONS

it has distinctive If thewholeof socialreality properties rendering forms of explanation, unamenable to simpledeductive-nomological ofall thesocial is especially ofpolitics this thecasefor thestudy which, on collective and adaptive focuses most sciences, directly goal-seeking withthe searchfor A political science processes. solelyconcerned whichconstrain choicewouldmissthedistinctive regularities aspect from toescape todiswhich is theeffort ofpolitical reality, constraints, to problems in the context of consolutions covervalue-optimizing W. Bennett recommends an approach to The anthropologist John which is oriented around andresearch theconanthropological theory ofadaptation: cept from likeculture or thereductive of abstractions Instead behavior, onhuman ofpsychology orgenetics, focuses actors formulas [adaptation] torealize andfind who while try objectives, satisfy needs, peace coping In their humans create the social future with conditions. present coping, newproblems orperpetuating ofgenerating oldonesand in thesense ofthepopulation construction in the thebiological even may modify ofstrategies, the factors that thechoice process....Byanalyzing guide
89 ThomasS. Kuhn, The Structure Revolutions of Scientific (Chicago: University of ChicagoPress1962). 90W. V. 0. Quine, Ontological Relativity (New York: ColumbiaUniversity Press

straints.

91Imre Lakatos, "Falsification and the Methodology of Scientific ResearchProand the Growth grammes," in Lakatosand Alan Musgrave, eds., Criticism of KnowlPressi97o). edge (Cambridge: Cambridge University 92 StephenToulmin, Human Understanding, I (Princeton:PrincetonUniversity and Understanding Press1972); Foresight (New York: Harperand Row i96i). 93 See Campbell (fn.71).

i969).

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

519

we gain knowledge of change of thepossibility and direction and the relation ofhuman tothemilieus.94 behavior

We wouldarguethat whatBennett has to sayaboutanthropology applieswithevengreater force to political science: "theimportant phenomena for an adaptational anthropology are dynamic human purposes, needs andwants... ." The emphasis ought to shift "toward strategic coping, thatis, theattempt to realizeindividual and social objectives through themobilization of socialand material resources. This category of humanbehavior has become dominant in thecontemporary world with itsinterdependence andgrowing constraints on free action."" DuncanMacRaeargues a similar thesis regarding thedevelopment ofthesocial sciences in thelastseveral decades.
They [thesocialsciences] haveevolved from an earlier form of social analysis, lessspecialized andrecondite, byimitating thenatural sciences tive pathto useful application liesthrough objective research andtheory construction, free from thecomplications ofideological andphilosophical dispute. Theyhavethusdeveloped distinct technical terminologies and of graduate methods of research, specialized journalsand programs the discourse instruction. thesedevices Through theyhave separated of ofthegeneral ofspecialists from andthecommunications that public, from of the social one another. The course the individual specialists decadeshas thusbeenguidedby the thepastseveral sciences during it to science-however distinct modelof natural they mayseemfrom natural themselves.96 scientists
. . . many social scientists have become convincedthat the most effec-

of thewithdrawal of thesocial to thisproblem MacRae'ssolution is tointroduce into social theuniversity sciences from problem solving which willcombine social a "discipline theories and ofpolicy analysis" discourse. that thepresent He believes with ethical analysis disciplined in thedisciplines of situation andvaluative ofcognitive fragmentation solution-the can onlybe overcome socialscience by an institutional of policy introduction of research and teaching departments analysis andapplied social science.97 in organizational and are lessfaith We havesomewhat solutions, ofpolitical science-which has tended that thediscipline to convinced toassign toa special MacRaenowwishes abandon thetask discipline
in Anthropology," and theConcept of Culture 94Bennett,"Anticipation, Adaptation, Science, Vol. 192 (May 28, I976), 847. 95Ibid., 850, 851. 96MacRae, The Social Functionof Social Science (New Haven: Yale University Press 1976), 3. 97Ibid.,277ff.

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

520

WORLD POLITICS

and evaluation a central ofreasserting rolein thestudy is still capable oftheexample attraction ofthenatural ofpublic policy. The powerful short ofouraspirahavefallen sciences hasbegun tofadeas ourefforts ofthetrend among ourmethodologists, tions. Despite theprominence ofourleading centers ofgraduate inourleading andin some journals, in theUnited majority oftheprofession instruction, theoverwhelming a sense resists States either themodel, and abroad actively experiences to it.Most or is indifferent ofobsolescence because ofitsprominence, inpolitical ofthe science, settles for goalslessambitious published work or hisThis workincludes descriptive thannomothetic explanation. framemaking limited useoftheoretical torical accounts orcasestudies and contributes to theaimsofunderstandworks and generalizations, alternatives political reality andpolicy ing,interpreting, andexploring topolicy which as crucial analysis. MacRaeidentifies in our forgreater rigor One might makethecasethatthesearch ifitsclaims might havemademore progress understanding ofpolitics lessexaggerated, lessdifficult and expectations had beenlessextreme, A more cautious toscienreality. approach tosquare with a recalcitrant andsocial thepeculiarities ofhuman tific reality, progress, recognizing in a moregeneralacceptance of appropriate might have resulted valueof formal-mathematical formulaoftheheuristic quantification, andthelike. methods, tion, experimental that a quarter of a century of It is ofinterest ago,in theaftermath science in thesocialdistoward WorldWar II, whenthemovement thisrelationship between thesearch for ciplines was justbeginning, solutions to to discover value-optimizing regularities andman'sefforts hispredicaments was moreclearly understood. One has onlyto combookwiththemorerecent ones parean early "scopeand methods" of thepioneers cited above.Sometwenty-five of the years ago,many in thesocialsciences movement contributed to a volume behavioral in Scope and The PolicySciences:RecentDevelopments entitled Method. In theleading HaroldLasswell stated hispriorities: chapter, whattopics "Ifourpolicy areto be served, ofresearch needs aremost of pursuit? . . . What are themostpromising methods of worthy and interpreting their significance forpolicy?How gathering facts in thedecision-making can facts and interpretations be madeeffective ?" The sameessay theintroduction ofscientific itself celebrated process intothe socialsciences-statistics, methods mathematical modelling, and related Butthisscientific of method was hardening approaches. of problem valueclarification, setin thecontext and theensolving, ofthehuman condition. Lasswell hancement looked as uponmethod

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

CLOUDS, CLOCKS, AND POLITICS

521

making possible actsof "creative whichmight move imagination" in constructive mankind directions awayfrom thetyrannies and catasThe connection betweenthe search for regularities and political creativity-clearly seenby that generation freshly from returned Washington and themilitary theaters of World War II-was gradually lost in the decades thatfollowed.The "methods"messageof Lasswell's sermonwas heard and acted upon with the mixed results we have reviewed, but the"policyscience"messagelargely fellon deaf earsfor reasons we have suggested above. What is underattackhere is the peckingorder,and the particular setofpriorities and resource whichhave come to dominate allocations, theprofession in thelastdecades. Thesepriorities and allocative policies, and thispeckingorder, are legitimated notby successes in theexplanation of politicalreality, but by the example and the demonstration effect of thehardsciences. A peckingorderin whichmathematization and sophisticated statistical analysis are viewedas the only sourcesof "real" or "powerful" theory, while theories producedfromthe interplayof imagination and induction are treated or "weak" as "heuristic" cannot be justified by the explanatory performance of the theory, former. Theories are inherently weak in the human sciences-both thosethatlook "strong" becausetheylook like thetheories of physics, or psychology, economics, and thosethat look "weak" because they derive from hypotheses theexamination ofindividual casesor historical Another aspectof thepeckingorderwhichis undercriticism hereis thedistinction between pure and appliedpoliticalscience.Even in the hardsciences, thecomparative intellectual of so-called payoffs pureand applied research are not at all clear-cut. Important discoveries often outofappliedresearch. In thesocialsciences, emerge including political losesitsmeaningsincethe specialcharacteristic thisdifference science, of social reality is man's adaptivebehavior. The partof the discipline which calls itselfpure politicalscience,searching for powerfuland has missedthe essential enduring regularities, pointof itssubjectmatthecontext of political ter.At bestit illuminates butit leaves decisions; theadaptive thepolicyoptions, unexplored and their searching process, of the as consequences. Surely to study public policy-viewed efforts and overcome constraints-is adaptto,cope with,modify, as basic and as is the search for constraining pure an undertaking regularities.
98 Daniel Lerner and Harold D. Lasswell, eds.,The PolicySciences:Recent Developments in Scope and Method(Stanford: Stanford PressI950), 3, I2. University

trophesof the I930's and I940's.98

experience.

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions

522

WORLD POLITICS

of political Indeed, science-insofar we might arguethat theessence it studies-is as it is to be defined of thepolitics the by theessence of constraints. analysis of choice in thecontext Thatwouldplacethe forsolutions forregularities, to problems, and the search thesearch evaluation solutions on thesamelevel. ofthese Theywouldall be parts man'spolitical of a common effort to confront fatewithrigor, with an inescapable ofidentification andwith sense thenecessary objectivity, matter which thepolitical scientist studies. with thesubject ofresearch andprofessional Ourpolicies needtobe support 'training fromimitating freed the hard sciences. Policystudies, institutional studies, and philosophically sophisticated evaluative studies areclaimwith much as is antson research as acsupport legitimacy currently cordedmathematical, and psychological and sociological statistical, reductionist studies. ofpolitical substance initsinstitutional, Knowledge hastobe re-established on an equal historical, andphilosophical aspects withsophisticated footing and reductionist methodologies knowledge in ourprograms ofgraduate A whole ofmeta-methodtraining. library ological handbooks and primers imposing themodelof hardscience on political hastobe re-evaluated in a newlight. reality Thesevolumes do notrepresent the"true path"to scientific progress; are rather, they a historical with mistaken a flirtation that deviation, metaphors tempoof socialscientists. the imagination Their historical rarily captured buttheir to practical importance is thus great, relevance research problemsin the socialsciences is limited. To progress the scientifically, socialdisciplines their own philosophy of science require based on and obligations to explanatory strategies, possibilities, appropriate human andsocial reality.

This content downloaded from 200.130.19.144 on Mon, 4 Nov 2013 07:17:12 AM All use subject to JSTOR Terms and Conditions