SOCIAL THEORY AND COMPARATIVE POLITICS Mark I, Licbbach It is nor che "actual" interconnections of "things" but the conceptual interconnections of problems which define the scope of the various sciences. Max Weber For the believer there are no questions; for the nonbeiiever there are no answers. Rabbi Menacbem Mendel We Live ... amid che debris of Reason. Adam Seligman INTRODUCTION Ebegin where inquiry should always begin; an assessment of the problem situa-... tion toward which inquiry is directed (Popper 1965).1 What is the currenc state of theory in comparative politics? Compared to twenty-five years ago, self-C-Conscious theoretical reflection finds almost no home in our field. We do not take our theories or our theorists seriously. l] wane co thank Robert Bates, Jeffrey Kopstein, Peter Lange, David Mape!, Michael McGinnis, 1 James Scarritt, James C Scact, Adnm Seligman, Sven Sceinmo, Nina Tunnenwald, and Alex Wendt; the ' Participants in the May, 199G, Btown University conference on "Interests, Identities and Institutions in ^Comparative Politics - Samuel Barnes, Peter Hall, Ira KatzneUon, Margaret Levi, Joel S. Migdal, Marc ^ Howard Ross, Sidney Tarrow, and Alan Zucketman; Barbara Geddes and che audience at the tm> panels 6"n "Theory in Comparative Policies" at the 1996 Annual Meetings of the American Political Science Association, San Francisco, Californa; and che graduate students in my 5075 course - Introduction to Polit- Science — for their lively and DtOVOCative mmmpnrq nn pfirlmr rlmfre rhic rhnnrer 240 CONCLUSION Evidence to support this harsh judgment comes from our leading journal's recent symposium on "The Role of Theory in Comparative Policies" (.World Politics, October 1995). The participants minimized the value of deductive, a priori theorizing of the sort chat is done within strongly defined research communities. While the symposium included widely acknowledged experts in specific research traditions, apparently no one viewed, for example, today's ra-tionahst—culturalist divide as theoretically interesting, exciting, and productive. Structural or institutional analysis was not even recognized as a distinc- C tive theoretical enterprise but rathet was thought to be part of the field's-" "messy center." Most participants feared that comparative politics might re-turn to the sort of Marxist—functionalist debate that characterized it in the ; 1950s and 1960s.Consequently, method — prediction, comparison, councerfac-ruals, history, quantitative and qualitative data, explanation, interpretation, causation, and generalization — was on everyone's mind. The "nomothetic" vs. "ideographic" divide was what really animated discussion. The consensus was that most comparativisrs are part of the consensus: Today's comparativists prac- .' tice "theoretically informed empirical political analysis" and adopt "divetse conceptual lenses" (2) and "eclectic combinations" (5). They are interested in "{• "questions" and "empirical puzzles" (10). Hence, "comparative politics is very much a problem-driven field of study" and comparativists are mostly interested Y in solving "real-world puzzles" (46). The flaw of this pragmatist, means-oriented heaven is obvious: "If the problem orientation of the field tends co relegate the role of theory mainly to that of a tool of empirical research, the quest for causal generalizations, by contrast, moves its role to the forefront" (47). Similarly, the conclusion from a mechods symposium on comparative (smail-n) studies in another leading journal {American Political Science Review, June 1995) may be stated as paraphrase of Kant: Good I theory without good research design is empty; good research design without good theory is blind (454). As Rogowski's (1995) important essay makes clear, one cannot begin inquiry with "evidence" derived from and used to test "theo- \ ry"; one must begin with theoretically embedded observations. The inevitable - j conclusion is that researchers must eventually reflect on the nature of that theory . — which leads co questions broadly defined as "social theory" or "philosophy of ; social science."2 World Politics's symposium did not contribute to the cause of theory in comparative politics because ics picture of cheocy in our field as dominated by a - -j "messy center" is inaccurate and self-defeating. This chapter seeks to refute that . perspective and advance theory in comparative politics in three ways. First, I recognize rhac three ideal-type research traditions — the rationalist, cultucalist, and structuralist — ate active in contemporary comparative politics, z Actually, not so inevitable. World Politia rejected a version of this chapter with the comment, B°c ~~ torn line: the exclusive focus on theory is not For us...." Can o discipline mature iFno one specializes in SOCIAL THEORY AND COMPARATIVE POLITICS 241 just as they are astir throughout the social sciences.3 Section 1 thus begins the analysis with three exemplary comparativists. Each thinks of himself or herself as a member of a strong research community. Robert H. Bates (1989) argues that he is a racionalist, James C. Scott (1985) identifies with the cukuralists, and Theda Skocpol's (1979) work is determined by structuralist principles. "While each recognizes the value of synthesis and the cross-fertilization of ideas, each is principally concerned with advancing a particular intellectual tradition and theoretical agenda that transcend comparative politics. Section 2 deepens this analysis by dissecting each research community's oncology, methodology, comparative strategy, lacunae, and subauditions. Second, I set the dialogue among the schools within che historical context of che development of social theory. Section 3 thus attempes to understand the thtee tesearch communities by tracing them back to Talcotr Parsons's (1937) effort to systematize several classic social theorists and thereby integrate social theory. I have modified his approach to take account of the scructure-action problem of reconciling individuals and collectivities. I call this modified approach the socially embedded unit act. Using this meta-framework to provide insight into the individual frameworks, I demonstrate the underlying unity and significance of the approaches for addressing questions of social theory. Finally, I set the dialogue among che schools within the historical situation confronting today's comparativists. Section 4 thus seeks an underlying unity in ; rationalist, culturalist, and structuralist thought by delving even further back to Max "Weber's master problem of a century ago. Weber studied the dialectic of v modernity in world historical and comparative perspective: how reason and non-rationality manifest themselves at individual and societal levels with great normative and empirical significance. The dialectic is important to contemporary , politics in che West. Due to the West's influence on the globe, the dialectic is equally important to the entire world community of nations. Seccion 5 is a summary of my theme about the problem sicuacion of contemporary comparacive policies: There are fundamental difficulties with a field that consists only of a "messy cencer" and basic virtues in a field that embraces creative confrontations, which can include well-defined syntheses in particular . research domains, among strongly defined research communities. Comparativists ' should explore che rationalist—culturalist—structuralist debate and thereby appreciate the different structute-action combinations of interests, identities, and institutions that guide inquiry. Even self-described "problem-oriented" compara-i'tivists — those who think of themselves as pact of a "messy center" - should be aware of the competing research traditions that have historically been a part of "the social sciences. We cannot remain theoretically challenged — a field of theoretical philistines — and actually solve substantive problems. Contemporary com- 3The current tevival of intetcst in the philosophy of che social sciences, which has centered on the *. significance of rationality, culture, and structure fot social theory, has become o miní-texebook industry. *My Graduate svllnbus. which i-nnrm'n^ nn lm-rn-rlnrp íHr nf r*.rprpTirrs. ij; available uoon reouest. 242 CONCLUSION parative policies therefore will be greatly enriched by a dialogue among the traditions, especially one chat is informed by self-conscious reflection about the enduring issues of social theory.4 Comparative politics needs strong and yet mutually sympathetic intellectual communities: believers who raise questions and nonbelievers who appreciate answers. SECTION 1: THREE EXEMPLARS In order co demonstrate lub-l uui n<^u ^^^^ ~.------------ j is necessary to examine comparativists who consciously specialize in specific re. rhp mnst widely cited and deeply respected ;: Robert H. Bates's (1989) Beyond the. Miracle of the Market, James works in contemporary comparative politics------ C. Scott's (1985) Weapons of the Weak, and Theda Skocpol's (1979) States and Social Revolutions. Bates explains how reason shapes the political economy of agrarian development in Kenya. He offers a materialistic theory of polirical preferences: An: actors location in Kenya's agrarian economy shapes his or her preferences about^; economic and political institutions. Bates also argues that institutions shape the calculations of political entrepreneurs and hence affect how material interesrs-; are defined, organized, and aggregated by vote-maximizing politicians. Inter-'-ests, in other words, are both materially and politically determined. The tragedy is that these reasoning voters, politicians, consumers, and producers create the drought, famine, and subsistence crises that plague the people of Kenya. Bates's*, book is therefore a seminal study of "the impact of economic interests upon pol-itics and the impact of institutions upon economic interests" (46), one that explores both the intended (and wanted) and unintended (and unwanted) coose.-?;. Scott's study of the peasant village or Sedaka in Malaysia tajtcs u vuy. different perspective: "The peasants of Sedaka do not simply react to objective i the interpretation they place on chose conditions as me- quences of reason. idy of the peasant village of Sedaka in Malaysia takes a very differ e: "The pe;-------rcj„i,„ j„ rParr rn objective candji; cions per se but rather to tne interpremuuu mcy piai-u uu u.u^ ,________ diated by values embedded in concrete practices" (305). He argues thate.the^_______ discourses and praccices of class conflict in Sedaka take the form of "everyday forms of peasant resistance" in which the poor and the well-to-do abide by di£--^ ,_'| ferent norms and rules. Scott's book is therefore a masterful analysis of ihe frag: ile ideological hegemony of the landed elite over the peasantry, one that traces. the basis of a reasoning and nonrational class order to the creation of iden«°csisjj, and communities. ''in the international relations field, the debate between neoliberais and neorealisrs helpsS£EUCtU^,^ inquiry and inform scholarly identities. Articles and books have thus evaluated the meaning ^ ^^gSflf cance of the controversy (Baldwin 1993; Kegley 1995; Keohane 1986). Because they participate!? ^ debate, out colleagues in the field of international relations are wcil aware of the value of social t'I*|^^e ----- -u„^„r;„i un,M (Weadt. forthcoming). The underlying purpose of this chapter nndyg^. SOCIAL THEORY AND COMPARATIVE POLITICS 243 Skocpol, in a comment that could have been directed at Bates or Scott, rejects a "purposive image" of social causation that "suggescs that revolutionary processes and outcomes can be understood in terms of the activity and initiation or interests of the key group(s) who launched the revolution in the first place" (17). Skocpol explains revolution by "rising above" the subjective viewpoints — the interests and identities — of the participants. She takes a structural perspective, or "an impersonal and nonsubjective viewpoint — one that emphasizes patterns of relationships among groups and societies" (18). Skocpol is especially interested in "the institutionally determined situations and relations of groups within society and upon the interrelations of societies within world-historically developing international structures" (18). Skocpol's book is therefore a classic comparative historical analysis of revolution in France, Russia, and China, one that traces the reason (e.g., the development of democracy, markets, and state bureaucracies) and irrationality (e.g., the blind violence and human costs) of revo-lution to an "iron cage" of forces5 that operate behind the backs of individuals. In sum, Bates offers a rational/social choice study of how interests produce .the dialectic of reason and irrationality in Kenya's political economy, Scott a cul-'. turalist/mterpretivist account of how communities and identities constitute the Sdialectic in Malaysia's class relations, and Skocpol a structuralist/institutionalist I analysis of how social forces drive the dialectic in the French, Russian, and Chi-1. nese revolutions. Even a cursory examination of our recent journals and books re-Is.veals that comparativists today have indeed coalesced around these three com-^.peting research schools: Social choice theories, culturalist approaches, and rl structural analyses offer competing visions of the field. As in political science !jS.rhore generally and social science even more generally, interests, identities, and || institutions coatend for theoretical primacy in comparative politics (Garrett and k Weingast 1993; Hedo 1994; Selznick 1992: 78) 'X, This significance of the rationalist—culturalist—structuralist dialogue in com-I parative politics is also demonstrated by the contributions to this volume that ex-'U amine substantive research areas. Peter A. Hall shows that the field of compara-smltive political economy involves a lively and fairly equal struggle among interest-, Ife'dea-, and institution-oriented perspectives. Doug McAdam, Sidney Tarrow, and ^.Charles Tilly indicate that the three main concepts used to explain contentious hpohtics — political opportunities, mobilizing structures, and cultural frames — •Qmbody rationalist, culturalist, and structuralist elements. Joel S- Migdal maintains that rational choice and culturalist perspectives have been marginal to comparative studies of the state but are now challenging the hegemonic structuralist ^Perspective. Finally, Samuel H. Barnes evinces that the survey research ttadition Uti the study of voter turnout and partisan choice has been affected by all the perspectives but has come to rely recently on the (declining?) rationalist approach. , j'Wehet's ({1904-05] 1985: 180 used of the term actually corresponds to an iron cage that is orifii-JP!% produced by actors and their ideas bur eventually becomes a set of material forces that externally con- 244 CONCLUSION If space permitted, the footprint of the rationalist—culturalist—structuralist dialogue could have been traced in comparative studies of democratization, globalization, modernization, and several other substantive domains. The intellectual problem, of course, is that as one contemplates the three exemplars from the competing research schools, one Is forced to recall a line from the old Monty Python show — "and now for something completely different." How are we to make sense of the fact that three such different theoretical perspectives coexist in the same field of comparative politics? Perhaps we should ask the question in ways that engage our three authors. Can rationalists, culturalists, and structuralists secure mutually profitably intellectual exchange, or is monopoly inevitable? What overall meaning can the thtee schools have, apart from a' teleology toward.ideological hegemony? How can a single discipline be structured such that the three perspectives coexist, or must one approach institutionalize its victory? In the face of this disorienting pluralism, partialism, and perspectivism, admirers of Weber can take comfort in one of his memorable lines: "It is not the 'actual' interconnections of'things' but the conceptual interconnections o( problems-which define the scope of the various sciences" ([1903-17] 1949". 68, emphasis in original). This nominalist proposition follows from the Kantian argument that' concepts or theories without empirical intuition or observation are empty phrases; empirical intuition without concepts is blind. Kant thus stresses the ordering ; function of theory and the impotence of experience without the guidance of theory. Combining Weber and Kant, the message to comparativists is clear: The-choice of a preconception or framework for ordering the chaos inherent in reality and hence for guiding empirical study is the fundamental analytical question.;. The rationalist—culturalist—structuralist dialogue indeed shapes inquiry. The question to be addressed now is how it does so. SECTION 2: THE THREE RESEARCH SCHOOLS Bates, Scott, and Skocpol can be best understood as exemplars of ideal-type re- . search schools in the social sciences. Each tradition shares an ontology, a methodology, and a philosophy of science. Each also faces characteristic lacunae which account for its historical development into subschoois. Similarities and differ- r_ ences among the approaches are summarized in Table 1. This table can be used in two ways: working down the properties and compare , ing rationalist, culturalist, and structuralist thought; or working across the ll"-Cj1tJi communities and comparing properties. Comparative case studies of the rationalist, culturalist, and structuralist research programs appear in the chapters in this voh'me^ by, respectively, Margaret Levi, Mate Howatd Ross, and Ira Katznelson. Rat het then ^ repeating their configurative discussions here, I will adopt the second approach-.^ With apologies to Ira Katznelson (this volume), I will "slice and dice" my variable^ SOCIAL THEORY AND COMPARATIVE POLITICS Table 1. Research Communities and Their Properties 245 Com munity Property Rationalist Culturalist Structuralist Ontology Methodology Comparison Lacunae Subtraditions Rational actors Intentional explanation Actions, beliefs, desires Methodological individualism Comparative statics Irrational social consequences of individually rational action Unintended, unwanted unavoidable, unexpected outcomes Positivism Generalization Explanation Instrumental rationality Mechanical-behavioral view of subjectivity Human nature rationalists Social situation rationalists Rules among actors Intersubjectivity Common knowledge Common values Relations among Actors Holism Meaning and significance Culture as cause/constitutes reality, identity, .action, order Interpretivism Case study Understanding Tautology, teleology in existence and causal impact on outcomes Subjectivists Intersubjectivists Exemplars Robert H. Bates James C. Scott Social types with causal powers Structures with laws of dynamics Realism Comparative history Causality Iron cage determinism Voluntarism absent State/society Pluralism-Marxism- statism Theda Skocpol ONTOLOGY „:Each school is founded on certain presuppositions about the way the world is constructed. Each perspective, that is, assumes something about the nature of existence: the entities and their properties that populate our lives. Rationalists like Bates axe methodological individualists who argue that collectivities have no status apart from the individuals who comprise them: Only actors choose, prefer, believe, learn, and so on. All explanations of groups, rationalists argue, must therefore be understandable in terms of individuals. People, in tUrn mncr hp nnrlprcrnnfl ti/il-h inrpnrinnul PVTilnnnrinn": nf rational choice: De- 246 CONCLUSION sires and beliefs direct action. In other words, if actions are taken for certain reasons, then the reasons motivate the actions. The concept of "interest" follows: If action A is in person P's interest, then P must be able to supply a reason for doing / A. Rationalists are therefore concerned with the collective processes and out-comes that follow from intentionality, or the social consequences of individually rational action. Often these consequences are quite irrational: They are unin- i tended, unwanted, unavoidable, and unexpected, albeit inevitable. For example, Bates's (19S9: 1) first lines are: "This book is about the political economy of development. It is about the politics and economics of agriculture. And it is about Kenya." Bates then indicates that he will use the Coase Theorem, a mictueco-nomic perspective that focuses on the transaction costs of individual exchange, to explore the efficiency of institutions, governments, and politics. In sum, ratio-nalist ontology depicts a world populated by rational individuals and possibly ir-rational collectivities. The rational pursuit of individual interest explains the all- S too-common occurrence of irrational social outcomes. Cuituraiists are methodological hoiists who think of norms as intersubjec-tive or transindividual: The members of a group or community have common, mutual, or shared ideas, orientations, or ways of looking at the world. These val- , ues are found in all of society's institutions - political, religious, economic, and social — and in society as a whole. Intersubjective consciousness is composed, mote specifically, of two elements: cognitions and conscience/Culture involves^ common knowledge — is's and not shouid's — about the construction of reality. • Culture also involves common understandings about the way the world should be. Common cognitions and conscience are constitutive of community. Hence; Scotr (1985: 234) refers to "the moral logic of tradition" in which custom, rim-..!, al, and norms define a community's meaningful roles or expectations. Culture and community — common cognition and common conscience — are . in many ways the bases of social order. First, they are needed for the practical management of daily social life. Collective action and social coordination require mutuality of information and values. Second, culture and community underlie the affective and emotional symbolism of daily life. The world is constituted by-. , social interactions and communicative acts endowed with meaning and signifi- -cance. Third, culture and community are the bases of social control. Roles dictate standards of social respect, recognition, "reputation, status, and prestige" (Scott 1985: 234). These, in turn, provide social sanctions that resttain self-seeking in-; dividualism, "dog-eat-dog" competition, and "beggar-thy-neighbor" strategies of % survival. As Scott (17) puts it, "For it is shame, the concern for the good opiniojJj, of one's neighbors and friends, which circumscribes behavior within the normal | boundaries created by shared values." Fourth, culture and community provideJ standards of individual and collective obligation "that lie beyond immediate re.-, rations of production and serve both to create and to signify rhe existence of com'-munity — one that is more than just an aggregation of producers" (169) Hence,. there is a "collective and public recognition that the village has an obligation topffia ......------1 a^-nprtatibns^ SOCIAL THEORY AND COMPARATIVE POLITICS 247 and preferences make the village, in effect, "one family" (196). Finally, culture and community underlie personal and group identities. The self is really a "communal self developed in interactions with others. Culture is thetefore both outside and inside individuals: external, in that it is materially real and transmitted from the past; internal, in that individuals are socialized into it. While culture constitutes social order, Scott argues that it is contested. In contradistinction to the Parsonians, he (xviii) offers a "'meaning-centered account of class relations" in Sedaka in which class consciousness is constitutive of class relations and class conflict. Hence, there is a "public symbolic order" (25) that is based on a "symbolic balance of power" (22). Class conflict thus turns out to be "a struggle over the appropriation of symbols, a struggle over how the past and present shall be understood and labeled, a struggle to identify causes and assess blame, a contentious effort to give partisan meaning to local history" (xvii). For example, the breaking of accepted social conventions and behavioral norms leads to the symbols and exemplars of "the greedy rich" and "the grasping poor" (18). Culture is thetefore constitutive of both consensus and conflict. On the one hand, class struggle "requires a shared worldview ... [it only makes sense] unless there are shared standards of what is deviant, unworthy, impolite" (xvii). On the other hand, class struggle is contingent on shared values that are betrayed: "What is in dispute is not values but the facts to which those values might apply: who is rich, who is poor, how rich, how poor, is so-and-so stingy, does so-and-so shirk work?" (xvii). In sum, culturalist ontology assumes that culturally embedded individuals follow social rules that are constitutive of their individual and group identities. : In conttast to the rationalists, interests are not merely given and/or random; reason is not necessary and universal but conditional and contingent; and the categories of rarional rhought and the nature of rationality vary by culture. Structuralists are also methodological hoiists. They study networks, linkages, interdependencies, and interactions among the parts of some system. A strucrural ..argument is therefore always concerned with the relationships — both static and .dynamic — among individuals, collectivities, institutions, or organizations.6 One vcan understand a thing, structuralists argue, only if it is related to other things of which it is a part. Hence, entities are defined in terms of relationships with other entities and not in terms of their own intrinsic ptoperries. Waltz (1979: 81) thus ■argues that "in defining structures the first question to answer is this: What is the principle by which the parts are arranged?" He (74) offers the example of George 'and Martha in Albee's Who's Afraid of Virginia WoolfP George and Martha are a pair ;.;6f individuals whose fortunes cannot be separated into individual-level components: Theit fate is rooted in their (marital) relationship. Structuralists thus focus on the political, social, and economic connections jiamong people. Historically rooted and materially based processes of distribution, 6Higher-!evel structures can, of rourse, be cnmposed of tower-level structures. For example, the state *]________ 248 CONCLUSION conflict:, power, and domination, thought co drive social order and social change, are their particular concern. Skocpol (1979), for example, argues that state break- ■ down, peasant revolutions, and state reconstruction have structural causes. Her structure—conflict—change approach to state and revolution emphasizes five structures. First, the international context: Skocpol focuses on international structures and relations, or war and trade, and the world-historical circumstances in which states find themselves. Second, intranational class conflicts: Skocpol is :; concerned with "historically specific institutional arrangements" (116, emphasis in original), such as "agrarian class and local political structures" (117), that affect intraclass and interclass relations.7 Third, the nature of the revolutionary crisis: : The processes of stare breakdown and of peasant revolts become the legacies of the Old Regime that affect state reconstruction. Fourth, the nature of states: i Skocpol explores states as "administrative and coercive organizations" (14) that , penetrate society and control people and territory. Finally, the relations of states ;: and classes: Skocpol is particularly interested in "the potential autonomy of the state" from dominant and subordinate classes in society (24).B In contrast to rationalists and subjective culturalists like survey researchers, 4 structuralists reject an agential and reduccionist focus on the actors themselves. ■ Skocpol is very forceful on this point. She (14) argues that "an adequate under- v standing of social revolutions requires that the analyst take a nonvoluntarist, . structural perspective on their causes and processes." Hence, "any valid explana- .: tion of revolution depends upon the analyst's 'rising above' the viewpoints of participants" (18) and taking "an impersonal and nonsubjective viewpoint" (IS). Skocpol (29) thus prefers to "emphasize objective relationships and conflicts among variously situated groups and nations, rather than the interests, outlooks;'; or ideologies of particular actors in revolutions." In other words, she focuses on i the "structural contradictions and conjunctural occurrences beyond the deliber- j; ate control of avowed revolutionaries" (291) Skocpol thus forcefully rejects a voluntarist approach to revolution based on ■ mobilizable groups and "the emergence of a deliberate effort" (15).9She argues" 7Intracloss relations include whnt she calls "peasant solidarity," or the peasantry's internal organization and resources. Hence, Skocpol (115) explores "the degrees and kinds of solidarity of peasant communities." Interclass relations include what she calls "peasant autonomy from direct day-to-day supervision v and control by landlords and their agents" (IL5). Skocpol hypothesizes that render agrarian class relations, those wich absentee nobles and a landed peasantry, beget numerous peasanc conflicts, while large estates : managed by nobles and worked by serfs or landless labor resist peasant rebellions. Pare of her explanation , also involves the political structures of local government and its relation to peasantry and national gov'-'; ernment: "Those vulnerable agrarian orders also had sanctioning machineries rhat were centrally and bu-. reaucratically controlled" (117). ......'. BShe is thus concerned with the relations between the state and the dominant (rural) classes. In sttt" rise societies" she suggests that the stare was autonomous from the nobility and hence better able to pusb through needed reforms {e.g., prerevolutionnry Russia modernized more than prerevolurionary China)., j Skocpol is also concerned with the relationship between the srate and the dominated (rural) classes Pea*.--* ant rebellions are a function, first, of whether the state penetrated the peasantry and consequently did not", rely on local elites for social control and, second, "the relaxation of state coetdve sanctions againsc peasant SOCIAL THEORY AND COMPARATIVE POLITICS 249 that analysts should not assume self-conscious and purposive revolutionary vanguards or movements whose members share grievances and goals. Hence, Skocpol minimizes the ability of revolutionary leaderships backed by revolutionary ideologies (e.g., Jacobinism, Marxism-Leninism) to transform the state. She prefers to focus on the structural conditions under which elites struggle to consolidate and use srate power, or the "specific possibilities and impossibilities within which revolutionaries musr operate as they try to consolidate the new regime" ■ (171). Revolutionaries thus do things they never intended, and preferences, goals, and ideologies are not a valid guide to outcomes.10 In sum, structuralist ontology explores how relations among social agents are concretely structured. Skocpol thus focuses on "objective structural conditions" rather than on "politically manipulable subjective conditions" (16). Rationalists therefore study how actors employ reason to satisfy their inter-- ests, culturalists study rules that constitute individual and group identities, and structuralists explore relations among actots in an institutional context. Reasons, rules, and relations are the various starting points of inquiry. I METHODOLOGY t Schools also have explanatory strategies. Each possesses a "positive heuristic" for §> ■ argumentation (Lakatos 1970). '~f Rationalists engage in vicarious problem solving. As Schelling (1978: IS) puts it, "If we know what problem a person is trying to solve, and if we think he can actually solve it, and if we can solve it too, we can anticipate what our subject will :;■ do by putting ourselves in his place and solving his problem as we think he sees ■it." Once they place themselves in aproblem situation, rationalists, as Bates (1989: 9-10) indicates, perform s.gedanken or thought experiment that involves two time '■■■periods. In the first period, their model is in equilibrium. The model is then perturbed by a series of exogenous shocks. In the second period, rationalists observe :the impact of the exogenous changes on the endogenous variables of concern. What kind of exogenous shocks are possible in these comparative static expertises? Since rationalists are intentionalists, variations in action can only be explained by variations in desires and beliefs. Rationalists, moreover, gravitate to- ■3: 9She maintains, more specifically, that voluntaristic theories of revolution go wrong in four ways. :. First, they get the process wrong: Revolutionary intentions develop in the course of revolution. Second, cthey get the counterfacruals wrong: Such theories imply a voluntaristic conception of political otdcr nnd ^stability. Third, they get the causes wrong: Revolutionary crises simply occur and are historically nonvol-Puntatistic, rather than being "made" by revolutionary movements. Finally, they get the outcomes wrong: ■States are not constructed by rhe revolutionary agency of vanguard parties. I '"Similar ro the rationalists, Skocpol looks at revolutions as the unintended consequences of che interaction of rational actors or sets of actors: "Revolutionary conflicts have invariably given rise to outcomes ^either fully foreseen nor intended by - nor perfectly serving the interests of- any of the particular groups Involved" (1979: IS). The term she (298, fil. 44) uses here is "conjuncture": "the coming together of separately determined and not consciously caatdinated (or deliberately revolutionary) processes and group ef-|§rts." Rather than focusing on explicitly formulated inrergroup coalitions, she thus prefers to fociis on 250 CONCLUSION fr SOCIAL THEORY AND COMPARATIVE POLITICS 251 ward materialise theories of pteference and cognition. The material constraints of the "objective external worid" are held to affecc action because they influence the .; desires and beliefs of the "subjective internal world" of the actor. In other words; •■. rationalists are positivists who restrict their comparative static exercises to hard ::: or "objective" shocks because they wish to avoid studying fuzzy or "subjective" ■ ones (Lichbach 1996: 233). Rationalists consequently explore the conditions of choice: the shadow, relative, or opportunity prices (in terms of forgone material i:| opportunities) of action. For example, Bates (1989: 10) indicates that "in the early portions of this work, the shock is the colonial incursion. In the later par^ cions, it is a failure of the rains. In the intermediate periods, the shocks include variations in access to land, cash crops, or productive ecological zones." While 4f Bates certainly understands the culture of Kenya, and he certainly factors it into ?I his equilibrium model, what ultimately drive his analysis ate independent and exogenous material shocks and forces. As Bates (153) reluctantly and revealingly puts it, he "has been driven co a materialise conception of politics." In sum, em- •.' pirically oriented rationalists are ultimately materialists in that they assume that material conditions drive subjective consciousness and ultimately rational choice. Rational actor theories are consequently parasitic on material structuralist ones. ■ Human actions are intentional: People express and act upon purposes. Cul-turalists like Scott (1985: 45) thus argue that the human sciences should concern;;| themselves with the emotions, attitudes, and other subjective dispositions chat " allow researchers to evaluate the meaning and significance of human interaction. Ic is only by penetrating the frames of meaning used by actors that analysts can explore how culture causes and constitutes reality, identity, action, and social -order. The methodology of interptetacion, hermeneutics, or verstehen makes four J fundamental assumptions. First, interpretive approaches are premised on the idea. "% that participants' understandings might not be the same as scientists' under- -standings.11 Interpretivists thus attempt to see things from the actor's point of;:.;,| view or in terms of his or her own self-understanding. Their goal is therefore to produce an empathetic awareness of the outlooks, feelings, motives, and expen-- ^ ences of another. Second, since the meaning of an action is comprehended in light , of the agent's particular situation, the norms, forms, and practices of his or her , | society are relevant. Interpretation thus involves value relevance: Meaning is eel-ative to culture. Third, interpretation involves a hermeneucic circle: The parts must be understood in terms of the whole and the whole must be understood-in j* terms of its parts. Meaning, in other words, must be established holistically oib}' • relating individual and society. Finally, comprehending the material world is not ■ the same as comprehending the social world. The social world must be under- ^ 5 stood from within rather than explained from without. In fact, the analyst should, limit himself or herself to comprehending the self-understanding of human be-; ings. He or she must go beyond establishing the materialistic causal connections^^ "Wefae'r feited in Calhoun 1995: 48) offers a classic rejoinder to this perspecrivisrn: "One does '"^^ It sought by rationalist gtdanken: Instead of seeking the external causes of behavior, analysts should seek the internal meaning of action. Understanding rather than explanation is therefore the goal: Positivists study cause-and-effect explanations, rooted in the nomothetic idea that recurrent law-like processes exist, and inter-ptetivists seek interpretive understanding tooted in the ideographic idea that societies are unique. In sum, culmralists reject materialistically oriented positivism and adopt an interpretive philosophy of science. They study how reason and non-rationality are constitutive of individuals and societies. Structuralists study structures and hence adopt a realist philosophy of science. Realism is characterized by two basic principles that are very compatible with structuralism. First, realists adopt an entity- rather than an event-centered ontology: "Entities (ontology) condition theories (epistemology)" (Wendc forthcoming: 20). Realists thus assume that objects and entities — perhaps known only by their effects — exist in the world. For example, the state is real and is not simply a police car; similarly, the international state system is real and is not just the U.N. charter. Structures are thus real entities or objects. Realists assume that mature scientific theories typically refer directly to this real world of (perhaps unobserv-■ able) objects and hence provide knowledge of reality "out there." Scientific theories are therefore about the basic building blocks of the world, including their properties and interactions. Scientists search for these fundamental entities, called "natural kinds" in theit particular domains of inquiry. Natural kinds have a difFerentiated structure and hence coherence and unity: They are forms or kinds of things with interconnected parts or elements. Hence, a natural kind is more than a heap of properties or an ad hoc collection of bundles of qualities. For example, a dog is a natural kind; a pile of sand or five randomly chosen objects on my desk are not. The implications for social science are that social structures are real and that social scientists should search for "social kinds." Little (1993'- 190) argues that "candidates for social kinds include 'riot,' 'revolution,' 'class,' 'religion,' 'share ■cropping land-tenure system,' 'constitutional monarchy,' market economy,' 'nationalist political movement,' 'international trading regime,' and 'labor union.'" •Skocpol thus focuses on historically concrete types of cases: "This book does not, of course, analyze in depth all available historical cases of social revolution. Nor :does it analyze a 'random' sample from the entire universe of possible cases. In :ftct, comparative historical analysis works best when applied to a set of a few cases that share certain basic features. Cases need to be carefully selected and the criteria of grouping them together made explicitly" (1979: 40). Hence, France, Russia, and China "have been grouped together as fundamentally similar cases of social revolution" (40, emphasis in original). Why? "It is the premise of this work that France, Russia, and China exhibited important similarities in their .Old Regimes and revolutionary processes and outcomes — similarities more than sufficient to warrant their treatment together as one pattern calling for a coher-_enr ......... 252 CONCLUSION Second, these perhaps unobservable natural or social lands — what I have termed structures - have, following Skocpol, causal powers. Wendt (forthcoming: 26) sug- ■ gescs that "the behavior of things is influenced by self-organizing or homeostacic internal structures, and the analysis of those structures should to that extent figure in explanations of behavior." Realists argue that scientists search for these real — albeir hidden — causal mechanisms. For example, chemistry looks for chemical elements and the laws of chemistry; physics looks for elementary particles and the laws of physics. Hence, structure, process, and outcome are linked: Natural bodies or lands have natural proclivities or powers which produce natural laws Df development. Pi-aget (cited in Lloyd 1986:257, emphasis in original) thus maintains that "there is no structure apart from constntction." Structuralists thus follow Aristotle and assume chat structures have actualities and potentialities and that form determines development. These law-like processes thus involve production and reproduction, stability and . change, growth and development, and maintenance and transformation. The implication for social science, as Ira Katznelson develops in his contribution to this volume, is that social structures must be analyzed in macro-historical and macrodevelopmental perspective. The substantive concerns in- ' elude state-building, war, capitalism, industrialization, and urbanization. Structural mechanisms that produce these historical dynamics include competition, conflict, consensus, division of labor, differentiation, diversity, distrib- v ution, inequality, stratification, polarity, size, density, and hierarchy. Still other structural logics that produce dynamics involve contradictions, paradox- : es, ironies, and unintended consequences. Examples abound. Adam Smith argues that in the pursuit of private gain society organizes itself and thereby pro- : duces a market governed by the laws of supply and demand. Karl Marx argues..; that capitalist societies have a different set of laws of development and ones that lead to their own demise. Max Weber argues that patriarchalism, domi-nation by notables, political pacrimonialism, feudalism, hierarchy, Caesarnpa-pism, bureaucracy, charismatic community, church, sect, household, neighborhood, kin group, ethnic group, oikos, and enterprise have characteristic patterns of development. Skocpol is thus concerned with the concrete historical dynamics of a certain type of state (1979: 304, fn. 1). What types of old-regime states that she studied ; were susceptible to social revolution? "Autocratic," "ptotobureaucratic," "imperial," "monarchies" that were "well-established," "wealthy," "politically ambi- ; tious," "historically autonomous," and in "noncolonial" states with "statist soa- , eties" and "agrarian" economies that faced "intense international military : competition" from "economically developed military competitors" underwent social revolution (1979:41; 161; 167; 285; 287-8; 304, fn. 4). What type of state was the outcome of the social revolutions she studied? Bureaucratic and mass-m-corporating states — rationalized, autonomous, and powerful — were the products of revolution. In sum, structuralists analyze real social types with causal powers and hence ■ : nf sfmrtniM. This imolies that thev are opposed^! SOCIAL THEORY AND COMPARATIVE POLITICS 253 .-.».»-. .-I -.T «-Vi Vitt lunamirc the rationalist's atomistic reductionism. Structuralists indeed reject the view that social life can be explained by particles of matter and their movement, which are subject only to the laws of motion and their own material narure. Skocpol thus opposes "strategies of analytic simplification" (294). She argues that analysts (5) should not concentrate "only upon one analytic feature (such as violence or political conflict" that characterizes major social revolutions. Rather, "we must look at the revolutions as wholes, in much of their complexity." Structuralists, moreover, view cause as natural necessity. This implies that they violate the rationalist's Hempelian deductive-nomological approach to explanation (see Alan Zuck-erman's chapter in this volume). Structuralists indeed transgress a positivism that sees cause as entailing only logical necessity. Rationalists therefore perform comparative static experiments, culcuralists produce interpretive understanding, and sttucruralists study the historical dynamics of real social types. Positivism, interpretivism, and realism are the possible philosophies of social science. COMPARISON Given a school's ontology and methodology, each develops an approach to comparison. All take a stand on the Jdeographic-nomothetic debate and on the question of covering laws and causal accounts raised in Alan Zuckerman's essay in this volume. Rationalise methodology involves the comparative-static experiments discussed earlier that link structure to action. This comparative static methodology sounds like the basis for generalization. It is indeed the ideal PGM — proposition generating machine. Rationalists see individuals as hardheaded scientists who ground their preferences and beliefs in the material world. Similarly, rationalists see themselves as hardheaded scientists who conduct gedanken and then evaluate their success. Rationalists are thus careful to specify what counts as decisive evidence against their experiments. They think in terms of observable implications that are falsifiable (i.e., rationalists suggest null hypotheses and counterfactuals). For example, Bates (1989: ch. 4) offers a series of regression equations that demonstrate chat political institutions and public policies affect Kenya's food stocks and thus stand between drought and famine. Refuting theories of unregulated markets offered by neoconservative development economists and of benevolently regulated markets offered by neoliberal development economists, he presents statistical evidence of a "policy-induced food cycle" (111). Institutions, that is, "may generate pressures that convert abundance into dearth and therefore translate droughts into food crises." Bates considers these lessons generalizable to agrarian politics in Dther Third World countries. As he puts it, "this chapter has caught us about subsistence crises" (115). In sum, rationalists are led, as if by an invisible hand, to quantitative methodologies and a positivist philosophy of science. They actempc to account for an explanandum (irrational social action) by fitting it into a structure of knowledge: Initial conditions (about rational desires 25'i CONCLUSION and beliefs) and general laws (abouc cheir operation) allow rationalists to deduce the anomalous (irrational) phenomenon in question.12 Culruralist methodology involves gaining interpretive understanding of meaning. Since meaning is peculiar to particular cultures, culturalists favor case studies. Moreover, they stress the uniqueness of cases. Individual cases are characterized by radical historical contingency. Individual developments are largely open-ended historical accidencs in four ways. First, outcomes are paradoxical: They are unintended and unwanted. Second, outcomes are path-dependent: Critical events and tipping points shape history. Locally, cases are temporally ordered and historically connected sequences of events. Globally, cases are subject to spatial and temporal diffusion as each case changes the cantexr within which subsequent cases operate. Third, outcomes are multifarious: Multiple equilibria are possible and counterfactuals are always relevant. Finally, outcomes are unstable and unpredictable: The forces chat produce any one outcome are finely balanced, and hence short-run. Culturalists are therefore suspicious of generalizing across cases. They reject the idea of nomothetic research conducted on random samples of the world's cut-rent population of srates in order to develop generalizations that hold independently of space, time, and context. All universals, uniformities, and invariants ate suspect. Norms, means, and averages are just that. Theories are never fulfilled 'j precisely the same way in all situations. Cases are not merely instances of general -things — lifeless variables, categories, and abstractions. Modifications, exceptions, 1 and qualifications are the rule. Hence, comparativists should not think in terms r > of ideal-type theories and concepts. These do not exist, have never existed, and will never exist. In short, all grand historical narratives and totalizing universal histories must be deconstructed. Rather than compare to establish vague similarities, culturalists believe thar.,vr:.&| comparativists should compare to establish sharp differences. Comparativists......- should thus be historical relativists. They should positively value diversity and . : multiplicity; expect historical particularity, specificity, and locality; understand...... individuality, singularity, uniqueness, and distinctiveness; appreciate deviants, ^ outliers, idiosyncrasies, unrepresentativeness, and anomalies; and hence study^ ^ variation, heterogeneity, fragmentation, differentiation, and plurality. _ 'i For example, Scott (1985: xviii) states that "a certain amount of storytelling .^J seems absolutely essential to convey the texture and conduct of class relations. "}^."^ Larger theoretical "considerations require, I think the flesh and blood of detailed. ^VJj instances to take on substance. An example is not only the most successful way ^ ^J, of embodying a genetalizacion but also has the advantage of always being richer -^J-jl and more complex that the principles that are drawn from it" (xviii). Hpnce^^|?f|| Scott opens his book with two wonderful stories of social outcasts: Razak, the,-....^ '^Margaret Levi, in her contribution to this volume, indicates thoc some comparatively nn^ ^'^'^JS cully oriented rationalists are moving away from positivisc comparative-static exercises and toward .fJJj^jEgg SOCIAL THEORY AND COMPARATIVE POLITICS 255 symbol of "the grasping poor," and "Haji 'Broom,'" the symbol of "the greedy rich" (18). In sum, culturalists stress configurative paths — there are as many paths as there are cultures. Hence, comparativists should compare in order to establish the differences among a set of developments.13 They should study phenomena in their local and concrete historical context, focusing on their origins and outcomes. Now consider the structuralists. Structures are patterned objects. There are obviously systematic similarities and differences among these patterns. Struc-tutalists thus divide objects into species and genera. Their theoretical generalizations and statements are confined to particular classes and categories of phenomena. Structuralists can therefore be located between the universalists (rationalists) and the particularists (culturalists): Between all and each lie some. Structuralists thus achieve generality by partitioning cases into subsets and establishing classificatory frameworks. While structures come in types and structuralists are basically classifiers, the way things are grouped by kind is very important co structuralists. As realists, they argue that scientists must take note of the real and objecrive divisions in the world: Analysts should divide nature at its joints. They reject the idea of nominalist, artificial, or subjective classifications that are merely imposed by observers and arbitrarily given by language. Divisions, in other words, are discovered and not invented. This is why Skocpol (1979), for example, focuses on the historically concrete forms of the state mentioned earlier. Struccuralisrs thus classify cases inco a number of categories, each fundamentally different from one another. They then investigate the historical dynamics associated with each class. Similar processes, sequences, and laws thus occur in similar structures; different processes, sequences, and laws occur in different structures. A small number of typical paths of development and change ate -thereby located. Structuralists therefore do comparative histories to discover the historical laws of structural development. Scate breakdown and peasant revolutions occur according to Skocpol (1979), for example, differently in agrarian bureaucracies than in postcolonial regimes. Because of differences in initial conditions, institutions, structures, groups, and contexts, similar causes or shocks (e.g., wars) produce dissimilar effeccs in different systems.1'1 On the other hand, different contexts within a similar overall type produce similar outcomes.15 This typological approach limits the generalizability of one's findings to the £type of cases examined. Hence, Skocpol (288) asks, "Can [these findings] be ap-,.plied beyond the Ftench, Russian, and Chinese cases? In a sense, the answer is unequivocally 'no': one cannot mechanically extend the specific causal arguments _'rthat have been developed for France, Russia, and China into a 'general theory of ~ revolutions' applicable to all other modern social revolutions. There are two im- ; 3This was one of Weber's principal methodological themes (Lichbnch 1995; 290-1). '^Examples include Brenner (1976, 1982) and Karzensrein {197B). : "Examples include Moore (1966) and Skocpol (1979). 25Ö CONCLUSION portant reasons why such a strategy would be fruitless." First, new cases might have new causes: "The causes of revolutions (whether of individual cases, or sets of similar cases) necessarily vary according to the historical and international circumstances of the countries involved" (288). Second, new cases might interact with old causes: "Patterns of revolutionary causation and outcomes ate affected by world-historical changes in the fundamental structures and bases of state power as such" (288). Skocpol (290) concludes that "other revolutions require analyses in their own right" because they occur in different types of structures (e.g., in different states and in different world-historical circumstances). In sum, structuralist comparison involves three steps. The first step involves classification: Structuralists locate different configurations of bounded and patterned action and interactions, The second step involves morphology: The principles that structure the relationships among the parts, or the theme, logic, or rules that establish the functioning of a configuration or form, are specified. The : final step involves dynamics: A structure's development, institutionalization, and change are studied. This involves a focus on origins, or how the structure comes into being; maintenance, or how the structure comes ro be stable; and transformation, or how the structure changes. Rationalists therefore generalize, culturalists particularize, and structuralists typologize. Comparativists can compare to establish similarities, differences, or both similarities and differences. LACUNAE A school's particular ontology, methodology, and approach to the ideographic-." nomothetic problem produce characteristic strengths and weaknesses. A compar- ; ative analysis of the approaches illuminates these virtues and vices. ■ The rationalist perspective is "externalist," "behaviorist," or "throughput"'; Given that rational actors attempt to efficiently adapt to their environment,; external conditions and not human consciousness are the focus of the theory..! Rationalists thus tend toward a mechanical-behavioral view of subjectivity and, adopt a particularly anemic or thin version of Intentionaliry, rationality, and interests. Actors ate thus left with an impoverished orientation to action: People-, are computing devices and mechanical robots who calculate their interests Ra-; tionalists who explain action in terms of exogenously changing prices thus m-; evitably slight the individual and group identity-formation question: Personal-and communal identities are treated as exogenous to rather than constitutive of- ^ stable and orderly social relationships and interactions. For example, Bit-tS j (1989-* 150) suggests that people are concerned wich efficiency and Pareto-op-timality because it can help everyone including themselves: "In an almost j Marxian manner, the theory concends that people devise institutions so as ia unleash the full productive potential of their economies." While this materi , and social order, dtf iTs»* ism might seem to be the basis of mutual cooperation nrnhlem is that neoDle are even more concerned with distribution, power,. SOCIAL THEORY AND COMPARATIVE POLITICS 257 property rights because these can help them most of all: "People see clearly where their interests lie. They invest in the creation of institutions in order to structure economic and political life so as bectet to defend their position within them. They invest in institutions so as to vest their interests" (151). Hence, rationalists like Bates ultimately offer a materialist theory of preferences16 under which interest is an obstacle rather than a basis for social order: Rationalists view ends as random in a positive sense and as equal in a normative sense, which means that values ultimately divide rather than unite people. In sum, rationalists sacrifice the subject and surrender the self, undoing the community and unmaking the collectivity17 Scott (1985: 27) asks, "Why ate we here, in a village of no particular significance, examining the struggle of a handful of history's losers?" Evidently, "the big battalions of the state, of capitalist relations in agriculture and of demography itself," which beget the metanarratives of large-scale peasant rebellion and revolution, "are arrayed against them." His answer is that while the material dimensions of class conflict and social change are undeniable, conflict and change must ultimately be understood interpretively. Other culturalists go even further than Scott, adopt an all-embracing Hegelianism, and argue that it is "ideas all the way down." Whether they are moderates or extremists, culturalists face the problem that the existence and causal impact of culture is difficult if not Impossible to investigate. There are major problems with testing arguments about the existence of norms because norms vary by people, contexr, time, integrarion, intensity, and completeness. For example, Samuel H. Barnes in this volume shows that partisan allegiance has temporal instabilities and comparative nonequiva-lences. Moreover, norms are not directly observable and are subject to the "owl of minerva" problem (i.e., they are easiest to discover when they are in decline). This leads to the second major problem faced by the culturalists: testing arguments about the consequences of norms. Do norms actually produce action and outcomes? When action and the material world are swept up into an all-embracing Hegelian idealism, teleology and tautology are inevitable. Hence, culturalists face the problem of eliminating plausible rival hypotheses. Their ideas are significant but nonfalsifiable. For example, while Scott (1985; 139) does suggests several "standards of evidence and inference" on which one interpretation is to be preferred to another, he does nor pretend to offer a research design capable of separating idealist from materialist forces. In sum, culturalists do not attempt to 'separate the material from the ideal because they assume that material must airways be interpreted in terms of the ideal.18 As indicated earlier, Skocpol (1979) minimizes the voluntarism of revolutionary masses and elites and slights the significance of their values, beliefs, and lfiA related criticism is that they also have a materialise theory of beliefs. 17A related criticism thus challenges the rationalist's methodological individualism. 1BOther critiques of cultural analyses of politics ore developed hy Marc Howard Rrss in his enntri-:;butian [Q [his volume. 258 CONCLUSION actions. Her purely structural theory emphasizes that structures, not actions, produce outcomes. She argues a rigid methodological holistic posicion: Structure is significant and individual actions, desires, and beliefs are not. In other words, individuals have no choices. They are all bur eliminatable, overwhelmed by structure. People are merely "bearers," "carriers," or "supporters" of functions determined by objective structures.19 Moreover, when structuralists consider individuals, they tend to homogenize them. All people within a category are the same, merely role players who lack individuality. Culruraliscs thus charge that : strict structuralists study history without a subject. Human beings are made into mechanical robots and dupes who are forced to comply with the dictates of some system. Structural theories, in other words, lack people with agency: actors who have choices and take meaningful actions. Structuralists thus produce a bloodless social science: People ate the victims of and silent witnesses to history. This : bloodless social science means that structural theories miss politics: che strategic interaction among goal-seeking individuals. They also miss human activity, ere- '■:< ativity, and ingenuity. Rationalists thus charge that strict structuralists miss collective action and coalicional processes. This bloodless social science also means;; chac scruccural theories ace deterministic: Given structure, outcomes follow. Structural causes are so powerful chat everything becomes predictable: There are -imperatives and not possibilities, dictates and not contingencies. To structutal- :: ists, in sum, structure is fate. This perspective leads to historical fatalism, an iron ■ cage determinism, and che absence of voluncarism. Racionalisc thinking cherefore culminates in materialism, cuiruralisc thought in idealism, and scruccural tenets in determinism. Hard-core rationalists';; lose values and contexts, crue-behever culturalists miss choice and constraint, and die-hard structuralists miss action and orientation. Bates (1989), Scott (1985), and Skocpol (1979) are well aware of these lacunae. In order co advance their te-: search communities, they willingly make these trade-offs. SUBTRADITIONS Each tradition specializes: Rationalists concentrate on action, culruralists focus on norms, and struccuraliscs center on conditions. "Thin" versions of programs stick closely to their cradicional cores. Consequently, one can cesc che program in: a very fundamenral way. The problem, however, is chac che program is easily fal- ,9In practice, of course, the level of constraints vaties from situation to situation and may produce more or less limited choices. Inglehart (1990: 18) wisely suggests that ; on one hand, one can conceive of situations so totally rigidly structured that virtually nothing the individual can do affects his orhet fate. The situation of a ptisonet in a concentration camp may be very near this extteme. On the other hand, one con also conceive of situations in which what happens mainly refleccs the individual's behavior, a libertarian society with lavish and well-distributed resources might approach this ideal. In the teal wotld, one is almost never at , cither extreme; outcomes reflect both internal orientations and external constraints. SOCIAL THEORY AND COMPARATIVE POLITICS 259 sified: Exclusivicy slights a great deal of che complex empirical world and hence produces unsatisfacrory explanations of che richness of social life. "Thick" versions of research programs thetefore begin to look empirically attractive to the members of each research community. Pragmatic researchers willingly add on elements from the other approaches. Consequently, a single tradition can subsume many specific theories, and one can test the program but not in any basic way. The chree research communities thus contain an internal struggle between che purists and che monopolists, or between chose who wish to develop thin versions of the program and those who wish co develop thick ones. Rationalists study individual action and social outcomes. Thin rationalists are pure incenrionallsts who see reasons as causes of accion. They have a reductionist view of conditions and culture chac understands chem as individual beliefs and desires. For example, economises who do public choice (e.g., Becker 1976) focus on a supposedly universal human nacure and its laws: diminishing marginal utility, irrelevance of fixed costs, substitutes and complements in choice, market equilibrium of supply and demand, etc. Hence, thin rationalists might be more accuracely called "human-nature racionaliscs." One can extend che boundaries of the rationalist approach by deepening the micro, and hence studying culture, and exploring the macro, and hence examining institutions (Lichbach 1995: chap. 10). Thick rationalists like Bates chus move coward scruccure by looking at conditions as both causes and effects (although they do not go all the way to the structuralist position and explore how structure affects the constitution of actors themselves). Thick rationalists like Bates also move toward culture by looking at preferences (although they do not go all the way to study how actors are themselves conscituted by values) and beliefs (although they do not go all the way and become cognitive psychologists) as both causes and effects. Hence, thick rationalists might be more accurately called "social-situation rationalists." Bates thus begins with the historically specific opportunity structute in Kenya which defines the desires, beliefs, and choices of Kenyans. In addition to exploring how che concrete situation in Kenya constrains or limits, and enables or empowers, individual Kenyans, Bates also examines how Kenyans determine their historically concrete situation: The economic, social, and political institutions and outcomes ■. of Kenyan political economy are endogenous, Culturalists study subjective and intetsubjective values and beliefs. Thin culturalists include the survey researchers who maintain chac accocs make cul-' rurally informed choices. They also maintain that material structures must always be filtered chrough ideas - values and beliefs. Culruralists broaden their perimeter by analyzing how culrure defines choices and structures. Thick culturalists thus explore the decision rules behind choice and how actors are constituted by culture. Intetsubjective approaches that take a thick view of culture include Gramscian hegemonic culcuraliscs and Parsonian functionalists. Sub-)ective and intersubjective subtraditions is a very significant divide. Samuel H. Barnes in this volume refers to it as the "I/we problem: Culture is what we be-(,'iieve, not what I believe." Marc Howard Ross's essay in chis volume surveys the 260 CONCLUSION Structuralists study civil society, the state, and the international system of states. Structuralists include die pluralísts, Marxists, and statists. Thin structuralists are materialists. They argue that a base or substructure drives a periphery or superstructure, They also minimize the significance of actors and their freedom to choose. Since they see choice and culture as derivative of structures, thin structuralists often do not even bother to examine them. Structuralists thicken their approach by studying how the reason and nonrationality contained in structures are manifested in actions and orientations. Thick structuralists thus explore the materially driven dynamics of structures of collective action and so- . , 20 cial notms. In sum, puxists/ttadirionalists and monopolists/synthesizers pull their research programs in opposite directions. Purists keep the approaches close to their traditional roots; they therefore rninirnize within-tradition variance and maximize between-tradition variance. Their extensions are usually trivial and theit arguments most often turn out to be wrong. Monopolists move their approaches beyond their traditional core to synthesize perspectives; they therefore maximize within-tradition variance and minimize between-tradition variance. Their extensions are usually more interesting, but it is usually hard to know whether the program is producing the really useful insights. SECTION 3: THE SOCIALLY EMBEDDED UNIT ACT These basic similarities and differences in rationalist, culturalist, and structural-^ ist thought raise deeper interpretive questions: What is the meaning and signify icance of the three approaches? How can we understand and appreciate the dispute among the three research communities that characterize contemporary comparative politics? And why, after all, is today's battle of the paradigms taking place among rationalists, culturalists, and structuralists and, unlike the , 1960s, not among functionalists, systems theorists, and Marxists? Such questions are best approached by setting the dialogue among the schools within the historical context of the development of social theory. The,., origins of social thought provide clues to contemporary understandings and debates in comparative politics. More specifically, compatativists can begin to appreciate the similarities and differences and the connections and disjunc-tures among the research schools by exploring how the approaches can be, traced to Parsons's unit act and Weber's paradox of moderniry. 30Ira Katznelson's essay in this volume offers another way to parse structuralism: One tradition (e fi i Moore, Skocpol) develops grand macraanalytic narratives of world-historical importance while another "smaller-scale historical tnstirutionalism" (e.g., the one described by Peter A. Hall in his essay in this v°' ume) is more empirically and theoretically restrained and makes the relatively modest claim that histories and institutions matter. SOCIAL THEORY AND COMPARATIVE POLITICS 261 Parsons (1937) used the conceptual device of the "unit act" to systematize or rationalize the ideas of several of the founders of social thought.21 His purpose was to unite the rationalist, culturalist, and structuralist foci on interests, identities, and institutions into one framework. The "action frame of reference," part of his voluntaristic theory of action, was the first attempt to end the war of the schools and integrate the conflicting paradigms,22 Building on Weber's ([1924] 1968) idea of social action, one can say that both acts and contexts matter and hence that all acts are socially embedded. I have therefore extended the unit act to take account of the structure-action problem of reconciling individuals and collectivities,23 The socially embedded unit act is represented in Figure 1. The diagram has three layers - an inner or individual layer, a middle or collective layer, and an outer or approach layer - which reveal important connections among the schools. THE INNER OR INDIVIDUAL LAYER The socially embedded unit act involves a hypothetical person in a siruation in which the world is at least partially under his or her control. The actor thus has some agency: He or she manifests subjectivity, has purpose, possesses free will, uses reason, and acts. The presumed result is human creativity and personal responsibility. Philosophical discussions of intentional explanation and technical discussions of individual decision making therefore stress that agents possess three important characteristics (Elster 1989). They have desires - goals, purposes, and ends — that they intend to satisfy. They have beliefs - infotmation and knowledge - about their situation. Finally, they make choices - act, do, and perfotm - in order to reach their goals. In sum, at the individual level desires and beliefs direct action. *' For nearly two decades, Parsons's unit act ant! relared conceptual schemes dominated a great deal of so-. cml science. Many of the paradigms thai became popular during comparative politics's earlier flirtation with , theory and generalization in the 1950s and 1960s were rooted in Parsons (e.g., srrucrural functionalist!!}.. ~2Farsons eventually moved from a genera! theory of action — a view of social order as resulting from the contingency of individualistic decision making and the voluntaristic interaction of isolated individuals in same larger framework of norms and values — to a structural-functional scheme — a systems theory of sncial order based on functional Dr systemic imperatives. This chapter does not consider the entirety of , Parsons's thought, including the strucrure-funccian scheme and "general action complexes" chat synche-sue social, cultural, personality, and behavioral aspects of modern societies, For a review of Parsons's work, see Sculli and Gerstein (1985). An important recent contribution is Camic (1989). "'Parsons's (1937) presenrarinn of the unit act in face mixes the individual and collective levels. Ac-, rors have goals. The siruation in which they find themselves is pared into two parts — conditions and means.. Conditions are the material elements which cannot be molded to the acror's purposes; they are the obstacles -. that cotisttain agency abour which actors develop beliefs. Means are the choices or actions undertaken by the actor that enable agency. Finally, actors approach the situation with certain norms. They use their own subjective judgments or standards to interpret or understand their situation. The pursuit of goals and the ,i; choice of means is therefore judged by normative considerations, ideal standards, or value expcranohs.;V 262 CULTURALIST APPROACH CONCLUSION STRUCTURALIST RATIONALIST Figure 1. The Socially Embedded Unit Act; The Banc Diagram THE MIDDLE OR COLLECTIVE LAYER The socially embedded unit act also involves sets of individuals who comprise some collectivity. People, in other words, are part of some social order. The structure-action or individual-collective problem involves linkages between the three properties of agency and three corresponding properties of soci- . ety. Individual desires reflect and produce social norms. Individual beliefs correspond ro and ultimately influence material conditions. Finally, individual action aggregates into and also responds to collective action. In sum, at the collective level cultural norms and environmental conditions affect social action. THE OUTSIDE OR RESEARCH COMMUNITY LAYER All grand syntheses, like Parsons's voluncaristic theory of action, become the object of close scrutiny. The intellectual division of labor takes its toll. Specialists have therefore appropriated each of the components of the socially embedded unit act and spawned a research community. There are now experts in action (the ---■---K~«.r-Y rtnmnc (i-Yir* ft~t I mi rtt 1T cft^ anrl rnnA i ri nn s (rhf srmrrnralists). 'is, t- SOCIAL THEORY AND COMPARATIVE POLITICS 263 Hence, the outer layer of the diagram indicates that each of the schools concentrates on one vertex of the triangle. Culturalists specialize in individual desires and cultural norms, structuralists in individual beliefs and environmental conditions, and rationalists in individual choice and social action.24 THE CONNECTIONS AMONG THE LAYERS The connections among these three layers problematize several important themes in social theory and hence reveal several significant relarianships among the approaches. First, Figure 1 clarifies why the debate among Marxisrs, structural functionalists, systems theorists, etc., .in the 1960s evolved into a debate among rationalists, culturalists, and structuralists in the 1990s. Theoretical thinking has sharpened and the issues are now crisper. Hence, there is now a certain symmetry among the competing positions, which is occurring throughout the social sciences,25 that was missing from the eatliet "war of the schools." Each contemporary school coalesces around a subject matter: Choice, cuirure, and context are the domains of study. Each adopts an ontology: Reasons, rules, and relations constitute the world. Each explores a key explanarory variable: Interests, identities, and institutions drive outcomes. Finally, each lends itself to a theory of social order: The intersection of strategies, symbols, and structures define society. In sum, our problem situations and research designs in comparative politics consist of three natural models and foils. The value in juxtaposing the approaches is that critical confrontations reveal the junctures where a school's lacunae are best addressed by the other schools. Second, the socially embedded unit act clarifies rhe cenrral issue in social thought: the structure-action problem of uniting micro, meso, and macro levels of analysis. The difficulty here is that human beings are the continually active subjects who make the eternally passive objects which limit their subjectivity.-6 Individuals are therefore more or less inrentional agenrs who make history, society, conditions, and rules and yet history, society, conditions, and rules make individuals. We are both autonomous creators and dependent creatures, innovators and prisoners. The world is both fact and counterfactual, constraint and construct. Some examples will drive home the point: Talcing and selling prisoners becomes rhe institution of slavery. Offering one's services to a soldier in return for his protection becomes feudalism. Organizing the control of an enlarged labour force on rhe basis of standardized rules becomes bureaucracy. And slavery, feudalism and bureaucracy be- MThere are several individualistic, micro, or action approaches to inquiry besides rational choice. When I considered culruralist approaches, for example, I briefly discussed subjectivist approaches (e.g., modern survey research) chat offer a richer focus on the cognition, reasoning, motivation, and existential meaning the individual actor attaches to his or her acrion. There are also richer inrersubjcccive approaches to consciousness choc emphasize praxis or the enactment and performance of social aces. ' See fbocnoce 3. 254 CONCLUSION come the fixed, eternal settings in which struggles for prosperity or survival or freedom are then pursued. By substituting cash payments for labour services the lord and peasant jointly embark on the dismantling of the feudal order their great grandparents had constructed. (Abrams 1982: 2-3) The structure-action problem is concerned, more specifically, with interrelating the three aspects of the micro (individual), meso (group and institutional), and macro (societal) levels of analysis. The first is the aggregation problem: how unintended, unwanted, unexpected, unpredictable and yet seemingly inevitable collective outcomes result from a set of more or less purposeful individual actions. The second is the institutionalization problem: how these emergent properties solidify over time into structutes. The third is the contextual problem: how this solidified social order comes to constrain and enable individual consciousness and action. Hence, the structure-action problem has important normative27 and positive28 implications, especially about freedom and determinism, and hence the possibilities of rationality and nonrarionaliry. A glance.at the socially embedded unit act diagram shows how the framework goes beyond ParsDns's individualistic unit act and clarifies these issues. Looking vertically, one discovers that there are actually two structure-action problems: culture and rational action, structure and rational action. Looking horizontally, there is one structure-structure problem: culture and structure. Legitimacy and social order thus test on the harmonization of institutions and identities, Identities and interests, and interests and institutions. In addition, the aggregation problem exists for all components of the socially embedded unit act,. — action (individual action and collective action), values (individual preferences: and collective values), and beliefs (individual cognitions and institutional development). Rationalization thus occurs in the action sphere, where individual and , collective action is reconciled through organization; in the ideal sphere, where the abstraction and systemarization of values (substantive rationality) proceeds;;, and in the material sphere, where bureaucratization (functional rationality) develops. The parts of the inner or individual layer of the socially embedded unit act (desires, beliefs, and action), are thus associated with the middle layer of spheres of society (ideal culture, material structure, and group action), something 27The individual-society issue is important normatively. All societies must deal with value conflicts among individuals and harmonize social values. All societies musr then establish principles of the good life and reconcile them with the principles of the good society. At! societies must create an ethical totality in the (ace of possible fragmentation, polytheism, and a relativistic "everything goes" mentality. Societies must, in short, establish legitimacy. However, our western values lead us to wish to preserve agency, as did Weber. The individual-society issue therefore cues to the core of the liberal agenda. It is intimately associated with an analyst's philosophical or value orientations. 38The individual-society prohlematique is related to all the major issues in positive political theory Quescians of personality (being, autonomy, existence, and alienation), culture (legitimacy, trust, morality justice, and ethics), economics (marker, socialism, equity, efficiency, and welfare), sociery (civil society, contract, cotporatism, community), conflict and cooperation (peace, war), and politics (che state, deraQC mrv nnrl lih^mtisTrO i»mnnle with the srrucrure-acrion nrnblem. - SOCIAL THEORY AND COMPARATIVE POLITICS 265 that was not clear in Parsons s individualistic unit act. Hence, an individual's desires are a reflection of his or her ideal culture; an individual's beliefs are founded on the material structure in which he or she exists; and an individual's actions are a parr of the activities of some collectivity. Thete are also individual-collective connections across the parts of the socially embedded unit act. One may investigate, for example, how individual actions teflect collective values. Hence, one may explore connections at the collective level, at the individual level, and across the individual-collective divide, both within and between parts of the socially embedded unit act. Third, the socially embedded unit act clarifies the competing perspectives on two other grand issues of social thought; social order and social change. Each school generates a theory of social order: Social order is based either on reason, common values, or a hierarchical structure's imposition of material rewards and punishments (Lichbach 1996). Each school also generates a theory of social change. Rationalists explore how individuals react to die unintended and unwanted social consequences of rational action and consttuct new institutions — which are, of course, subject to dysfunctions and hence institutional change. Cul-turalists focus on how cultute institutionalizes structures and hence how cultural change tenders situations obsolete. Institurionalists explore the historical dynamics of structures. Finally, consider how the diagram relates to the nominalism vs. realism debate. I employ a Weberian nominalist rather than a structural realist orientation to the three research schools. Since Weber's "ideal types" address general methodological problems in comparative politics, and since the nature of ideal types sets the boundaries of this conceptual exercise, Weber's ideal-type analysis bears elaboration (Weber {1903-17] 1949). One cannot study all the theories and approaches that exist in contemporary comparative politics in all of their complexity and flux. The chaos of the theoretical world is as severe as the chaos of the empirical world. Hence, we need orienting models and guiding frameworks that define and frame questions and problems. These ate of necessity selective working heuristic tools rather than exhaustive and exact depictions of reality. They embody, in other words, a rather one-sided picture of the "reality" of the theories in comparative politics that exaggerates, accentuates, intensifies, highlights, and dramatizes certain features. The socially embedded unit act is therefore a typology or classification scheme designed to juxtapose the three "ideal types" (not all possible types) of social theory now found in our most prominent research communities. What are the central and significant features of the theories that deserve emphasis in these ideal types? One must take account of the goals, perspectives, and ideas of the theorists themselves. I also have interests and values — what Weber calls value relevance - in exploring those features that I consider significant for contemporary social theory. The ideal types of rationalist, culturalist, and structuralist thought that I have developed mix both sets of concerns. 266 CONCLUSION SOCIAL THEORY AND COMPARATIVE POLITICS 267 Each ideal type theory is thus logically coherent: ontology, methodology, comparison, lacunae, and subtraditions form a whole. This demonsttates that there is a certain internal logic to my interpretation of the value systems held by each set of theorists. The set of ideal types - the classification scheme - reveals certain contrasts that allow the exploration of the alternatives and conflicts inherent in the differentiation among the types. More concretely, the typology can be used to generate four fruitful types of comparison among types (general theoterical traditions) and cases (particular theories). First, one can compare among the types to show their similarities and differences. Weber thus explores types and subtypes, sets and subsets, of cases. I have drawn numerous comparisons in rationalist, culrur-alist, and structuralist thought. Second, one can compare among cases. Weber thus argues that one needs clear concepts before one can show the similarities and (especially) the differences among cases. I have drawn numerous comparisons among the theories developed by Bates, Scott, and Skocpol. Third, one can compare a type with a case within its range (to show its applicability). Weber thus uses an ideal type as a yardstick to define individual cases: It is a standard against which empirical cases can be measured. By comparing the real with the ideal, Weber is able to throw a case into relief, highlighting errors and assessing deviations. I have drawn comparisons between rationalism and Bates, culturalism and Scott, and structuralism and Skocpol. Finally, one can examine a single case from the point of view of several ideal types. Weber thus tries to explain particular cases by applying a battery of ideal types and theories. Actual cases, in other words, are unique, specific, and distinctive combinations of the ideal types. The next section will therefore analyze Weber from tationalist, culturalist, and structuralist perspectives. These four sets of contrasts, analogies, and juxtapositions allow compara-tivists to fruitfully mix the general and the specific, engaging theories and the cases. They allow Weber to place cases and situate developments within general contexts; they facilitate general comparative and typological explication of particular events and situations; and they enable him to focus his analysis on those cases that had great substantive and theoretical significance. Moreover, ideal type analysis makes comparativists aware of the pitfalls of moving from theory to case. Weber argues that ideal types are only useful fictions that do not exhaust and exactly depict reality. Hence, they should not be reified into something "real rather than something "nominal": They cannot be used to "deduce" cases and therefore they are not "falsified" by locating deviations from real cases. I have therefore combined various properties of social theories (e.g., ontology, methodology) to produce three types of social theories. The ideal-type rationalist, culturalist, and structuralist research communities developed here should be judged on pragmatic grounds: They are useful or not useful for this or that probf. lem from this or that conceptual point of view; they should be displaced by another sec of ideal types that highlight different and, from another point of view, m r->r-o c I rrn rant- aCnfm n In sum, a consideration of three ideal-type research schools forces comparativists to confront some fundamental issues in social theory: the nature of the competing paradigms and the structure-action, social order-social change, and nominalism-realism debates. The socially embedded unit act helps clarify, albeit not solve, these enduring issues. SECTION 4: MAX WEBER, MODERNITY, AND COMPARATIVE POLITICS TODAY Parsons's failed synthesis thus holds the key to a deeper understanding of the connections and disjunctures among the three research communities that dominate contemporary comparative politics. Any satisfactory substantive explanation in our field will employ one or more of these sets of "nuts and bolts" (Elster 1989). But what would be the purpose of such a theory? What is the historical situation in which social scientists find themselves that requires understanding? Max Weber29 argues that social scientists should begin inquiry by analyzing the value relevance of the problem situation in which they find themselves. Weber thus explores the world-historical significance of his own circumstances and then evaluates his state of affaits from the point of view of a normative theory of politics: how people ought to live, the definition and implementation of the common good, and what is the best or right life and regime. This reflection led Weber to identify his central problem, and surely the central problem of his age, as the dialectic of modernity, or how modernity emancipates and exploits.30 Medieval thought centered on the word of God as revealed through the teachings of the Roman Catholic Church. As Hawthorn (1976: 8), paraphrasing Hobbes, puts it, "Reason was always and necessarily subservient to revelation, which alone could reveal God's purpose." Weber's concerns, and indeed the concerns of all of the founders of social thought, can be traced to a process, begun in the Renaissance and continuing in the Enlightenment, that disputed medieval thinking. The rigid but stable old order of communities and hierarchies, based on ecclesiastical, feudal, and monarchical authority, was challenged by cognitiveiy and morally reasoning individuals (Nisbet 1966). These individuals attempted to create a new rational order in all spheres of social life. *sIt is fascinating to recall haw much of che language of contemporary compararivisrs can be traced to Weber: ideal interests and material interests; ideal type and ventibea; class, status, and patty; traditional, charismatic, and rational legitimacy; fotmal and substantive rationality; the iron cage of reason and the disenchantment of the world; bureaucracy and the modern state; the ethic of ultimate ends and the ethic of responsibility; and the Protestant ethic and a calling. Behind this vocabulary lie Weber's methodology and substantive theories which have also had a lasting impact on comparative politics. 3DWeber's master problem has also been identified as the origins of capitalism and the West, the nature of domination and the state, and the science of culture (Hennis 1983; Nelson 1974; Schroeder 1992: chan. l-TKtnhni^U 1000*1 268 CONCLUSION SOCIAL THEORY AND COMPARATIVE POLITICS 269 Modernity thus involved the growth of reason that culminated in a series of macro changes involving the rationalization of social structures. In politics, liberal democracy, the rule of law, and state bureaucracy were created. In international relations, globalization and internationalism culminated in a world system. In the intellectual world, science and technology were founded. In the economic world, bourgeois capitalism — markets, industrialization, mass consumption, and manufacturing — were developed. In civil society, specialization, division of labor, complexity, and pluralism were fashioned. Finally, the culture sphere saw the growth of liberalism, individualism, universalism, egalitarianism, humanism, secularism, materialism, and the idea of ptogress. The projecr of modernity thus involved reason at two levels. At the individual level, individual rationality and moral autonomy were to be constitutive of identities. At the societal level, the rationalization of structures and institutions were to be constitutive of social order. Weber (1946: 117) is a rationalist in that he warns about the paradoxical consequences of this modernity: "The final result of political action often, no, even regularly, stands in completely inadequate and often even paradoxical relation to its original meaning." The macro transitions and transformations, in other words, produce new irrationalities, instabilities, inefficiencies, and contradictions that challenge social order. Modern western (occidental) rationality, he maintains, is concerned with means-ends calculations (formal, rationality) rather than with reasoned judgments about the value of ends themselves (substantive tationality). The drive to control all aspects of the natural and human worlds has several structural consequences: imperialism and dependence, or the ruthless expansion and exploitation of the planet and all of its peoples, which inevitably produce destructive wars; the bureaucratization of everything, or the growth of rationalized and anonymous administrative systems that regulate all forms of modern life, which ultimately conttols the body and sexuality; and the struggle by all peoples for nationalism and democracy, which finally cause intractable and deadly conflicts. The policy consequences are equally dramatic: Humans cteate an artificial world of irrationalities; this begets further rational means-ends calculations, which result in another round of policy interventions designed to rid the wotld of the first-order irrationalities; in the end, newet and deeper irrationalities result. The drive to control the natural and social worlds also has major consequences for values: the secularization of culture (what Weber calls the "disenchantment of the world," or the expulsion of magic, myths, and all forms of it-rational social life) and the standardization of culture (what others have called massification and homogenization). Finally, the supposedly rational individuals / suffer the most dire consequences of all: the loss of certainty about the meaning and purpose of their lives. Moral individualism, the radical isolation, separation, and divorce of individuals from all social ties, is responsible for several evils: The commodification and depersonalization of social relationships, and hence the deterioration of common values and the shrinking of a shared ethical space, are pet-innTic Pvnprrprl rmrrnmpť rhp prnwth of feelines of alienation and powerlessnešs,; and the resulting failure of the will to power (the drive to self-perfection and self-promotion, or the desire to perfect and extend the self by relying on personal creative power rather than depending on anything external), are perhaps paradoxical outcomes; solipsism, relativism, and perspectivism are perhaps logical outcomes; and since people need meaning, the rise of fundamentalisms, myths, and superstitions to replace the traditional values that have been lost is perhaps the inevitable outcome of moral individualism. In the twentieth century, these structural, policy, cultural, and individual developments culminated in pathologies unimaginably worse than the greatest evils of the premodern world: two world wars and totalitarianism. The logical consequence of reason applied to means and not ends was the Holocaust: The trains ran on time but to a place inconceivable to all but moderns. Rational means had furthered irrational ends, and an unimaginably pathological and barbaric irrationality at that.31 Weber is also a culturah'st in that he explores a fascinating paradox in the origins of modern rationality: how the irrational quest for meaning and salvation helped create the rational individuals and institutions of the modern world. He hypothesizes that "Calvin's doctrine of predestination resulted, among his followers, not in fatalism, not in a frantic search for earthly pleasures, but curiously and counterintuitively — in methodical activity informed by purpose and self-denial" (Hirschman 1977: 130). In other words, the Protestant Reformation produced ascetic Protestantism which, in turn, actualized the individualism that encouraged a rational social order: The spirit of capitalism motivated bourgeois capitalism; an inquisitive scientific outlook inspired the Cartesian-Newtonian scientific framework; moral individualism encouraged political liberalism; and methodical patterns of action galvanized state bureaucracies. Rationality, according to Weber, thus depends upon such irrational motivations as the Protestant doctrine of ptoof and the idea of a "calling." Swidler (1973: 41) concludes that "the values which motivate rationality, the control of ideas over action, must themselves be non-rational. There is always a sphere of social life which is non-rational, and it is on the preservation of this sphere that the rationality of the rest of the system depends." Only meaning, faith, and a calling can save us from the irrationalities of reason.3" Finally, Weber is also a structuralist in that he studies how the institutional dynamics of state and society cage individuals in the dialectic of reason and Irrationality. He explores the institutionalization of three types or systems of domination or authority - rational, traditional, and charismatic legitimacy ([1924] 1968) - and the logics of three types or systems of stratification - class, status, 3'The postmodernists challenge the "logic" of such grand historical narratives. "Science, according to Weber, is perhaps the best example of how the basis of rationality is an irrational commitment to ultimate values. He suggests that values guide the choice of scientific problems and that the commitment to science is itself 0 value commitment based on the desire to shape the wotld. Hirschman (1977: 38) concludes chat "unintended consequences flow from human thought (and from the shaDe it is niven thrnni*h lt!n