Return to Home Page

Interest, Class, and Identity

Microfoundations for Asian Studies

Daniel Little

Department of Philosophy and Religion

Colgate University

Hamilton  NY 13346

Association for Asian Studies

April 3, 1992

Abstract.  Social explanations require microfoundations: accounts of the local-level processes and circumstances that lead individuals to behave in ways that give rise to society-wide patterns of organization and change.  This paper examines several fundamental theoretical assumptions in terms of which researchers have attempted to explain patterns of social organization and change in agrarian Asia: rational-choice models, class-conflict models, and identity-based interpretations.  To what extent is it possible to account for patterns of agrarian change in terms of the economic interests of participants?  To what extent do the social-property arrangements that determine local class divisions affect the processes of agrarian change?  And how much influence can be assigned to features of local group identity (ethnic, religious, national, or class) in processes of collective action and local politics?  The paper will draw on prominent explanations of agrarian change in China and India. 

            Asianists, like other social scientists, are concerned to put forward explanations of various types of social phenomena: the occurrence of rebellion, the persistence of traditional agriculture, the persistence and change in distinctive patterns of family structure, and many other phenomena.  How should such explanations be constructed?  What theoretical resources are available in terms of which to explain social patterns?  This paper focuses on explanatory frameworks that have two features in common.  First, each respects the need for what I will call "microfoundations" for social explanations—explanations that account for social outcomes in terms of the individual-level actions that participants perform.  And second, each depends on analysis of features of individual agency; analysis of features of purposive, meaningful activity by individual agents stands at the center of these explanatory models.  Where these models differ from each other has to do with the level of abstraction at which proponents want to describe facts about individual agency. 

            The frameworks that will be considered here derive from rational choice theory, contemporary analytic Marxism, and contemporary analyses of ethnic politics in terms of a notion of ethnic identity.  Putting the frameworks loosely: The rational choice paradigm attempts to understand social outcomes as the aggregate effect of rationally self-interested individuals pursuing their private interests.  The analytic Marxism framework understands social outcomes of individuals acting deliberately on the basis of interests as well; but individuals are conceived as having affinity with class-defined groups, with the result that self-interest includes some characterization of group interest.  And the ethnic politics paradigm understands social outcomes as the result of highly specific cultural identities of participants.

            There are strongly disputed disagreements among these frameworks, to be sure.  But it is important to recognize as well the affinities among them, and the disagreements each represents with other more holistic or structuralist approaches to social explanation.  For each of these approaches casts doubt on the idea that there are autonomous social entities or structures whose properties determine the behavior of social systems.  Each emphasizes the importance of identifying the individual-level mechanisms through which social changes occur.  And each represents an empirically sophisticated alternative to the somewhat simplistic empiricism of some quantitative work in the social sciences.

            The discussion below will be organized around examples of popular politics, in order to give a basis for comparison of the explanatory resources made available by these several frameworks.  Underclass collective actions—riots, strikes, jacqueries, protests, and rebellions—are a recurring thread in human history.  Chinese peasants support millenarian Buddhist revolts against the Qing state; Vietnamese rural poor support violent mass demonstrations against the French colonial government; Hindu nationalists come together in large demonstrations at the site of the Babri mosque.  These are all instances of popular politics: collective actions supported by large numbers of ordinary people in pursuit of some shared goal.  The dynamics of popular politics are critically important in the process of historical change: regimes rise and fall, national political movements find support or wither, wars are won or lost, and colonial powers survive or retreat, depending (in part) on the political behavior of the masses.  Each of the frameworks under scrutiny in this paper provides a basis for attempting to explain instances of popular politics, and each raises its own set of methodological problems.

What are microfoundations, and why are they needed?

            A central contribution to contemporary philosophy of social science is the idea that macro-explanations in social science require microfoundations.[1]  This doctrine maintains that macro-explanations of social phenomena must be supported by an account of the mechanisms at the individual level through which the postulated social processes work.[2]  More specifically, the thesis holds that an assertion of an explanatory relationship at the social level (causal, functional, structural) must be supplemented by two things: knowledge about what it is about the local circumstances of the typical individual that leads him to act in such a way as to bring about this relationship; and knowledge of the aggregative processes that lead from individual actions of that sort to an explanatory social relationship of this sort.[3]  This doctrine may be put in a weak and a strong version: weakly, social explanations must be compatible with there being microfoundations of the postulated social regularities, which may, however, be entirely unknown; and more strongly, social explanations must be explicitly grounded on an account of the microfoundations that produce them.  Here I will assume the stronger version: that we must have at least an approximate idea of the underlying mechanisms at the individual level if we are to have a credible hypothesis about explanatory social regularities at all.  A putative explanation couched at the level of high-level social factors whose underlying individual-level mechanisms are entirely unknown is no explanation at all.

            This perspective may be put to particularly useful work in treatment of the problem of class politics: under what circumstances are classes capable of achieving collective action in pur­suit of shared interests?  Classical Marxism holds that classes tend to become class-conscious (that is, aware of shared interests), and class-active (that is, motivated to act as a group in pursuit of shared interests).  But once we adopt a microfoundational perspective, this assumption is suspect, since it takes no account of public goods problems (free-rider problems and collective action problems).  If we are to put forward a theory of class politics at all, it must include an account of the processes through which classes are capable of constituting themselves as political agents; and this means we need an account of the micromechan­isms of collective action within a class society.  This means that it is necessary to provide an account of the microfoundations of collective action, since most instances of polit­ical collective action involve the pursuit of public goods.  It is not sufficient, therefore, to refer to the shared interests that members of a class have in the attainment of a political end; it is neces­sary to identify as well the individual-level circumstances that give potential participants an in­centive to involve themselves in the collective action.  Otherwise we should expect free-riding and prisoners' dilemmas to make collective action unattainable.[4]

            Do explanations in Asian studies require microfoundations?  They do, for reasons that exactly parallel arguments about structuralist Marxism.  Let us focus on one topic that has been central in both China studies and analytical Marxism: the explanation of peasant collective action.  In Understanding Peasant China (Little 1989) I survey several explanations of peasant rebellions in Qing China: Susan Naquin's treatment of the White Lotus rebellions, Elizabeth Perry's treatment of the Nian rebellion, and Robert Marks's analysis of peasant politics in the Canton Delta.  These explanations all proceed from different assumptions about peasant motivation; Naquin emphasizes millenarian religious beliefs, Perry emphasizes rational strategies of subsistence security, and Marks attends to features of class identity.  But in each case the author puts forward an analysis of the local circumstances of motivation and mobilization.  And the central interest of these explanations is the detailed analysis the authors give of the local mechanisms of political mobilization: the militia organizations, the bandit gangs, the itinerant martial arts instructors, or the lineage groups that transform general political factors into specific processes of political action and inaction.[5]  In each case, then, we find explanations that succeed because they conform to the microfoundations dictum: they provide local- and individual-level accounts of the processes through which collective action proceeded.

            There are also important instances of research within Asian studies that violates the requirement of methodological holism.  For example, consider the various efforts to apply Wallerstein's world-systems theory to the historical development experience of East and South Asia.  As David Washbrook's trenchant criticisms show, this model is highly unconvincing as a basis for explaining the patterns of economic development in China, Japan, or India [Washbrook, 1990].  And an important element of the failure of the framework is its neglect of the local and regional processes through which very distinctive patterns of economic development took place in these regions; instead, Wallerstein and his followers offer explanations that depend only on global structural characteristics, without showing how these global structures disaggregate into local-level interests and constraints.

Interest-based explanations

            Let us turn now to substance: what are some of the explanatory frameworks within the context of which it is possible to formulate adequate microfoundational explanations?  The microfoundations thesis aligns closely with emphasis on the tools of rational-choice theory in social explanation.  The microfoundations approach directs the social scientist to turn to the individual-level processes that produce social out­comes; and rational‑choice theory offers a general account of what those individual-level pro­cesses are.  Rational-choice theory thus functions as a research program for social science: to ex­plain social outcomes as the aggregate result of individuals' calculating efforts to pursue their in­terests given their beliefs about the environment of choice.  This program is plausible because human beings are purposive beings capable of forming beliefs and choosing actions on the basis of their goals and beliefs.  This is not to say that human beings are perfectly or always rational, but it does imply that rational-choice theory provides a common starting-point for analysis of social phenomena.

            The rational-choice paradigm of explanation rests on one central premise and a large set of analytical techniques.  The premise is that individual behavior is goal-directed and calculating.  Individuals are assumed to have a set of interests in terms of which they evaluate alternative courses of action; they assign costs and benefits to various possible choices and choose an action after surveying the costs and benefits of each.  Rational choice explanations thus depend upon the "means-end" theory of rational action.  An action is rational just in case it is an appropriate means of accomplish­ing a given end, given one's beliefs about the circumstances of choice.  On this account, to explain an individual's action is to identify his or her background beliefs and goals, and to show how the action chosen is a reasonable way to achieve those goals given those beliefs.[6]

            This account of rationality may be described as a "thin" theory of human action.[7] It depends on an abstract description of goals in terms of interests, utilities, or preferences, and postulates a simple mode of reasoning—utility maximization, for example.  On the basis of these simplifying assumptions the means-end theory hopes to be able to explain a variety of forms of human behavior.  The advantage of this approach is explanatory parsimony and power; to the extent that these assumptions bear some discernible relation to human behavior, they provide the basis for explaining a wide range of social phenomena in a variety of cultural settings.

            Rational-choice approaches have sometimes tended to give insufficient attention to the institutional context of action.  Taking neoclassical economics as a paradigm, rational-choice theorists have sometimes treated the institutional context as analogous to the marketplace—a homogeneous and impersonal arena for social transactions within which maximizing individuals make their choices.  But this tendency is not inherent in the rational-choice approach.  It is perfectly consistent with this approach to accept the point that institutions matter: different institutional arrangements produce different patterns of individual behavior.  And they do so by altering the circumstances of choice of the typical agent—altering the opportunities, constraints, information, and incentives that define the choice space for the agent.  Greater attention to institutional arrangements has led to important new insights within rational-choice analysis of social action: the possibility of cooperation among rational agents (Axelrod 1984, Taylor 1982), the possibility of community-based management of public goods (Ostrom 1990), the implications of different systems of revenue collection (Levi 1988), the bias of agricultural policies in tropical Africa (Bates 1981), and the implications of transaction costs for social action (North 1990).

            The rational-choice approach generally gives short shrift to the workings of norms and values in human action.  However, much recent work within the applied rational choice field gives more attention to the role of norms and values in motivating or constraining individual choice.  Whereas some social scientists within a rational-choice framework have sought to minimize the role of norms and val­ues (e.g. Samuel Popkin in The Rational Peasant; 1979), thinkers such as Elster have come to recognize that refer­ence to normative systems has a place within an individualistic theory of social action.[8]  This is a step forward in the direction of a more empirically adequate theory of individual motivation, and one that can eventually be deployed to produce more complex models of social processes that reflect both prudential and normative motivations. But it is perfectly consistent for an individual­ist theory to introduce social norms into its explanations; contrary to Durkheim's familiar view that norms have a supra-individual status, it is plain enough that norms can only be embodied in the actions, sanctions, gestures of approval and disapproval, etc., of particular individu­als.

            So far we have examined the concept of individual rationality.  How does this concept give rise to explanations of social phenomena—the occurrence of collective action, the stability of social institutions, or processes of social and economic change? The rational-choice approach attempts to explain social outcomes as the aggregate result of large numbers of individuals acting on the basis of rational calculations.  Why do strikes often collapse before they gain their objectives?  Because of the advantages of defection for individual strikers.  Why do prices tend to oscillate around the cost of production plus an average rate of profit?  Because rational entrepreneurs enter and exit industries according to the rate of profit in the industry.  Why do arms agreements tend to break down?  Because participants fear unilateral defection by their opponents.  What these hypotheses have in common is an explanatory strategy: to explain a social pattern as the aggregate consequence of the rational actions of a large number of participants, given the circumstances of the social and natural environment within the context of which agents deliberate. 

            It is at this point that the analytical tools of rational choice theory become an explanatory resource.  The tools of microeconomics, game theory, collective action theory, decision theory, and Marxist economic and political theory are now available as models that may be applied to particular social phenomena.  Each represents an analytical technique that works out some of the implications of rational decision-making within stylized contexts of choice.  If we assume, for example, as Theodore Schultz (1964) does, that peasants make cropping decisions on the basis of profitability, we can predict that the portfolio of crops sown will be responsive to relative prices.  If we assume that individuals engaged in non-zero-sum competitive activities make choices based solely on their own interests, we may infer that prisoners' dilemmas will occur.  And so on for an open-ended range of analytical techniques.  These techniques permit us both to make predictions about individual behavior, and more importantly, to derive the aggregate consequences of individual behavior.

            These efforts may be described as aggregative explanations.  Thus Philip Huang (1985:108) explains cropping portfolios mixing cotton and sorghum chosen by North China peasants as the result of a rational appraisal of risks and benefits; Samuel Popkin (1979) explains the failure of collective action in village societies as the effect of freerider choices; and Robert Brenner (1976) explains the stagnation of French agriculture as the absence of incentives and opportunities towards technological innovation on the part of landlords and peasants.  In each case the author identifies a pattern of rational individual behavior responding to a particular set of incentives and constraints, and then attempts to show how this pattern of individual behavior aggregates into the observed macro-pattern.[DL1] 

            The rational-choice approach has been applied fruitfully to problems of popular politics in Asia by a number of scholars.[9]  On this approach, the goal is to explain peasant rebellion and political behavior on the assumption that peasants make deliberative decisions about participation.  The guiding thread of these theories is that rebellion is a form of deliberate collective action that originates in the local interests of the individuals who participate and is facilitated by the local political resources (organizational forms, militias, kinship organizations, etc.) available to potential rebels.  Individuals are assumed to have a clear idea of their interests and opportunities, and engage in collective action as a deliberate strategy of furthering their individual interests.  The primary emphasis in this approach is on uncovering the local processes through which these general factors come to fruition.  A distinctive feature of these theories is the observation that largescale rebellions are often only unintended, unforeseen consequences of essentially local processes.  The local political process is rational-intentional, whereas the global process is unintentional.[10] 

            Elizabeth Perry's Rebels and Revolutionaries in North China 1845-1945 (1980) is an important recent example of this sort of approach.[11]  Perry analyzes the Nian Rebellion (1851-63) and the Red Spear uprisings (1911-49), both in the Huai-pei region on the North China plain.  The Nian Rebellion represented the gradual transformation of local bandit gangs into loosely-coordinated regional forces capable of defeating regular Qing armies.  By 1856 small Nian groups had been brought together into a loose confederation of five "banners" under the nominal leadership of a bandit, Chang Lo-hsing; in fact, however, leadership and power within the Nian remained at the community level rather than the regional level.  Nian groups were generally adept at cavalry warfare, and practiced mobile strategies that were difficult for the more static Qing military forces to counter.  The Nian groups retained close connections with their local communities, which often took the form of walled villages and towns; these villages provided both an economic base and a defensible retreat for Nian groups.  At its peak the Nian may have had as many as 100,000 men under arms, and held sway over large parts of Huai-pei.  Regular Qing forces were supplemented and finally replaced by Manchu and Mongol cavalry, and after years of inconclusive fighting the Nian armies were destroyed in 1868.[12]

            Perry's account of rebellion in Huai-pei gives central place to the ecological circumstances that surrounded peasant life in North China.  She notes that the ecology of the North China plain was extremely harsh, with regular flooding, drought, and famine. The region had only a low level of commercialization, given low agricultural productivity.[13]  As a result peasant lives were both poor and insecure, and there was precariously little surplus available to tide over periods of disaster.[14]  Perry puts the point this way:  "In sum, interactions between people and nature rendered Huai-pei a highly precarious ecosystem.  During the century under consideration (1845-1945), nature wielded the upper hand" (16).  Perry thus puts the problem of survival at the center of peasant behavior in this region and time; and she explains the extensive rebellions that occurred during the period as the results of several different types of survival strategies by peasant actors.  She argues that peasant actors sought out a variety of means of survival, some individual and some collective, through which to ensure their continuing welfare and security.  She distinguishes broadly between predatory strategies and protective strategies.  Predatory strategies include smuggling, robbery, and banditry, while examples of protective strategies include crop-watching societies, local militia, fortification, and tax resistance (58-95).  She holds that each strategy gave rise to distinctive forms of collective action. Further, the forms of collective action inspired by each strategy influenced the subsequent development of the forms of collective action inspired by the other.  The Nian rebellion grew out of the formation of more and more powerful bandit gangs (predatory strategy), while the Red Spears period reflected the autonomous power of local militias directed by local notables (protective strategy).

Marxism and popular politics

            Turn now to the explanatory framework of analytic Marxism.  Classical Marxism offers a simple yet powerful theory of popular politics.  According to that account, exploitation and class are the central factors that explain processes of underclass politics.  Class society—capitalism, feudalism, slavery, or agrarian landlordism—rests upon a system of exploitation in which surpluses are extracted from producers for the use and benefit of an elite group.[15]  This system is constituted by a set of social relations of production.  The social relations of production are the relations of power and authority that govern the use and enjoyment of the forces of production—in brief, the property relations.[16]  These relations define the objective interests of members of the various classes.  This system thus constitutes a set of classes within society whose most basic material interests are defined by the particulars of the property system.  Further, Marx holds that there is an endogenous tendency for members of a class to come to accurately perceive their shared material interests.  Finally, Marx supposes that, as rational agents, members of exploited classes will come to support a revolutionary movement to overthrow the system of exploitation. Members of exploited groups have the capacity to come to recognize the nature of the system of exploitation that constrains them, and as rational agents, they adopt political means for struggling against that system.

            Classical Marxism thus identifies classes as the politically salient groups in society.  It holds that such classes come to identify themselves as such, and that members of classes engage in collective action in defense of their material interests. Marx believed that revolution is the natural culmination to the historical development of a class society: class society involves the exploitation of the many by the few, and eventually the many will acquire both the will and the means to overthrow the economic relations that govern them.  Thus classical Marxism advances a fairly simple macro-level causal claim: the social tensions created by an exploitative economic system, and the conflicts engendered between landlord and peasant, capitalist and worker, master and slave, give rise to underclass collective action (food riots, social banditry, rebellion).  And the mechanism establishing this causal relationship is described by an abstract model of rational political calculation: the structure of the system of exploitation defines the material interests of the various classes; and members of those classes, as rational agents within a structured environment of choice, act in a variety of ways so as to defend their interests.  There is a straightforward connection between the objective conflicts of interest embodied in the property system, and the political behavior of the various participants.  Rebellions and popular collective action are rational strategies of collective self-defense on the part of subordinate classes.[17]

            This theory has much to recommend it as a theory of the causes of revolution and rebellion.  However, I will argue here that it fails to provide analysis of a number of factors that are critically important to the occurrence of popular collective action.  Whereas this class-conflict model tacitly presupposes a transparent and direct relation between class interests and class behavior, study of any of a number of class societies shows that this is not the case.  Instead, class structures, and the group interests they define, are mediated in their effects on collective action by a large number of other social circumstances that are causally independent from the class structure.  Thus the behavior of persons and groups within a given class structure is indeterminate; we need to know much more about them and their environment before we can arrive at an expectation of rebellion, jacquerie, private strategies of survival, or passive acceptance.

            The chief deficiencies of the classical Marxist theory of underclass politics fall in several areas.  The notion of objective class interests is ill-defined because it ignores an important array of local and regional variations in group interests.  And this account has virtually nothing to say about the role of the political organization and leadership of the revolutionary class.  It is plain from historical experience, however, that mass militancy, even if present, does not automatically translate into successful political action without organizational resources—formation of strategies and tactics, gathering of information, collection of resources, mobilization and motivation of widespread support for their program.[18]  What is needed, then, to complement the classical Marxist theory is a detailed analysis of the microfoundations of class politics: the features of local organization, group formation, and individual behavior that lead in the aggregate to the occurrence of largescale collective action by members of classes.  And in fact scholars working within analytical Marxism have begun to provide such foundations.

Level of group aggregation

            The assumption that class membership is politically salient amounts to the claim that members of classes so defined will come to identify themselves as such and to acquire a disposition to act in accordance with shared class interests.  Class is salient to politics, then, because material interests are fundamental to individual political action, and because the class structure defines people's material interests. Groups will tend to coalesce around class lines (rather than ethnic or religious identity, kinship group, patron-client relations, etc.).  This formulation is radically incomplete, however.  As a start, we need to have an account of the processes through which individuals come to adopt a collective identity.  The formulation above refers to a highly abstract group: a spatially and socially dispersed group of persons sharing an abstract characteristic (position in the property system).  Members of this group do not share a set of homogeneous material interests; and they are not bound together by shared circumstances of life, patterns of everyday association, or uniform material interests.  It is of course true that various subgroupings of classes are bound together in these ways. Workers in a particular factory or industrial city or miners in a complex of mines are in everyday contact with one another in ways that may plausibly lead them to identify themselves as a cohesive group.  But coal miners in Pennsylvania, farm laborers in California, and industrial workers in Detroit, form a highly diverse and heterogeneous group; and yet they share the position of "wage-laborer" in the property system, and they are all exploited by ownership of the means of production.  So why should we suppose that the abstract feature of class membership is likely to become politically salient for members of these groups?  The class-conflict model thus encounters problems concerning the proper level of aggregation in its description of groups, group interests, and group identity.  This point involves two considerations.  First, "class" interests may be specified on a variety of levels, from local to global and from specific occupation to abstract class.  Second, group identity may be expected to take shape in different ways at different levels of group definition.

            Consider first the spatial side.  The class-conflict model does not incorporate any regional specification; it treats a class as an undifferentiated whole over the full society.  But it is plain that there are politically significant differences of interests within classes over space.  For example, peasants of late-Qing Henan had an interest in famine relief that peasants of the Yangzi did not share (because food supplies were more stable in the latter region), whereas the latter group has a crucial interest in water works not shared by the former; these variations derive from differences in the farming systems and ecologies of the two regions. Regional differentiation thus imposes interests on segments of a class that may be politically significant.[19]

            Class interests may thus be viewed from a range of perspectives from the local to the global.  For example, tenant farmers in a given village or marketing community share certain interests in common (e.g. water rights, access to firewood, etc.), which give them a material basis for engaging in collective action together.  But class interests may be defined on a more comprehensive geographical scale as well; thus, for example, tenant farmers throughout all of north China share certain material interests—economic and political—with each other.  (A concrete example of shared interests is imperial tax policy—should taxes be remitted in times of crop crisis, should taxes be assessed to land or to farmer, etc.)  Once we recognize the range of levels at which group interests may be defined, however, it becomes clear that there may be fundamental conflicts of interest within groups defined at the higher levels.  Thus a strategy of demanding tax relief for farmers in Henan may impose greater tax burdens on farmers in Sichuan. There is thus a material basis for mobilization of local politics on the basis of local interests in direct opposition to global class interests.  This point makes it plausible to suppose that local material interests might mask global class interests—local elites and peasants may join together in violent action against neighboring villages in conflict over water rights, disputed land rights, etc.[20]

            Turn now to occupational diversity.  The structural definition of class distinguishes among a small number of social categories, depending on position within the property relations.  On this criterion, all tenant farmers belong to the same class; all industrial wage laborers belong to a single class; all owners of industrial wealth constitute a class; and so forth.  However, it is obvious upon closer analysis that there are important forms of diversity within classes so defined that may produce importantly different interests for the various sub-groups.  Tenant farmers in a wet-rice region, for example, have different interests with regard to the state than do tenant farmers in dry cropping areas, because they are more dependent on state-financed water works. Unskilled workers, semi-skilled workers, and skilled workers, to take a different example, have different interests concerning technological change; as a result, these different strata of the proletariat may be mobilized in support of very different economic policies and political actions.[21]

            There is a tendency among class-conflict theories of underclass politics, however, to consider class interests on too high a level and on the basis of too homogeneous an account of the circumstances identifying material interests.  But it is entirely possible that vertical local interests may loom larger in the material welfare of members of a micro-class than horizontal regional interests—thus making it difficult to secure collective action around regional class interests.  A policy or strategy may be prudent at one level of interests and counterproductive at higher or lower levels.  These considerations suggest that class-conflict theories confront a serious difficulty in arriving at an analytically justifiable level at which to characterize group interests and identity—which must be done if we are to speak meaningfully of class and class conflict.

Group identity formation

            Turn now to the problem of group identity formation.  What mechanisms would lead a group characterized in terms of its shared material interests to come to identify itself as a political agent?  The simplest and most plausible case is that at the low end of both spectrums above: groups that are geographically compact and occupationally homogeneous.  Members of groups of this type have a set of prominent material interests in common, and they have concrete opportunities for developing political activity together, through shared organizations and acquaintances and a shared life experience.  But most analyses of class politics proceed at a higher level of aggregation.  When Robert Marks writes that the peasants of Haifeng County began to conceive of themselves as a class in the middle and late nineteenth century, does he mean that poor peasants of a given village came to recognize their joint interests, or that poor peasants throughout the region came to recognize their joint interests with other peasants beyond their own acquaintance and social intercourse?  At the local level it is possible to imagine the social mechanisms through which such group identity might emerge through normal social contact (though it is also possible to identify mechanisms working in the opposite direction—e.g., competition for the right to rent a piece of land or cross-class allegiances based on lineage or patron-client relations).  But it is more difficult to conceive of mechanisms at work in local society that would lead to a substantially broader sense of group identity, in which poor peasants of many lineages, many villages, many market systems, and several different ethnic groups should come to regard themselves as a cohesive class, sharing important interests and disposed to engage in political activity in support of those interests.

            Finally, it is an important truth that much observed collective action in the non‑industrial world does not occur along clean class cleavages; instead, much collective action involves alliances between local elites and lower classes.  This is true, for example, in the context of intervillage conflicts of interest: e.g., water and land rights. Prasenjit Duara describes some of the forms of organization through which North China villages and market towns mobilized cross-class support for the purpose of defending water rights (Duara 1988).  Second, Chinese village self-defense and militia organizations typically had a cross-class character with elite leadership (Kuhn 1980).  Third, local society was capable of providing collective goods—e.g. village water resources (flood control and irrigation); and, once again, these forms of collective action typically proceeded under elite leadership.  Peter Perdue's historical treatment of the organization of waterworks in Hunan illustrates this point (Perdue 1987).  These examples suggest, however, that class factors were often eclipsed by economic interests shared across village society as a whole.[22]  In each of these cases we find that collective action occurs under the leadership of village elites—landlords, village headmen, local literati.  But this poses a problem for the class-conflict thesis; if much traditional collective action occurs through vertical organizations and with elite leadership, then what resources are available to create horizontal organizations and leadership?

            These considerations present the class-conflict theory with an unresolved but deep problem: what social mechanisms would permit a geographically dispersed class to arrive at a group identity based on its shared material interests in opposition to other, regionally-based interests that may be shared with locals from other classes?  What would lead the landless peasants of the Vendée to identify their fortunes and goals with those of the Languedoc, rather than with local elites?  Why should we expect textile workers in Manchester to identify their interests with metal-workers in Lancashire?  The apparent answer to this question unavoidably involves reference to supra-local organizations—a topic I will consider in the next section.  But it would appear dubious that there is an endogenous tendency for dispersed groups to come to identify their interests as classes rather than as groups of other sorts.

Organization and leadership

            Suppose, finally, that a group possesses both a set of objective material interests in common, as well as the motivational resources needed to permit mobilization of collective action in support of those interests.  The political behavior of the group is still indeterminate because successful political action requires competent organization and leadership, if it is to go beyond short-term and sporadic outbursts.  There is a tradition within Marxist historiography, however, which tends to downplay the importance of organization and leadership, and to emphasize instead the potential for spontaneous mobilization on the part of exploited classes.  Thus Mark Selden (1971) and Robert Marks (1984) both emphasize that the impulse for revolution in rural China came from below—not from a radical intelligentsia or party.  Likewise, Marxist scholars have paid little systematic attention to the problem of leadership within political action.  It is thought that emphasis on the role of leadership is inconsistent with materialism and reflects an inappropriate attention to subjective rather than objective factors.  Revolutions occur, on this account, because of objective class conflicts, not because of the particular personalities of the leadership.  I will argue here that effective leadership and organization are necessary (though not sufficient) conditions for successful popular political movements.  It is therefore important to have a basis for analyzing the features of leadership and organization that contribute to successful and unsuccessful popular movements.  The presence of mass militancy is likewise a necessary causal condition, and one might plausibly hold that the latter is more critical because more difficult to provide if lacking (whereas new leaders can be recruited).  Attention to organization and leadership is in no way inconsistent with historical materialism, however; rather, it reflects recognition that social processes depend on complex social mechanisms, and an adequate theory of popular politics must take into account all the chief causal factors.  Leadership and organization are among the "micro-mechanisms" that underlie popular politics.

            Let us turn, then, to the problems of organization and leadership within a revolutionary movement.  What is the systemic role of the revolutionary organization in the processes of revolution?  In what ways can appropriate use of organizational resources facilitate mobilization and action?  And how important is leadership in the process of popular politics?  What motivates leaders; must we suppose that leaders are simply selfless agents of class interests, or should we attempt to understand their behavior and choices in terms of some complex mixture of private and public motivations?

            The experience of underclass collective action over the past century shows that the availability of competent organization is a necessary condition for the success of a revolutionary movement, for a variety of fairly obvious reasons.[23]  First, revolution demands effective mobilization of followers that, on a large scale, requires competent organization.  In order to create a revolutionary movement that is self-sustaining, it is necessary to recruit followers, maintain the level of commitment of current followers, raise funds, and create groups specializing in planning, communication, training, and so forth.  But mobilization of funds and followers depends critically on the ability of the organization to make its competence credible to potential or current followers.  Given the inherent risks of rebellion or revolution, potential followers will be calculating in their judgment of the likelihood of success; and the competence of leadership and organization will be central variables in their calculations.

            Second, revolution requires complex coordination of activities by large numbers of persons, which likewise requires effective organization.  In order to seriously challenge the state it is necessary for a revolutionary movement to undertake complex collective actions based on accurate information and sound strategic analysis; and it is necessary to be able to collect resources to fund such actions.  All of these features of successful revolution demand organization, however, and it is difficult to see how they could emerge from spontaneous local militants.  In order for such complex collective actions against the state to be successful, there is needed a high level of coordination among spatially separated groups of rebels.  This requires access to a reliable form of communication among groups.  It also requires discipline; the leadership needs to be able to count on different persons and groups carrying out their assigned tasks.  Finally, planning of such an action requires accurate information—for example, concerning the disposition of the regime's local military forces.

            Finally, organizations are needed to overcome the inherent localism of spontaneous underclass struggles.  Lucien Bianco (1986) holds, for example, that the Chinese revolution would not have occurred—in spite of the class tensions present in Chinese rural society—had the CCP not been present.  This is so in large part, Bianco holds, because spontaneous peasant activism is invariably localistic.  It is concerned with improving local conditions—resisting taxes, reducing the power of local elites, enforcing charity in times of famine, defending a village against bandits or soldiers, etc.  Once the immediate occasion for protest and resistance has passed, however, this localism dictates that peasants will attempt to return to their ordinary lives.  This means that a peasant movement will be continually on the brink of dissolution into its component parts.

            How do organizations work?  How do they manage to mobilize support, keep followers, raise funds, and formulate plans?  These are critically important questions that cannot be adequately addressed here.  But several points are fairly clear. Consider one important recent example: Samuel Popkin's analysis of the role of organization in rural Vietnam during the anti-colonial struggles of the twentieth century. Popkin gives the problem of organization a good deal of attention.  First, an organization must disaggregate its goals into a series of realizable sub-goals.  This is true for at least two reasons: potential followers are not likely to be moved by grand  ultimate purposes, and potential followers need to be convinced of the attainability of the movement's goals.  Second, successful organizations need to provide ongoing benefits to followers.  In his analysis of the popular movements of early-twentieth-century Vietnam, Popkin refers to adjudication services, provision of literacy and bureaucratic competence, and protection from bandits, as some of the ongoing benefits provided by the Catholic Church, the Viet Minh, and the Hoa Hao sects.  To an important extent, then, the organization needs to "piggy-back" its longterm goals on more immediate interests and services for local followers.[24]  Finally, it is clear that organizations need resources, financial and intellectual, in order to do their work of planning, implementing, coordinating, and mobilizing; and this requires effective resource-gathering on the part of the organization.

            It emerges from this discussion that class is at best a latent factor in political behavior on any but the local level.  It is possible to mobilize members of classes around their class interests, and to cultivate a class identity among members of classes.  Class membership constitutes a possible basis for mobilization because it does in fact identify a set of interests that are shared by members of the group; and these interests are, as Marxism postulates, particularly fundamental.  But there is no reason to expect that either group identity or political action will emerge spontaneously around class position.  The prominence of purely local interests and issues threatens to swamp the emergence of more global concerns and identity.  Thus latent class interests can only be converted into effective political motivations for a dispersed group through skillful organization and mobilization on a regional or national level.

Identity-based explanations

            Interest-based theories of popular politics—whether rational choice or Marxist—rest on the assumption that individuals make calculating choices in their political behavior, based on an appraisal of their interests.  But there appear to be some instances of political behavior that fail to reflect this model of calculating choice.  Particularly salient among these are the many occasions of ethnic politics and inter-ethnic conflict that have come into such prominence in various areas of the world since World War II.  In India, Sri Lanka, Indonesia, Nigeria, and Yugoslavia we are faced with violent conflicts organized around groups defined in terms of religious, ethnic, or national identity; and these instances often appear to involve conflicts over the politics of symbol rather than material conflicts of interest.  In these instances the phenomena seem to imply a strikingly different pattern of individual and collective behavior from that described by interest-based theories.  Individuals appear to have an extremely powerful affinity with their group identities; they seem strongly motivated to support collective actions on the basis of these identities, even at great cost to their individual material interests; they are easily mobilized by appeals to these identities; and these conflicts appear intractable in face of conflict-resolution strategies that ought to be successful if interest-based explanations were true (e.g. compromise over conflicts of material interests).  Given free-rider problems and the difficulties of organization and mobilization, such conflicts ought not arise so frequently; and given the common divisibility of economic resources, they ought to be more readily resolvable than they have proven to be.  So the question before us is this: what modifications of our theories of collective action are forced by examples of ethnic politics?  This question in turn disaggregates into two simpler problems.  To what extent do ethnic affinities constitute a powerful source of individual agency?  And what factors explain the mobilization of largescale ethnic violence?  Note that this way of analyzing the problem fits closely with the microfoundations approach outlined above.  It asks, first, what are the individual-level features of agency that govern individual actions; and second, what are the aggregative processes through which collective behavior emerges out of individual agency?[25]

Ethnic affinity and individual agency

            Let us consider how best to interpret an identity-based theory of political motivation.  There are several possibilities.  First, an identity-based theory may be construed as a psychological theory of the workings of a set of norms.  Individuals are assumed to have absorbed very powerful motivational constraints and imperatives, and these constraints and motives lead to actions that are not directed by self-interest or other forms of calculation.  On this account, individuals' behavior is construed as the causal consequence of a powerful set of group-specific norms.  This approach treats an individual's affinity with an ethnic, religious, or national group as a primordial motivational spring.  The most striking defect of this approach, however, is the violence it does to the notion of individual agency and choice.  The individual is construed as something of a robot, acting out of causal factors that are outside of his or her control.

            A second possible interpretation preserves the individual's agency and choice.  This position may be called the Scott-Bourdieu line.  On this account the individual acts in consideration of his or her beliefs and intentions—thus rationally; but these beliefs and purposes are themselves fundamentally shaped by the strands of group identity.  The individual's actions are construed as being strategic, calculating, and considered—not merely the causal consequence of an underlying normative program of behavior.  But the agent's choices are affected by norms, symbols, commitments, perceptions of the possible, and worldview.  On this approach, ethnic identities do sometimes become politically salient; but they serve to refract agency and choice rather than replace them.

            James Scott's (1976) analysis of peasant politics in The Moral Economy of the Peasant—though not a theory of ethnic politics—illustrates this general view.  Scott's analysis may be schematically represented in these terms.  Peasants in a given traditional society (Annam or Burma) share a specific subsistence ethic—a vision of their rights of subsistence and of the norms that ought to govern the behavior of other agents (landlords and the state).  The ethic serves as a basis of collective identity.  Peasants are led to violent protest when institutions offend the subsistence ethic in the context of a material crisis of subsistence.  When subsistence shocks occur which badly effect the economic security of large numbers of peasants, and when parties engaged in surplus extraction (landlords, notables, and tax collectors) fail to respond in accordance to the subsistence ethic, large numbers of peasants will be disposed to engage in violent protest (194).  Tax rebellions occur, for example, not when taxes are most onerous, but when they are perceived to be the most inflexible within the circumstances of economic crisis; attacks on landlords occur, not when rents are high, but when they are collected without regard to how much is left for the peasant cultivator.  Whether rebellion in fact occurs in such circumstances depends on other factors as well; particularly central is the presence of appropriate organizations and leadership.  Scott writes,

            We can learn a great deal from rebels who were defeated nearly a half-century ago.  If we understand the indignation and rage which prompted them to risk everything, we can grasp what I have chosen to call their moral economy: their notion of economic justice and their working definition of exploitation—their view of which claims on their product were tolerable and which intolerable. (Scott 1976:3)

            Scott's more recent books offer a more developed interpretation of the role of culturally specific identities within peasant politics.  In Weapons of the Weak (1985) he emphasizes the power of underclass conceptions of self and society—a kind of counter-hegemony —as a basis for everyday resistance of exploitation.  Here Scott holds that it is essential to explore the subjective side of class relations--the experience of subordination and the cultural vocabulary in terms of which subordination is lived in particular social circumstances—if we are to understand peasant politics.  Domination and the Arts of Resistance (1990) extends this perspective by providing a textured reading of the experience of domination and indignity that power relations impose upon the powerless.  Scott's central innovation in this work is his distinction between public transcripts ("the open interaction between subordinates and those who dominate;" p. 2) and hidden transcripts ("discourse that takes place 'offstage,' beyond direct observation by powerholders;" p. 4).  Postulating that there is a sharp divide between the behavior, language, and customs that dominated groups assume in public, and the language, jokes, and criticisms that structure their lives within the back streets, slave quarters, or rice paddies of their within-group experience, Scott attempts to provide tools for excavating the latter. 

            Both public transcripts and hidden transcripts have effects on the everyday politics of power.  The public transcript is by way of a libretto for the dominated, a stylized public performance through which they adopt the forms of deference and respect for the powerful that are needed to avoid punishment.  But Scott maintains that this performance is only skin-deep. The dominated are by no means taken in by their own affirmations of the justice and good manners of their masters, and behind the scenes we may expect to hear much raucous laughing, merciless lampooning, and bitter criticism.  "Offstage, where subordinates may gather outside the intimidating gaze of power, a sharply dissonant political culture is possible.  Slaves in the relative safety of their quarters can speak the words of anger, revenge, self‑assertion that they must normally choke back when in the presence of the masters and mistresses" (p. 18).  It is Scott's aim to shed light on this hidden transcript, with the idea that an understanding of this level of consciousness of the dominated is much closer to the reality of their lived experience and provides a better basis for understanding their political behavior.

            How do the constraints and impulses of ethnic affinity affect the agent?  Pierre Bourdieu's Outline of a Theory of Practice (1977) provides a fruitful framework in terms of which to think of socially constrained action.  Bourdieu challenges the idea that much human behavior is rule-governed (e.g. by a system of normative restrictions on marriage).  Instead, Bourdieu emphasizes the centrality of practice—practical, calculating action within the context of social, cultural, and material constraints.  As the model of human social behavior, he offers the analogy of an extended interaction between two intelligent actors (for example, two boxers jockeying for advantage within the context of the rules of a prize fight) in place of the idea of an authoritative system of rules (1977, p. 11).  Bourdieu's conception of the habitus provides a resource in terms of which to understand ethnic political behavior.  Bourdieu defines the habitus as "the durably installed generative principle of regulated improvisations" (Bourdieu 1977. p. 78).  This conception emphasizes both the cultural constraints and impulses, on the one hand, and the strategic improvisations of the agent, on the other.  Like Bourdieu in his analysis of individual behavior within the context of Berber kinship relations, then, we may attempt to understand the outcomes of ethnic politics as the aggregate effect of purposive individuals strategically interacting within a system of norms and expectations that they share.  This position accepts, then, that individuals come to identify with ethnically defined groups, and these identities both shape the perceptions of their members and provide motivational impulses that can be mobilized by ethnic politicians.  But individuals are no less agents because they possess ethnic identities.  And this means that they have their own independent purposes and goals that they pursue through the resources and opportunities available to them.  They make calculating decisions; they respond selectively to opportunities and risks in their environments; they pursue strategies designed to bring about the outcomes they care about; and so on.  This means, in turn, that the machinery of rational choice theory is applicable in this range of phenomena as well; but the set of purposes and beliefs that we must attribute to individuals require thicker description than is characteristic of rational-choice accounts. 

The aggregation of ethnic politics

            The considerations discussed in the previous section do not resolve the problem of ethnic politics.  For it is no simpler to construct a theory of collective behavior on the basis of this theory of individual agency than it is in the case of the interest-based approach.  The processes of aggregation through which individual actions come together into coordinated collective action are not transparent; the fact that many Hindus share a powerful affinity with a nationalist conception of their identity does not guarantee that there will be largescale collective actions in support of that identity.  Rather, it is necessary that this individual-level identity be mobilized through some aggregative process—organizations, communication, and leadership—if there is to be collective action.  And at this point the discussion of the need for a theory of leadership and organization within Marxist political theory is equally pertinent here; presumably the same processes of organization and leadership are the instruments through which individual-level identities are transformed into largescale collective movements.

            At this point, then, we need to shift our focus from the states of mind of typical potential followers to the interests, motives, and identities of leaders.  Given that it is possible to mobilize large numbers of people around an ethnic cause, what would lead a political leader to make a sustained and organized effort to do so?  And there appear to be two general possible answers to this question.  First, the leader too may possess the strands of ethnic affinity that we postulated for followers above.  Leaders and followers share a vision of what is desirable, and leaders undertake the task of organizing and mobilizing collective action in pursuit of this vision.  The other possible answer interprets leaders' behavior as politically rational and opportunistic.  It regards leaders as having material political interests—winning elections, advancing the influence of the party, reducing the influence of other parties, retaining office, keeping effective control of the army or police—and making appeal to ethnic identities as just one out of a spectrum of mobilization appeals.  On the approach, the problem confronting the leader is to mobilize support for his political organization and agenda.  There are various ways of accomplishing this goal: appealing to the material interests of potential supporters, appealing to regional or sectoral alliances, or appealing to religious or ethnic identities.  And the choice of appealing to ethnic identities is explained as the consequence of a calculation on the part of the leader that this appeal will be more successful than other possible mobilization appeals in furthering his political interests.[26]

            Many observers have argued that numerous contemporary examples of ethnic conflict derive from the latter process.  Thus Paul Brass argues that much Hindu-Muslim conflict in India in the 1980s reflected the effort by Muslim and Hindu politicians to further their own political interests (Brass 1989).  And Brass emphasizes the centrality of the state in creating politically salient ethnic identities (Brass 1990).  Atul Kohli argues that the increase in inter-caste conflict in India reflects a political agenda of various parties whose central concerns are material conflicts over a system of domination and privilege (Kohli 1990, pp. 192-96).

            On this approach, then, both ethnic affinities and rational calculations are at work in typical ethnic conflicts.  At the level of leadership and organization we may argue that agents are working within a rational framework of political calculation: clearly identified political ends and rational assessment of the means available to bring those ends about.  From this perspective, the temple is a convenient instrumentality through which to mobilize support, raise funds, and keep the organization's agenda in the public view.  When we turn to the rank and file followers, by contrast, we may wish to turn to a psychological theory of mobilization, according to which individuals are construed to act out of a set of attachments and commitments of varying strengths.

Conclusion

            The three frameworks discussed here emphasize different aspects of human agency and collective action.  The rational-choice framework emphasizes the strategic rationality of agents.  It offers a fairly thin account of the individual's goals and purposes and attempts to show that various social outcomes are the equilibrium result of an n-person strategic interaction.  The class-conflict theory accepts the idea that individuals have material interests; but it adds the observation that there are fundamental conflicts of interests that are defined by the lineaments of the property system; with the inference that groups who share a set of material interests will tend to come together in collective action in support of their interests.  This approach provides at least one important correction to the rational-choice approach: it draws attention to the critical importance of the particular institutional arrangements through which social and economic life is structured.  And identity-based theories emphasize the importance of group affinities as a basis of collective action.  The upshot of this view in relation to interest-based theories may be seen in a more or less radical way.  Most extremely, the identity-based theory might be thought to imply that many instances of collective action are not deliberative or rational at all; they spring instead from non-rational impulses of group loyalty and affinity.  The Scott-Bourdieu line offers a less radical implication, however: not that ethnic politics do not reflect deliberation and rational choice, but rather that the agent's purposes are often symbolic as well as material.  On this approach, we can interpret the participants as pursuing rational strategies to bring about ends whose importance is symbolic rather than material.

            We may draw several tentative conclusions on the microfoundations that are needed in support of adequate explanations of popular politics.  Most importantly, an adequate theory of popular politics must provide an account of the local processes through which group identity is formed and through which members of groups come to identify themselves as political actors.  Agents act deliberately, on the basis of their own understandings of their interests, allegiances, rights, and the like; consequently, in order to explain the political behavior of a group it is necessary to give some account of the processes through which this particular group identity and set of collective goals have been formed.  It is essential to analyze the processes of political identity-formation through which a group is transformed into a political agent with shared goals, beliefs, values, commitments, and plans.  And these processes are typically local if spontaneous, and dependent on competent organizations and leadership if supra-local.  For spontaneous group identity, it would seem, emerges from contiguity and shared perceptions of the social world; it requires a common history of struggles, demands, successes, and failures.  If a geographically and socially dispersed group is to acquire such a self-conscious identity, it is difficult to see how this could occur without the deliberate efforts of a competent regional or national organization. It is necessary to formulate a diagnosis of the social world and a political program that will permit dispersed members of such a group to come to regard themselves as part of a meaningful political agent; and this vision must be communicated to members of the group through competent local cadres.  All of this requires organization, however; and without such, it is unlikely in the extreme that class consciousness on a national scale would emerge spontaneously.

References

Arrigo, Linda Gail.  1986.  "Landownership Concentration in China: The Buck Survey Revisited."  Modern China 12 (3) : 259-360.

Axelrod, Robert. 1984.  The Evolution of Cooperation.    New York: Basic.

Bates, Robert H. 1981.  Markets and States in Tropical Africa.    Berkeley: University of California Press.

Bianco, Lucien.  1986.  "Peasant Movements."  In The Cambridge History of China volume 13.  Edited by J. K. Fairbank and A. Feuerwerker.  Cambridge: Cambridge University Press.

Bourdieu, Pierre. 1977.  Outline of a Theory of Practice.  Cambridge: Cambridge University Press.

Brass, Paul.  1989.  "Discourses of Ethnicity, Communalism, and Violence in India."  Seminar Paper, Center for International Affairs, Harvard University.

Brass, Paul R. 1990.  The Politics of India Since Independence.  Vol. IV-1.  The New Cambridge History of India. Cambridge: Cambridge University Press.

Brenner, Robert.  1976.  "Agrarian Class Structure and Economic Development in Pre-Industrial Europe."  Past and Present 70 : 30-75.

Buchanan, Allen. 1982.  Marx and Justice.  Totowa, NJ: Littlefield, Adams.

Chen, Yung-fa. 1986.  Making Revolution: The Communist Movement in Eastern and Central China, 1937-1945.  Berkeley: University of California Press.

Chesneaux, Jean. 1973.  Peasant Revolts in China 1840-1949.  New York: Norton.

Cohen, G. A. 1978.  Karl Marx's Theory of History: A Defence.  Princeton: Princeton University Press.

Croix, Geoffrey De Ste.  1984.  "Class in Marx's Conception of History."  New Left Review  146.

Duara, Prasenjit. 1988.  Culture, Power and the State: Rural North China, 1900-1942.  Stanford: Stanford University Press.

Elster, Jon. 1985.  Making Sense of Marx.  Cambridge: Cambridge University Press.

Elster, Jon.  1986.  "Three challenges to class."  In Analytical Marxism.  See J. Roemer. 

Elvin, Mark. 1973.  The Pattern of the Chinese Past.  Stanford: Stanford University Press.

Feuerwerker, Albert. 1975.  Rebellion in Nineteenth-Century China.  Ann Arbor: Center for Chinese Studies, The University of Michigan.

Gellner, Ernest. 1983.  Nations and Nationalism.  Ithaca: Cornell University Press.

Guha, Ranajit, and Gayatri Chakravorty Spivak, ed.  1988.  Selected Subaltern Studies.  New York: Oxford University Press.

Hilton, Rodney. 1973.  Bondmen Made Free.  London: Maurice Temple Smith Ltd.

Hookway, Christopher, and Philip Pettit, ed.  1978.  Action and Interpretation: Studies in the Philosophy of the Social Sciences.  Cambridge: Cambridge University Press.

Horowitz, Donald L. 1985.  Ethnic Groups in Conflict.  Berkeley, California: University of California Press.

Huang, Philip C. C. 1985.  The Peasant Economy and Social Change in North China.  Stanford: Stanford University Press.

Huang, Philip C. C.  1990.  The Peasant Family and Rural Development in the Yangzi Delta, 1350-1988.  Stanford: Stanford University Press.

Johnson, Chalmers A. 1962.  Peasant Nationalism and Communist Power The Emergence of Revolutionary China 1937-1945.  Stanford: Stanford University Press.

Kohli, Atuhl. 1990.  Democracy and Discontent: India's Growing Crisis of Governability.  Cambridge: Cambridge University Press.

Kuhn, Philip A. 1980 [1970].  Rebellion and Its Enemies in Late Imperial China Militarization and Social Structure, 1796-1864.  paperback ed.  Cambridge: Harvard University Press.

Levi, Isaac. 1986.  Hard Choices: Decision Making Under Unresolved Conflict.  Cambridge: Cambridge University Press.

Levi, Margaret. 1988.  Of Rule and Revenue.    Berkeley: University of California.

Lewis, John Wilson, ed.  1974.  Peasant Rebellion and Communist Revolution in Asia.  Stanford: Stanford University Press.

Little, Daniel. 1989.  Understanding Peasant China: Case Studies in the Philosophy of Social Science.  New Haven: Yale University Press.

Little, Daniel.  1991.  Varieties of Social Explanation: An Introduction to the Philosophy of Social Science.  Boulder, Colorado: Westview Press.

Margolis, Howard. 1982.  Selfishness, Rationality, & Altruism A Theory of Social Choice.  Chicago : University of Chicago.

Marks, Robert. 1984.  Rural Revolution in South China: Peasants and the Making of History in Haifeng County 1570-1930.  Madison: University of Wisconsin Press.

Marx, Karl. 1977 [1867].  Capital.  Vol. 1.  New York : Vintage.

Nash, Manning. 1989.  The Cauldron of Ethnicity in the Modern World.  Chicago: University of Chicago Press.

North, Douglas C., and Robert Paul Thomas. 1973.  The Rise of the Western World: A New Economic History.    Cambridge: Cambridge University Press.

Ostrom, Elinor. 1990.  Governing the Commons: The Evolution of Institutions for Collective Action.    Cambridge: Cambridge University Press.

Paige, Jeffrey. 1975.  Agrarian Revolution.  New York: Free Press.

Perdue, Peter. 1987.  Exhausting the Earth: State and Peasant in Hunan, 1500-1850.  Cambridge: Harvard University Press.

Perry, Elizabeth J. 1980.  Rebels and Revolutionaries in North China 1845-1945.  Stanford: Stanford University Press.

Popkin, Samuel L. 1979.  The Rational Peasant.  Berkeley: University of California.

Przeworski, Adam.  1985.  "Marxism and Rational Choice."  Politics & Society 14 (4) : 379-409.

Roemer, John. 1982.  A General Theory of Exploitation and Class.  Cambridge: Harvard University Press.

Sabel, Charles F. 1982.  Work and Politics.  Cambridge: Cambridge University Press.

Schelling, Thomas C. 1978.  Micromotives and Macrobehavior.  New York : Norton.

Schultz, Theodore W. 1964.  Transforming Traditional Agriculture.  New Haven: Yale University Press.

Scott, James C. 1976.  The Moral Economy of the Peasant.  New Haven: Yale University Press.

Scott, James C. 1985.  Weapons of the Weak: Everyday Forms of Peasant Resistance.  New Haven: Yale University Press.

Scott, James C. 1990.  Domination and the Arts of Resistance: Hidden Transcripts.  New Haven: Yale University Press.

Selden, Mark. 1971.  The Yenan Way in Revolutionary China.  Cambridge: Harvard University Press.

Sen, Amartya. 1987.  On Ethics & Economics.  New York : Basil Blackwell.

Sen, A. K.  1982.  "Rational Fools."  In Choice, Welfare and Measurement.  See A. Sen. 

Sen, A. K.  1982b.  Choice, Welfare and Measurement.  Cambridge: MIT Press.

Shaw, William.  1984.  "Marxism, revolution and rationality."  In After Marx.  See T. Ball and J. Farr. 

Skinner, G. William.  1964-65.  "Marketing and Social Structure in Rural China."  Journal of Asian Studies  24 (1-3).

Skinner, G. William.  1977a.  "Regional Urbanization in Nineteenth-Century China."  In The City in Late Imperial China.  See G. W. Skinner. 

Skinner, G. William.  1977b.  "Cities and the Hierarchy of Local Systems".  In The City in Late Imperial China.  See G. W. Skinner. 

Skinner, G. William, ed.  1977c.  The City in Late Imperial China.  Stanford: Stanford University Press.

Skocpol, Theda. 1979.  States and Social Revolutions.  Cambridge: Cambridge University Press.

Tai, Hue-Tam Ho. 1983.  Millenarianism and Peasant Politics in Vietnam.  Cambridge: Harvard University Press.

Tambiah, Stanley.  1990.  "Reflections on Communal Violence in South Asia".  Journal of Asian Studies 49 (4) : 741-760.

Tambiah, Stanley J. 1991.  Sri Lanka: Ethnic Fratricide and the Dismantling of Democracy.  Chicago: University of Chicago Press.

Taylor, Michael. 1982.  Community, Anarchy and Liberty .    Cambridge: Cambridge University Press.

Taylor, Michael, ed.  1988.  Rationality and Revolution.  Cambridge: Cambridge University Press.

Tilly, Charles, Louise Tilly, and Richard Tilly. 1975.  The Rebellious Century 1830-1930.  Cambridge: Harvard University Press.

Vlastos, Stephen. 1986.  Peasant Protests and Uprisings in Tokugawa Japan.  Berkeley: University of California Press.

Wright, Erik Olin. 1978.  Class, Crisis and the State.  London: Verso.

Wright, Georg Henrik Von. 1971.  Explanation and Understanding.  Ithaca: Cornell University Press.