About five years ago a group of American sociologists formed a group they called MESO, endowing it with its literal meaning, middle. The group is an informal one; it does not publish a journal. By now it has about two hundred members, and meets once a year for presentation and discussion of papers. It grew out of a dissatisfaction among a number of students of formal organization with the micro-macro distinction, a distinction that gained currency in the 1980s (see Collins 1981; Alexander et al. 1987). They felt that that distinction distorted their world of study—the middle—and that the middle constitutes a crucial link between the psychological and the societal. The focus of the group is—though not in a very deliberate or coherent way—on the meso-level phenomena identified at the beginning of the first chapter: groups, formal organizations, social movements, anti some aspects of institutions.

Of the four levels that constitute my subject matter, the meso-level is the most vague. It seems most helpful to delineate it by instances rather than by formal definition. It concerns what Tocqueville ([1835] 1945) referred to as "associations"; it includes that level of society identified by mass society theorists (Kornhauser 1959) as "intermediary"— community life, voluntary associations, trade unions, and political parties; it overlaps with what political scientists and others refer to as civil society, that complex of political groups and institutions that mediate between the citizenry (micro) and the polity (macro) (Putnam 1993). Because of this vagueness of reference of "meso," a few clarifying theoretical remarks are in order at the outset.

Analytic Levels and the Problem of Reduction

Early in the last chapter I mentioned the danger of reifying the organizational principles on which these essays are based. Even though the micro, meso, macro, and global levels can be identified, it must be remembered that in any kind of social organization we can observe an interpenetration of these analytic levels. This can be illustrated by reference to a "meso-level" structure, a bureaucratic organization. Evidently such an organization lies "between" interacting individuals anti larger societal structures. Any bureaucratic organization is populated by individuals (micro level) and is regulated by laws and other normative systems—for example, laws of charter and incorporation, standards of accountability—and legitimated by at least implicit reference to cultural standards and values (macro level).

Despite the soundness of this observation, we social scientists appear to be programmed with a certain bias when relating different analytic levels to one another: the bias of methodological individualism. We live in the Western cultural tradition, which has exploited the cultural values of individualism. As children of that tradition, we are most comfortable taking the individual person as the starting point of analysis. Put another way, that cultural tradition "tilts" us toward assuming that the natural unit for the behavioral and social sciences is the individual. The same tilt informally discourages the recognition of other levels of social organization as equally natural. There is reason to believe, however, that other levels of reality are analytically as important as—more important for some purposes—the person. I now venture a few observations on this score.

For those lines of inquiry rooted most firmly in the individualist tradition—I have in mind psychology (the study of the person) and the Anglo-American discipline of economics—the individual is the basic unit of analysis, and the movement to higher levels of organization is frequently a matter of aggregation of individuals. That has been the main mode of transition between microeconomics and macro-economics, with markets and whole economies (e.g., gross domestic product) being treated as summations of thousands or millions of individual transactions. In survey research we add and percentage individual responses to survey questions and imagine that we have measured "public opinion." In a social psychological expression of this principle of methodological individualism, Floyd Allport (1924) argued that the crowd mentality and crowd behavior are nothing more than the aggregation of individual characteristics. Interestingly, Simmel flirted with the same notion in one of his definitions of society. Society, he said, in "only the name of the sum of [social] interactions. . . . It is therefore not a unified, fixed concept by rather a gradual one, . . . a constellation of individuals" (quoted in Frisby 1990: 17). This observation did not exhaust Simmel's treatment of society, but it is a vivid statement of the logic that the whole is the sum of its parts.

Another intellectual strand in the sociological tradition has attempted to establish the analytic value of more comprehensive levels of reality. One crude attempt was that of Gustav Le Bon, who asserted that the crowd exhibits a qualitatively new mentality from that of its individual members. The notion of a higher sociological reality is also at the heart of Durkheim's Rules ([1895] I958), which constituted simultaneously a claim that there exists, sui generis, a supraindividual society with distinctive characteristics and a claim on behalf of sociology as the science of that society. Simmel also developed a version of this appeal. In his identification of prototypical sociological forms— for example, dominance or competition—he argued that such forms were analytically independent of both the psychological characteristics of individuals involved and their cultural context. Finally, in a formulation influenced by Simmel's idea of form (Kaern 1990), Weber attempted to build a supraindividual level of reality in his conception of "ideal types," though he regarded these as abstractions from the individual meaning-experiences of actors. All of these supraindividual formulations have enjoyed only fragile, impermanent status, and all have been subject to various forms of criticism, stemming, I submit, from the fundamental preference for methodological individualism in our traditions of social thought. Accordingly, the temptation to fall into psychological—or rather, individual—reductionism is alive and well in the social sciences.

My own effort to resolve this problematic has always been to insist on the conceptual validity of higher levels of formulation—interaction, group, organization, institution, society, even multisociety—not on grounds of any absolute philosophical claims to reality but on essentially programmatic grounds: it is impossible to understand and explain events, situations, and processes of "lower" units without appealing to some higher order of organization by which they are constrained. Physics requires its chemistry, chemistry its biochemistry, biochemistry its biological organism, biological organism its integrative mental processes, and individuals their social organization, if we are to proceed beyond atomistic characterizations and understand more complex behaviors and sequences. This acknowledgment does not call for any special assertions about reality, but rests on the need for higher-level organizing constructs necessary for comprehensive explanations.

Mechanisms Linking the Individual and Meso-Level Structures

One argument for focusing on the meso level is that structures at that level constitute the primary bases for organizing the routines, interactions, and affective linkages of individuals' daily lives. As individuals we connect daily with the larger society via the groups, organizations (places of employment, unions, churches), associations, and social movements of which we are members. This range of life is what Simmel

(1965) had in mind in his concepts "circles" and "web of group affiliations." Through these linkages social life becomes real to the individual, certainly more real than his or her relationship with institutions, systems of institutions, and social classes, to say nothing of the state, the society, and the international order. This point makes general the assertion, familiar to political scientists, that the success of political democracy depends as much on the specifics of civil society (that network of intermediate, or meso, organizations lodged between individual and polity) as it does on the formal institutions of the polity. It is in the more intimate structures that the civic culture is learned and given vitality.

This observation leads immediately to the question of the mechanisms that bind individuals to the groups, organizations, and associations. Why do they attach to them? This is simultaneously a question of motivation and a question of incentives, or, in a phrase familiar to the sociologist, a question of socially structured motivation: learned and normatively articulated orientations of individuals toward their group anti organizational environment.

In keeping with our inherited individualist-utilitarian frame, it seems—but only seems—easier to think about this problem in some contexts more than others. If we ask, for example, what ties individuals (workers) to organizations that employ them, we typically turn to the following kind of explanation: the employer offers wage payments to individuals, who, in return, provide labor and cede to the employer a measure of control over their time and independence. Whether this constitutes a mutually beneficial contractual agreement (as in classical political economy) or an instance of exploitation (as in Marxian economics) seems secondary. The actual mechanism is identical, whatever the interpretation assigned. Furthermore, we are comfortable with the idea that reference to these mechanisms constitutes a sufficient account of the motivation-and-incentive situation at hand. I would suggest, however, that the reason we are comfortable with such an account is that this interplay of motivation and incentives is deeply institutionalized in the money-market complex of contemporary society.

Furthermore, so embedded is this complex of institutions and assumptions that it sometimes creates intellectual puzzles that are analytically unnecessary. Consider, for example, the issue of why people join social movements. We will observe later that in the study of such movements many "reasons for joining"—imitation, contagion, suggestibility, ideological commitment, expressive gratification, the need for solidarity, among others—have been generated. If, however, we approach the problem within the individualistic-utilitarian perspective, we are likely to generate unwanted paradoxes and unnecessary resolutions of those paradoxes because—within that framework —individuais are seen as having no reason to join social movements since a cost-benefit analysis yields no plausible motive to participate. It is only in this context—not as a completely general matter—that the "free rider" and related paradoxes arise, anti induce scholars operating within the individualistic-utilitarian tradition (e.g., Olson [1965], Oberschall [1973], and Coleman [1990]) to generate complex cost-benefit schemes to account for why individuals affiliate with and participate in social movements and social movement organizations.

It might prove worthwhile to break from the individualistic-instrumental-rational set of assumptions that generate that statement of the problem and its attempted solutions. We might then turn to a different definition of the situation, to identify other socially structured types of motivation and treat these as equally valid bases for generating models linking individuals to meso-level structures.

One starting point would be to revive earlier efforts by Talcott Parsons and his associates (e.g., Parsons and Smelser 1956; Parsons 1963a, 1963b, 1968) to identify "generalized media," of which the main types are money, power, influence, and value-commitments. Parsons treated these media mainly as mechanisms that facilitated exchange and equilibrium at the social system level. They can, however, also be regarded as socially structured motivational complexes that form the basis for individuals' affiliation with groups, organizations, and movements. They constitute simultaneously types of motivations and types of incentives or rewards. For understandable reasons, Parsons's formulation of wealth and power attained greater analytic clarity than did the treatment of influence and value-commitments. It is clear, however, that influence is a motivation that combines sociability, expressive gratification, and appeal to common membership; and that value-commitment is a motivation that links the continuity of individual identity, attitudes, and values to cultural patterns. These constitute motivational-incentive complexes for linking with social organizations as much as wealth and power do. Furthermore, there is no reason for sacrificing analytic power in appealing to them. By positing models based on these complexes, it remains possible to generate rigorous explanations of individual participation in the more expressive and affective sides of social life.

Meso- Level Structures

In the remainder of the chapter I will touch on the four sets of structures identified as "meso"—groups, formal organizations, social movements, and institutions—and point to a number of problematics for each.

Groups

Groups, especially expressive groups, have become something of a casualty in the recent history of sociology, despite their notable place in both sociology and social psychology. Simmel's pioneering work on group size and group process (Wolff 1950: 87-177) is itself a notable part of that tradition.

The golden age of the group was the 1940s and 1950s, when the informal group was recognized as a salient force in industry (Roethlisberger and Dickson 1939), in the military (Shils and Janowitz 1948), in community action (Lewin 1948), in market and voting behavior (Katz and Lazarsfeld 1955), in disaster behavior (Killian 1952), and in sociological theory generally (Homans 1951). This concern has, by and large, fallen by the wayside. The efflorescence of innovative experimental studies on group interaction and process (e.g., Bales 1950; Leavitt 1951) has likewise subsided. Similarly, the family as site of group process has given way to the themes of family as institutional victim (in the family literature) and family as vehicle for dominance and subordination (in the feminist literature).

This is not the moment to develop a sociology of knowledge about this decline, but the following factors may be mentioned.

•     The surge of macrosociological interest (mainly neo-Marxian, neocritical, and neo-Weberian) in the 1960s and 1970s, with its attendant focus on macro-level domination.

•     The failure of the "microsociological revolution" (symbolic interactionism, ethnomethodology, phenomenology generally) in the 1970s and agency theory subsequently to move beyond the person and personal interaction and to revitalize an interest in group processes as such.

•     The "system" focus that dominates feminist sociology and race relations research.

•     The institutional—not the group—focus of the "new institutionalisms" in economics, sociology, political science, and history.

The group does survive as a tradition in experimental social psychology, and Habermas's theoretical interest in the life-world processes reminds us of the importance of the face-to-face group, but his interest has been in group interaction as a counterforce to the colonizing tendencies of the market-state-bureaucratic apparatus, not in group life as such.

One additional factor might be evoked in explaining the decline of interest in groups, and this factor points toward a new line of inquiry. I suggest that, because of accelerating economic and social developments in the world, the group as it was conceptualized in the "golden age"—a relatively stable, enduring, face-to-face, cooperative unit— has, as a matter of institutional fact, receded dramatically by the end of the twentieth century. The developments that have occasioned this recession, moreover, are those that Simmel originally identified in his imaginative depiction of the metropolis—the enhanced "individuation of the individual" through social and economic mobility, fleeting contacts dictated by the pace of life, segregation of social circles through further differentiation, and the resulting experiences of isolation, freedom, and a blasé mentality. All these have eclipsed the primary group as we knew it. And that is one of the major reasons we pay less attention to it.

That is not the final answer, however. The group remains important in contemporary society, but operates according to a different principle. The image I have in mind is the "fission-fusion" principle that has been identified as a principle of bonding in the group life of seagoing mammals and primates (Wrangham and Smuts 1980; Symington 1990; Smolker et al. 1992). That principle involves the frequent coming together of social groups in apparently meaningful form for the animals, but it is accompanied by an equally apparent tendency for these groups to dissolve—or dissolve partially—only to re-form in new but also impermanent combinations. This fission-fusion principle has accelerated dramatically in human life as well at the end of the twentieth century. The instability of group life—in the workplace, on the street corner, in the office, and in the family—is now more the rule, and stability is more the exception.

To acknowledge this is not to say that groups have receded in importance in the human condition. They remain central and crucial. They still express the fundamental—perhaps genetically fixed—tendencies to bond with others and to be socially dependent in the human condition. These tendencies need now to be studied more intensively, however, in their relatively fleeting—rather than permanent—form.

New understandings and new models of this accelerated fission-fusion principle are required. In rising to this new kind of understanding, we also need to understand whether—and if so, in what ways—the human predisposition for bonding is being taxed to the point of generating serious social costs. In the following chapters I will underscore the increased salience and assertiveness of subnational groupings based on race, ethnicity, region, language, and other "local" forces. It may well turn out that these tendencies are expressions, in part, of a reassertive reaction to the erosion of human bonding occasioned by the acceleration of the fission-fusion principle in contemporary social life.

Formal Organizations

Max Weber established the principle that formal organizations are the structural signature of the rise of industrial capitalism. Weber also affirmed that bureaucracies were not a child exclusively of capitalism, but predicted, correctly, that socialism would only further the march of bureaucracy, largely, we suspect, because of the premium that socialism gives to government ownership and management. We might even extend Weber. Whatever the transition from industrial to post-industrial might mean, it certainly has brought the further consolidation of formal organizations in the lives of individuals and societies. If anything, the past decades have seen the transition from discrete organizations to multiorganizational systems —expanded civil service bureaucracies, multinationally coordinated corporations, multicampus universities, and ecumenical formations of churches.

In the scholarly work on organizations we discern three recurrent and overlapping problematics: organizations as efficient or inefficient, organizations as adaptive or maladaptive, and organizations as closed or open systems.

•     Two major traditions of organizational study—the theory of the firm in economics and Weber's theory of bureaucracy—established the notion that bureaucracies are efficient, though the logic leading to that conclusion was different in each tradition. In classical economics, the firm was a kind of black box, essentially without internal organization, that responded rationally to the markets for the factors of production on the one side and the markets for the firm's products on the other, producing equilibrium market solutions in the process. Weber turned to the internal organization of bureaucracies—hierarchy, authority through rules, division of labor, and written procedures—to locate their comparative advantage over staff organizations based on charisma and tradition. Two subsequent lines of inquiry have challenged the efficiency assumption—the insight that informal groups can systematically undermine the formal purposes of organizations and the long-standing popular and scholarly appreciation of the debilitating power of bureaucratic encumbrances such as red tape and procedures for procedures' sake (Parkinson 1957).

•     Most traditions in organizational sociology treat the organization as adaptive, or as at least striving to be adaptive. However, the field has proceeded beyond earlier assumptions, built into both the classical economic and the Weberian traditions, to an extensive literature that takes adaptation as problematical, and considers conditions such as information, technology, competitive environment, organizational culture, age and size of organization, and internal structure as determinants of adaptation or mal-adaptation (Aldrich and Marsden 1988). One notable model is the "garbage can model of organization choice" (Cohen, March, and Olsen 1972) that treats decision making as calling up strategies selectively from a loosely organized reservoir of criteria and possibilities. (The model foreshadows Swidler's [1986] "reservoir" theory of culture.) I call attention to this model because, although it lies in the "adaptive" tradition, it breaks from the dominant assumptions of rationality that have dominated organization theory. The "garbage can" or "reservoir" models may be appropriately criticized on one set of grounds: although eminently realistic as an account of decision makers' activities in organizations—for what are decision makers if not people who try to make use of all resources, strategies, and tactics that they believe to be at their disposal?—the model is also highly indeterminate because it incorporates only "flexibility" as preference function. It seems an appropriate strategy for theorists of organization to build such flexibility into their psychological assumptions about decision making, but they should also take the opportunity to generate submodels for the organizational and environmental conditions under which different strategies might be selected.

•     The earlier "closed systems" approach to organizations, also associated with the classical schools, gave way in midcentury to a stress on "natural systems." By now most research assumes that formal organizations are implicated in complex environments composed of differentially available technology, different qualities of information, other organizations, and legal and regulative systems.

I mention the dimension of "openness" because two of the most important recent trends in the organizational literature focus on the institutional, competitive, and technological environments of organizations. The first goes by the name of "the new institutionalism" in sociology (Powell and DiMaggio 1991), which has revived and modified the notion of the penetration and reproduction of institutional forces in organizations. Its major focus is on the cultural routines and scripts that are invoked as orienting symbols, constraints on choice and rationality, and stabilizing forces. The second approach involves the idea that new competitive and technological forces—especially in the global setting—are pushing toward radically different forms of organization to such a degree that formal organizations—like primary groups—are undergoing such fundamental changes that they demand completely new loci of analysis. Transaction cost analysts have raised the question whether hierarchy (authority) in organizations— the hallmark of classical Weberian theory—is not in many instances too costly a structure (Williamson 1985). There is also a small, enthusiastic literature on new organizational forms that has produced a flurry of catchwords—"flexible specialization" (Piore and Sabel 1984), networks, self-managed teams, "adhocracy," franchise organizations, consortia, partnerships, and even "virtual organizations"—all of which suggest that loose cooperative, informal, continuously recreated organizations are coming to replace authority-based organizations with a specialized, detailed, and fixed division of labor (Fordism). One influential model to have emerged from this line of thinking is the "contingency model," which also is meant to describe the weakening of hierarchy, authority, anti specified rules and procedures: "Rapidly changing environments and uncertain technology, such as characterized the electronics industry, . . . appeared to produce organizations with adaptive, free-flowing, 'organic' structures. The organic structure emphasized employee interactions, horizontal as well as vertical communication, and greater professional autonomy in which employees 'discovered' rather than were assigned to their jobs" (Dill and Sporn 1996, after Lawrence and Lorsch 1986). While most of these new trends—and the literature that describes them—refer to the corporate world, some analysts believe they constitute a model for such unlikely candidates as university organizations in the postindustrial world (I)ill and Sporn 1996).

I believe we should not be swept away by either corporate or scholarly enthusiasts who believe that the days of organizational hierarchy arc numbered and that in the interests of efficiency, the infrastructure of economic and other organizational life will be supplanted by a mix of market, monitoring, network, coordinating, and individual selfregulating mechanisms. However, it seems clear that we may expect a major reconceptualization of received notions of division of labor, hierarchy, commitment, and incentives in light of ongoing changes in organizations in postindustrial society.

Social Movements

Social movements lie at the meso level of social organization because they are phenomena to which individuals forge direct ties as participants, in which they interact directly with others, through which they seek to realize their collective aims and effect changes in their social environment, and in which, as meaningful points of social reference, they often find personal identities as well as day-by-day rhythms in their lives.

It is instructive to call to mind some features of the history of the sociological study of social movements. In the nineteenth century macrosociological theorists—notably Marx and Tocqueville—recognized revolutionary movements as an integral part of convulsive historical change. But the social psychology and sociology of social movements began properly with the work of Gustav Le Bon, the French journalist, toward the end of the nineteenth century. Le Bon's ([1895] 1952) analysis was irrationalist in the extreme, treating crowds as unreasoning, impulsive, emotional, swayed by suggestion and demagoguery, dissolving individuals' self-control, and capable of the most extreme destructiveness and idealism. Furthermore, he abhorred the crowd as destructive of institutions, and attributed the rise of the "era of crowds" to a pathology unleashed by the decay of traditional feudal and religious institutions.

Le Bon's social psychology dominated the field for some decades, providing the major underpinnings of the psychological theories of Sigmund Freud ([1922] 1955), William MacDougall (1920), and the American sociologist E. A. Ross (1916). By the middle of the twentieth century this irrationalist and negative assessment of collective behavior was attenuated, as these phenomena became the object of what may be called "naturalistic" inquiry, that is, as the object of scientific inquiry and "to be explained" as a matter of scientific interest. Investigators like Herbert Blumer (1951) dealt primarily with the mechanisms involved (e.g., milling), Ralph Turner and Lewis Killian (1957) considered processes by which groups came to define their situation and develop normative understandings, and my own work (Smelser 1962) dealt with the ideologies guiding collective behavior and social movements and identified a diversity of social conditions—including macrosociological structures—that operate as determinants in the development of social movements. In this process the dominant imagery of collective behavior and social movements as "irrational" and "threatening" receded in favor of a certain attitude of dispassion.

In the 1960s, a decade notable for the proliferation of social mover ments (the civil rights, student, antiwar, feminist, and countercultural movements, for example), the literature on social movements took a dramatic turn. For one thing, advocates of and sympathizers with those movements were among those who contributed to the literature, and they understandably regarded them as setting the world right and thus as far from irrational and threatening. One extreme statement (Skolnick 1969) treated the movements of the day as fundamentally rational, that is, as containing a correct diagnosis of the ills of contemporary society, and treated authorities and others who opposed the movements as irrational in their defense of a corrupt and unjust status quo. In accord with this orientation, contributors to this literature tended to regard all past theories of social movements as "irrationalist," conservative, and apologetic for one establishment or another.

Since the 1960s two main lines of analysis have come to dominate the study of social movements. The first is "resource mobilization theory." It crystallized in the work of scholars such as Meyer Zald and Roberta Asch (1966) and William Gamson (1975). Its basic theoretical thrust is that social movements are not to be explained by the recruitment of the alienated and the disaffected to irrational or nonrational "ideologies." In that respect the resource mobilization approach resembled the view that rose in the 1960s. Instead, social movements are better regarded as purposive, directed enterprises whose success or failure depends on their effectiveness in mobilizing resources (financial support, existing groups, and recognition by political parties, for example). It is apparent that this kind of interest in social movements marked a turn in a "rationalist" direction, even though resource mobilization analysts tend to maintain a neutral stance with respect to the larger social significance of social movements in society. This theoretical orientation also led to a focus on social movement organizations (SMOs), those organized groups that make it their business to mobilize resources on behalf of the movement. By this circumstance the study of social movements moved closer to—and, in a certain sense, became part of—the study of formal organizations, those special organizations dedicated to mobilizing resources, holding adherents' loyalties, and gaining political successes for the movement. That focus excited, in turn, the study of strategy, tactics, and decision making. That framework continues to dominate the literature on social movements, though it also has come in for its share of criticism for down-playing the ideological, social psychological, and cultural aspects of movements. A revived interest in the role of ideas and ideology has developed around the idea of "framing," or the active efforts on the part of social movement organizations and actors to produce and maintain "meaning for constituents, antagonists, and bystanders or observers" (Snow and Benford 1992).

The second development, largely European in origin and interest, is called "new social movements." Its starting point was the recognition by European intellectuals and social scientists that "old social movements"—working-class union and revolutionary movements understandable in the context of a Marxian worldview—were by and large spent, as was the Marxian analysis of society. The "new" social movements were not especially class based; included among them were regional, racial-ethnic, and language movements; antiwar and antinuclear movements; the feminist movement; and various countercultural and lifestyle movements. Most interpreters of the new social movements retained a neo-Marxist or neocritical note in their explanations, however, in that they interpreted those movements as a kind of generalized protest against an oppressive capitalist-state-bureaucratic-technological-media complex in postmodern society. The new social movements impulse has diminished in the past decade, and although those who wrote in that tradition accurately described a historical change in the pattern of social movements, their literature can best be understood as a dialogue among those interested in the Marxist and critical traditions of sociology.

This brief review of the sociological study of social movements is of some interest in itself, but for present purposes, I include it because it throws light on three problematics in sociology as a discipline, as follows:

•     At the meso as well as the micro level the problem of assigning motives, reasons, and understandings to the people and the organizations we study is a recurrent methodological concern. Le Bon and those he influenced faced the concern directly, by endowing participants in social movements with irrationality, if not derangement; those who single out the alienated or the estranged as candidates for social movements also have a "theory" of why certain social circumstances predispose individuals to be attracted to ideologies of social movements; and even those resource mobilization theorists who tend to regard motivation as secondary have not been able to escape the issue of why people are predisposed to being mobilized. The problem of understanding "other minds" thus manifests itself at analytic levels higher than that of social psychology and social interaction.

•     The history of the study of social movements underscores a special vulnerability of sociology in general—how difficult it is, when studying a subject matter of charged moral and political significance, to maintain a posture of neutrality and dispassion toward it. Many of the scholars mentioned had no hesitation about evaluating their subject matter. Perhaps more significant, even when a scholar makes a good faith effort to remain neutral about dramatic and publicly controversial phenomena, others in subsequent generations will locate some bias—real or imagined—no matter how successful or unsuccessful that scholar was in his or her own scientific intentions.

•     This tradition of study also underscores the difficulty that social scientists have in coming to terms with the nonrational aspects of social life. They find them difficult to formalize theoretically, so they are forever being consigned to some kind of residual status. Or, alternatively, social scientists give in to the temptation to make rational that which, on its face, is not. I regard these tendencies as occupational hazards facing social scientists. After all, all of us are intellectuals and trained professionals, and the major institutional commitment in those universities and colleges in which we have been formed is still to the pursuit of the truth, which means the pursuit of the rational. Especially in the late twentieth century, when the nonrational impulses I have documented are in full sway, we are still prone to interpret the world in our own rationalist image. It would behoove us to engage in a campaign of self-examination to recognize and perhaps break ourselves of that tendency.

This last point leads me to identify a paradox in our contemporary situation as social scientists. It is evident that the resource mobilization and related approaches to the study of social movements are of a rationalist stripe (i.e., calculative, purposive, understandable-in-our-terms). They have more or less consigned the nonrational to a position of residuality or nonstudy, despite the minor comeback of interest in ideology in the resource mobilization literature. At the same time, the late twentieth century has produced a range of social movements—roughly speaking, those identified in the literature on the new social movements—that possess elements that are not readily understandable, or if understandable only by stretching, in terms of our dominant conceptions of rationality. The evidence of absolute ideologies, commitment without apparent calculation as well as primordial imagery and behavior, stands out in many social movements of our time. Does it not strike you as odd—as it strikes me—that we as social scientists interested in social movements should, in the late twentieth century, be so preoccupied with the rational aspects of social movements, precisely when the nonrational elements are so self-evident? I remind you that what I have just noted is a general problematic that has and will run through these essays: to take cognizance of the nonrational in social life, to recognize it as such, and to take it as deserving substantial attention in our enterprise.

Institutions

As we consider institutions, we begin to stretch the limits of the meso level. I regard institutions as lying at the core of social structure, and social structure belongs—as the next chapter will show—at the level of the macro. For that reason I will make only two observations about institutions, both of which touch the meso level, reserving the fuller discussion of social structure until later. By institutions I understand those complexes of roles, normative systems, and legitimizing values that constitute a functionally defined set of activities that gain permanence through the very processes of institutionalization. A concrete listing of institutions reveals a conventional inventory: family, education, religion, medicine, science, business, law, government, and others.

My first point is that institutions—structures at a general level of societal organization—are in large part "imagined," much as societies themselves are imagined communities (Anderson 1983). This means that they are not "seen" in any immediate sense, in the way that neighbors, policemen on the beat, the corner grocery store, and the local school are seen. At the same time, these institutions are "public," in the sense that they appear as nouns in language, and are spoken of as if they enjoy an empirical existence—as implied, for example, in the question, What is happening to education these days? This simultaneous invisibility ("imaginedness") and reality means that the agents or spokespersons who represent the institution assume a special sociological significance. With regard to the family as an institution, for example, these agents are vocal parents, psychologists and psychiatrists, educators, social workers, advocates for "family values," and others, including sociologists. They speak for, define, and represent the institution in the public and political arenas. These processes of representation are not well understood and merit understanding. They link the institution with the microworld of individual understanding and the macroworld of politics and public policy.

By the same token, individual persons do not interact with "institutions" per se, but with persons who represent the institution in day-by-day interaction. These representatives and their interactions shape institutions as well, but in ways different from those of the public spokespersons working on their behalf. These persons—lawyers, teachers, physicians, and so on—also hold the fate of the institution in their hands, because they are the ones who put forward the day-by-day presence of it and define, correct, modify, or reinforce the "folk" understandings of the institutions. This aspect of institution representing also deserves more systematic study.

Two Points in Conclusion

I have already moved into the supraindividual world in this chapter, and I will move even further in the chapters to come. This raises a long-standing question in sociology, that of the "group mind" or of "supraindividual" levels of reality. I do not wish to enter into all the ranges of controversies and misunderstandings that have surrounded these issues over time, but to make only one comment. At a certain point in the study of characteristics, attitudes, and behavior, we must turn to the involvement of individuals in higher levels of social organization—meso, macro, and global—that constitute a clear set of determinants. We may or may not want to describe these levels as real in some epistemological sense, but to proceed without taking into account the constraints of higher levels of social organization is to fail as sociologists.

On an entirely different note, let me suggest that in the contemporary world we face what might be described as a crisis at the meso level of social organization. Why should this be a crisis? On the one hand, the advanced nations of the world confront a situation in which the historically important meso levels of social integration—the immediate family, the extended family, the community, the neighborhood, the church, the tavern, the club, and, more recently, the political party—have declined and are continuing to decline in their significance as mechanisms of social integration. This decline is undeniable. At the same time, those social forms that might be regarded as taking their place—the fissions and fusions, the situationally based groups, the formal organizations, the social movements, for example—either have not done so or are in such a stage of transition that they cannot be considered to be adequate functional substitutes. We might then ask: Where is the meso in contemporary society, and where is it going?

That question looks both downward and upward. We know that mesostructures are important from the standpoint of the psychological continuity and identity of individual persons. What is the future of the person if we do not know the nature of the mesostructures in which the person is involved? Also, we know that the mesostructures—the heart and soul of our civil society—affect the character and effectiveness of the social integration of the larger society. To pose this question is not to answer it. But, speaking as a sociologist, I have to say that if we do not keep our eye on the meso level, we are likely to ignore the most problematic feature of society of the coming decades.