Introduction
Prophet: the one who speaks before
Movements in complex societies are disenchanted prophets. The charmed universe of the heroes has definitively dissolved under the impact of an era taking cognizance of itself as a planetary system, riven by molecular change, as a system which constantly generates tensions and then in turn adapts to them by striving to control them. Movements are a sign; they are / not merely an outcome of the crisis, the last throes of a passing society. / They signal a deep transformation in the logic and the processes that guide ; complex societies.
Like the prophets, the movements 'speak before': they announce what is taking shape even before its direction and content has become clear. The inertia of the old categories may prevent us from hearing the message and from deciding, consciously and responsibly, what action to take in light of it. Without the capacity of listening to these voices, new forms of power may thus coalesce, though multiple and diffuse and no longer reducible to any linear and easily recognizable geometry.
Contemporary movements are prophets of the present. What they / possess is not the force of the apparatus but the power of the word. They • announce the commencement of change; not, however, a change in the ' distant future but one that is already a presence. They force the power out into the open and give it a shape and a face. They speak a language that seems to be entirely their own, but they say something that transcends their particularity and speaks to us all.
This book was born over the last twenty years as an attempt to listen to the voices and read the signs of precisely that which collective action is proclaiming. But the mind that sets about to regard the societal actors today must in a similar manner proceed within a disenchanted framework. The
2 introduction
Introduction 3
intellectuals who claim to represent the good conscience or the true ideology of a movement have always participated in preparing the way for the advent of the Prince, only to end up as either his victims or his courtiers. The contemporary transformations of social actors parallelling the shift in the focus of conflicts and the changes in the forms of power have rendered the situation even more problematic. Both passionate and critical, involved and detached, the analysis of collective action is confronted with new challenges it itself must recognize, lest 'those who speak before' should go unheeded and the walls of stone or of silence muffle their message.
When looking at contemporary movements, we can assume one of two different attitudes - that of 'resolving' or that of 'listening.' Modern technology with its practice of intervention, wherein success is measured in terms of the efficacy of the given technique, claims victory for the "resolu-tionary' approach and renders listening impossible. Under the influence of the general predisposition to immediate remedial action, social movements are taken into consideration solely on account of their capacity (or lack thereof) to modernize institutions or to produce political reform. But this is to forget, or to ignore, that the reduction of contemporary social movements to their political dimensions alone is tantamount to solving the 'symptom', to suppressing the message contained in their specifically communicative character ('symptom' literally means 'to fall together') and simply moving about the problem in the background.
Reflection on the analysis of social movements, however, is not warranted for the sake of scholarship only. At the same time, it may become a topical antidote in society: the work of analysis can contribute to the culture of the movements themselves, enhancing their resistance to the illusion that the word they bear is sacred and undermining the urge to totality that will swiftly turn them into churches or new powers that be. Heightened awareness of the possibilities and constraints of action can transform the word of the movements into language, culture, and social relationships, and may out of collective processes build a practice of freedom.
The continuum which ranges from protest and rebellion by a social group to the formation of a mass movement and a large-scale collective mobilization comprises a huge variety of intermediate forms of action, and any attempt to classify them seems at first sight all too formidable an undertaking. Indeed, one doubts whether such an operation might even reward the effort, since it remains questionable whether any continuity or homogeneity among the phenomena considered can actually be found. Here, more than in any other field of sociology, misunderstandings reign supreme. Terms such as 'collective violence', 'collective behaviour', 'protest', 'social movements', or 'revolution' often denote diverse phenom-
ena and generate ambiguities, if not outright contradictions. It is not by chance that this confusion rotates around phenomena which closely involve the fundamental processes whereby a society maintains and changes its structure. Whether wittingly or not, the debate on the significance of collective action always embraces the issue of power relationships, and on closer examination derives its energy from defending or contesting a specific position or form of dominance. But the increasing prominence of the problem does not first and foremost stem from an ideological confrontation. It is social reality itself which presents us with a variety of collective phenomena, of conflictual actions, of episodes of social revolt which evade interpretation guided by traditional political categories, thus calling for new tools of analysis. Behind random protest or manifestations of cultural revolt in our complex planetary society - which by now also includes the developing societies of the 'South" - there of course always lie diverse prob- , lems and social structures. In this situation, the increasing diffusion of i these phenomena and their diversification is, paradoxically, matched by the / inadequacy of the analytical tools available to us. /
In a certain sense, then, this book constitutes a venture into the uncertain terrain of a theory still to be constructed. In this search - which at the present stage can only proceed by trial and error - the capacity of a theory to rely exclusively on its own analytical foundations is necessarily limited. From this fact derives the importance of the growing body of research into cases of social movements and episodes of collective action, which in recent years has enriched theoretical analysis with a large quantity of empirical material relating to actual behaviour in society. From this point of view, the nonlinear progress of any analysis that attempts to come to grips with the theme of social movements and collective action is also understandable, obliged as it is to rely upon overspecific observations to fill gaps in the theory, just as it is, by the same token, forced to run the risk of general hypotheses where empirical material is scarce or nonexistent on the other hand.
In the last thirty years, analysis of social movements and collective action has developed into an autonomous sector of theory formation and research within the social sciences, and the amount and quality of the work in the area has grown and improved. Not incidentally, the autonomy of the conceptual field relating to the analysis of social movements has developed parallel to the increasing autonomy of noninstitutional forms of collective action m complex systems. The social space of movements has become a distinct area of the system and no longer coincides either with the traditional forms of organization of solidarity or with the conventional channels of political representation. The area of movements is now a 'sector' or a 'subsystem' of the social.
4 Introduction
Introduction 5
Recognizing this autonomy forces us to revise dichotomies like 'state' and 'civil society', 'public' and 'private', 'instrumental' and 'expressive'. The crisis of such polar distinctions signals a change in our conceptual uni- j verse. The notion of 'movement' itself, which originally stood for an entity I acting against the political and governmental system, has now been rendered inadequate as a description of the reality of reticular and diffuse forms of collective action. ;
Contemporary 'movements' assume the lorm of solidarity networks entrusted with potent cultural meanings, and it is precisely these meanings that distinguish them so sharply from political actors and formal organizations next to them. We have passed beyond the global and metaphysical conception of collective actors. Movements are not entities that move with the unity of goals attributed to them byIdeologues. Movements are systems of action, complex networks among the ditterent levels and meanings of social action. Collective identity allowing them to become actors is not a datum or an essence; it is the outcome of exchanges negotiations decisions, and conflicts among actors. Processes of mobilization, organizational forms, models of leadership, ideologies, and forms of communication -these are all meaningful levels of analysis for the reconstruction from the within of the action system that constitutes the collective actor. But, in addition, relationships with the outside - with competitors, allies, and adversaries - and especially the response of the political system and the apparatuses of social control define a field of opportunities and constraints within which the collective action takes shape, perpetuates itself, or changes.
Contemporary forms of collective action are multiple and variable. They are located at several different levels of the social system simultaneously. We must therefore begin by distinguishing between the field of conflict on the one hand and the actors that bring such conflict to the fore on the other. In the past, studying conflicts implied analysing the social condition of a group and submitting what was known of that condition to deductive reasoning in order to wrest the causes of the collective action from it. Today, we must proceed by first singling out the field of conflict, and then explain how certain social groups take action within it.
Since no actor is inherently conffictual, the nature of action assumes a necessarily temporary character, and it may involve different actors and shift its locus among the various areas of the system. This multiplicity and variability of actors makes the plurality of the analytical meanings contained within the same physical phenomenon even more apparent. The totality of a given empirical collective action is usually attributed a quasi-substantial unity, when it is instead the contingent outcome of the interaction of a multiple field of forces and analytically distinct processes.
The inner differentiation of action is reinforced by the fact that in a plan-cian system social reality becomes synchronic: in the contemporaneity created b> the media system, all the 'geological strata' of human history are iimishaneously present. In the unity of the present, movements thus contain in one problems and conflicts that have different historical roots. Adding to this, movements attract the forms of discontent and marginal-■azúon 'híij; the social system generates, while the forming elites exploit conflict to seek opportunity to affirm themselves or to consolidate their positions.
An analytical perspective that draws on these insights helps us clarify one of the issues recurrently debated over the last decades. It concerns the 'newness' of contemporary conflicts: What is 'new' in the 'new social movements' is still an open question. Bearing the responsibility of the one who introduced the term 'new social movements' into sociological literature, I have watched with dismay as the category has been progressively reified. 'Newness', by definition, is a relative concept, which at the time of its formulation in the context of the movements research had the temoorary function of indicating a number of comparative differences between rhe historical forms of class conflict and today's emergen i forms of collect ive action. But if analysis and research fail to specify nit- dis dut live featuies of the 'new movements', we are trapped in an arid dehate between the supporters and critics of 'newness'.
On the one hand, there are those who claim that many aspects of the contemporary forms of action can be detected also in previous phenomena in history, and that the discovery of their purported newness is in the first place attributable to the bias shown by numerous sociologists blinded by emotional involvement with their subject matter. On the other hand, the defenders of the novel character of contemporary movements endeavour to show that these similarities are only formal, or apparent, and that the meaning of the phenomena is changed when they are set in different systemic contexts.
However, both the critics of the 'newness' of the 'new movements' and the proponents of the 'newness paradigm' commit the same epistemologi-cal mistake: they consider contemporary collective phenomena to constitute unitary empirical objects, seeking then on this basis to define the substance of their newness or to deny or dispute it. When addressing empirical 'movements', one side in the debate sets out to mark out differences with respect to the historical predecessors, the other stresses continuity and comparability.
The controversy strikes one as futile. In their empirical unity, contemporary phenomena are made up of a variety of components, and if these
6 Introduction
Introduction 7
elements are not analytically separated, comparison between forms of action that belong to mutually distinct historical periods becomes an idle activity, it will be extremely difficult to decide, for instance, the extent of the 'new' in the modern 'women's movement", as a global empirical phenomenon, compared with the first feminist movements of the nineteenth century. Paradoxically, the result of the debate on 'new movements' has been the accelerating decline of the image of movements-as-entities. Through comparative work on different historical periods and different societies, we know now that contemporary movements, like all collective phenomena, bring together forms of action which involve various levels of the social structure. These encompass different points of view and belong to different
1 historical periods. We must, therefore, seek to understand this multiplicity of synchronic and diachronic elements and explain how they are combined in the concrete unity of a collective actor.
Having clarified this epistemological premise, we may however still ask ourselves whether a new paradigm of collective action is not at the moment taking shape: not in the empirical sense - that is, in terms of the observed phenomenon as a whole - but analytically, in terms of certain levels or ele-
: merits of action. It is thus necessary to inquire as to whether there are dimensions to the 'new' forms of action which we should attribute to a systemic context different from that of industrial capitalism.
This question is dismissed by critics of 'new movements', who trace such phenomena on an exclusively political level. The resulting reductionism dispenses with the question of the emergence of a new paradigm of collective action without, however, having first provided any answers as to its pertinence. Moreover, it ignores those specifically social and cultural dimensions of action that feature so significantly in the 'new movements'. This gives rise to a different bias, to the exclusive concentration on the visible and measurable features of collective action - such as their relationship with political systems and their effects on policies - at the expense of the production of cultural codes; but it is the latter which is the principal activity of the hidden networks of contemporary movements and the basis for their visible action.
Do contemporary collective phenomena comprise antagonist conflicts that are systemic in nature, or do they rather belong to the phenomena of social emargination, of aggregate behaviour, of adjustment by the political market? So general a question can only be answered by first exploring alternative explanations of collective action, formulated for example in terms of dysfunctions or crises, or with reference to political exchange. Many of the contemporary conflicts can be explained through recourse to the workings of the political market, commonly as the expression of excluded social
groups or categories pressing for representation. Here, however, there is no antagonistic dimension to the conflict; there is only the pressure to join a system of benefits and rules from which one has been excluded. When the confines of the political system are rigid, such a conflict may even turn violent. However, this needs not necessarily entail antagonism towards the logic of the system; it may, instead, express a simple demand for a different distribution of resources or for new rules. Similarly, a poorly functioning organization may be subject even to an intense conflict, the aim of which, however, is not to dismantle that organization but rather to restore it to its normal state.
After exhausting the explanatory capacity of these dimensions, it, still remains to be asked - and this is important - whether there is anything left to account for. And here we must preserve a sufficient theoretical space in which to formulate the question of systemic conflicts; otherwise the issue will be glossed over without answers being provided or the questions themselves having been shown to be pointless. Today, we refer to the changes under way in contemporary systems using allusive terms (complex, post-industrial, postmodern, late capitalist society), the implicit assumption being that they follow a logic significantly different from that of industrial capitalism. But to do so is to neglect or to suppress the theoretical problems this very assumption raises.
The question of the existence of antagonistic conflicts of systemic scope, however, keeps open a number of issues with which theoretical analysis must now come to grip: for example, whether one can conceive of a dominant logic that disperses itself over a variety of areas of the system, producing thereby a great diversity of conflictuai sites and actors.
'If God gave me the choice of the whole planet or my little farm, 1 should certainly take my farm', wrote Ralph Waldo Emerson. Today we can no longer take the farm, since we have already been obliged to take the whole planet by virtue of the fact that the planet has become a whole. The Gulf War of 1991 has been the most recent and shocking demonstration of the global interdependence of our destiny as human beings on this planet and of the crucial role of information in shaping our reality. While we might not yet be fully aware of the reality of this fundamental change, contemporary social movements act as signals to remind us that both the external planet, the Earth as our homeland, and the internal planet, our 'nature' as human beings, are undergoing radical transformations. The reality in which we live has in its entirety become a cultural construct, and our representations of it serve as filters for our relationship with the world. For the first time in the history of the human species, this assertion is also true in a literal sense. In fact, the world of which we speak today is a global world of planetary scale,
8 Introduction
Introduction 9
and this is made possible only by information, or the cultural processes with which we represent our world to ourselves. The consequences of this change are enormous. But the emergence of the transnational dimension to issues and social actors, more than a political question, is in the first place a sign of the fact that human action by now is capable of culturally creating its own space. The planet no longer designates just a physical location; it is also a unified social space which is culturally and symbolically perceived.
Interest in cultural analysis has grown in the last two decades alongside the extraordinary cultural transformation of planetary society. We are witnessing, with mixed feelings of amazement and fear, the impressive development of communication technologies, the creation of a global media system, the disappearance of historical political cleavages, the collision of cultural differences within national societies and at the world scale. Never before have human cultures been exposed to such a massive reciprocal confrontation, and never has the cultural dimension of human action been as directly addressed as the core resource for production and consumption. It therefore comes as no surprise that social sciences are rediscovering culture, that a new reading of the tradition is taking place through the lens of this key concept, and that a wave of interest in cultural analysis is bringing a new vitality to theoretical debates in sociology.
Social movements too seem to shift their focus from class, race, and other more traditional political issues towards the cultural ground. In the last 5 thirty years emerging social conflicts in complex societies have not expressed themselves through political action, but rather have raised cultural challenges to the dominant language, to the codes that organize information and shape social practices. The crucial dimensions of daily life have been involved in these conflicts, and new actors have laid claim to their autonomy in making sense of their lives. Contemporary society with its tightly woven networks of high-density information requires for its proper functioning the development of a distinct degree of autonomy of its component parts. It must presuppose and depend on individuals, groups and subsystems, which act as self-regulating units capable of sending, receiving, and processing information. To this end, development of formal skills of action, decision-making, and continuous learning is encouraged. However, increasing systemic differentiation simultaneously threatens social life with fragmentation, lack of communication, atomized individualism, and calls
' for deeper integration of individual and collective practices. The key focus of control shifts from the manifest forms of behaviour to motives and the
' meaning of action, to those hidden codes that make individuals and groups
i predictable and dependable social actors.
Social conflicts tend to emerge in those fields of social life which are
directly exposed to the most powerful and intense flow of information, and where at the same time individuals and groups are subject to the greatest pressure to incorporate in their everyday behaviour the requirements and the rules of systemic normality. The actors involved in these conflicts are transient, and their action serves to reveal to and caution the society of the crucial problems it faces, to announce the critical divisions that have opened up within it. Conflicts do not express themselves through action taken m accordance with the purposive norms of efficacy. The challenge is : made manifest in the upsetting of cultural codes, being therefore predominantly formal in character.
In contemporary systems, signs become interchangeable and power ' operates through the languages and codes which organize the flow of' iuto'-mution. Collective action, by the sheer fact of its existence, represents in f' • r t y form and models of organization a message broadcast to the rest o: society. Instrumental objectives are still pursued, but they become more precise and particular in their scope and replaceable. Action does still have effects on institutions, by modernizing their culture and organization, and by selecting new elites for them. At the same time, however, it raises issues that are not addressed by the framework of instrumental rationality. This kind of rationality is devoted to the effective implementation of whatever has been decided by anonymous and impersonal powers operating through the apparent neutrality of technical expertise.
Actors in conflicts recast the question of societal ends they uldress the differences between the sexes, the ages, cultures; they probe into the nature and the limits of human intervention; they concern themselves with health and illness, birth and death. The action of movements dehbeiately differ- / entiates itself from the model of political organization and assumes increasing autonomy from political systems; it becomes intimately inter-weaved with everyday life and individual experience.
Increasing control is applied to people's routine existence by the apparatuses of regulation which exact identification and consensus. Conflicts involve the definition of the self in its biological, affective, and symbolic dimensions, in its relations with time, space, and 'the other'. It is the individual and collective reappropriation of the meaning of action that is at stake in the forms of collective involvement which make the experience of : change in the present a condition for creating a different future. Movements thus exist also in silence, and their presence is fundamental for the vitality of information societies. The challenge embodied in the movements' action keeps raising questions about meaning, beyond the technical neutrality of procedures which tends to install itself in institutions and governs their role in the society.
10 Introduction
This dimension, however, docs not exhaust the significance of collective action. Contemporary collective action weaves together its different roots in multiple meanings, legacies from the past, the effects of modernization, resistances to change. The complexity, the irreducibility, the intricate semantics of the meaning of social action is perhaps the most fundamental theme of this book. Only a society that is able to accommodate the thrust of the movements by providing an unconstrained arena for the fundamental issues raised by collective action, as well as democratic channels of representation and decision-making, can ensure that complexity is not ironed out, that differences are not violated. Keeping open the space for difference is a condition for inventing the present - for allowing society to openly address its fundamental dilemmas and for installing in its present constitution a manageable coexistence of its own tensions.
PART I
Theory of collective action
1
The construction of collective action.
Traditions
When talking of social movements and collective action, one is usually referring to empirical phenomena with a certain degree of external unity. Movements are often described in terms similar to those used in addressing personalities or personages in tragic theatre, characters with a distinct and coherent role. Yet what in fact is in question are heterogeneous and fragmented phenomena, which internally contain a multitude of differentiated meanings, forms of action, and modes of organization, and which often consume a large part of their energies in the effort to bind such differences together. Movements, characteristically, must devote a considerable share of their resources to the task of managing the complexity and differentiation that constitutes them.
It is, furthermore, customary to refer to movements as the effects of a particular historical situation, or as an outcome of a particular conjuncture (such as an economic crisis or contradictions within the system), in doing so, however, one ignores the motives for, and the meaning and components of, collective action, by assuming that the ways in which such action comes into being and persists over time are irrelevant when compared to the interplay of 'structural' variables. These manners of considering social movements as either historical characters or results of structural determinants are not just commonplace notions of everyday discourse; they also stamp many of the current analyses of contemporary collective action.
They indeed demonstrate how wide the gap still is that separates the
established linguistic convention, or the political interest that issues in the talk of 'social movements', from the possibility of giving an adequate theoretical basis to the analysis of collective action. Too often, a movement is still portrayed as the incarnation of an essence or the secondary effect of
14 Theory of collective action
The construction of collective action 15
the 'tendential laws' of a structure. The collective action of a movement is thus always related to something other than itself; properly speaking, it does not even exist.
It is important to react against such theoretical liquidation of an object so salient in daily discourse and theoretical debate, and with so crucial a role in contemporary social processes. Beyond linguistic convention, only a theory of collective action can provide a meaningful basis for analysis of social movements. A discipline that sets out to study social movements can accomplish its task meaningfully only if it starts out from a theory that can account for the specificity and autonomy of social action, and can give a foundation to its collective character as something different from the sum total of aggregate individual behaviours.
Up until the 1960s, those interested in these issues in the sociological field drew, directly or indirectly, either on Marxist theory or on the sociology of collective behaviour. One has not much to say about the former, for I believe that, strictly speaking, there exists no specifically Marxist branch of analysis of social movements today in the proper sense of the term, only studies (sometimes very accurate) of the crisis of the capitalist mode of production and of its transformations. Marxism has provided a theoretical framework for the historical analysis of class action, but its explicit contribution to the theory of social movements has been poor, indirect, or frankly derivative (see Calhoun 1982; Pakulski 1995). On the other side, one finds the scholars who in the 1960s ventured to examine collective behaviour within the functionalist and interactionist traditions, the most influential among them being Smelser and Turner (Smelser 1962, 1968; Turner 1969; see also Turner and Killian 1987). Even though many differences divide and sometimes oppose to each other the functionalist and the interactionist perspectives, they both rely on a theory of shared beliefs, applied to various kinds of collective behaviour ranging from panic to revolution. The great spectrum of behavioural phenomena drawn to these analyses likewise dissolves the object 'social movements'; it now becomes a particular case of generalized belief, a specific way of restructuring the field of collective normative patterns. When norms or shared values are threatened by some form of imbalance or crisis, the response through which an attempt is made to reestablish social order is centred around a common belief which, while often fictitious, mobilizes collective energies.
In the legacy of these intellectual traditions, two ingenuous epistemo-logical assumptions still persist that have left their mark on the study of collective phenomena. The first one is the supposition that factual unity of the phenomenon, as perceived or believed to be there by the observer, actually exists. The proximity in space and time of concomitant forms of individual
and group behaviour is elevated from the phenomenological to the conceptual level and thus granted ontological weight and qualitative homogeneity; collective reality, as it were, exists as a unified thing. A second assumption now enters into the process of reification: the collective dimension of social behaviour is taken as a given, as a datum obvious enough to require no further analysis. How people actually manage acting together and becoming a 'we' evades the problematic as it is taken for granted.
However, in contemporary societies affected by accelerated change and permanently on the brink of a catastrophe, it has in the meantime become evident that social processes are products of actions, choices, and decisions. Collective action is not the result of natural forces or of the iron laws of history; but no more is it the product of the beliefs and representations held by the actors. On the one hand, research traditions have located the roots of all conflicts in the social fabric (in the economic structure in particular) and explained them in terms of an historical necessity of some sort. The most significant example of this tendency is given to us in the dilemma that, at least since the Second International, has divided Marxist scholarship: is class action born out of voluntarist orientation, emerging spontaneously from the condition of the proletariat, or is it a necessary effect of the contraHirtions of a capitalist production system marked by fa+e for a collapse? This question has remained unresolved in the Marxist traditions, and the fact bespeaks all the difficulties that arise when collective action is taken to be a phenomenon without its own autonomy from 'structural determinants'. Various attempts have been made to bridge this gulf between the contradictions of the capitalist system and class action, sometimes by emphasizing the determinism of structural laws, at others by stressing the voluntarism of mobilization. This dualistic legacy is still alive 1 in current debates on the relationship between structure and agency (Sewell: 1992; Berejikian 1992).
On the other hand stand those who seek to explain collective behaviour/ in terms of the beliefs held by actors, such as are manifest, for example, in common objectives or shared values. Actors, it is claimed, respond to certain dysfunctions of the social system by creating a collective set of representations which fuel action. In this case, too, the problem of how the collective subject of action comes about and persists in time is left unresolved. The actors' own beliefs will not provide a sufficient ground tor an account of their actions, for such beliefs always depend on the broader relations in which the actors are involved. Analysis cannot simply identify action with that which the actors report about themselves, without taking into account the system of relationships in which goals, values, frames, and discourses are produced.
16 Theory of collective action
The construction of collective action 17
Thus, explanations based on the common structural condition of actors take for granted the actors' ability to perceive, evaluate, and decide what they have in common. That is, such explanations ignore the processes which enable actors to define a 'situation' as a field of shared action. On the other hand, actors' motives, beliefs, discourses and individual differences again are never enough to provide an explanation of how certain individuals or groups recognize each other and become part of a 'we'.
Between these two poles of the dualism bequeathed to us by research tradition there stretches an open, still unexplored theoretical space: it concerns the ways in which actors construct their action, During the last twenty-five years, to be sure, some progress has been made towards resolving the evident impasse created by the dualistic tradition. European authors on one hand have contributed to a better understanding of the process through which collective action is formed in highly differentiated -or postindustrial - systems (Touraine 1977, 1981, 1985; Habermas 1984, 1987, 1990; Giddens 1984, 1987, 1990). American proponents of the Resource Mobilization Theory, on the other, have provided a framework for the analysis of the actual mobilization process ((available resources, entrepreneurs, opportunity structures) (McCarthy and Zald 1973, 1977; Zald and McCarthy 1979, 1987; for a review. Jenkins 1983). Other authors have extended in original ways this paradigm (Oberschall 1973; Gamson 1990; Gamson, Fireman and Rytina 1982; Tilly 1978; McAdam 1982: Klandermans 1984; Tarrow 1989a). In my previous work I have tried to bridge these approaches by stressing the constructive dimension of collective action (Meiucci 1980, 1984, 1988, 1989; and other authors have increasingly supported the necessity of reducing the gap between European and American tradition (Cohen 1985; Tarrow 1988b; Klandermans, Kriesi and Tarrow 1988; Klandermans and Tarrow 1988; McAdam, McCarthy and Zald 1988; Gamson 1992a). More recent contributions are building on these advances and are explicitly addressing the processes through which actors give meaning to their action (Klandermans 1989a, 1992; Tarrow 1992,1994; Morris and Mueller 1992; Mueller 1992, Larana, Johnston and Gusfield 1994, Johnston and Klandermans 1995). Today we are in a better position to build a new framework for the analysis of collective action by both acknowledging the legacy of the past and overcoming its deficiencies.
Collective action as a construct
Should we want to draw up a balance sheet summing up the contribution of classical and recent sociology to the study of social movements and collective action, one can point out a number of fundamental insights which
c,>nsf itiuc indispensable points of reference for the ongoing debate on, and theoretical analysis of. collective action.
The edition of Marxism has taught us that collective action cannot be anaKsed without addressing its relationship to a "structural' (or, better, structured') field of relationships which provides resources and constraints for the action itself. Moreover, it has persuasively demonstrated the importance of social conflicts and the fact that some of them are of an antagonist nature. It is within this legacy that recent European contributions (Touraine 1988a, 1994a; Habermas 1989, 1990; Giddens 1990, 1991) have tried to understand the changes that modern, postindustrial systems are undergoing today.
Within the classic functionalist approach, Merton's well-known distinction between deviance and nonconformism goes beyond the limited and sometimes ideological perspective from which Parsons examines social conflicts. This distinction - with whose terminology one may or may not agree - raises a crucial problem for analysis of social movements. It rejects any reduction oi collective action to the status of a mere symptom of the degradation o! die ocial system (not by coincidence, identification of every form of collective action with deviance is a feature typical of the ideology of the dominant groups). It also permits a further distinction to be drawn between collective processes that stem from disaggregation of the system and those which rather seek to rebuild that system on a different basis.
The analyses by the Chicago School and the contribution of the sociology of collective behaviour (particularly such authors as Smelser and Turner) have taught us that it is not possible to distinguish, to use the common label in the dominant discourse, "normal' social behaviour from a 'pathological' social behaviour, of which the forms of collective action would be an index. Analysis of collective action must be conducted using the same categories that are applied to other components of the social system: the tools employed in analysis of collective phenomena must be framed by some general hypotheses on the social system.
Symbolic interactionism building on the work of B! timer, on the other hand, has taught us that collective action is not the expression of irrationality or of psychological suggestion that the crowd psychology of the nineteenth century assumed (LeBon 1960; Tarde 1969): it is, instead, meaningful behaviour (see Turner 1983): there is a logic of collective action which entails certain relational structures, the presence of decision-making mechanisms, the setting of goals, the circulation of information, the calculation of outcomes, the accumulation of experience, and learning from the past.
Finally, resource mobilization theory in all its variants has demonstrated
18 Theory of collective action
The construction of collective action 19
to us that collective action does not result from the aggregation of atomized individuals. Rather, it must be seen as the outome of complex processes of interaction mediated by certain networks of belonging. Collective action, therefore, is not unstructured behaviour in the sense that it would not obey any logic of rationality. It involves an articulated structure of relations, circuits of interaction and influence, choices among alternative forms of behaviour. It only appears unstructured when set against the dominant norms of the social order, and against the interests which that order wishes to maintain (as in the discourse that labels collective action as marginal, deviant, rootless, irrational).
But beyond the specific contribution of the sociology of social movements, an understanding of contemporary collective action could hardly take place without referring to the implications of cultural changes for a theory of social action. The central role of culture in shaping social action has been one fundamental reminder of the recent developments in sociological theory by authors different in many respects as Alexander (Alexander 1988a, 1988b, 1989) and Bourdieu (1977, 1984, 1990a, 1990b). Within a paradigm that stresses the capacity of human action to construct meaning and making sense of reality, my particular understanding of the cultural dimension of collective action builds on. the work of Norbert Elias (1991, 1994), Mary Douglas (1970, 1986, 1992) and Clifford Geertz (1973, 1983).
1! one discards the simplistic image of collective action as the upshot of irrational and perhaps suggestive processes, along with the naive assumption that collective phenomena are simply empirical aggregations of people acting together, it becomes of critical importance to develop a theoretical model to account for the nature and the emergence of this type of action through the identification of the general and specific factors of its formation. We can now identify at least five distinct problems upon which to concentrate our inquiry. A first level of analysis concerns the definition of collective action and involves both the devising of analytical criteria and the empirical delimitation of the field. Another issue is establishing the processes that give rise to collective action, its, formation in the social structure. At this level, it will be important to distinguish between structural conditions and conjunctural factors. Thirdly, analysis is called for of the components that structure collective action, that is, of the system oi iciauotis... which which give it continuity, adaptability, and effectiveness. Closely connected with this level of analysis is the problem of the forms assumed by collective action (and particularly by contemporary social movements). Finally, the field of collective action must be examined, as the set of social relationships providing resources for and constraints to the action. In this
chapter I will address the first of these analytical problems, while the remaining chapters of part I (the theory of collective action) will be (jevoted to a closer examination of the questions involved in the second and third issues. The third issue will also be extensively addressed in part IV (internal dimensions of collective action). Part II (actors) of this book will discuss in detail contemporary social movements, while Part IV touches upon the forms of collective action in general. Part III (the systemic field) addresses the environmental conditions for collective action and in particular lays out the political system and the state as that particular field against which all collective action must measure itself in a concrete society and through which other systems of opportunities and constraints become evident.
The first question to be addressed is whether, and to what extent, it is possible at all to identify the analytical criteria which enable us to make more specific distinctions within the general category of collective action. Is it possible to establish a homogeneity of pattern between a panic and a revolt? Or, conversely, what is it that authorizes us to talk of social movements as sociologically specific phenomena? The sociology of collective behaviour, for instance, moves within the perspective of the former problematic. It defines collective behaviour as a general analytical level of social action which enables panic, fashion, crazes, and revolutions to be explained altogether using the same criteria. All that differs from one phenomenon to the other is the degree of generality in the components of action affected and restructured by collective behaviour. Smelser s theory, for example, is the first explicit attempt to develop an analytical framework which encompasses all the different forms of 'collective behaviour'. What in 'crowd psychology' was confused and implicit becomes, for Smelser, methodological requirement in the construction of theory. For early researchers, the irrationality of deep urges was the implicit analytical level at which to situate crowd behaviour. According to Smelser, however, it is generalized belief which is the feature common to all collective forms of behaviour, enabling us to decipher their analytical significance.
But is the category of generalized belief analytically precise enough to allow distinctions to be drawn among the various empirical forms of collective action? For this purpose, 'collective behaviour' is too general a container, bringing together under its categorial unity a great multitude of different empirical phenomena ranging from 'spontaneous' panic to planned revolutions. The only common feature shared by this heterogeneity is ultimately the 'collective' character of each phenomenon in concern, which simply describes an empirical contiguity but remains unsatisfactory for us set about to analytically differentiate among these phenomena.
20
Theory of collective action
The construction of collective action 21
My perspective builds on a strictly phenomenological point of departure: collective phenomena are those sets of social events that comprise a number of individuals or groups exhibiting, at the same time and in the same place, behaviours with relatively similar morphological characteristics. These phenomena are variously defined in sociological literature as collective behaviour, social movements, protest events, crowd behaviour, and the like, but the assumption that all these social practices share some common features stops short at the mere recognition of their common 'collective' character (for an example, see Hardin 1982). Beyond this phenomenological trait it is difficult to assume some kind of qualitative unity or homogeneity without making a conceptual assumption about the analytical nature of the phenomena. Even the choice between collective 'behaviour' or collective 'action' implies a different theoretical frame which needs to be explicitly addressed. Consequently, as a starting point, 1 will define collective action as a set of social practices (i) involving simultaneously a number of individuals or groups, (ii) exhibiting similar morphological characteristics in contiguity of time and space, (iii) implying a social field of relationships and (iv) the capacity of the people involved of making sense of what they are doing. This definition already contains some of the theoretical assumptions which will be discussed in the following pages, but it is also the minimal and the most general starting point for a different approach to the empirical phenomena that are usually referred to when speaking of collective action, social movements, and other similar com-monsense notions.
First of all, escaping the dualistic inheritance of the sociological tradition in the study of collective phenomena will only be possible if we reverse the naive assumption regarding collective action as a unitary empirical datum. Instead of taking it as a starting point, we should examine that very datum in order to discover how it is produced, and disassemble its unity so as to reveal the plurality of attitudes, meanings, and relations that come together in the same whole of the phenomenon. Addressing the problem of Low a collective actor takes shape requires recognition of the fact that, for instance, what is empirically called 'a movement' and which, for the sake of observational and linguistic convenience, has been attributed an essential unity, is in fact a product of multiple and heterogeneous social processes. We must therefore seek to understand how this unity is built and what different outcomes are generated by the interaction of its various components.
This approach signifies a real change of perspective. Historical studies and the sociology of work, for example, have shown the multiplicity of
levels present in what, almost by linguistic convention, is called 'the
porkers" movement' and which, despite its analytical heterogeneity, -aenimed from a common, underlying social condition (Calhoun 1982;
Fmitu:«a 1988; Hirsch 1990a). Strikes have never been homogeneous phenomena, for, internally, they have brought together a host of mutually conflicting demands, including those aimed at the organizational system of the
firm those addressed to the political system, and elements of class struggle aa-il.isi the capitalist mode of production as such (Badie 1976). This differentiation of objectives and interests is even more evident in contemporary collective phenomena, which moreover are not rooted in a shared social condition.
I propose to differentiate the general category of collective action and to shift from an empirical to an analytical point of view. Within this broader framework, as we will see in section 3.4 of this chapter, I specifically propose to use the notion of 'social movement' not as an empirical categorization of certain types of behaviour but as an analytical concept: understood this way, it addresses a,particular level of collective action that should be distinguished from other levels present in the empirical collective phenomena. No phenomenon of collective action can be taken as a global whole since the language it speaks is not univocal. An analytical approach to those phenomena currently called 'movements' must be firmly placed within a theory of collective action, and it must break down its subject according to orientations of action on the one hand and the .system of social relationships affected by the action on the other. For example, campaigning for functional changes in an organization is not the same thing as challenging its power structure; fighting for increased participation in decisionmaking is different from rejecting the rules of the political game. Only by distinguishing among the different analytical meanings and relational fields of the collective action under consideration can we begin to understand the contents of a concrete 'movement' as the vehicle for multiple and often contradictory demands.
Thus conceived, the concept of social movement, along with all the other concepts to be presented for analytical purposes in the following section, are always objects of knowledge constructed by the analyst; they do not coincide with the empirical complexity of the action. The study of collective action is still prisoner of an 'objectivist' assumption about his categories and it seems rarely aware of the epist^mological turn introduced by hermeneutics (Ricoeur 1974, 1976, 1981, 1984; Gadamer 1976) and the cognitive revolution (Bruner 1986, 1990). An awareness of the constructive operation of our conceptual tools is today an epistemological requirement if we are to abandon for good the naive assumption that social phenomena are 'out there' existing in full independence from our point of view, and if
22 Theory of collective action
The construction of collective action 23
we wish to be able to assume responsibility for the role that knowledge plays, and can play, in contemporary social life.
Principles for analysis of collective action
Analysis must distinguish between a reaction to a crisis and the expression of a conflict.
As stated, the appearance of collective action has often been linked to a crisis in one sector of the system or the another, the crisis denoting breakdown of the functional and integrative mechanisms of a given set of social relations. Collective action has thus been often viewed as a pathology of the social system. A conflict, on the other hand, is defined by a struggle between two actors seeking to appropriate resources regarded by each as valuable. The actors in a conflict join battle in a shared field for control of same resources. For an event to constitute a conflict, the actors must be definable in terms of a common reference system, and there must be something at stake to which they both, implicitly or explicitly, refer. Without a distinction between conflict and crisis it would be impossible to make sense of many historical and recent forms of collective action. Had working-class struggle, in the history of capitalism, been nothing more than a reaction to economic exploitation and cyclic crises, it would have been over as soon as the workers won better pay and improved working conditions. But the con-flictual character of the workers' movement derived rather from the fact that it was a struggle against the very logic of industrial production under capitalist conditions (Katznelson and Zolberg 1986).
Conflicts, therefore, are not only conceptually distinct from crises, but among them are included those in which the adversaries enter the strife on account of the antagonistic definitions of the objectives, relations, and means of social production they assert and defend. A conflict of this kind within a social system may be brought to the surface by particular situations of crisis internal to the system itself. But when a collective actor by its action makes visible a conflict which is antagonistic in nature, this should not be confused with a simple reaction referring back to the crisis that, at this particular juncture, has provoked or accelerated that action.
A crisis always arises from the processes of disaggregation of a system, having to do with dysfunctions in the mechanisms of adaptation, imbalances among parts or subsystems, paralyses or blockages in some of these, difficulties of integration. The scope and the intensity of a crisis naturally depend on the particular levels of the system affected. A crisis provokes disintegration and the subsequent reaction of those who seek to redress the balance, whereas an antagonistic conflict makes manifest a clash over the
-ontrol and allocation of crucial resources (Collins 1975). In the history of ,nv particular society, these two dimensions are often meshed together, ren-jering the analysis of the processes of collective mobilization even more difficult.
The difference between a crisis and an antagonistic conflict, then, emerges as a distinction of great consequentially. In practical reality, it is pjwed out in the fact that the dominant groups always tend to define movements as simple reactions to crises, that is, to a dysfunctional mechanism of the system. Admitting that they are something else would entail recognition of collective demands that challenge the legitimacy of power and the current deployment of social resources.
Analysis should distinguish among different orientations of collective action.
We can discriminate between sets of basic orientations of collective action that are helpful in establishing analytical distinctions among various kinds
of behaviour. They are as follows:
1 Some collective phenomena involve solidarity; that is, the ability of actors to recognize others, and to be recognized, as belonging to the same social unit. In other cases, collective action arises as an aggregation of atomized behaviours (Alberoni 1984). This latter orientation I will designate as aggregation: (a) Aggregative orientations do not involve solidarity and they only express spatio-temporal contiguity; (b) they can be broken down to the level of the individual without the loss of their morphological features; and (c) they are wholly oriented towards the outside rather than inwards the group.
Collective orientations of this kind usually form in response to a crisis in the social system or to accelerated change, and they result from the aggregation of atomized individuals through a generalized belief, in the sense given to the term by Smelser. The operation of such a belief - which is not a system of solidarity but an object of affective identification by individuals -joins together actions which in themselves are separate. An aggregate results from the temporal and spatial proximity of the repetitive multiplication of individual behaviours.
The phenomena which can be most readily assigned to this category are those that the sociology of collective behaviour has studied closely (crowd behaviour, panic, booms, crazes, fashion) (Smelser 1963; Turner and Killian 1987; Weller and Quarantelli 1974; Marx and Wood 1975; Aguirre et al. 1988). One should not forget, however, that these empirical phenomena likewise have different analytical meanings: a fashion, for example, is
24 Theory of coiled ive action
I he construction of collective action
25
never an aggregate phenomenon pure and simple, since it is also the result of changes in production modes, of the workings of the market, and of the emergence of new needs. On the other hand, even the most highly structured social movements contain aggregate elements which manifest themselves, for example, m rituals, in the broadcasting of symbols, in mass events, and so on. Thus the empirical object should always be broken down analytically to reveal the multiple meanings it contains within itself,
2 Some collective phenomena involve conflict, that is, the opposition of two (or more) actors who seek control of social resources valuable to each of the protagonists. Others, again, come into being thiough c onsensus over the rules and procedures governing the control of valued tesoirces.
3 Lastly, some collective orientations involve a hi ecu It oi the limits of compatibility of the system of social relationships within which the action takes place. 1 define 'the limits of compatibility' as the range of variability in systemic states that enables a system to maintain its structure (or the set of elements and relations that identify the system as such). Orientations of collective action break the limits of compatibility when they are propelled beyond what is covered by the range of such variations that the system can tolerate without altering its structure. Other kinds of collective action have order-maintaining orientations, in that their effects remain within the limits of structural variability of the given system of social relations.
A simple breach of the compatibility limits of the reference system is not enough for an action to signify social conflict: it merely signals the disruptive character of the action. A breach of the rules or the rejection of the shared norms do not necessarily imply a straggle between two actors over something at stake, but is instead symptomatic of deviant behaviour: here the actor is defined by marginality with respect to a system of norms, reacting to the control that such norms exercise without nevertheless challenging their legitimacy - without, that is, identifying a social adversary and a set of contested resources or values. Deviance, as the product of breakdown in the order or as the inadequate assimilation of norms by individuals, resolves itself into the search for particularist rewards outside accepted norms and behaviour. In this case, too, I treat deviance as an analytical category endowed with autonomous weight. The empirical analysis of those forms of behaviour that are commonly classified as deviant is, then, a different matter altogether. The criticism of the functionalist framework longtime employed by sociological analysis of deviance has allowed numerous misunderstandings. It has been rightly pointed out that deviant behaviour cannot be reduced to social pathology, and that such behaviour is often implicitly critical of the dominant normative system. But preoccupation with specifying labelling processes and the processes of social production
of deviance (Spector and Kitsuse 1973; Kitsuse 1975) has often obscured the best achievements of the functionalist paradigm: its focus on phenomena engendered by dysfunctions in the integrative mechanisms of a social system. Only by preserving the analytical distinction between simple disruptive behaviour and conflictual processes can we avoid both the reduc-tionism that treats all forms of dissent as social pathology (as in the classical version of functionalism) and the attribution of an innovative or even revolutionary potential to every act that breaks the order (as in some radical extension of labelling theory).
On the other hand, if a conflict is not pushed beyond the limits of the reference system, what is in question is an opposition of interests within a certain normative framework. In such a case, action seeks to improve the relative position of the actor, to overcome functional obstacles, to change authority relationships. Conflict observes the limits set by the partners' joint preoccupation with ensuring the compatibility of the system as defined above and respecting the rules of the exchange. This kind of behaviour - common in large organizations, systems of industrial relations, and in the political systems of complex societies - can be defined towpon-turn; its analytical content concerns the presence of contending mteiesis and acceptance of set 'rules of the game'.
These basic orientations can be plotted as axes along which the various forms of collective action can be arranged and identified (figure 1).
The analytical field of collective action depends on the system of relationships within which such action takes place and towards which it is directed.
The reference systems of collective action should not be confused with the concrete sites of social praxis in which action effectively takes place (institutions, associations, organizations, and the like). The physical workings of a certain social arrangement always combine a number of different processes: the school, the factory, the city are all the result of the interaction of productive structures, of systems of stratification, of decisionmaking processes, of symbolic systems, of forms of power, and so on.
The reference systems of collective action should therefore be understood as analytical structures, as specific forms of social relationships which can be -differentiated in terms of the nature of the social link binding individuals or groups together. Any analysis that, implicity or explicity, introduces the notion of the 'breaking of limits' must define a reference system. Sociologists, however, have often failed to recognize the full importance of this imperative ----- for example, when they have looked at 'protest' and usually
26 Theory of collective action
The construction of collective action 27
Solidarity
Maintenance of systems limits
Breaching of systems limits
Aggregation
Consensus
Figure
denned it as a form of disruptive action (Lipsky 1968; Eisinger 1973; Di Nardo 1985; Lofland 1985; Epstein 1990a). But what are the confines that the protest breaks or transcends ('disrupts')? Without a definition of the reference system, the notion of protest is, analytically speaking, meaningless.
The first question is: What is meant by a 'system'? An approach to the social reality in terms of systems refuses to characterize this reality as any kind of essence or a metaphysical entity, and instead considers it to be a coincidence of interdependent relationships. A system is simply the complex of the relationships among its elements. A system does not possess a privileged nnrleus that would contain the meaning of the whole. Each
: element stands for itself in relation to the others, and each variation in these relations affects the whole. To analyse society as a complex of social relations is tantamount to declining the invitation both to reduce the social to the natural and to turn it into an expression of essence (of man, of Spirit, of morality). Social action is not the effect of mechanical laws or natural determinism, but nor is it the incarnation of the spirit or a progeny of
I values; it is the result of relationships which tie together a plurality of social
factors producing meaning for what they do (Alexander 1988a; Collins
: 1981, 1988, 1989; Schelling 1978).
Different systems which may be defined according to the specific types of relations that characterize them. Minimally, we must thus distinguish jxtween (1) the system that ensures the production of a society's resources; (2 j [he system that makes decisions about the distribution of these resources; l3i the system of roles which governs the exchange and deployment of the lattei; and (4) the lifeworld or the system of reproduction in everyday life:
1 1 lie first of these systems consists of antagonistic relationships that comprise the production, appropriation, and allocation of a society's basic re^mi rces. This level of relations defines the modes by which society ^produces and appropriates its basic resources, incorporating imbalances of power and manifesting a basic conflict over the means and orientation of sociul production. As regards production, we should remember that the historical experience of the era of industrial capitalism has supported the ascendance of the reductivist identification of the mode of production with economic activity. The social relations of capitalist production, which in industrial society were culturally defined in material terms, have been over-generalized to stand for production as such, obscuring thereby the understanding of the cognitive, symbolic, and relational components that have always given the social activity of producing its very character.
2 The political system (see chapter twelve) constitutes the level of a society at which, within a framework of shared rales and through processes of representation, normative decisions are made between competing interests. This analytical level not only coincides with political systems in the strict sense but can today be identified in all complex organizations, decentralized administrative systems, and the like as well.
3 The third system, the organizational system, comprises the relationships whose purpose it is to ensure the society's interna! equilibrium as well as its adaptation to the environment through processes of integration and exchange among different parts of the system (in particular through exchange among roles, or systems of normatively regulated, reciprocal expectations of behaviour). This analytical level applies equally to a global society and to an individual organization or institution.
4 The lifeworld, or the reproductive system, is that level of social relations within which the basic requirements of social life are maintained and reproduced through interaction and communication. In everyday life intimate interpersonal relations allow individuals to make sense of their world. Physical reproduction and affective primary bonds rely on face to face relationships governed by the fundamental dynamics of identification and differentiation.
From this multiplicity of the systems making up the social structure, it is clear that each one of such systems is 'incomplete' in itself, and that each of
28 Theory of collective action
The construction of collective action
them reacb.es out to other systems, to relations and meanings, to goals and interests beyond their individual confines. There is a hierarchy whereby one system imposes on others a greater burden of limitations than what the others may accomplish with respect to it; this, however, is not a mechanical, predetermined relationship, but one of autonomy and dependence. Dependence is manifest in the fact that the possibilities and the limits in the functioning of one system are determined by another. Autonomy arises where each system has developed processes and rules of its own, and each has the capacity to create constraints on the system upon which it depends. Thus social production sets specific limits on the functioning of the political system; and the political system, in turn, establishes the rales for social organizations and everyday life. Each of these systems, however, is also governed by its own internal logic and constituted by specific relations (opposition between different societal ends, the play of pressure and influence in decision-making, exchange and interaction between roles, interpersonal and affective communication in Hfeworlds). Moreover, each system can affect the others, among them even those with respect to which the system's balance sheet of mutual constraint remains negative: for example, the meaningfulness or the emptiness of primary relations, and the equilibrium, or lack thereof, of the role system can affect the political system and the mode of production, just as openness or rigidity in political decision-making mechanisms can retroact on the relations of production and on the appropriation of social resources.
This set of analytical distinctions enables us to differentiate among the multiple fields of collective action that combine in various ways with the orientations listed above in the concrete phenomena that are currently called by the observers, or call themselves, 'social movements'. Through this set of analytical categories, competition regulated by interests that operate within the confines of the existing social order can be distinguished from forms of solidarist action which force the conflict to the point of breaking through the system's compatibility limits; the atomized sum of individual behaviours present in certain aggregate phenomena can be differentiated from deviant behaviour which pushes beyond shared rules but does not reach conflictual dimensions, and so forth.
The notion of a social movement is an analytical category. It designates that form of collective action which (i) invokes solidarity, (ii) makes manifest a conflict, and (Hi) entails a breach of the limits of compatibility of the system within which the action takes place.
Within the framework of the principles just laid out, I propose to transform the notion of social movement from an empirical generalization into an
;i_> ci.eot. As an empirical generalization the notion of social ,11V o( ently applied to various empirical phenomena of collects ,v- , i .! p ig from political protest to different kinds of 'disruptive' behaviour. Tne empirical features selected by the observer normally lead to differing definitions of what a social movement is, with a low and erratic deeree of comparability among the various definitions (for recent examples .,;Lh different theoretical backgrounds, see Boggs 1986; Diani 1992; Epstein 1990b; Tarrow 1994). 1 propose, instead, to define the concept of •soci.il movement' through certain analytical dimensions which indicate specific qualities within the broader field of collective action. One can sptof a 'social movement' only when these analytical conditions required by the definition are met. Or, better yet, one can employ the concept as an analytical tool to detect in the variety of empirical behaviours the presence of those analytical dimensions that identify a specific type of collective action.
The cpistemological shift I thus propose implies an equivalent shift in the attitude of the observer-analyst: that from simply mirroring empirical reality under the assumption of its 'objective' existence, towards a more explicit and conscious acknowledgment of the active role of our analytical tools in selecting among the mass of empirical 'data' and in constructing our 'objects' of knowledge.
In this specific case, such a shift nevertheless involves a linguistic problem. In our ordinary language, we still hold on to the notion of social movement to indicate various empirical collective actors (as when we speak of "youth movement", 'women's movement', 'peace movement', and the like). Confusingly, thus, the same term is used to designate at once an analytical concept and a variety of empirical phenomena. The persistence of this linguistic ambiguity, however, depends on the life expectancy of the notion of a social movement itself. Its crisis is related to, and settled together with, the crisis of the general paradigm to which it belongs and which gave birth to it: that of the industrial capitalism. We cannot rid ourselves of old languages as long as we remain imbedded in the old paradigm of which they are an organic part; and at the establishment of a new paradigm the old problem ceases to exist altogether as it comes to be defined in a different way, generating thereby entirely new concepts. At the present, I see no alternative but provisionally to accept the uncomfortable linguistic ambiguity while at the same time intensifying the efforts to push the notion of social movement towards a creative self-destruction.
Under this conscious theoretical discomfort, I wish to define a social movement as a concept that comprises three analytical dimensions. A movement is the mobilization of a collective actor (i) defined by specific
30 Theory of collective action
the construction of collective action 31
solidarity, (ii) engaged in a conflict with an adversary for the appropriation and control of resources valued by both of them, (iii) and whose action entails a breach of the limits of compatibility of the system within which the action itself takes place. A movement, therefore, does not just restrict itself to expressing a conflict; it pushes the conflict beyond the limits of the system of social relationships within which the action is located. In other words, it breaks the rules of the game, it sets its own non-negotiable objectives, it challenges the legitimacy of power, and so forth.
In order to identify a movement - as a category of analysis rather than as an empirical phenomenon - we therefore have to verify three conditions, each one of which must be met before we can speak of a 'social movement' in any analytical sense. These dimensions also allow a clear distinction from other kinds of collective action theoretically bordering on social movements.
A 'social movement' refers to just one specific form of collective action among many others that combine orientations and fields of different kinds.
My purpose here is to suggest elements of a method rather than engage in a typological exercise. We must at all times keep in mind the limits of any typology: they depend on the dimensions of action that are originally selected by the observer. A refinement or improvement in the analytical
procedures along the lines I have proposed would produce different typologies. For that reason, my remarks are intended to address questions of
method instead of aiming to contribute to production of a comprehensive summary of the various forms of collective action.
From this starting point, we may proceed to next examine more closely some of the possible observable combinations of the categories presented above that refer to actual forms of collective action bordering on social movements.
(i) Social movements were already defined as those forms of action analytically implying conflict, solidarity and a breaching of the system limits.
(ii) In general terms, one may talk of competition when conflict and solidarity are confined within the boundaries of the given system.
(iii) At the opposite pole, forms of behaviour which breach these compatibility limits without, however, implying solidarity and constituting a conflict can be identified as deviance.
(iv) Cooperation designates the area of collective action that is based on
solidarity but not oriented towards a conflict, and which is entirely located within the limits of compatibility of the system.
(v) Upt0 this point, social movements literature, my own work included, has been mainly devoted to oppositional movements, revealing an explicit bias of the majority of the students of collective phenomena. With some important exceptions, much less attention has been paid to what we can call reaction, that area of collective action where solidarity is employed to defend social order even by breaching the system limits. The literature on right-wing movements and counter-movements (see Mosse 1975; Billig 1978; Lo 1982; Zald and McCarthy 1987; Blee 1991) provides good examples of such an orientation. These forms of action turn increasingly towards an explicit fascist character as they move from the organizational level to the mode of production.
Other areas of collective action are less prone to a categorization that would inevitably imply a detailed typology. My methodological purpose is fulfilled here by the simple warning that what we empirically call 'social movements' are in fact composite phenomena of collective action comprising a multiplicity of analytical dimensions. The specific level of collective action that I have analytically called 'social movement' is empirically surrounded by and intertwined with many other forms of action implying differing orientations and affecting different fields. However, one should never forget that collective action takes place not only where it manifests itself in visible mobilizations against public authorities. Collective action is also present in forms of behaviour that apparently never reach any comparable prominence:
(vi) Individual resistance, like the slowdown of work rhythms or sabotage in capitalist factories (Dubois 1976), is not just an individual behaviour. As shown by many studies, such seemingly atomized behavior is a primitive form of conflictual resistance to capitalist power in the workplace, an embryo of class consciousness without which the more visible forms of collective action could not be explained. It is an action which expresses a conflict and breaks the system limits, but which takes the form of an aggregate behaviour. Other forms of elementary resistance which precede more organized forms of behaviour have been analysed in rural societies (see Hobsbawm 1959; Scott 1986; Colburn 1989; Abu-Lughod 1990).
(vii) On the other hand, individual mobility is sometimes an alternative to collective competition, when the channels for the improvement of individual conditions are open and the costs for mobilization are high
32 Theory of collective action
The construction of collective action
(see the classic exit-voice model by Hirschman 1975). Individuals express a conflictual orientation within the limits of the system in an aggregate form that does not reach the level of solidanstic action and looks for atomized individual advantages. (viii) At the opposite pole, collective rituals that publicly celebrate and reinforce social order may sometimes be carriers of social movements, their womb or their mentors (Turner 1969, 1982; Ozouf 1988; Mosse 1975). They are aggregate phenomena that imply consensus and take place within the limits of a given system.
All these levels of collective action should be of main interest for the students of social movements because some of them are always associated with the big processes of collective mobilizations and can provide useful hints for the understanding of the multiple meaning of collective action (figure 2 provides a summary of the present discussion in a graphic form).
A further step in the differentiation of analytical levels of action consists in articulating them with the different systemic fields. This exercise could end up in a rigid typology, which as already stated is not my purpose. I will therefore just give some examples to show the possible applications of the criteria adopted here, knowing that none of these forms of behaviour is by definition 'pure'.
1 Where forms of competition are concerned, claimant action and political competition can be used to address those conflicts that lie, respectively, within the confines of an organizational and a political system. Cultural innovation is a form of action which is in conflict with the bases of the mode of production, but which, at least for a while, keeps within the system's compatibility limits.
2 In the case of deviance, behaviours can be examined, first of all, at the lifeworld and the organizational level. This is the case with most of the behavioural forms that the classical literature on deviance identifies as the product of dysfunctions in integrative mechanisms, in processes of social-ization, and in the agencies of social control. At the political level, we may refer to forms of political violence which break the rules of a political game without any reference to institutional change or to the modification of power relationships. Many forms of expressive violence and terrorist action (which often come to coincide) assume this feature. Certain forms of extreme alienation seek to jolt the fundamental logic of the mode of production by totally, and typically through violent means, rejecting it. However, such action does not develop into conflict, lacking as it does the identification of the stakes and/or of the social adversary. Here the presence of an unborn conflict is detected by its absence, by its negative imprint.
Social movements Competition
Deviance Cooperation
Reaction Individual resistance
Individual mobility Ritual
Figure 2
34 Theory of collective action
The construction of collective action 35
as it were; precisely for this reason such a form of behaviour constitutes a signal that should be read all the more carefully (for a significant example in the case of youth, see Dubet 1987).
As for cooperation, a distinction among the levels of action is easily exposed to the risk of confusing analytical categories with natural phenomena. Many forms of altruism and community action correspond to these analytical orientations at the everyday life and organizational level respectively. Voluntary action can manifest this orientation in the political system, particularly in campaigning, fund-raising, and lobbying activities. Press and intellectual campaigns are approximate examples of activities carrying this orientation and affecting the mode of production.
Distinguishing among the various levels of aggregate behaviour, here too, is made more difficult by their nebulous and scattered nature, and by the increased risk of confusing analytical categories with natural phenomena. None the less, drawing on the existing literature on collective behaviour, at the lifeworld or organizational level we can identify those kinds of behaviour which link up most closely with a crisis or a change in either the functional processes or the instrumental apparatuses of a system (panics, booms). As regards the political system, one may speak of crazes and riots as phenomena that indicate aggregate response to a crisis or a change in the decision-making apparatus. Finally, bearing in mind the conceptual difficulty inherent in referring aggregate behaviour back to a mode of production, fashions are probably forms of behaviour which, at the aggregate level, are one reaction to a crisis or transformation in the production and appropriation of social resources.
Social movements can be distinguished according to the field of their action.
We can now apply the same general criteria to the specific category of social movements. They too can be classified into four analytically different types of behaviour according to the system invested by collective action.
(a) If the conflict and the breaking of the rules take place at the lifeworld level, we can talk of a conflictual networking. Molecular action is taken against the rules governing social reproduction in everyday life through the creation of networks of conflictual social relations. Forms of popular resistance are always present in society, creating a free space that precede visible action (Evans and Boyte 1986; Fantasia 1988; Colburn 1989; Scott 1986, 1990b).
(b) Within an organizational system characterized by roles and functions,
action may be appropriately called a claimant movement. The collective actor presses for a different distribution of resources within an organization and strives for a more efficient functioning of the apparatus. Such action, however, clashes with the power that imposes the rules and decides on the division of labour. The action taken may be in defence of the advantages enjoyed by a distinct category, it may mobilize a group of underprivileged workers, or it may seek to bring about a different distribution of roles and rewards. In doing so, however, it tends to exceed the established limits of the organization and its normative framework. The conflict moves beyond the operative level to affect the production of norms.
(c) Á political movement expresses conflict by breaking the confines of the political system. It campaigns to extend the criteria for participation in decision-making and fights against the bias in the political game that always privileges some interests above others. It seeks to improve the actor's influence over the decision-making processes, or to ensure its access to them, and endeavours to open up new channels for the expression of previously excluded demands, by pushing in any case participation beyond the limits set by the existing political system.
(d) An antagonist movement consists of collective conflictual action aimed at the production of a society's resources. It not only contends the specific way in which resources are produced, but equally challenges the goals of social production and the direction of development as such.
Antagonist movements are by definition the most abstract of the categories proposed so far, since no collective actor can ever be wholly 'antagonistic'. Set within a concrete society, what is currently called a 'movement' operates through everyday networks, organizational systems and the mechanisms for political representation and decision-making. What, then, is the meaning of making this distinction?
There are two points. Firstly, the dominant groups in a society tend to deny the existence of conflicts which involve the production and appropriation of social resources. At the very most, they acknowledge the existence of grievances or political claims, seeking however then to reduce all conflictual phenomena to these only. Secondly, we must acknowledge that not all forms of collective action are antagonistic in their nature and that the functional or political problems of a. society have their own autonomous existence.
Moreover, the degree of autonomy or specificity of political systems and organizational mechanisms vis-á-vis the constraints of social production is a key factor in assessing the impact of antagonistic demands within such
36 Theory of collective action
The construction of collective action 37
systems. There are no antagonist movements in undiluted form, unmedi-ated by the political system or the social organization. A 'pure' antagonist movement, unprepared to equip itself with an instrumental base and without any relationship with the mechanisms of representation and decision-making, tends to break up and disintegrate along the two dimensions that define its action. Conflict and the breach of the compatibility limits are divorced, and the conflict loses its social rootedness and its antagonistic connotations, terminating in a mere symbolic search for alternative - a search easily assuming the features of an escapist and marginal counterculture unable to exert any influence on the crucial mechanisms of the society. Or, on the other hand, limit-breaking action becomes deprived of any conflictual referent (adversary and stake) and turns into that obsessive rejection which, as an end m itself, finds its only form of expression in violent alienation.
In the more undifferentiated societies of the past where the functions of unification and centralization were performed by the state, social movements were unable to express themselves without the mediation of collective action tied to the social organization or the political system. As a result of the increasing differentiation of societies and the greater autonomy of the various systems that constitute them, it is now easier to pursue antagonist action without the mediation of organizations or institutions. Thus we can today witness the appearance of forms of antagonist action which state the issue of control over key collective resources in directly cultural terms. Complete lack of any kind of mediation, however, renders these forms of action extremely fragile. In any case, they probably anticipate, in an embryonic form, the advent of antagonist action that is less constrained by organizational and political mediation, and hence more likely to 'explode' in the two directions I described above.
The movements of the late 1970s and the 1980s were the first signs of the transition from movements as organizational or political actors to movements as media. The movements of the nineteenth century were at the same time social actors - class actors - and political actors, acting for the inclusion of the working class within the bourgeois political system and the bourgeois state (Tilly 1975, 1990; Katznelson and Zolberg 1986). In complex systems these two aspects are breaking apart, creating different and separate processes. On one hand, there are political actors, engaged in action for reform, inclusion, new rights, the opening of the boundaries of the political systems, redefinition of the political rules, and so on. And on the other hand, there are actors addressing the issues in a pure cultural form, or in pure cultural terms - bringing the issue to the fore, to the public. When the issue is named, it can at once be processed through political
means but if it is not named, and until it is not named, it is simply acted tlir, .i'-ih structures, powers, imbalances, domination, and so forth. A movement :ts a pure medium is a form of action that simply brings to the light {1,.; fact that there is a societal dilemma and a conflict concerning some ha-orientations of society. Of course, the observable empirical forms of 4Cii;ni never reach such 'purity', but we can still expect these two trends to ,vr»u;'.n increasingly separate.
Tins transition will not be linear, since political actors will still be needed :i- we continue living in historical societies with political systems, within borders of states, and so on. Without political action, change cannot be institutionalized in complex societies, but movements increasingly act as jiev. media by their very existence. When they escape the risk of pure symbol i( counterculture (see, for example, Marx and Holzner 1975) or mar-uioiil violence, they fulfil their role and transform themselves into new inn*-lotions, providing a new language, new organizational patterns and w,, personnel. This outcome, however, depends on what it is possible to uri'i i fa politically and on the degree of openness and flexibility of the given ni'ii- n al systems.
Every concrete form of collective action has a plurality of analytical meanings.
1 \v jam of our discussion so far has been to suggest elements of a method for the analysis of 'social movements' more than to provide an exhaustive empirical account of them (which at this point could only be descriptive and classificatory). The distinctions I have drawn are analytical. That is to say, they are conceptual instruments to be used in analysing empirical phenomena. A concrete collective actor is always a complex and heterogeneous process which unfolds in reality and which contains meanings of action thai are addressed by the various analytical categories 1 have set forth in what goes before. A collective actor operates within various organizational systems at once; it lies within one or more political systems; it acts within a society comprising various coexisting modes of production. Its action therefore involves a whole range of problems, actors, and objectives.
One dimension may outweigh the others and thus give a particular character to a movement. Alternatively, the dimensions may combine in different ways. An empirical 'movement' within an historical society is often the confluence of the marginal and deviant groups present in a system, and aggregate behaviours form and coagulate within it. On its borders, action dissolves into mere negotiatory behaviour or violent rupture according to whether it becomes wholly integrated within sphere of the rules and limits
38 Theory of collective action
1 he construction of collective action 39
of the system or loses its capacity to locate an adversary and a common field for conflict. For the meaning and direction of a collective actor to be understood, this magma of empirical components must first be deconstructed by analysis and then reconstructed into a system of meaningful relations.
This operation, however, does not proceed in a straightforward fashion. In particular, it is not easy to identify the elements that allow one to speak of an antagonistic orientation of collective action:
(i) The way in which a system affected by collective action responds to the conflictual impulse is a first indicator of the meaning of that action. The adversary - controlling more resources and having thus more to lose -will not permit itself a broad margin of error. The interests under attack react in those areas where they are perceived to be threatened the most, and when a movement is antagonistic this response usually occurs at a systemic level higher than the one that is directly affected by the movement's action. Protest arising within an organizational system and directly challenging the setup of power provokes intervention by the political system and the repressive apparatus of the state. A political movement which pushes beyond the allowed limits of participation arouses a reaction which involves the mode of production (for example, economic crisis, a halt in innovation, the rise of new elites).
(ii) Secondly, an antagonist movement, using the language of its own cultural system, tends to describe the situation as a struggle between those who produce crucial social resources and those who appropriate them. Whether directly or indirectly, what is at stake in this struggle is always the control over these resources - that is, the society's mode of production.
(iii) Finally, in moving from everyday networking to a claimant movement, to a political movement, and then to an antagonistic one, action passes along a spectrum consisting of the following dimensions:
Increasing symbolic content. An antagonist movement campaigns for objectives that always concern the fundamental identity of the actors. This is no longer an issue of control over immediate resources or of acquisition of material advantages, but of the fundamental nature of social production itself; what is at stake is the possibility of giving a different form to, and profoundly reorganizing, the structure and goals of the appropriation of social resources. In this sense, antagonistic conflict strikes at the heart of the cultural foundations of a society.
Decreasing divisibility (or negotiability) of goals. When conflicts are internal to an organization or a political system it is easier to adopt partial strategies and to negotiate about intermediate goals. Antagonist movements embody goals and forms of action that are not negotiable with the existing arrangement of social power and with the forms of political hegemony exercised by dominant interests.
Decreasing reversibility of conflicts. Conflict resolution becomes progressively more difficult as action completes the passage from the claimant movement type to an antagonist movement, and the stakes become increasingly more important for the group concerned.
Decreasing calculability. The ratio between the costs and benefits of the action is clearer, and calculation of the effects of various courses of action is easier, when that which is at stake is more readily quantifiable and when it is possible to identify various alternative solutions. When the stakes concern general cultural orientations of society not everything can be calculated and the affective and emotional dimension (which is not irrational!) becomes ever important.
Solution tending towards zero sum. The further one moves along the spectrum towards antagonist movements, the closer the conflict approximates a zero sum solution. In struggles for the control of social production the stakes are indivisible, whereas in an organization or a political system alt parties to the conflict may hope to gain a partial advantage, and victory for one of them induces only a relative imbalance of gains and losses.
Every form of collective action is a system of action.
The collective action of an empirical 'movement' is the outcome of purposes, resources, and limits. Put differently, it is a purposive orientation built on social relations within a field of opportunities and constraints. It therefore cannot be considered as either the simple effect of structural preconditions or the expression of values and beliefs. Individuals and groups acting collectively construct their action by means of organized investments: in other words, they define in cognitive and affective terms the field of possibilities and limits which they perceive, and they simultaneously activate their relationships to create meaning out of their joint behaviour, so as to give sense to their 'being together' and to the goals they pursue.
40 Theory of collective action
The construction of collective action 41
The empirical unity of a social movement should be considered a result rather than a starting point, a fact to be explained rather than sor thing already evident. Collective action is a multipolar system of ach which combines different orientations, involves multiple actors, a encompasses a system of opportunities and constraints which shapes " actors' relationships. Actors produce collective action because they are a to define themselves and their relationship with the environment (otl actors, available resources, present opportunities and obstacles). 1 process of creating such definitions is, however, not linear: the events which a number of individuals act collectively are the product of the int action, negotiation, and opposition between different action orientatio The actors construct a 'we' (more or less stable and integrated according the type of action) by rendering common, combining and then painst; ingly adjusting three different kinds of orientations: those relating to ends of the action (to the meanings that the action has for the actor), to means (that is, to the possibilities and limits of action), and finally to re tionships with the environment (to the field in which the action takes plac
The multipolar system of action of a collective actor thus organizes its- V-along a number of polarities: the three axes (ends, means, environme constitute a set of interdependent vectors in reciprocal tension. In fact, c lective action has to be able to handle within its own field mutually c< flicting needs, it has to meet multiple and contrasting requirements in ter of ends, means, and environment. It is never the simple expression of <*m» goal-directed impulse; it builds itself out of the resources available to du-actors and located within the field of possibilities and limits of a parti' lar environment. Collective mobilizations can occur and even contii because the actor has succeeded in realizing, and in the course of the act: continues to realize, a certain integration between those contrast requirements. Constant tensions arise among ends, means, and environ. ■ ment: Goals no longer match means or vice versa; the environment is ciť poor or rich in the requisite resources; the means are more or less congruent with the field of action. Even within each one of these three axes, tensions are continually generated: over the definition of ends, between she and long-term ends, over the choice of means, over the choice between al eating resources for the pursuit of efficiency and for building solidarity, over relationships with the environment, between internal equilibrium exchange with the outside.
Collective actors constantly negotiate and renegotiate these aspects their action. This 'social construction' of the 'collective' through repea negotiation is continually at work when a form of collective action occi A failure or a break in this constructive process makes the action impo:
ble. Leadership and organizational forms represent attempts to give a more liable and predictable structure to to this multipolar system, which is permanently subject to stress (see chapters 16 and 17). usually, when one considers collective phenomena, attention is focused on the most visible aspects of the action (events, mobilizations); but these presuppose the generali}/ ignored analytical level to which I have already drawn attention iscc chapter 4 for further development). Visible action is born and persists over rime because the actor manages to achieve a certain degree of integration among the various orientations just described. Undoubtedly, the emergence of concrete actions is aided by conjunctural factors (such as the structure of political opportunities, the existence of entrepreneurs, the extent of equilibrium or crisis in the environment). But it would be impossible for these factors to exert any influence were the actor not able to perceive them and to integrate them into the system of orientations which frames the action.