1A review of the social movements literature reveals that social anthropologists have generally not played a prominent role in theoretical and conceptual debates within this field of research. What has prevented anthropologists from engaging in a theoretically-informed analysis of contemporary social movements? Why is there an established sociology but not an anthropology of social movements? What does this absence tell us about the politics of anthropology and the anthropology of politics? This article addresses these questions and considers a range of possible reasons for the 'invisibility' of social movements in anthropology. It is argued that anthropology's failure to study social movements is largely attributable to how political anthropology constructs its object, and particularly to the weakness of its concepts of politics and practice. The article's conclusion is, therefore, that the development of an anthropology of social movements will depend on a more general re-orientation of the discipline's approach to politics.
2Since the 1960s, the field of social movements research has undergone rapid expansion. An important aspect of this development has undoubtedly been the emergence of clearly identifiable « schools » or « traditions » of research and analysis. This has stimulated debate and acted as a motor for theoretical and methodological innovation. As della Porta and Diani (1999: 3) have noted in their recent introduction to the social movements literature, four main approaches can currently be distinguished: the collective behaviour perspective, resource mobilisation theory (RMT), new social movements theory (NSM), and the « political process » model. Drawing on symbolic interactionism, the first of these stresses above all the importance of collective action in producing and establishing new social norms. Social movements have thus been presented by collective behaviour theorists as relatively loosely-structured, informal initiators or opponents of change at the level of a society's value system (Turner and Killian 1957: 3-19, 307-330).
3The focus of resource mobilisation theories (RMT), in contrast, is on the organisational structure of social movements and on the rational or strategic logic of collective action. In the analysis of large-scale mobilisations, proponents of such theories insist on the central importance of « objective » factors such as recruitment networks, links with pre-existing organisations and the availability of financial resources and professional expertise. This is characteristically combined in RMT with an emphasis on the logic of instrumental rationality which, it is claimed, governs the cost-benefit calculations performed by collective actors in pursuit of their interests and objectives (Jenkins, 1983).
4The new social movements (NSMs) perspective, on the other hand, places processes of identity formation and the creation of solidarity, rather than strategic interaction and organisational resources, at the centre of analysis. For NSMs theorists, contemporary forms of collective action involve the articulation of novel identities and conflict over cultural orientations through complex interactional processes which cannot be understood simply in terms of a logic of instrumental rationality. As Cohen (1985: 690-705) indicates, the nature of the relationship between social movements and large-scale societal or cultural changes such as a transition to postindustrialism or postmodernity has also been a central issue in « European » theories, notably in the work of Touraine (e.g. Touraine, 1978).
5The fourth approach currently dominant in the analysis of social movements is represented by « political process » theories (della Porta and Diani, 1999: 9-11). Those associated with this perspective are critical of previous resource mobilisation and new social movements theories for their « neglect of politics » (Tarrow, 1988: 423), in particular their failure to examine the relationship between social movements and the state (Birnbaum, 1993: 166). In contrast, political process theorists have highlighted the crucial role played by social movements in bringing about political change and the implementation of new policies, as well as the importance of the state in shaping forms of collective action.
- 1 See, for example, Escobar and Alvarez (1992), Fillieule and Péchu (1993), McAdam, McCarthy and Zald (...)
6Until the late 1980s, the four schools of social movement research and analysis which I have just described developed separately, with little cross-fertilisation or even mutual awareness. Surprisingly perhaps, it is only within the past decade that researchers have really begun to debate the relative strengths and weaknesses of the different perspectives. The outcome of this has been a widespread recognition that « each of these approaches showed but one side of the coin » (Klandermans, 1991: 17). As a result, the 1990s witnessed an increasing number of attempts to link together elements from the collective behaviour, resource mobilisation, new social movements, and political process schools in an « integrated » theory of social movements1. This was accompanied by the emergence of substantial agreement among scholars from the four theoretical traditions over a definition of the concept of social movement itself. Thus, most scholars currently working in the field would probably find little to disagree with in della Porta's and Diani's recent characterisation of social movements as: « (1) informal networks, based (2) on shared beliefs and solidarity, which mobilize about (3) conflictual issues, through (4) the frequent use of various forms of protest » (1999: 16).
7This definition emphasises, firstly, that social movements are not organisations, like political parties or interest groups, but rather are networks, composed of a diverse range of interconnected and interacting individuals, groups and organisations. It follows, as della Porta and Diani point out, that 'a single organisation, whatever its dominant traits, is not a social movement', although it may form part of one (1999: 16). A second characteristic of a social movement is the presence of a collective identity. This refers to the sense of belonging and the shared beliefs and values which movement participants develop in the course of interaction. A social movement's collective identity links together individuals and groups in a way which transcends specific organisational or group identities and provides a sense of continuity during periods of less intense activity.
8Thirdly, social movements are characteristically engaged in political or cultural conflict with other actors over a range of issues. The latter may include the control and distribution of resources, and the meaning of core cultural or political values, as well as social changes of a more « systemic » nature, involving the transformation or defence of structural relationships of domination. The conflictual action of sociopolitical movements typically involves regular recourse to forms of public protest (for example, demonstrations, occupations or strikes): in della Porta's and Diani's view, this is the feature of such movements which perhaps most clearly separates them from sociocultural movements (1999: 15).
- 2 An important collection of essays on Social Movements and Culture (Johnston and Klandermans, 1995a) (...)
9From the above brief account it should be clear that social movements research and analysis has become a dynamic field of inquiry within the social sciences over the last thirty years. As Arturo Escobar has perceptively observed, however, « (a)nthropologists have been largely absent from this extremely active and engaging trend » (1992: 396), in marked contrast to their colleagues in sociology, political science, women's studies and history. Writing in the early 1990s, Escobar lamented the « invisibility » of social movements in anthropology and the discipline's failure to contribute significantly to debates about contemporary forms of collective action. Nearly a decade later, there is little evidence to suggest that Escobar's call for anthropologists to « pay serious attention » (1992: 396) to social movements has been heeded. Although some anthropologists (e.g. Bergendorff, 1998; Nash, 1992) have recently begun to explore this topic, they have generally failed to relate their work to the theoretical debates mentioned above. Even the present trend towards a « cultural analysis » (Johnston and Klandermans, 1995b) of social movements appears to be passing anthropologists by as well as bypassing anthropology2.
10Given the steady growth of interest in social movements within the social sciences generally, it is important to examine the reasons why anthropologists have made such a limited contribution to our knowledge and understanding of this phenomenon. After all, there is no shortage of anthropological work on millenarian and other religious movements. What, then, has prevented anthropologists from engaging in a theoretically-informed analysis of contemporary sociopolitical and sociocultural movements? Why is there an established sociology but not an anthropology of social movements? These are the questions which the next section will attempt to answer. In the article cited above, Escobar explains the invisibility of social movements in anthropology in terms of five different factors and I will use them to structure my own discussion here. Following Escobar I argue that anthropology's failure to study social movements is partly attributable to the weakness of its concepts of politics and practice. But I also suggest more positively that recent debates about the future of political anthropology indicate possible ways forward for those currently involved in social movements research.
- 3 Escobar's argument here echoes Kathleen Gough's earlier discussion of the factors lying behind the (...)
11According to Escobar (1992), a convergence of five factors has been responsible for the ‘invisibility’ of social movements in anthropology3. The first of these is the prominence of issues relating to textuality and representation in anglophone anthropology during the 1980s and early 1990s. One of the first indications of anthropology's « literary turn » (Scholte, 1987) was the publication of Writing Culture (Clifford and Marcus, 1986), an edited collection subtitled « The Poetics and Politics of Anthropology ». As many critics have since argued, however, the essays contained in this volume tended to devote considerably more attention to textual or literary questions than to the subject of politics as such. In particular, most of the contributors approached issues of power and domination in terms of the construction of the textual authority of the ethnographer rather than through an examination of, for example, the material basis of ethnographic production or the politics of knowledge. Escobar claims that this led to a focus in subsequent debates on the politics of representation. While not denying the importance of this development, he contends that it has produced a rather narrow definition of ‘the political arena’ which has directed attention away from issues such as collective political practice and the relationship of contemporary social movements to political processes (Escobar, 1992: 398, 401).
- 4 It should be pointed out here that Marxist anthropologists (e.g. Terray, 1972; Godelier, 1977; Bloc (...)
12An inadequate conceptualisation of practice is, Escobar maintains, a second factor contributing to the absence of social anthropologists from current debates on social movements. He accepts Sherry Ortner’s (1984) assertion that the concept of practice has become increasingly important in anthropology since the 1960s. One aspect of this is a greater awareness of the need to examine the role of everyday as well as ritualistic practices in the construction and reproduction of social and cultural formations (a point which had, of course, already been emphasised by Malinowski [1922: 24] in his comments on « the imponderabilia of everyday life »). Escobar claims, however, that social anthropology as a discipline has few theoretical or conceptual resources with which to study collective political action and its part in creating the world(s) in which we live. The collective production of social life by social actors, he argues, has been rendered invisible in anthropology by the prevalence of « an individual-oriented notion of practice » (Escobar, 1992: 401). A more satisfactory conceptualisation of practice which recognises its collective dimension can, he proposes, be derived from the work of de Certeau and theorists of popular culture such as John Fiske (Escobar, 1992: 409)4.
- 5 See Kuper (1999: 68ff.) for a discussion of the separation of the study of culture from that of soc (...)
13Thirdly, Escobar states that « divisions of labour within the academy » (1992: 401) have also prevented a recognition of social movements as a topic of anthropological inquiry. There is no further elaboration of this comment, however, and its precise meaning is unclear. Nevertheless, a possible indication is provided by Escobar's argument later in the article that social movements are 'relevant' to anthropology because they involve conflict over cultural meanings as well as social and economic resources (1992: 412). This implies that in the past anthropologists have regarded social movements for the most part as socio-economic struggles and, as a result, of interest primarily to sociologists and political scientists. Although it must be emphasised that Escobar does not develop the point explicitly, his view appears to be that a distinction (or « division of labour ») between anthropology as the study of ‘culture’ and sociology as the study of ‘society’ may previously have functioned to inhibit anthropological research on social movements5.
- 6 The production of knowledge in anthropology may indeed depend, as Escobar claims, on 'dominant mode (...)
14A fourth factor contributing to the paucity of anthropological research on social movements is, in Escobar's view, academic anthropology's detachment from the interests and concerns of the wider society. Escobar argues that the discipline operates within an epistemology « a western will to knowledge » which renders it « abstract, disembodied and disembedded from popular social contexts, [and] accountable primarily to the academy » (1992: 419). In other words, anthropology's ways of constituting and knowing social reality are the product of a particular historical process (Western modernity) which has characteristically involved the separation of academic from other social practices. These « modes of knowledge » have defined anthropology as an academic or scientific discipline, Escobar acknowledges, but they have also « [made] unlikely certain styles of research » (Escobar, 1992: 401)6.
15The marginal place occupied by action research within the social sciences is a good illustration of this last point. As Gerrit Huizer has indicated, the adherence of many researchers to a conception of objectivity understood as « non-involvement » has frequently led to an eschewal of action research as well as a more widespread reluctance to study social conflict at all. Many social scientists have perceived action research as necessitating a personal commitment or partisan involvement which would undermine the scientific or objective status of their work (Huizer, 1979: 396-406). The implication of Escobar's argument is that similar concerns may also have prevented anthropologists from studying social movements.
16The final factor adduced by Escobar to explain the scarcity of anthropological research on social movements is « the decline of collective action » (1992: 401) in society – in this case the United States – during the 1980s. To be fair, Escobar is cautious about asserting too direct a correspondence between waves of social movement activity and the degree of academic interest in the phenomenon. Nevertheless, given that social movements research has « flourished » (Escobar, 1992: 396) in Latin America, Western Europe and North America over the past twenty years, it is still necessary to ask why social anthropologists in these places have been so reluctant to enter the debates. In other words, even if current levels of collective action are comparatively low (which is in fact debatable, see Tarrow ), this does not explain the specific absence of social anthropologists from social movements research and analysis.
17To my knowledge, Escobar's work represents the first sustained attempt to identify the underlying causes of the invisibility of social movements in contemporary anthropology. Although not all of the five « factors » which he discusses are entirely convincing, for the reasons suggested above, there is no doubt that he raises fundamental questions about the current state of the discipline and, in particular, the adequacy of its concepts of politics and practice. It is disappointing, therefore, that the line of argument which he develops has subsequently attracted little (if any) critical comment from other anthropologists. Escobar's programmatic discussion of the relevance of social movements theory and research for anthropology (Escobar, 1992: 402-412) has not prompted a significant debate within the discipline. Similarly, the key conceptual and theoretical issues which he highlights have not been addressed further, even by those anthropologists who have recently turned their attention to the empirical investigation of social movements.
18The theoretical and methodological foundations for an anthropology of social movements thus remain to be established. The present article is intended as a contribution to such an enterprise, and seeks to extend and develop Escobar's discussion of the invisibility of social movements in anthropology. Rather than pursue an assessment of the five factors in terms of which Escobar himself explains this absence, however, I would like here to approach the question from a slightly different angle. As noted earlier, certain aspects of Escobar's analysis appear to apply more to anthropology as it has developed in the United States than to its European branches. This suggests that it may be instructive to « localise » the problem by considering the possible reasons for the failure of anthropologists based in Britain to study social movements. It is to this issue which I want now to turn.
19The importance of examining social movements research (or the lack of it) in a national context has been emphasised by the sociologist Paul Bagguley in a recent article (1997). The initial point which he makes is that there is no sociology of social movements in Britain equivalent to that which has emerged in the US and in other European countries such as France, Germany and Italy. While interest in the topic currently appears to be growing in Britain, Bagguley maintains that:
- 7 The « fragmentation and particularism » of social movements research in Britain has also been noted (...)
20In the early 1990s the area [of social movements] was neither established as a topic of theoretical work, funded empirical research nor teaching within Britain. There is no clearly identifiable « school » of British social movement research and analysis […]. (Bagguley, 1997: 149)7.
21As Bagguley acknowledges, the absence of a sociology of social movements in Britain is surprising for a number of reasons. In the first place, comparable 'sub-disciplinary areas' such as women's studies have developed within British sociology in conjunction with the other social sciences. Secondly, the tradition of sociological theory in Britain is « exceptionally vigorous » as a result of its exposure to both North American and European currents of thought. Thirdly, levels of social movement activity in Britain were high compared to the US, West Germany and Italy in the 1960s and 1970s, the period when the social movement field began to flourish elsewhere (Bagguley, 1997: 151-2).
22According to Bagguley, the limited contribution of British sociologists to social movements analysis is related to the dominance of a « class-theoretical paradigm » in political sociology during the late 1960s and 1970s. He claims that the emergence of a sociology of social movements in Britain was effectively « blocked » by the ascendancy of a theoretical model which interpreted such phenomena primarily as the expression of diverse class interests. The social movements which arose during the 1960s, for example, tended to be regarded simply as « middle class movements ». Bagguley suggests that this « class-reductionist political sociology » prevented British sociologists from asking « the right questions ». The complex meaning of contemporary forms of collective action was obscured and social movements analysis remained marginal to the development of the discipline as a whole. As a result, the theoretical schools and research centres which provided an intellectual and institutional space for the study of social movements in the US and Europe were never established in Britain (Bagguley, 1997: 149-151).
23The argument developed by Bagguley, therefore, is that the class-theoretical approach of political sociology in Britain from the 1960s onwards became a major obstacle to the expansion of social movements research. This is an important thesis, which in itself would merit more detailed examination. However, the particular question which interests me here is whether a similar line of reasoning can help to explain the absence of an anthropology (as opposed to a sociology) of social movements in Britain. As I have already indicated, Bagguley's suggestion is that the class paradigm which once dominated British political sociology had the effect of marginalising the sociology of social movements. In a parallel fashion, I want now to consider the possibility that the failure of anthropologists in Britain to make a significant contribution to social movements analysis is attributable, at least in part, to the shortcomings of post-war political anthropology. Recent critiques of the subdiscipline of political anthropology, I will argue, echo many of the points raised by Escobar with respect to social movements research and offer a way out of the current impasse.
- 8 As both Vincent (1990: 390) and Collier (1997) have indicated, however, the recent decline of polit (...)
24For the purposes of the present discussion, one of the most useful commentaries on the development of political anthropology in Britain since the 1940s is to be found in an essay by Jonathan Spencer on « Post-Colonialism and the Political Imagination » (1997). As Spencer explains, the thirty years which followed the publication of African Political Systems (Fortes and Evans-Pritchard 1940) can be viewed, in retrospect, as the 'heyday' of political anthropology in Britain. Since the 1970s, in contrast, the subdiscipline has « remained obstinately out of fashion » (Spencer, 1997: 1), in spite of growing wider interest in such topics as power, post-colonialism and nationalism. In order to explain the continuing unpopularity of political anthropology, Spencer maintains that it is necessary to re-examine the ways in which anthropologists have traditionally conceptualised the political. He suggests that anthropological approaches to politics have tended to be underpinned by a number of problematic assumptions. These have ultimately contributed to the decline of political anthropology (Spencer, 1997: 3)8.
25The account of « anthropology's problems with politics » presented by Spencer (1997: 3) is an important attempt to prepare the ground for the task of reorienting the anthropological study of politics generally in more productive directions. However, his argument can also be used to throw light on the more specific question of the absence of social movements in anthropology. In other words, the critique of political anthropology which he outlines can help to explain why the analysis of social movements has not figured prominently on the discipline's agenda. Spencer makes two points which are particularly relevant in this regard. Firstly, he shows how political anthropology, from Fortes and Evans-Pritchard to Leach and Bailey, was constructed on the basis of a radical distinction between the political and the cultural. The functionalist comparison of political structures developed by Fortes and Evans-Pritchard, for example, required these to be « stripped of their cultural idiom » (quoted in Spencer, 1997: 4). The subsequent emergence of structuralism did not significantly alter this emphasis on the complete separation of the two categories. As Spencer notes, the end result was an extremely narrow understanding of politics itself:
26Where others spoke of cosmologies and modes of thought, ritual and symbol, unconscious structures and implicit meanings, political anthropology became determinedly unexotic, anti-cultural and dull. By 1970 all the richness and complexity of actually existing politics had been reduced by anthropologists to the micro-study of instrumental behaviour … Political anthropology, so conceived, was the subdiscipline that died of boredom. (Spencer, 1997: 5).
27Writing in 1967, the French anthropologist Georges Balandier had already drawn attention to the limitations of both functionalist and structuralist approaches within political anthropology. Although not referring explicitly to the importance of culture, he nevertheless criticised formalist models for denying the dynamism, instability and antagonistic nature of political systems. Presaging the turn of events later discussed by Spencer, he warned that the dominant types of analysis then employed by political anthropologists would lead only to intellectual « dead ends (voies sans issue) » (Balandier, 1995: 224).
28Returning to the main theme of this section, I would suggest that classic political anthropology's adherence to a rigid distinction between the political and the cultural may have produced an inability to comprehend the nature of postwar social movements. As Escobar has emphasised, there is a widespread view among theorists in the other social sciences that « social movements cannot be understood independently of culture » (Escobar, 1992: 405). The sociologist Alberto Melucci, for example, argues that today's social movements are engaged in conflict over « symbolic resources » (1985), while Touraine maintains that actors are struggling to (re-)define society's « great cultural orientations » (Touraine, 1978: 42). Such writers have thus drawn attention to the fact that the movements which emerged during the 1960s (civil rights, feminism, ecology, gay liberation) were concerned not only with social and economic transformation but also with culture and identity. This frequently involved a redrawing of the boundaries of politics itself and the creation of new forms of political practice. With its tendency to abstract politics from culture, however, contemporary anthropology would have been ill-equipped to appreciate the significance of these developments.
- 9 While this is a plausible explanation, rather different reasons for the paucity of anthropological (...)
29The second point I want to consider from Spencer's article is his contention that anthropological studies of politics have typically excluded the empirical investigation of large-scale institutions such as the state and political parties. According to Spencer, there has been an « unspoken assumption that modern political institutions are either pre-eminently rational and transparent, or anthropologically irrelevant and intellectually unchallenging » (1997: 3). Using one of Geertz's essays in comparative politics (Geertz, 1973) as an example, he argues that anthropologists have tended implicitly to regard « the state » and « civil society » as relatively unproblematic phenomena which do not require investigation in their own right9. Spencer suggests that surprisingly little attention has, as a result, been devoted to the structure and workings of the post-colonial nation-state, even in more recent work on nationalism (Spencer, 1997: 6-7).
30In a passage which echoes Escobar's critique of the literary turn, Spencer continues that a concern with the state or political institutions more generally has also been missing in the post-Writing Culture literature on « power ». Associating this with the « theoretical looseness » with which the terms « politics » and « power » have been used in recent debates, he poses the following question:
31…if everything is « political », what word can we use to mark out that special area of life which people themselves refer to as « politics » (…)? The problem is real enough because, for whatever reason, mass politics–parties, elections, the state [and social movements?]– has been more often than not absent from this literature. (Spencer, 1997: 13, one reference omitted).
32The upsurge of interest in « power » within anthropology over the past decade has not, in other words, led to a greater emphasis on the examination of what Spencer refers to as « the institutional context of modern politics » (1997: 3). If anything, there is a danger that the inflation of the meaning of the political will hinder the future development of this type of analysis.
- 10 It is clear, however, that relating the absence of anthropological research on social movements to (...)
33The key point to emerge from the above comments is that the anthropological study of mass politics, and particularly its institutional aspects, is still at an embryonic stage. As noted earlier, the argument advanced by Spencer is that anthropologists have tended to treat the ritualistic or symbolic dimensions of post-colonial politics as their primary concern; apparently more « rational » elements, such as the state itself, have attracted considerably less attention. While this is undoubtedly one factor, I suspect that another may simply have been a perception (in my opinion, erroneous) that traditional anthropological methods were inappropriate for the investigation of these phenomena. Whatever the precise reasons for the discipline's failure to address issues of mass politics, I would suggest that the invisibility of social movements in anthropology can be viewed as an example of this wider problem. Although social movements are not strictly speaking part of the formal political system, they nevertheless interact in complex ways with the state and political parties, and play a crucial role in shaping and mobilising public opinion. Given that anthropologists have devoted relatively little attention to the state, parties and elections, as Spencer has indicated, it is then perhaps not entirely surprising that they have also rarely investigated social movements10.
34In this section I have suggested that the absence of an anthropology of social movements can be attributed, at least in part, to the way in which post-war political anthropologists in Britain have constructed their object. As Spencer has argued, classical political anthropology tended both to define politics in opposition to culture, and to ignore the institutional or organisational aspects of mass politics. My contention is that one effect of this narrow conception of politics was to deflect anthropologists' attention away from social movements, at a time when interest in the topic was increasing within other social science subjects. Although issues of power and resistance subsequently came to occupy a central place in anthropological debate during the 1980s, I would argue (with Escobar and Spencer) that definitions of the political sphere remained problematic, and served indirectly to perpetuate the marginalisation of social movements research within the discipline.
- 11 See Abélès and Jeudy (1997: 13) for a similar argument.
35The implication of the preceding argument is that the future development of an anthropology of social movements in Britain (and elsewhere) will depend on a more general transformation of the subdiscipline of political anthropology. One way forward would be for anthropologists to display greater sensitivity to what Spencer has termed « the empirical unpredictability » (1997: 9) of the political, by which he means the diverse and sometimes unexpected (to the anthropologist) types of behaviour which people themselves understand as « politics »11. A political anthropology seriously committed to understanding the full range of action which people describe as « political » would not be able to ignore social movements and their relationship with political parties and the state. However, it is also the case that anthropologists involved in researching social movements or collective action more generally can no longer adopt an attitude of « naïvety » (Devons and Gluckman 1964) with respect to the multi-disciplinary body of theoretical and empirical work which now exists on the subject. It is only through a critical engagement with this literature that a theoretically-informed anthropology of social movements will eventually emerge.