Gramsci, movements and method: the politics of activist research

Paper to 4th "Alternative Futures and Popular Protest" conference
Manchester Metropolitan University, April 1998

- Abstract -

The western Marxist tradition identifies the active engagement of human beings with their environment and with each other as a central ontological category. This physical, verbal and cognitive engagement is embodied through skill: the practical availability of what are often prediscursive modes of action, generated in collective learning processes such as conflict or alliance, materially sedimented in experience, practices, language, networks and so on, and thus continually subject to transformation or loss, but also constantly available as a resource for creative action. Movements, from above or below, are then different possible "proto-hegemonic" attempts at developing this potential from different starting-points and mobilising it around shared social projects and against others.

Strategies of research into movement contexts parallel these possible organising modes: given the diversity of participants' orientations and of external interventions, there is necessarily a politics of research characterised by collusion with some participants' knowledge interests and conflict with others. The paper draws on Gramsci's conceptualisation of class consciousness to argue for a critical realism that extends the logic implicit in participants' skilled activity to a more comprehensive standpoint, using the researchers' own standpoint and knowledge interests critically as a part of this dialogue. The use of metaphor, illustration and other "hegemonising" strategies are geared to developing this two-way communication between different knowledge interests, which remains precarious unless it is developed into the coordination of shared activity.

Such a politics of knowledge makes sense only given particular starting-points. A concrete example is given in the case of my own research, which moved from a participant's developing choice of priorities to a traditional intellectual's attempt to relate the milieu to externally-determined projects. The class and other relations involved in this process are examined critically, with a view to bringing out the ability of participants to "locate" the researcher and fit my activity in turn into their own perspectives and projects. The cognitive implications of this analysis enable a more complex understanding of such research activity and point to important political and ethical issues around the potential value and limitations of research for participants and researchers alike.

Gramsci, movements and method: the politics of activist research

This paper begins with a paradox in social movement theory, moves backwards to an ontology of human existence, then forwards to outline a political epistemology, and concludes with the history of a research process. Worse, the maps it uses to explore this complex territory are those of an idiosyncratic collection of theorists who are on the whole neither the focus of much serious intellectual attention nor commonly seen as relevant to the discussion of social movements (with one or two exceptions in each case). To complicate matters further, these maps are themselves based on my speculative reconstruction of what I take to be the shared assumptions of this unlikely and heterogenous group of thinkers. The only justification for this paper is that it attempts at least to sketch out a possible perspective on some theoretical and methodological problems raised by research into contemporary social movements. If this perspective is convincing, it may help think about some old problems in a new light; if not, the questions asked may be large enough to provoke better answers from other points of view (1).

The case of the missing theory

The paradox in question is that of the Marxist theory of social movements, or more accurately the lack of such a theory. It is not, of course, that Marxist writing on social movements does not exist; but rather that to the best of my knowledge no systematic attempt has been made to formulate a Marxist theory of social movements. What is by now the standard analysis of the field identifies an "American" mode of theorising, with roots in rational choice theory, and a "European" mode of theorising, normally seen as "post-Marxist" in its stress on the development of "new social movements" (cf. Cohen 1985, Diani 1992); it is commonly argued that these perspectives are now converging, though what this means theoretically is far from clear (Melucci 1989). These are not, of course, the only options on offer; recent years have seen the publication of Weberian (Scott 1990), cognitivist (Eyerman and Jamison 1991), culturalist (Eder 1993) and state-centred (Foweraker 1995) analyses, among others. Yet, acknowledging a steady stream of Marxist critiques of the concept of "new social movements" (Bagguley 1992, Barker and Dale 1997), the only systematic theoretical formulation on social movements from anything like a Marxist position would seem to be the body of writing associated with contemporary critical theory (Habermas 1984, 1987, Offe 1985, Cohen 1982, 1996), which asserts an essentially liberal view of social movements as the defence of civil society and of the life-world.

The paradox here is of course that Marxism has frequently identified itself as a theory from and for social movements: at once a theoretical reflection on the experience of the workers' movement and a source of analyses for the use of that movement (2). How is it, then, that this "social movement theory" possesses no "social movement theory" of its own, no separate and coherent body of theory which could define the nature of social movements, explain their existence, analyse their development and theorise their effects? What can one say of a theory which is no longer capable of reflecting on its own conditions of existence, and whose contribution to the analysis of contemporary movements is limited to the chimerical pursuit of a homogenous and objective class basis to their existence and the drawing of undemonstrable assumptions about the revolutionary potential or otherwise of such movements?

I think one might say that both this imagined critique and much actual Marxist writing on social movements are, to a greater or lesser extent, missing the point. It is a mistake to think that there is no Marxist theory of social movements, and that Marxists therefore have the choice between ignoring such movements, denying their relevance, reducing them to something else or adopting mainstream theories either wholesale or piecemeal. The argument I want to develop in the first part of this paper is rather that Marxism is a theory of social movements; and, perhaps, nothing else.

The Heraclitean perspective

A first glimpse of what this might mean can be offered by the first section of the Communist Manifesto, with its dramatic claim that "The history of all human society, past and present, has been the history of class struggles" (cited from the Ryazanoff edition in Mills 1962: 47). This claim is developed into an analysis of the revolutionary role of the bourgeoisie in the destruction of feudalism and the creation of a new world order, transforming economics and technology, national and international politics, communications and cognition; following this, by the analysis of the development of the workers' movement from the experience of misery to the struggle against oppression, aided by growing concentration and communication, into a complex learning process of increasing political self-confidence and clarity towards another and final revolution. It would be more than possible to distil from these few pages the presuppositions of a general Marxist theory of social movements which was not other than the Marxist theory of history - but paying perhaps more attention to the discussion of the nature of movement activity, its preconditions and the context of its development towards the reshaping of society than has sometimes been the case.

For the purposes of this paper, however, I want to draw on what for lack of a better phrase I shall call western Marxism (Jay 1984, Gottlieb 1989): a body of thought which distinguishes itself within the Marxist tradition by its strongly activist and humanist turn of thought, and its consistent rejection of structuralist and fatalist points of view (3). I am not, obviously, claiming that this is the only perspective from which one can see Marxism as a theory of social movements; but simply that it is a position from which it is easy to see what this might mean in practice.

As the title of this paper indicates, the most developed expression of this point of view is Gramsci's (1971, 1991), and my reconstruction of western Marxism as a theory of social movements depends in particular on my reading of Gramsci and the associated positions of Williams (1980, 1985). The other defining author within this tradition is the young Lukács (1971), who develops a movement-centred ontology and epistemology which is elaborated further by Touraine (1981), despite the latter's abandonment of the Marxist periodisation of history. I would add two other sources for this reconstruction: Thompson's (1963, 1993) analysis of working-class history, theorised further by Vester (1975), and Wainwright's (1994) movement-oriented theory of knowledge (4).

The guiding thread which I think runs through these theories is a commitment to a view of history as nothing other than the product of human activity; and, more specifically, as the product of collective human action, articulated in conflicts which encompass the totality of society and in turn define that totality; conflicts which are not only grounded ultimately in the material activity of human beings but are at the same time conflicts over how that activity is to develop. In other words, I am arguing that western Marxism, so defined, is a theory of social movements, and one which elevates social movements to the central, perhaps the only, feature of the historical process and the social structure.

It could be argued that this in fact represents a correct reading of Marxism in general, but this is not central to my argument either way. It is sufficient to note what this perspective does, and does not, involve. Clearly it does not leave any space for an analysis which sees economic or social structure as anything other than the result of human practices, which would treat them as somehow extra-social. Similarly, it excludes the possibility of reifying power, the state, rationality, discourse etc. as pre- or supra-social. In other words, it is a thorough-going historicism and humanism which treats all features of the social world as in the last analysis the product of collective and conflictual human practices, or in other words of social movements.

Class and hegemony

Within the western Marxist tradition, two names in particular have been given to these practices, concepts which I am arguing represent the Marxist version of a theory of social movements. These concepts are social class, in particular class-for-itself or class culture, and hegemony. In one formulation, which can best be identified in Lukács and Touraine, social movements are class movements in the sense that they are essentially movements of one class only; they represent a subordinate class coming to consciousness of its own situation and interests and expressing that consciousness in conflict with a dominant class which has already achieved this level of self-awareness and self-organisation. In Gramsci's formulation, however, social movements are class movements in the rather different sense that they are movements led by a single class or social formations representing that class; they entail an interaction between the way in which a given class organises its own activities and the way in which it organises the practices of other social classes. Can these positions be reconciled?

One possible answer is that they can be combined if Gramsci's formulation is weakened to the point of representing a purely external alliance between the formal organisations and leadership élites of essentially separately organised social classes. Such situations do undoubtedly occur; a case in point might be the alliance between the Swedish workers' and peasants' movements in the 1930s, an alliance which entailed agreement on the basic outlines of the new social order but left separate parties which even identified themselves with different possible governnments. At this level, however, the "hegemony" of the Swedish workers' movement over the peasant movement was no different from its relations with the representatives of capital - a relationship which might be identified as compromise or even consensus, but hardly as hegemony in Gramsci's sense.

A more plausible and consistent answer would be a historicising one. On this view, it might be said, it is unlikely but not impossible that a class could develop into a "class-for-itself" without at the same time achieving some measure of hegemony over other social classes. This double articulation is of course the normal situation for a ruling class; its active side is expressed by Gramsci's concept of a "passive revolution" or "revolution from above". A ruling class which fails to maintain this hegemony is almost by definition in deep crisis. From the other side, Thompson's account of the "making of the English working class" identifies the important role played in this process by a "demotic culture", including important elements of the petty bourgeoisie; Gramsci's analysis of the need to build links between the working class and the peasantry points in the same direction. (Everything depends, of course, on the question of who is exercising hegemony over whom in such situations, as the post-war history of the PCI illustrates.) It is, however, not impossible under unusual circumstances for a class to attain a high level of self-organisation in isolation, as the example of the SPD in Bismarck's Germany suggests; the case of the PCF in contemporary France points to the possibility of such a situation arising precisely as the result of a loss of hegemony.

If nothing else, I think the problematic nature of these examples illustrates the way in which western Marxist theories of social movements raise central questions about the totality of power relationships and modes of social organisation within particular states. When they are extended, as I think they must be, to the level of an entire society or "economic world-system", to borrow Wallerstein's phrase, we are starting to ask the kinds of questions which should be central to a Marxist understanding of contemporary society. To name contemporary attempts at answering even some of these questions is simultaneously to identify how complex the challenge is: for movements from above, Harvey (1990), Lash and Urry (1987), or Sklair (1995); for movements from below, Katsiaficas (1987) and Arrighi, Hopkins and Wallerstein (1989) offer starting points. In particular, Katsiaficas' discussion of "world-historical movements" (1987: 6) is an ambitious pointer to how much remains to be done, not only theoretically but practically.

Implications for social movement theory

If we give the name of social movements to these collective practices, we are taking quite a different tack from conventional "American" theories of social movements in three respects at least (5). This suggests, incidentally, that attempts to combine theories of "strategy" and "identity" (Cohen 1985, Melucci 1989, Diani 1992) are fundamentally flawed, as are attempts to compare them as if they were talking about the same thing (Scott 1990). The field identified by the two approaches is only partially comparable, if at all: the strategic approach assumes the stable continuation of existing social categories - something which is at least placed in question by the critical theory approach. The western Marxist analysis, however, sets itself the goal of understanding the totality of social conflict, placing any particular manifestation within this broader context and asking after its ability to transform social categories and the relations between them.

This is then the first difference to be mentioned: social movements are not seen as unusual phenomena in need of particular explanation, occasional blips on the otherwise passive or institutionalised landscape of society. Rather, situations of passivity and institutionalisation just as much as situations of activity and unconventional practices need to be seen as part and parcel of an actively created and maintained dynamic tension between opposing social forces. One of the great merits of this perspective has been to open up new areas of social life for analysis: to mention Gramsci, Thompson and Williams is simultaneously to point to the intellectual preconditions for the cultural studies project. Social movements, then, are the way in which human practices are socially articulated.

Secondly, social movements include not only the actions of the dominated and exploited, but also the actions of those who dominate and exploit - including, centrally, the practices of exploitation and domination themselves. The changing relations of ownership (identified in feudal society by Marc Bloch (1961)) and the changing form of the state in capitalism, just as much as the developing forms of political and cultural organisation from below, are forms of collective practice geared to maintaining or transforming social relations. Social movements, then, come not only "from below", but also "from above" - and the presence of the latter is rather more systematic than that of the former.

Thirdly, social movements are not identified with any one kind of social phenomenon. They are neither specific features of a political subsystem, for example, nor particular forms of unconventional organisation. Or rather, they may at times be expressed in these ways, but they may equally be found in the normal movements of capital, the everyday organisation of needs and desires, the thoroughly institutionalised relationships of corporatism. A good example of this openness of form, I think, can be found in the juxtaposition of papers from a session at last year's conference. Colin Barker's (1997) discussion of "moments of collective effervescence" examined those powerful moments during which social movements from below are capable of mobilising vast masses of people in dramatic challenges to the status quo. Mike Waite's (1997) analysis of "flecks and carriers" included among other things a discussion of how movement ideas and experiences survive even in the worst periods of drought and on the stoniest ground. My own paper (Cox 1997) discussed relatively stable "movement milieux" in a time of active, but limited, social movements. From the perspective of the movement as a totality, all of these are important "moments" of a given history. Thus this perspective historicises movement activity over the lifetime of any given movement; it also historicises it, however, over the longer term, as against analyses of supposed "cycles" of movement activity (Brand 1982) or inherent "logics", for example of institutionalisation (Scott 1990), which attempt to insulate the categories of movement activity from longer processes of historical change. Social movements, then, are not static forms, but change in both short and long historical movements in interaction with their opponents.

Researching social movements

If, then, we cannot know a prior what form social movements take, if they can neither be identified exclusively with unconventional political activity nor with politics from below, what is movement research to look for? The logic of the argument I am outlining is that we need to start from more general categories and work our way towards specific analyses of the shape movement activity takes in particular times and places. I want to suggest two such categories in particular. If social movements are the way in which human practices are socially articulated, they can and perhaps must be approached both from the foundational level of the practices being articulated and from the viewpoint of the totality within which, and oriented towards which, this articulation takes place. One way of making this connection, which I have presented elsewhere (Cox forthcoming), is in terms of "local rationalities" elaborated in specific movement milieux. Such rationalities represent an elaboration, a formalisation and a decontextualisation of particular practical (material and social) skills developed in particular social locations; this decontextualisation enables the generalisation of such rationalities as means of articulating multiple social milieux dispersed spatially, socially and even temporally. One example of such a rationality - an extremely powerful one - is the abstract form of capital, which moves from particular forms of local calculation to a "capitalist rationality" capable of coordinating a global economic system. Another such rationality is that known within the Marxist tradition as working-class consciousness, whose formalisation and generalisation of course includes Marxism and the workers' movement.

One difficulty with this language, however, is that "rationality", within the Weberian tradition, has the implication of a formally-given possibility and a logic capable of self-generation and self-reproduction, rather than - in keeping with the argument I am outlining - the hegemonic project of specific social actors, and one which is capable of being both logically and practically self-contradictory (6). I therefore suggest the following working definition: A social movement is the organisation of multiple forms of locally generated skilled activity around a rationality expressed and organised by would-be hegemonic actors, and against the hegemonic projects articulated by other such actors. To avoid confusion, I propose to reserve the word hegemony to the analysis of dominant social projects, and to describe challenges from below as "counter-hegemonic" to express the normally weaker and more fragile articulation of subordinate projects.

Before moving to a discussion of research on hegemonic and counter-hegemonic projects as seen from the "viewpoint of totality", however, we need an analysis of what it is that these projects mobilise, organise and express: a heuristic means of theorising the foundational level of human practices. I want to propose the category of skill to approach this problem, and show (A) how it connects to western Marxist theorisations of human practice; (B) how it is materially embodied; and (C) how it leads directly into the generation of social movements.

The nature of skill

(A) My analysis of skill derives from Gramsci's discussion of the nature of human activity:

"The active mass human being acts practically, but does not have a clear theoretical consciousness of this activity, which is however a knowledge of the world, in that it transforms it. In fact, their theoretical consciousness can be historically in contrast with their practical activity. It can almost be said that they have two theoretical consciousnesses (or one contradictory consciousness): one implicit in their activity and which truly unites them with all their collaborators in the practical transformation of reality; and one which is superficially explicit or verbal, which they have inherited from the past and have accepted without criticism. Nevertheless, this 'verbal' consciousness is not without consequences: it connects them to a given social group and influences them in their moral conduct and in the direction of their will, in more or less energetic ways, which can lead to a point in which the contradictory nature of their consciousness does not permit any action, any decision or any choice, and produces a situation of moral and political passivity. Critical self-understanding thus comes about via a struggle of political 'hegemonies', of opposing forms of direction, first in the field of ethics, then in that of politics, to arrive at a superior elaboration of their own conception of the real." (Gramsci 1991: 13)

Using the category of "skill" to analyse this practical activity has the advantage of representing the engagement of human beings with their environment and with each other as something active, practical and creative, something which cannot be taken for granted or automatically reduced to some externally-given feature of the natural environment; to put this another way, it identifies a "mode of production" as an active, collective (and conflictual) way of doing things, something learnt and developed over time - and by implication something open to challenge.

(B) This enables a connection to be made between this general ontology and a specific historical analysis of institutions. Since Berger and Luckmann (1967), it is clear that institutions cannot exist without knowledgeable human activity; it might be better to say that institutions or practices are a means of making practically available given modes of skilled activity, in the sense that they make the reproduction of production possible. These modes of skilled activity, as Gramsci observes, can be prediscursive in the sense that they contradict currently hegemonic means of talking and organising; they can also, however, become prediscursive through retraditionalisation, in the sense that they are sufficiently institutionalised no longer to need verbal explanation, and in some cases form such a part of "taken-for-granted" common sense that they become abstracted from communicative contexts altogether (consider Habermas' (1984, 1987) analysis of how money and power become sufficient to themselves, beyond the reach of any "lifeworld" understanding). This need not be the case, of course: language is one form in which skilled activity can be sedimented (consider the practical orientations embodied in the languages used to discuss emotion, ethics, and kinship), but only one form among several.

On an aside, this analysis of institutions as skilled activity has interesting implications. It makes it possible to discuss the ways in which such skill can be lost - not only in contexts of deskilling and the obsolescence of traditional forms of skill, but also for example in periods of reaction, which consist among other things in a sustained assault on the institutions that embody the skills of subordinate movements and classes: from political parties and the movement media through to what Gramsci defines elsewhere as the basic mode of reception of a social movement: "a conception of the world with a corresponding ethics". Even the us / them distinction, and basic ethical categories such as solidarity, then, are sedimentations of skilled ways of understanding and responding to the world, and as such subject to erosion and attack.

(C) This points to the third element of the analysis, which is to see human activity as practical learning activity. If skill can be lost, it can also be developed; whether practically, in direct interaction with the natural and social world, or indirectly, for example by transmission of particular modes of organising social movements and of thinking about politics. The point of Marxist theory, and socialist organisations, within the workers' movement is arguably precisely to enable such indirect learning, to avoid having to reinvent the wheel. Social movements are a privileged case of such learning, as Vester's (1975) analysis of Thompson's The making of the English working class seeks to establish. Vester argues that social movements represent "collective learning processes", in which the elements Marx analyses as key to class conflict - an increasingly clearer self-understanding, a fuller grasp of social structure and historical process, and an increasingly adequate mode of organisation and struggle - are generated in the conflict with a movement's opponents. The history of recent decades suggests that skill can be lost as well as developed. Hilary Wainwright's (1994) analysis of the "politics of knowledge" of social movements also points, I think, in this direction, as does, from an earlier age, Banks' analysis of social movements as a form of "social technology" (1972). As we shall shortly see, this is not all social movements are; but these points should be enough to establish an internal link from the bases of skilled activity to the articulation of social movements.

A practical illustration of the nature of this category of skill in social movement contexts can be given from my own interviews into social movement activists. Four different institutional locations were particularly mentioned in these interviews: the Dublin movement scene for its opportunity to learn from other people's experience, the London squatting scene for the development of practical skills, literature on the American 1960s as a source for indirect experience of social change, and interaction among engineering, computer and physics students geared towards solving technical problems. One particular discussion centred on the book Ideal Home (Suspect 1986), produced by London anarchists as a guide to squatting and travelling, and described by Irish squatters in the following terms:

Mick: It's a remarkable book, you really should read through it. "How to break and enter." [laughs] Cheers! [laughter] Legally. You know, it's like, covers the complete legal situation on it, the com, everything.
Shane: What's it, it's just called Ideal Home?
Mick: It's called Ideal Homes, yeah. It's er, you know, "how to squat: the law". "Thankyou!" [laughs]
Shane: D'you remember that -book?
Mick: I forget who used to, it's Crowbar
Shane: Right
Mick: It was a squatters' organisation. -Crowbar used to distribute it.
Shane: -Three of them, there were Crowbar, ASS and BSA
Mick: Yeah. [laughs]
Shane: Cause we had
Mick: We had leaflets from all of them. [laughs]
Shane: Yeah. Em
Mick: Very useful stuff that was disseminated around the place, actually.

Such literature is by no means unusual in this milieu, typically (like Ideal Home) focussed entirely on the practical (from advice on particular acts via discussion of eviction proceedings and details of how to defend a squat through to histories of successful squats and lists of contacts). Ideal Home thus fits into a history of systematic attempts at stabilising and developing particular forms of knowledge, which in the case of contemporary movements goes back at least to Nicholas Saunders' Alternative England and Wales (1975) and Abbie Hoffman's Steal this book (excerpted in Hoffman 1989). Comparable literature exists for continental Europe as well (from details of how to set up alternative radio stations (Network Medien-Cooperative 1983) to details of how to carry out actions against armaments firms (Maass 1983)), and of course much of the alternative press is devoted to such matters, from computer encryption to details of forthcoming demonstrations (for which Green Anarchist was recently closed down by the police). Skill can of course take less tangible or abstract forms than this; but I think this sufficiently illustrates the material and institutionalised nature of skill in social movement contexts.

This ontology of skilled practical activity as the starting-point of human society, then, offers a more general and I think more fundamental starting-point for social movement research than one which assumes specific institutions and practices as defining; it directs attention precisely to the historical question of how skill is embodied in particular places at particular times; and it offers a direct connection to the other available starting-point for research, that of the conflictual social totality, to which I now turn.

Movements, "imputed consciousness" and totality

The passage from Gramsci I quoted earlier continues as follows:

"The consciousness of being part of a given hegemonic force (that is, political consciousness) is the first phase in a further and progressive self-consciousness in which theory and practice are finally unified. The unity of theory and practice, then, is also not a given mechanical datum, but a historical becoming, which has its elementary and primitive phase in the sense of 'difference', of 'distance', of barely instinctive independence, and develops up to the real and complete possession of a coherent and unitary conception of the world" (1991: 13 - 14).

By implication, this political epistemology does not expect to find, in the normal state of affairs, fully-formulated movements from below that are capable of mounting a fully-fledged practical and theoretical challenge to the social totality; such a situation represents rather the limiting case of a counter-hegemonic project on the brink of revolution - as well, of course, as the normal situation of hegemonic movements from above. Nevertheless, it suggests that this limiting case is of great importance in representing the fullest possible self-expression of a given movement, as well as illuminating most clearly its relation to the social totality and to its opponents. What can this mean in practice for research on social movements?

If our categories are to be historical, if they are to be geared to movements as they develop and are eroded over the short and long timescales of conflict, they must be oriented to the whole history of a movement, not simply to its current appearance at a single point in time. But how is this to be done? The western Marxist tradition offers two related ways of thinking the problem. The first is that outlined by Lukács, in his discussion of "imputed class consciousness". It is interesting to note, given the disfavour into which the concept has fallen, that Lukács himself thought that the concept was similar to Max Weber's "ideal type"; in other words that it was oriented to asking what, all other things being equal, one could expect the interests and self-understanding of a particular social class to be: "class consciousness consists in fact of the appropriate and rational reactions 'imputed' to a particular typical position in the process of production" (1971: 51; cf. note 11 on p. 81 for the reference to Weber). The problems with this point of view hardly need to be stressed; ah it is interesting that the obvious criticism - that this legitimates virtually any external imposition in the name of the "true" interests of the working class - is frequently made when these interests are identified as revolutionary; rather less frequently when social interests are identified in more conservative terms.

A second, less "contemplative" approach, to quote Lukács' own later critique of this theory (1971: xix), is that offered by Alain Touraine (1981), in his methodology of "sociological intervention". This is geared to discovering, in his case through a complex dialogical procedure between the researchers and specially constituted focus groups drawn from movement activists and opponents, the highest possible self-expression of a given movement; in other words, starting from its actual position, to see how far and in what directions it is capable of understanding its own nature and interests, those of its opponents, and of articulating an independent social project. The underlying methodology has come in for strong criticism, which I think misinterprets its goal. Touraine is not trying, and the theory I have sketched out here is not trying, simply to describe the specific situation of particular organisations or social groups. Rather, it is trying to identify both the local rationalities which are at the root of a movement's support and the directions in which those rationalities are articulated, theoretically and practically. In other words, it is an attempt to extend the logic implicit in participants' skilled activity to a more comprehensive standpoint. To be sure, one is assuming that such a thing exists in some way. But it is worth remembering that this definition of movement leads us to look, not at all the membership of the Lower Uppington NIMBY Association, but rather at phenomena which are spread across considerable areas of time and space, which express themselves in a great variety of forms of social and political activity, which thematise many different issues and are capable of making some kind of bid for hegemony, however unlikely - and within these, at the "hegemonic" core which both makes possible this breadth of manifestations and which gives it coherence and direction.

Two related assumptions have to be made to buttress this research strategy. These add up, I think, to what can be defined as a critical realist methodology (McLellan 1981; cf. Cox 1994). Such a methodology operates on the realist assumption that there are underlying patterns to the immediately discoverable empirical world, and that these are at least indirectly knowable. It is critical in two respects: first, as we shall see shortly, in that it stresses the social relations of knowledge as a key element in its account of the process of discovery of this underlying real. Secondly, and perhaps more unusually, that its aims are explicitly interventionist. In Touraine's account, the point of such research depends on a rationalist strategy of bringing the movement - or at least the research participants - to a greater level of self-knowledge along the dimensions already mentioned; somewhat like psychoanalysis for social movements. But if we consider what kind of knowledge is gained, and what its use might be, we can perhaps see another point. Clearly to discover this hegemonic project or highest potential of a movement is not to predict that it will in fact achieve this level of articulation. It is rather to say that this is the highest level of articulation that it is capable of achieving as a movement, and thus to say something both about its limits and its potential. This is the kind of knowledge which is useful to movement activists, and in particular to those movement activists who are capable of thinking not only strategically (in terms of the options available given a fixed situation) but of taking the "point of view of totality": of seeing the movement, its opponents and the social totality as all open to intervention and transformation.

The politics of research

But how is such knowledge generated? In the past two decades, the concept of reflexivity in methodology has gained considerable ground. One way of phrasing it is that there is no extra-social means of gaining knowledge of the social world, so that the means whereby that world is known - the social interactions entered into during the research process, and the situation and interests of the researcher themself - are a necessary part of that knowledge, not an unscientific accretion. Curiously, movement research has tended to make little or no reference to reflexivity. Yet two features of the concept are particularly interesting in the light of what I have said so far. The first is that its immediate antecedents are, precisely, movement-linked social theories: Marxism and feminism. The latter case is, perhaps, well-known; the former less readily recognised. And yet the proposition that all knowledge is socially located, that our understanding of the world is closely tied to our interests and our experience, and that there is nowhere outside this "radical sociality" where we could stand, is a basic presupposition of the Marxist theory of knowledge (e.g. Goldmann 1969). It is of course anathema to positivist theories of science; in subtler ways, it is rejected by both Weber and Mannheim. Secondly, and more interestingly, in a western Marxist context this proposition has a political and active edge which it tends to lose rapidly in other formulations. It is not a theory of a passive relation to knowledge given by virtue of simple oppression; rather, it has tended to be formulated in terms of the relationship between given theories and doctrines and their authors' political positions, organisational strategies and the social groups they sought to appeal to. Buried in the internecine polemics of the workers' movement is a sophisticated, and thoroughly "reflexive", conception of knowledge politics.

Perhaps the most interesting feature of this conception is that it draws an implicit parallel between organising modes and strategies of research. Indeed, in western Marxist formulations the two are not necessarily distinguishable, and for good reasons: to know the world, in a critical realist perspective, entails a practical intervention, in that we only know the world insofar as we engage with it, and we only engage with it from the point of view of particular interests. The "politics of research" and the politics of social movements, then, are not two separate things. In each case, a movement is being "constructed" by the bringing together of spatially, socially and temporally scattered practices, by the attempt to relate the understandings of participants to one another and to a more general idea of the nature of the movement, one which also identifies the nature of the opposition and the political implications of the demands being made. To do this is then also to collude with the knowledge interests of some participants and to conflict with that of others; just as activists do, in other words, so researchers find themselves in practice agreeing to a greater or lesser extent with the way some participants see the movement, and organising their research accordingly, and disagreeing to a greater or lesser extent with other views, and organising the research in ways which tend to exclude these other definitions - of the boundaries of the movement, for example. This is, I think, inevitable if the research process is to involve any identification of an object of research, if it is to involve any method of engaging with that object, and if it is to result in any analysis whatsoever. Yet if this is the case, it becomes crucial to be able to give a clear account of the politics of a specific research process.

How can we understand the practice of movement research, then? What are the knowledge interests of researchers, and how does their research activity position them in relation to the movement as a whole? It is conventional to start with discussions of who the researcher is and move "outwards" from this towards a discussion of how they do their research, while ignoring partly or entirely the question of where they do their research. This order suggests a peculiarly contemplative image of research, in which the researcher is essentially unaffected by their interaction with the people they are researching, and naturally avoids the political and activist implications I have just sketched out. Reversing this ordering may be of interest.

Firstly, it is a strange feature of experience that reasons for selecting particular fields of study, within movement research at least, are rarely discussed. Often, it would seem, subjects for study are chosen pretty much "because they were there", with the implication that any other subject for study would have done equally well. The image this conjures up of a ready-made theory and method which can be made to fit any particular point on the social map is perhaps unfair; in some cases at least it is clear that topics for study are chosen because of personal interest or access, yet here again this is rarely theorised, and it is unclear how far this is a question of where people happened to find themselves and how far it is a matter of a personal commitment to a particularly important project. The image conjured up by the former, at least, is of an entirely flat social landscape, where the character of the knowledge gained in one location and of that gained in another is essentially identical. Yet precisely those arguments which ground a reflexive methodology suggest that this is not the case. Perhaps the most charitable suggestion is that the pressure to "do research" and the drive to credentialisation tend to make such questions seem naive. Social movement research may even be fortunate in the proportion of researchers who do have genuine reasons for the work they are doing and have made serious commitments to the areas they are studying. This is not simply an ethical question, but also, very obviously, an intellectual one: not only in that there is no intrinsic value in "lots of research being done" (7), but that if the researcher does not themself engage with the movement in question seriously, they are likely to learn as little as would participant observers who refused to take anything they saw seriously. Thirty years ago, Glaser and Strauss (1967) made the suggestion that the selection of areas for research should be based on the criterion of their possible contribution to the development of theory; with a more political tone, that still seems a useful encouragement. This is not, by any means, to suggest that all research should concentrate on the elites of proto-hegemonic movements; for one thing, such elites are probably capable of doing their own theorising. It is perhaps more to suggest that research should develop in a dialogue with movements, even perhaps to the extent of directing research into areas that the movements themselves are interested in rather than areas decided by the "traditional intellectuals" of the academy, whether that means young postgraduates or widely published researchers.

A good example might be the German Green Party. Unusually within the ecological movement, it had during the 1980s a fair smattering of academics of its own, where the new movements are more commonly dependent on sympathetic and often over-enthusiastic outsiders. Nevertheless, it made extensive and systematic use of outside researchers, particularly during its internal crisis in the late 1980s and early 1990s, publishing collections of contributions from sympathetic political scientists (Frank 1991) and inviting them to address party meetings. One published a selection from his ongoing research as an explicit contribution to the discussions of that time (Raschke 1991). Over a somewhat longer period, members of the party elite (Antje Vollmer, Wolfgang Thierse) were and are involved in the journal of new social movement research, the Forschungsjournal neue soziale Bewegungen. This situation is perhaps unusual in terms of the level of competence and the scale of resources available to the party, but not otherwise. As Tomás Jones has pointed out, there are strong dangers in a situation where research is guided by purely external criteria: the politics of European social movements research - and its funding - has shifted rightwards over the last two decades (1993: 7 - 8). To take the most alarming example, Diani and Eyerman's otherwise fascinating volume (1992) on the methodology of social movements research came out of a European Consortium for Political Research session jointly sponsored by the Italian Consiglio Nazionale delle Ricerche and ... the North Atlantic Treaty Organisation, a fact mentioned without comment in the introduction to the volume.

This discussion of the purposes of research obviously ties into the question of how research is done. This is rather more familiar ground in discussions of reflexivity, and I only want to make two points. The first is to add to the discussion of research methods, and the kinds of social and power relations these set up between researchers and participants, a discussion of the conditions for the employment of these research methods. The application of particular "methods" always takes place in a context of power and inequality; in the case of a social movement, typically one of conflict as well. What situation does the researcher have to be in to use particular methods? I want to consider two examples, from opposite ends of the spectrum. In one case, the use of in-depth interviews with members of radical political and social groups, it is clear that to get anything more than "official stories" or even blank refusals a reasonable level of trust is needed. This is frequently gained in practice by researchers participating and contributing to the movement in ways that can convince participants of their bona fides: whether this is a matter of going on demonstrations, living on sites, or running the office. In the other case, that of the elaborate methods recommended by Touraine (1981) and Melucci (1989), the business of organising large focus-group type events, with video recordings in Melucci's case, and in Touraine's case multiple iterations, seems clearly to rest either on high levels of prestige and resources on the part of the researcher (the methods call for several interviewers) or for explicit deals with some of those involved (which Melucci discusses at some length). Examples could be multiplied, but clearly these general research situations - not simply the "methods", abstracted from the actual social context - need to be examined critically.

My second point is perhaps simpler. It is not, I think, enough to describe the research process, even in these expanded terms, and leave it at that. What is important, for much the same reasons as in participant observation, is a way of bringing research to engage with the politics of the participants. The point in discussions of reflexivity is that the researcher's own standpoint and knowledge interests cannot be separated from the research process. Yet if all that is done is to describe retrospectively what they did, and not why, little has been gained. It has frequently been suggested that more dialogue between researchers and participants would not be a bad thing; how much more than lip service is paid to this is hard to tell. One problem with it, as I can attest from my own research, is that it runs up against the pressure both parties are under (for rather different reasons), and seems to be superfluous to the central interests of both. A subsidiary problem is the question of comprehension. Both participants and outsiders are liable to find social theory in some respects difficult to grasp. This is not, I think, so much a question of vocabulary and training as it is of content. When presented with documents written in jargon, we tend to look for the point, as we conceive it; if we find none, we give up on comprehension. To engage in dialogue with movement participants, then, researchers would need to have something to say to them that participants would recognise as a pointful statement. In other words, to solve the subsidiary problem we need to tackle the main one: to find ways of communicating that are not superfluous to what participants are engaged in, thus to produce research whose politics are of interest to the participants. Obviously the scope for this is immense, from actively taking sides in movement politics to writing for the alternative press. This engagement, however, also needs to be valuable to the researcher within the terms of their own research activity.

One, very obvious value is that there is no better way to improve your thinking than to have it criticised by people who know the situation you are talking about - and those are often few and far between in academia. Secondly, and perhaps less crudely, there is the issue of how we formulate and communicate our understanding. It is well known that examples, illustrations, quotes, metaphors and so on can make a text much more readable, but why? In this case at least, I think it is because this kind of struggle to communicate our meaning to the widest possible spectrum of viewpoints is effectively a "hegemonising" skill. It is not so much by offering an external interpretation of a movement, as through providing an image or a phrase which connects up both multiple knowledge interests within the movement and the knowledge interests of the researcher, that researchers can ground their research in the totality of a movement. In other words, through communicating with movement participants we come to engage more fully with the movement as a living problem and not simply as an isolated data set. Obviously this communication is extremely precarious, and remains so to the extent that it does not develop into the coordination of shared activity. A brief research project does not necessarily contribute much in this direction. There are, then, some intellectual advantages to research on movements we are committed to and people we live with: the more important our communication with the people we are researching is to us, the more we will work on it and the harder we will think about it. "Smash-and-grab research" is precisely the activity of researchers who have no intention of maintaining contact with participants, and who only have to convince academic peers of the value of their research. This is, I think, a stronger form of reflexivity than the simple sharing of backgrounds.

Finally, what can research contribute, and to whom? It should at this point be fairly obvious that to think of it in these terms is slightly misleading. Research being a political activity, it is (in some ways) good news to some participants and (in some ways) bad news to others. Perhaps more importantly, it is immediately received, translated and made use of by different participants for different purposes. Yet not all such purposes are equal; and the point of the "western Marxist" research strategy I have outlined is geared precisely to this fact: to contributing a picture of the movement as a whole, to elaborating its practical rationalities and strategic directions, and to noting its limitations and its potential. Other research strategies have rather different purposes, ranging from simple dismissal of awkward phenomena to organisational discussions likely to be of most use either to the police or to movement bureaucrats. The western Marxist strategy has at least this much to recommend it: that its aims in effect orient it to the interests of the movement as a whole and to those of its participant core; and that it tries to come to the most complex and powerful understanding that can be gained from research on the movement.

History of a research process

I want to conclude this paper with a brief analysis of my own research process as an example of some of the issues I have covered in this second half of the paper. The danger in "reflexivity" of this kind, of course, is that it can easily become self-indulgent, if the claim that our own situation as researchers is an important point of analysis is not taken seriously. To take it seriously, of course, is to subject such analysis to the same kinds of theoretical and political criticism as any other statements, rather than to shield them with claims to personal authenticity, identity, and so on.

My current research (Cox 1997), now coming to a close, started from an interest in understanding and locating a "counter-cultural" network of friends based in Dublin, but including emigrants in Britain, Europe, America and Australia, formed in Irish student politics and London squats, and regularly involved in social movement activity of different kinds. The impulse for the research came from my own association with them, and my own activities as one of the "intellectuals" of the group. Attempts at developing this kind of understanding were and are absolutely normal among this well-read, if largely self-educated, group, so that the research in effect consisted in following a line traced from within this milieu. Pursuing this project, I went first to Hamburg to carry out research; as a postgrad, the research became a defining feature of my situation as researcher and teaching assistant, and came to involve attempting to bring the activities I was already familiar with into the externally-defined categories of the conventional social movements literature. At the same time, this "theoretical" history was paralleled by a political one. In Dublin, as an active member of the milieu, I had been involved in what were essentially its self-controlled activities: a college occupation, a student peace society, a semi-anarchist group, later street theatre and the attempt to set up an infoshop among others. In Hamburg, however, as part of my research I joined the German Green Party, coming to act as liaison between the Party and a peace camp during the Gulf War, and later helping run a local section. Returning to Dublin, the distancing of my theoretical understanding from the local intellectual forms of the milieu was paralleled by increased involvement with the Irish Green Party: running the college branch, editing the party's theoretical journal and eventually coming to act as the party's European representative (8). There is, the, at least a structural parallel between the theoretical and organising relationships involved at each point in time, and one which suggests that an account of the latter is by no means irrelevant to the discussion of the former.

This history also strikes me as a disturbing contrast to Gramsci's discussion of traditional and organic intellectuals. Gramsci's traditional intellectuals were those belonging, in training and particularly in their social situation, to already-existing social movements: priests and doctors, for example. While individuals drawn from these groups - such as himself, whose father had been a disgraced bureaucrat and whose incompleted education was in areas such as linguistics and philosophy - would necessarily form part of the first intellectual cadres of new movements, eventually those movements would form intellectuals of their own, capable of speaking "in their own language". This was clearly a necessary part of the development of hegemony. Yet the example I have just given seems to me far more characteristic of the situation since the 1960s at least, in that those who have some intellectual and organising status in the movements of this period are likely to be students, teachers and other state employees. There are clearly areas where this is less true, but if we are thinking in Gramsci's terms of the development of hegemony, the extent to which even contemporary Marxism and feminism depend on the academy, and the lack of control over institutions of their own, is deeply worrying. A situation where the bulk of social movement research is carried out within the framework of traditional intellectual activity - in terms of the categories I have outlined earlier, within the institutional forms of social movements from above - should, I think, be cause for concern.

Thankfully, however, the situation is not one where participants are purely passive and only intellectuals (organisers, theorists) are active. Participants, rather, are themselves creative agents who are thoroughly capable of "locating" these intellectuals and fitting their activity in turn into their own perspectives and projects. A few obvious examples can be given from my own research. One, relating to movement organisation, is that of participant discourses around what they perceive as formally or artificially organised events. Despite both their participation in externally organised events such as demonstrations and in internally organised parties, squats, projects and so on, a number of participants made comments in interviews such as "nothing happened unless you organised it", suggesting here a distinction between the "organic" and the artificial, and also discussed the value such events had for them in their own terms - as occasions to meet one another and exchange news, to organise other things, and so on. A second, relating to research, is that of interview transcripts, copies of which I have given back to participants as the transcripts have been completed, and papers based on them. Unlike what might be expected from movement organisers and formal groups, participants showed no interest in making changes - other than corrections of transcription - to the transcripts, of controlling the presentation of the group or of challenging the analysis in any way. What they were deeply concerned about was their own anonymity, partly in relation to the police, but also in relation to each other. That this latter concern was justified is shown by the fact that on most accounts one of the most abiding features of interest in copies of papers I have given them is identifying the pseudonymous authors of particular statements. Participants, then, are fully capable of locating the activity of intellectuals and using it for their own purposes. This does not, of course, mean that participants have no interest in the issues they are themselves working on or in the possibility of social change. Rather, they formulate these possibilities in ways which are so different from those preferred in academia and political parties that they tend to see activity in these fields as largely irrelevant, and find what they see as more practical ways of resolving the problems in question.

Given this, what if anything can research contribute, and what are its limits? It is perhaps inevitable that theorists feel they do have something to contribute, and even that they feel it is more valuable than the understandings of other participants. For obvious reasons, it may be worth while suspending judgement on this claim, and instead concluding with three concrete examples from this research that may make it easier to answer the question. One, the most basic, is the example of interviews and transcripts. With only one exception, all participants have expressed a great deal of interest in their own interview transcripts, and in at least one case have circulated the transcript further (despite earlier concerns about anonymity!) It seems that the lapse of time between interview and transcription (1 - 2 years) and the unusual experience of seeing accurate transcriptions of one's own speech, made reading transcripts something which could be done with sufficient "distance" to be useful and "objectifying", in other words to help participants gain a clearer sense of their own self-understanding and history: an important part of "intervention" research. A second example, at the other end of the spectrum, is that participants have given me considerable assistance in producing an alternative magazine that grows out of much the same roots as the research. It is hard to know precisely what value they feel it has, but help has often been given unsolicited, which suggests that it is contributing to the shared project. Thirdly, as I have mentioned earlier, prior to the specific methods and process of research are the social conditions of research. To research a milieu means among other things to find a way of taking part in it and contributing in whatever way one can to that milieu. Thus, for example, one part of this role was for several years to maintain as an open room a college society which served as a crashpad, a drop-in centre, a library and other things for members of the milieu.

Such contributions do not, I think, add up to a picture of the leading role of Theory and Organisation; despite proceeding out of them, they suggest that their main contribution has to be analysed from the point of view of movement participants. Touraine suggests at one point that the process of intervention research is concluded when participants have come to adopt a more adequate understanding of the movement; I want to conclude with the suggestion that we might also ask whether the researchers have come to engage more closely with the participants.