book cover.jpg


Social Anthropology

French republicanism as an anthropological object


Published as Candea, M 2006. 'Anthropology of cross-channel debates'. A response to Fassin (AT22[1]) and Bazin et al (AT22[2])Anthropology Today. 22:4, 24 In his last preface to the oft-republished book Orientalism, a few months before his death, Edward Said reiterated his commitment to the humanistic task of ‘opening up the fields of struggle’ (Said 2003). More specifically, he once again called for his famous book to be read not as an invective against the West (as it had so often been understood by detractors and supporters alike), but as an attempt to challenge the West/Orient and West/Islam divides themselves. Particularly in a post-9/11 world, he argued,

a special intellectual and moral responsibility attaches to what we do as scholars and intellectuals. Certainly I think it is incumbent upon us to complicate and/or dismantle the reductive formulae and the abstract but potent kind of thought that leads the mind away from concrete human history and human experience and into the realms of ideological fiction, metaphysical confrontation and collective passion. [...] Our role is to widen the field of discussion, not to set limits in accord with the prevailing authority. (ibid.: xvii-xviii)

Regardless of whether Orientalism itself has actually helped or hindered this process, Said’s plea to rethink the supposed opposition between ‘Islam and the West’ points to one of the most pressing issues of concern for anthropologists in the present day. But besides this major and far from straightforward task (cf. Tarlo, AT 21[6]), recent contributions on the subject of France in Anthropology Today suggest another, smaller yet possibly essential, item on our anthropological to-do list: to rethink the rift between so-called ‘multiculturalist’ and ‘French republican’ models of society.

Didier Fassin (AT 22[1]) and Laurent Bazin et al. (AT 22[2]) have made it clear: those who wish to gain an anthropological understanding of the French banlieues will need to discard (or suspend) the French republican (from now on, “republican”) notion according to which ethnicity does not form a valid category for analysis. The republican approach makes ‘visible minorities’ invisible – but unfortunately so far to statisticians,1 rather than to racists. An informed anthropology of the banlieues cannot be done from a dogmatically republican framework.

But from what position can we undertake an anthropology of the republican framework itself? When republicanism is described as a ‘political’ or a ‘state myth’, the assumption is usually that it is the misrepresentation of an underlying multicultural reality. In Fassin’s contribution, this is accompanied by a suggestion that France is due finally and inexorably to open its eyes to this reality, which has long been evident to foreign researchers working there. French proponents of republicanism are likely to object, with Keith Hart, that English-speaking observers’ distaste for ‘republican’ policies ‘reflects an unthinking multi-cultural liberalism’ (AT 21[1]), rather than ontological far-sightedness.

But mine is a different objection: firstly, republicanism in France, and multiculturalism in the UK and US, are not straightforward national orthodoxies, but rather matters of ongoing debate (cf. Modood and Werbner 1997, Eller 1997, Hewitt 2005). In France, the republican model has its detractors and its supporters, and both camps are furnished, as such camps often are, with a mix of the reasonable, the well-informed and the crudely simplistic and, I might add, with committed humanists and rather more unsavoury characters.

Secondly, what makes such debates fascinating is that they are centrally concerned with the definition of reality. Take for instance this characteristic exchange between the European Commission against Racism and Inequality (ECRI) and the French state. In a recent report, ECRI admonished France to recognize that it was a ‘multiracial’ entity. The ‘French authorities’2 responded:

‘Although ECRI feels it must consider that “de facto, [minority groups] exist” [...] it must be pointed out that there is no consensus of opinion on this assessment of French sociological reality in the country itself’ (European Commission against Racism and Intolerance 2000).

They went on to suggest not, as some might expect, some kind of ‘integrated’ France, but a complex ‘sociological reality’ made up of individuals with multiple, partial and conflicting identities, which any ‘multicultural’ or ‘multiracial’ reading could only misrepresent.

Some will dismiss this as a cynical post-modern cover for institutional racism, others will hail it as the accurate statement of the complexity of individual self-understandings which multiculturalism tends to forget. But my point is that neither reaction leads us to an informed anthropological study of the debate as a whole. And I contend that such a study is overdue for anthropologists who, as Bazin et al. point out, should be truly at home with anthropology ‘at home’ (AT 22[2]) – including the anthropology of the very debates they are engaged in. This understanding of the debate is not the opposite of engagement, but a precondition for it.

Since what is at stake is precisely the definition of reality, such a study is unlikely to be particularly enlightening if it works from the framework of false consciousness. If the starting point of the analysis is that republicanism is merely a myth covering a multicultural reality (or the reverse), then there is little left to elucidate.

A more promising starting point is the principle that reality is not independent from the tools which are used to describe it. Both multiculturalism and republicanism could be seen as performative attempts to establish a certain kind of reality (Austin 1975, Pels 2002). This means that, rather like the ethnic or cultural differences which are its main focus, the republican/multicultural difference is in a powerful sense both ‘there’ and ‘not there’. Like cultural difference, or the so-called ‘clash of civilizations’ (Tarlo AT 21[6]), it is both an obviously ‘constructed’ and constantly transgressed abstraction, and a performative principle in a constant process of self-realization.

Of course, in order to understand the French riots of Autumn 2005, we cannot take for granted the republican version of sociological reality, in which ethnic groups are a forbidden unit of analysis. Processes of ethnicization and racism undoubtedly played a part in the riots, all the more so for not being officially recognized. But neither can we overlook the fact that this republican notion has shaped reality to the extent that the riots never became in any straightforward sense ‘race riots’, for instance.

I doubt we will understand the multiculturalism/republicanism debate if we persist in thinking of either position as either truth or myth. If, as I have suggested, they are not just accounts of, but also operations upon reality, then we should attempt to study them ‘symmetrically’ (Callon 1986, Latour 1991) that is, to study the difference between them without first taking it (and the respective reality posited by each approach) for granted. In other words, we need to bring the multiculturalism/republicanism difference itself into the ethnographic frame. This does not involve a disengagement from these debates, or some flight to a neutral position: it just means being fully aware of the performative nature of our own contribution, and that of others.

Who knows but that this small item on the to-do list might help us along to Said’s larger one.

Matei Candea

Cambridge University


  1. Or rather it did so until recently. The famous 1978 law – held up in France as a central element in the state’s anti-racist policy – which renders illicit the collection or treatment of data concerning ‘racial or ethnic origins, political, philosophical or religious opinions’, has recently been amended. Data of this kind can now be collected for medical or legal reasons, as well as in the name of ‘public interest’. Political, philosophical and religious groups are entitled to collect data relevant to their denomination (although from their members only). Finally, according to this ruling, the national statistics agency (INSEE) itself is now entitled to collect and treat such data (2004; 1978).
  2. No more specific authorship is available. The following quotations are drawn from the appendix to the ECRI’s ‘Second report on France’, which the reporters themselves introduce in the following manner: ‘In the course of the confidential dialogue process between the French governmental authorities and ECRI on the draft text on France prepared by ECRI, a number of comments of the French governmental authorities were taken into account by ECRI, and integrated into the report. However, following this dialogue, the French governmental authorities expressly requested that the following observations on the part of the authorities of France be reproduced as an appendix to ECRI’s report’ (ECRI 2000).



Austin, J.L. 1975. How to do things with words. Cambridge, Mass.: Harvard University Press.

Callon, M. 1986. Some elements of a sociology of translation: Domestication of the scallops and the fishermen of St Brieuc bay. In Law, J. (ed.) Power, action and belief: A new sociology of knowledge?. London: Routledge & Kegan Paul.

Eller, J.D. 1997. Anti-anti-multiculturalism. American Anthropologist 99(2): 249-256.

European Commission against Racism and Intolerance (ECRI) 2000. ‘Second report on France’.

Hewitt, R. 2005 White Backlash and the politics of Multiculturalism. Cambridge: Cambridge University Press

Latour, B. 1991. Nous n’avons jamais été modernes: Essai d’anthropologie symmétrique. Paris: La Découverte & Syros.

— 2004 Loi n°2004-801.

— 1978 Loi relative à l’informatique, aux fichiers, et aux libertés. 78-176.

Modood, T. & Werbner, P. (eds) 1997. The politics of multiculturalism in the new Europe. London: Zed Books.

Pels, D. 2002. Everyday essentialism: Social inertia and the ‘Münchausen effect’. Theory, Culture and Society 19(5/6): 69-89.

Said, E.W. 2003. Orientalism. London: Routledge & Kegan Paul.