The uses of universalism. ‘Diversity Statistics’ and the race issue in contemporary France

Measuring diversity

The French legislation is quite restrictive with regard to the constitution of statistical data related to race or ethnicity: the eighth article of the 1978 bill on computing and public freedom ⁴ prohibits any kind of computerized collecting or recording of ‘personal data that can divulge, directly or indirectly, racial or ethnic origins, political, philosophical or religious opinions, union membership, sexual life or health information’. The Commission nationale de l’informatique et des libertés ⁵ (CNIL) was created by the 1978 bill as an independent administrative authority in charge of its enforcement; its mission is, therefore, to ensure that the statistics and data collected and recorded by public administrations or corporations are not prejudicial to citizens, and public freedom. The 2007 report of the CNIL, Mesure de la diversité et protection des données personnelles (‘Measuring diversity and protecting personal data’), aimed at identifying legal as well as political issues that the setting up of ‘diversity’ demographic tools would face. Recalling the legal constraints around personal data, the report formulates nine suggestions that include the recourse to ‘objective’ genealogical data (the nationality and parents’ place of birth) or the onomastic method (that allows deduction of a cultural origin from the linguistic analysis of a name). The report also preconized more flexibility in the CNIL’s general conditions along with reinforced guarantees of the anonymity and confidentiality of data. Be that as it may, the present legislation and general conditions of the CNIL still restrain significantly the possibility of statistical studies or demographic census that could establish factually the range of discriminations in France.

Exceptions can be made to these restrictions on grounds of ‘public interest’ (Article 42 of the 1978 bill). This invocation of ‘public interest’ must be approved by the CNIL and edicted by the Conseil d’État (State Council). Only three fields now benefit from a special authorization to collect ‘sensitive data’ on skin color: medical research, companies recruiting on the basis of physical appearance (like modeling agencies or production companies in the entertainment business) and finally intelligence agencies and judicial police (police judiciaire). The ‘public interest’ of these two last statistical files has been edicted by the State Council on a CNIL recommendation and allows for ‘the recording of any “specific physical traits, that are objective and inalterable”’ (Debet, 2007: 11). What is remarkable in the motive of this decision is the definition of such data as ‘objective’. In the French context, among the most prominent arguments against the recording and treatment of data on physical traits and skin color is the ‘subjective’ nature of such information, given that on one hand, and according to public administrations such as the CNIL, ‘the notion of race has no [biological] scientific value whatsoever’ (Debet, 2007: 4) and, on the other hand, that there is no national racial statistical nomenclature. In the CNIL arguments, ‘sensitive data’ is often opposed to ‘objective data’ – which can explain its position in favor of the onomastic method or genealogical investigations that rest on objective information such as the place of birth and name, state-controlled through the état civil or birth certificate. On the other side, ‘sensitive data’ such as sexual orientation, religious beliefs, political opinions or physical traits appear categorized as such on the basis of their subjective and even potentially changing nature. Regarding race and ethnicity, the creation of a racial/ethnic nomenclature for the national census would probably diminish its subjective aspect, especially if combined with a collecting of the data based on self-declaration (citizens ticking the box they see as most accurate to identify themselves on ethnic and/or racial grounds).

The main actors of this debate (scholars, head-masters of public administrations, politicians and civil associations) are, however, quite divided on the question of setting up such a nomenclature. The Institut National d’Études Démographiques ⁶ (INED) has conducted a preliminary study to evaluate the French population’s position on the subject, it showed that 72% of the respondents would agree on answering questions regarding ‘ethnic origins’ in public scientific inquiries. The CNIL, the INED and the Institut national de la statistique et des études économiques ⁷ (INSEE) are nevertheless hostile to the idea of introducing a national nomenclature for ethnic and/or racial identifications. On this issue, the CNIL report gives rather surprising information. Quoting the INED study, the CNIL asserts that if the ‘migrants’ and their ‘direct descendants’ are mainly reluctant to ethnic and/or racial questions in the national census, ‘it is not the case of “blacks”, which contradicts the idea that the minority groups most vulnerable to discriminations are also the most opposed to categorizations’ (Debet, 2007: 40). What is surprising here is the apparition of the word ‘black’ to designate a population in an official report. As made clear by the debates around ‘diversity statistics’, in the French context such designation would usually be seen as a subjective information, disrespectful of the consensual scientific criterions of objectivity: the idea being that since race is a notion to be understood has having no biological value whatsoever, it cannot be used to construct a population group in statistical studies. But what strikes the most is probably the assertion that ‘blacks’ constitute the minority group ‘most vulnerable to discriminations’: in a country where migrants and French citizens of multiple origins live, and that prohibits diversity statistics, such assertion cannot be grounded on accurate empirical data. Clearly, the CNIL itself struggles uneasily with the deep damages that the restraints on ethnic and racial statistics cause to social sciences research and to public policies and administrations. Furthermore, one could say that the absence of statistics even lays the ground for a continuing racialization of the public sphere: the lack of factual data allowing for all kinds of groundless assertions about the social conditions of ethnic and racial minorities. One such case has been particularly discussed in France, and created a national polemic, when TV journalist and right-wing columnist Éric Zemmour asserted that ‘most drug dealers are Blacks and Arabs … it is a fact’. ⁸ If the words are outrageous, they nonetheless led to endless debates on the potential accuracy of the statement; however, the absence of any statistics based on the distribution of convicted dealers according to racial identity or identification could not, in any way, lay the ground for a factual discarding or endorsement of the assertion. The arguments disapproving Zemmour’s words were, therefore, themselves trapped in the hazards of moralism: voicing a protest against the sheer racism of the assertion, but actually unable to demonstrate it. Far from being grounded on what Jürgen Habermas (1985) described as ‘a discursive rationality’ supporting validity claims (p. 72), public deliberations on racism as they take place in French mediated arenas rather seem grounded on highly unstable moral grounds unable to provide for conditions of acceptability and recognition of antiracists’ claims.

In December 2008, the debate on ‘diversity statistics’ took a new turn with the nomination of Yazid Sabeg as the High Commissioner on Diversity and Equal Opportunities (Haut-Commissaire à la diversité et à l’égalité des chances). Sabeg argued for the development of new statistical tools fit to evaluate the make-up of French society, as well as the improvements and drawbacks in terms of diversity in all the realms of social life (Coroller, 2009). He created a Committee for the Recording and Evaluation of Diversity and Discrimination (Comité pour la mesure et l’évaluation de la diversité et des discriminations; known as COMEDD) headed by François Héran, president of the INED. Before it even finished its report, the COMEDD had met a fierce opposition, mainly coming from the Alternative Think Tank on ‘Ethnic Statistics’ and Discriminations (Commission alternative de réflexion sur les ‘statistiques ethniques’ et les discriminations; known as CARSED). This alternative commission was headed by Hervé Le Bras, renowned scholar and key player in the so-called demographers’ controversy (la ‘controverse des démographes’) in which he opposed the use of ‘ethnic categories’ in an INED study on geographical mobility and social inclusion. In the wake of this late 1990s controversy, the CARSED underlines the prescriptive nature of public statistics and aims at warning public opinion on the risks of an ‘ethnicization or racialization of French society’ (CARSED, 2009). However, how can one think that French society is not already filled with forms of racializations where reminiscences of scientific, cultural and vernacular racisms superpose? The idea that any ‘measure of diversity’ would contribute to ‘racialize’ French society seems, in a way, grounded on the premise that France is already free of any racial discriminations and social tensions. In an interesting formulation, the CNIL stated that in France measuring diversity could prove ‘uneasy’ because ‘it can contradict the French model, equalitarian and universalistic’ (Debet, 2007: 4). This assertion, which is particularly important – as it constitutes one of the main reasons of the CNIL’s reluctance to agree on the recording of racial and/or ethnic data – is also quite ambivalent. Does it mean that as a mode of particularization, ethnic and racial statistics contradict the principles of universalism? Or does it mean that the data these statistics could provide, asserting the actual state of discriminations in the country, would refute universalism and expose it as a pure ideal with no social reality whatsoever? The stance of the CNIL and of the CARSED appears as a form of wishful thinking, taking the abstract republican value of equality for granted, before any sociological or demographical public evaluation. ‘Does fighting against discriminations’, says the CARSED report, ‘necessarily entails the brutal definition of ethnic and racial identities that would inevitably end up constituting closed, rival and artificial minority groups?’ (p. 8). The premise of the CARSED position is obviously that of the nonexistence of racial minority groups and identities. Logically, under these conditions, a statistical treatment of discriminations seems rather useless: if there are no ethnic and racial minorities in France, what power relations could these data record? To oppose any ‘diversity statistics’, the CARSED presents the French republican universalistic model as efficient … in fact, so efficient that it may not suffer any public evaluation. By underlining the possibility of a ‘communautarisation’ of French society (an old trope of the French public sphere, expressing its anxieties vis-a-vis the multicultural US model), ⁹ the CARSED chooses to highlight the theoretical ‘indivisibility’ of the French Republic and to put aside the material social divides produced by racial and ethnic discriminations.

Indirect discriminations

France’s reluctance to step forward on ‘diversity statistics’, whether expressed by scholars’ associations or public administrations and state legislation, is particularly surprising considering the fact that, following the launching in 2001 of a European community program against discrimination and the introduction of the new legal category of ‘indirect discrimination’ a year before, the Council of Europe highly recommends that European Union States produce statistical data to map out discriminations. Until then, the French legislation favored the notion of ‘direct discrimination’, a legal approach grounded on the necessary demonstration of the defendant’s explicit intention to discriminate or use of prohibited criterions of selection (Simon and Stavo-Debauge, 2004). Besides the fact that such demonstration often proves impossible due to the lack of documents or testimonies to attest it, this approach tends to isolate the incriminated person from the network of collective frames and processes of action that produce discriminations. On the opposite, the notion of ‘indirect discrimination’ specifically aims at revealing the potential prejudiced nature of policies, criterions or practices, even as they may appear neutral and whether they were conceived with any explicit intention to discriminate at all. According to the 29 June 2000 European directive – the aforesaid ‘Race directive’ – indirect discrimination can be found when such ‘policies, criterions, or practices may disadvantage persons of a specific race or ethnic origin’. The notion was designed as to identify the social fields where racial and ethnic minorities are de facto excluded or marginalized. France introduced the legal notion of ‘indirect discrimination’ in its Labor Code (in Articles L.122-45 on discriminations and L.123-1 on gender equality in the workplace) but a European Commission report on the ‘limits and potentials of the notion of indirect discrimination’ observed that France did not include a ‘clear definition of the notion on its national legislation’ (Tobler, 2008; see also Hummel, 2008). In a way, it may be said that if the French definition of indirect discrimination does respect the letter of the European directive, it does not, however, respect its spirit. The reluctance to provide statistical data that could record de facto discriminations – which would open the door to racial and ethnic statistics and census – seems particularly to limit the application of a notion defined in the European directive as a type of discrimination that ‘must be established by every means possible, including by providing statistical data’. For the notion of ‘indirect discrimination’ must indeed lean upon statistical tools; tools that do not aim at ‘measuring discriminations’ as such – given that discriminations do not constitute facts prior to statistical treatment – but rather at exposing significant gaps due to the variables ‘race’ or ‘ethnicity’. Statistical data can therefore allow for a factual rendering of indirect discriminations: in other words, it can help provide an objective account of situations of unequal treatment that social actors may not perceive as such (Simon and Stavo-Debauge, 2004).

If the decision of creating a national nomenclature for ethnic and/or racial categories lies in the hands of the legislator – under the monitoring of the Conseil Constitutionnel, responsible for the strict respect of the Constitution – Patrick Simon and Joan Stavo-Debauge (2004) remark that there are nonetheless possible legal alternatives to statistical studies on ethnic and racial discriminations. ‘The French legislation, they assert, is [ … ] equipped to authorize such operations’ (p. 75). On one hand, the notion of ‘public interest’, which appears in Article 8-III in the 1978 bill on computing and public freedom, could justify the production of statistical data on race and ethnicity. The CNIL itself underlined and criticized its definitional flaws (Debet, 2007), but the ‘public interest’ of such studies might, as a matter of fact, be politically and legally asserted considering the European directives pointing in this direction. On the other hand, the recourse to such statistical studies and data has been authorized in several judicial cases, when courts estimated that such data were ‘necessary to certify, to exert or to defend a legal claim’ (Article 8-II, paragraph 4). ¹⁰ According to law, the collect and statistical treatment of data relative to race and ethnicity could be authorized in some judicial cases, if it can help prove discrimination and therefore bring about a compensation for the complainant.

The current overlap of the French ‘integration model’ and European ‘diversity model’ produces multiple incoherencies and paradoxes in the public struggle against discriminations. It particularly tends to euphemize the ‘race issue’, bringing to the front of public policies expressions such as ‘visible minorities’ (minorités visibles), ‘foreigners’ (étrangers), ‘migrants’ (immigrés) or even ‘migrants’ descendants’ (descendants d’immigrés). If the recourse to such terms is motivated by the necessity to avoid any direct reference to race or ethnicity in a country that prohibits so strictly their uses in public studies and policies, they nonetheless prop up strong stereotypes. The case is made especially clear by the category ‘migrants’ descendants’ that tends to partially denationalize French-born citizens, or with the category of ‘foreigners’, often improperly used to describe non-white populations in a profoundly multicultural country. Such public policies and studies categories appear both as stigmatizing and inefficient to establish factually the state of racial and ethnic discriminations (Simon and Stavo-Debauge, 2004: 78–80). They also bring to the front the complex interplay of language and power, and in particular ‘the significance of language in the production, maintenance, and change of social relations of power’ (Fairclough, 1989: 1). Tracking the trends and changes in public lexicons, in the mediated public sphere as well as in institutional language, therefore allows us to account for the contingent and dynamic nature of power relations as they are articulated through specific arbitrary closure of meaning, as Stuart Hall (1985) put it. Through the method and theory of articulation in particular (Slack, 1996), Hall suggested that the close analysis of discursive formations can lay the ground for an understanding of the conditions by which meaning and power relations come, through articulation, to fix subject positions. This process of ‘temporary suture’ by which subjects are caught in discursive formations can be observed in a French context in which the euphemizing of the ‘race issue’, which is the inability to name the specifically color-based nature of continuing discriminations, ultimately produces vicarious forms of essentialism fixing and trapping racial minorities in the figure of the ‘eternal migrant’ or ‘stranger’.

An ideological debate as fierce as that on ‘diversity statistics’ rather seems to require a pragmatic perspective based on the legal, political and social aspects of the issue of the proper ways to measure racism. France’s state of blissful ignorance in the face of continuing racial and ethnic discriminations asks for a better understanding of these damaging social phenomena. If the basic principles of scientific carefulness must be respected – the idea that ‘in categorization lies knowledge as well as oppression’ as Colette Guillaumin (1972: 183) puts it – the development of targeted, coherent and coordinated public policies against discriminations must necessarily rest on statistical treatments that can objectify, or render visible, direct and indirect discriminatory processes as actual social facts. Furthermore, as Didier Lassale (2011) expressed it ‘for European countries, there is now little scope left for truly national policies to be designed and implemented in the field of “integration”’ (p. 241), pointing out that the times when the French State bluntly opposed to multiculturalism in defense of its approach based on ‘social cohesion’ through universalism seem to be over, while post-Blair Britain may be perceived as having turned multiculturalism into ‘a ghost of the past’ (p. 240) in favor of a new emphasis on ‘integration’. In this changing context, the French social model of intégration and the European Union model of ‘diversity’ do not appear antagonistic as much as overlapping in a complex interplay. If the national debate on ethnic and racial statistics seems far from over, it must be noted that the current public redefinitions of antidiscrimination policies in France now rest on discursive repertoires intertwining the ‘diversity’ and ‘integration’ political vocabularies.

Absolute universalism

In this ongoing debate, the discourses and beliefs of social actors are heavily influenced by the political doctrine of universalism. Yet there seems to be two different types of universalism conflicting: an ‘absolute universalism’ and a ‘reflexive universalism’. Two types that should be understood as ideal types in the Weberian sense, designed to capture conceptual relations without mechanically reflecting empirical reality.

The epithet ‘absolute’ defines an abstract and theoretical type of universalism, confusing an ideal social state with the realities of French society. This absolute belief in universalism seems to exempt the republican ideal from any sociological confrontation – as exemplified by the opposition to ‘diversity statistics’. By rendering absolute the constitutional principle of citizens’ equality – so much that material inequities may be ignored – this type of universalism mixes up equality de jure and de facto. Absolute universalism can constitute the ground on which, in the French context, racism is defined as a simple acknowledgment of ‘differences’, whether that acknowledgment takes the shape of statistics, quotas or even identities. This definition of racism can sometimes operate as a safeguard against the risks of racialization of the nation-state (Balibar, 1992: 241–256). But somewhat paradoxically, it also prevents any sociological or political acknowledgement of the racialization processes that do produce symbolic as well as material differences. This important limit of absolute universalism is particularly striking in the opposition it fuels against the use of the notion of ‘race’, even in contexts of antidiscrimination studies and policies and where it is understood as a social construction.

This opposition to a constructivist use of the notion of ‘race’ takes its roots in a confusion between racialism and racism: by thinking race only as an obsolete biological and anthropological concept, any conceptualization of racism as a multifaceted social fact that has been partially des-articulated from the realm of natural sciences is rendered impossible. In the perspective of absolute universalism, the fall of the biological and anthropological concept of ‘race’ in the wake of World War II seems to ratify its pointlessness to address social relations, thus giving, quite ironically, natural sciences a sort of epistemological advantage on social sciences. Critical voices leaning on absolute universalism to oppose the use of the notion of race thus relegate thought in a kind of monism that does not allow for the acknowledgment of the relative independence of social and natural forms. This monist tendency bears a crucial political and epistemological risk: that of the biologization of the social. For example, does the scientific assertion of significant genetic variations within the human genome mean it could, in the future, ratify the social relevance of ‘race’ as an organizing feature? This means that this kind of monism does not only prevent any sociological account of the symbolic dimension of race and of its material effects on French society (Bonniol, 1992: 11–53) but also that it may submit social sciences to the hazard of biological models and theories.

Europe and France have faced in the last 10 years a rise of racial violence and discrimination (European Network against Racism (ENAR), 2008; European Union Agency for Fundamental Rights (EUAFR), 2011) while Europe has seen a dramatic political turn toward extreme-right parties, with parties such as The Northern League in Italy or the Freedom Party of Austria participating in state government coalitions, and while France’s President Nicolas Sarkozy has been heavily criticized for having broken the ideological dam that used to set apart the right conservatives from the xenophobic National Front. Within the European Union, the constructivist notion of ‘race’ has been used to design a community program against racism and discriminations (see the UE ‘race directive’, 2000). In such a conjuncture, the appropriate question that the opposition to the notion of ‘race’ addresses is: what other notion could be used to apprehend sociologically and politically fight against such discrimination and violence? A similar question was addressed by the French academic journal Mots in 1992, in a published debate on the problems raised by the use of the word ‘race’ in the French Constitution. (The word appears in the first paragraph of Article 2 stating that the French Republic ‘guarantees the legal equality of all citizens without distinctions of origin, race or religion’.) Philosopher Étienne Balibar (1992) has taken position in favor of maintaining the word ‘race’ in the Constitution, asserting that ‘things need to be called “by their name”: not their exact name, as a meticulous scientific nomenclature would establish, but their historically determining name’ (p. 249). This debate is indeed still a very lively one, as it resurfaced during the 2012 presidential campaign, when soon to be elected François Hollande declared that as President he would ‘suppress the word “race” from the Constitution’ because ‘in the Republic, there is no place for “race”’ (Le Monde and Agence France Presse, 2012a), provoking a large public discussion on the dreadful history of the word as well as on the potential consequences of the dissolution of what still remains the main constitutional protection against racism.

Besides grounding the oppositions to the constitutional or sociological use of the notion of ‘race’, absolute universalism also tends to abstract social actors from race relations. This abstraction is discursively produced through the construction of republican universalism as a ‘neutral’ position. Yet, just as Barthes’ (2002) neutral, this neutral does not entail a pure withdrawal, but a responsive third term opposed indiscriminately to racism and antiracism. As it have appeared in a sociological field study conducted in France through semidirective interviews and focus groups (Cervulle, 2013a, 2013b), this third term corresponds to a type of whiteness defined by a suspension of action. Social actors embodying this type of whiteness situate themselves in a so-called neutral position, away from both racism and antiracism. Interestingly, this suspension of action can precisely be seen as that which allows for an untroubled enjoyment of the social gratifications that may benefit white social actors in a society structured by systemic racism. It can indeed dissipate guilt, reassure oneself in its feeling of embodying ‘the universal’ (and therefore the French ‘republican values’). But it prevents any critical interrogation of whiteness and of the social gratifications of systemic racism, as the imperative of neutrality does not allow for an acknowledgment of the racial dimension of those gratifications.

Reflexive universalism

‘What one means by “the universal” will vary … ’ writes Judith Butler (1995):

This is not to say that there ought to be no reference to the universal or that it has become, for us, an impossibility. On the contrary. All this means is that there are cultural conditions for articulation which are not always the same, and that the term gains its meaning for us precisely through the decidedly less-than-universal conditions of its articulation. (p. 129)

This definition of a contingent universalism, dependent on its conditions of emergence, could resonate with that of what we choose to call ‘reflexive universalism’ to describe a certain mode of understanding universalism in the French context. It signals a strategic use of the contingency of ‘universalism’ described by Butler. The epithet ‘reflexive’ describes a type of pragmatic universalism, whose articulation may be changing to adapt to the historical conjuncture and social context where it is formulated. Unlike absolute universalism, which can lay ground for a suspension of action in the face of systemic racism, the reflexive vantage negotiates the terms of action accordingly to context and conjuncture. It can, therefore, serve as a mode of legitimizing ‘diversity statistics’, with the notion that the universal value of equality is precisely endangered by the historical and social effects of continuing racializing processes. The use of a ‘sensitive category’ – such as race – for the purpose of diagnosis could, therefore, be accepted precisely in the name of universal values. Such a case has been exemplified by the COMEDD report. Taking a stand in favor of ‘diversity statistics’, and trying to conform such a tool to the scientific, legal and political constraints, the report struggles with the ‘absolute universalism’ vantage that asserts the inadequacy of the notion of ‘race’ to understand discriminations. By replacing the absolutism of this vantage by a more pragmatic approach, the COMEDD report asserts that

the position claiming that ‘race does not exist’ may be absolutely right, scientifically grounded and politically necessary, it however does not exempt us from studying this particular social relation that racism is and to see how it affects the trajectory of the people it targets.

Therefore, ‘the sociologist must … produce a social reading of racial relations, when the discriminator produces a racial reading of social relations’ (Héran, 2010: 18). Interestingly, a 2011 tract of the antiracist and postcolonial political movement Les Indigènes de la République – distributed during a demonstration against ‘State Racism’, after the radicalization of the Sarkozy administration on immigration, and multiple polemics ¹¹ – echoes the COMEDD report from a different vantage, by asserting that ‘race does exist’, emphasizing the continuing racialized processes of exclusion, violence and social inequities in spite of the claims and celebration of a said discrimination-free France supposedly illustrated by the recession of the word ‘race’ in the public sphere.

In these examples, reflexivity is directed toward universalism – identifying what it can enable and what it can obstruct in precise contexts – as well as toward the state of French society. Republican universalism and social relations are, therefore, weight against each other: one revealing the flaws of the other. Universalism understood as open to negotiation, as well as useful to diagnose the ills of French society appears no longer as an absolute doctrine, but rather as an ethical principle that can help measure the gap between political values and social facts. This resignification of universalism can be seen as a minimal definition of its terms that allows for its translation in diverse contexts.

By accepting the widest postulates of the hegemonic definition of universalism – equality, freedom, State neutrality – the reflexive vantage enables a participation in the public sphere without threatening the social beliefs that resist its absolutism; considering that the adoption of universalism as a discursive mode appears as a condition of intelligibility within the French public sphere. The reflexive vantage, therefore, carries out a double process of translation: on one hand a local and contingent translation of the ‘republican universalism’, that can guarantee a negotiated access to the public sphere and, on the other hand, the translation into the hegemonic discursive regime of a discourse often marked by an experience of minoritization. This reflexive perspective may manifest itself politically when minority groups situate their claims in a slightly extended universalist frame or reclaim the universalist frame as a ground for antiracist politics. It has, for example, been the case of the association Collectif contre l’islamophobie en France (Cif), which has carefully articulated its discourse around the notion that ‘national cohesion’ and ‘republican values’ were not put in danger by identity politics but rather by the persistence of racism and anti-Muslim discriminations. ¹²

Conflicting universalisms

By enabling a negotiation of the terms under which universalism comes to articulation, this reflexivity paves the way for an acknowledgement of the myriad ways by which ‘the universal’ is perceived, interpreted and experienced. It could, therefore, usefully be read through the prism of William James’ pluralist perspective. James’ radical empiricism

leads directly to a negation of monism, to pluralism. Because if there is one empirical fact, it is the plurality, the diversity, the multiplicity of phenomenon and the impossibility to reduce them to this absolute unity claimed by monist philosophers. (Flournoy, 1911: 95)

James’s defiance toward philosophies of the Absolute introduced the notion that ‘the sundry parts of reality may be externally related’ (James, 1909: 127). In a sense, the British Cultural Studies’ theory and method of articulation (Slack, 1996: 112–127) might be seen as walking in the footsteps of this pragmatist emphasis on the reciprocal exteriority of the components of reality. The theory of articulation concerns the contingent modalities of social construction. One could, therefore, in a cultural pragmatist mode, analyze the plurality and contingency of the articulation of universalism: how come disparate discourses came to be articulated to constitute such a hegemonic discursive regime in the French context? This perspective implies that a hegemonic regime – material and/or discursive – is the product of internal antagonisms. The reflexive readings of ‘universalism’ deployed by institutions and social actors illustrate in a way its built-in antagonisms. This reflexivity underlines the degree by which the discursive repertoires of universalism rest on a social consensus around its terms that emerge from conflicting interpretations of ‘the universal’.

The temptation might be strong to read the conflict between ‘absolute universalism’ and ‘reflexive universalism’ as a return in disguise of the opposition between universalisme et relativisme. It might, however, be better understood as a form of popular echo to the philosophical tensions between essentialism and constructivism. In the wake of the debates around the conditions of articulation of human rights, in the context of a critique against eurocentrism and the Western bias in international politics, the opposition between universalism and relativism has often been dismissed as inhibiting action. In this dismissal, both notions are understood as abstract and absolutist notions: universalism is seen as a tendency to Westernize ‘the universal’, while relativism is perceived as reifying cultures (Panhuys, 2004). In a way, the opposition between universalism and relativism could, therefore, be considered as entirely built into ‘absolute universalism’ as it manifests itself in the French public sphere. ‘Reflexive universalism’ could, on the other hand, be conceptualized as a type of radical pluralism – what came to be known in the field of human rights as ‘pluriversalism’ (Eloi, 2009). The notion of pluriversalism helps take into account the plurality of human experiences as well as the necessity to conceptually unify the experience of human commonality. Therefore, it constitutes a constructivist model of ‘the universal’, underlying how much universalism rests on an a posteriori construction process, that articulate in a common horizon the most disparate modes of being.

What would a ‘pluriversalist diversity politics’ mean? Such politics would probably first and foremost allow for an acknowledgment of the antagonisms built into the very notion of ‘diversity’. From its corporate readings to its tendency to euphemize ethnic and racial discrimination, the French uses of the word ‘diversity’ are filled with social tensions. These tensions should not be obscured, but rather confronted in a constructivist perspective in order to expose the conflicting social beliefs it carries. Unfolding the layers of antagonisms that the word ‘diversity’ carry precisely means to map out the antagonisms and social tensions of French society toward the urgent questions the political register of diversity addresses. To study the competing and conflicting interpretations of universalism in France, as we attempted to, is to try and analyze how a society thinks itself.

In a political conjuncture defined by continuing processes of racialization and exclusions, and in which the leader of the extreme-right party Marine Le Pen obtained a record vote rate of 17.9%, the times seem more characterized by a renewal of the public lexicons of racism than of a sharpening of the governmental tools to fight it. The election of President François Hollande in 2012 have inspired hope that, after a Sarkozy administration marked by numerous racial polemics and a troubling stasis on questions of discrimination, public policies would embrace the path of diversity. However, the issue of racism still seems not to be on the agenda, as Yazid Sabeg former High Commissioner on Diversity and Equal Opportunities declared when he resigned a few months after the election. ¹³ Furthermore, Hollande pointed toward what appears as a step back from diversity policies, declaring during the presidential race that once elected he would remove the word ‘race’ from Constitution. This negative politics, defined by the erasure of both the word for oppression and the constitutional protection against racism, seems once again to bolster the representation of France of an already discrimination-free country. A new articulation of the State discursive constructions of the ‘race issue’ that may produce new difficulties, for scholars and activists as well, for charting the sinuous movements of racism as it pervades French society.