Representation as imposition

6Far from signaling a radical constructivist turn in the study of representation, the idea that the representative constitutes the represented by her claim to representation directly meets the conception of representation developed by Thomas Hobbes in the famous pages of Leviathan that he dedicated to this notion. In the sixteenth chapter, which closes the first part of the book, Hobbes defined the relation of representation as a relation in which the acts of one person – the representative – are attributed to an author – the represented. But while for Hobbes the sovereign results from a social pact, an authorization given by those who are represented rendering them the authors of her acts, it should be added that it is not a people that establishes this pact, but isolated individuals, a multitude. Thus, according to the concept of representation developed by Hobbes, “[a] multitude of men, are made one person, when they are by one man, or one person represented […] For it is the unity of the representer, not the unity of the represented, that maketh the person one”. [10] Every individual that makes up the multitude authorizes the representative to act in his/her name, but the people, as a collective unified entity, does not exist prior to the representative. This is what justifies, for Hobbes, the absolute power of the sovereign: the people’s only rights are those granted by the sovereign, because it does not exist before having been represented by the sovereign. Instead of being the reflection of the people, for Hobbes the representative gives the people its image; it shapes the people. [11]

7As Quentin Skinner has shown, this conception of representation is then developed by Hobbes for a specific political intention: to counter the claims made by the English members of parliament to attribute a representative capacity to themselves, in the context of the first English revolution. [12] Carl Schmitt makes a similar use of representation against liberal parliamentarism. [13] However, the idea that the representative creates the represented is not necessarily used as a justification for absolutism. It could even be said that it was at the heart of the founding gesture by the deputies of the Third Estate, deciding in 1789 to establish themselves as a National Assembly, [14] or of the stance taken by the writers of the United States’ Constitution when they began their text with “We, the People of the United States”. [15] In each case, persons speaking in the name of a people caused it to become a political entity, and in so doing, authorized themselves to determine its primary characteristics. This reasoning can be extended to representative claims that do not concern a people, but a class, a social group. Thus, the manifesto “For a Women’s Liberation Movement”, published in 1970, began by symbolically forming women into a collective subject (“We, from time immemorial, live like a people colonized within the people […] We are the class that has been oppressed for the longest time”), before calling for the formation of a “women’s liberation movement” aiming to justify “a political takeover to represent, in our turn, our interest as being the universal interest”. [16] The activity of women’s representation by the Women’s Liberation Movement, of which this is one of the founding documents, is caught up in an explicit work of constructing this group as having a unique interest that can be made universal.

8It is clear that – whether the aim is to justify a coercive authority, to establish a new political order, or to enable the action of an emancipatory collective subject – the idea that representative claims participate in the constitution of the represented group (imposing upon it an identity as a group) has existed for a long time. This conception of representation as imposition is not just a theoretical or activist position: it had also inspired a range of studies in the social sciences, especially in political science. In particular, Pierre Bourdieu’s research on political representation, beginning in 1979 in Distinction and followed by several articles at the beginning of the 1980s, certainly constituted one of the most complete attempts to subject such a conception of representation to empirical analysis. By drawing primarily on the study of relationships between the working class and its spokespersons, Pierre Bourdieu brought out several characteristics of the relationship of representation. It is created and deployed in a political field – a competitive space structurally homolog to the social space – where political enterprises or mechanisms struggle using various forms of political capital, forcing social groups, especially the most destitute, to give themselves over to these enterprises for the defense of their interests. [17] But the most interesting for our purposes is that, according to Pierre Bourdieu, the relation of representation, usually seen as a delegation process, is in reality a process whereby the represented is instituted or constituted. This is clear in cases where a group depends on its representatives for its existence – Pierre Bourdieu used the example of the Church (also used by Hobbes) to prove that certain inanimate objects could be personified. However, the argument can be extended to the representation of social groups, especially when it involves dominated groups: then, “the act of symbolization by which the spokesperson is constituted, the constitution of the ‘movement’, happens at the same time as the constituting of the group […] The signifier is not only that which expresses and represents the signified group: it is that which signifies to it that it exists, that which has the power to call it into visible existence, by mobilizing it.” [18] This is what Pierre Bourdieu calls the “mystery of ministry”, which enables the spokesperson to act for the group, or the “oracle effect”, a kind of “ventriloquism” that causes the representative to speak for the represented. [19] It is not simply a matter of dispossession: the power held by the representative to form the group goes along with the power to define what the group does, and the social characteristics that constitute it as a group. Indeed, “the objects of the social world […] always include a certain indeterminacy and vagueness”, [20] opening up the possibility for a “symbolic struggle […] for the monopoly of legitimate naming” [21] of these objects, in particular of social groups. Therefore, groups are always constructed by their representatives, and the representations (in the symbolic sense) of these groups always depend on their representation (in the institutional sense). [22]

9Thus, in these very diverse works, both in their method and political aim, there is a shared idea that representation is above all a relation of imposition, in which the representative imposes onto the represented their characteristics and imposes herself as their spokesperson. According to this first ideal-type of representation, representative claims play a primordial role, in both senses of the word: it is both the first chronologically as well as the most important for understanding what takes place in the activity of representation.

Representation as composition

10Why then does Michael Saward’s text seem innovative and why has it led to real theoretical and empirical renewal? The first reason is that the conception of representation as imposition has for a long time been played down or even eliminated from political theory, especially in the UK and the United States. This is certainly due to the success of Pitkin’s book, The Concept of Representation, which appeared in 1967 and which provided a largely hegemonic theoretical framework for the study of political representation. In it, Pitkin refuted the Hobbesian analysis of representation in terms of authorization, in which she justly noted an attempt “to explain and justify political obligation”. [23] More significantly, the entire book was based on the idea that to represent meant “making present again”, or more specifically, the “making present in some sense of something which is nevertheless not present literally or in fact”. [24] Yet, this “metaphysical concept of representation” [25] implies that the represented exists prior to the activity of representation – just as the conception that Pitkin then developed concerning political representation as responsiveness presupposed that there was a represented that could express its desire independently from the representative. At most, Pitkin recognized the possibility that the representative participated in the construction of the represented in her chapter dedicated to symbolic representation, but she did so in order to immediately disqualify this conception of representation as characteristic of fascism. For Pitkin, thus, the whole question was to know what elements of the represented composed the representative, how the latter defended the interests of the former and responded to its wishes.

11The idea that the represented exists prior to the activity of representation is of course not Pitkin’s invention, even though it certainly was her book that contributed to its dominance in political theory. It originated, at least in modern political philosophy, in liberalism, more specifically in John Locke. Indeed, in the Second Treatise of Civil Government, Locke took the opposite view to Hobbes in attributing rights to individuals, from the state of nature, rights that do not disappear with the social contract. Representation does not create the people as political subject; on the contrary, by building a political society, people voluntarily form a body politic, and since “[it is] the interest as well as intention of the people to have a fair and equal representative”, [26] “the people […] [have] reserved to themselves the choice of their representatives”. [27] There is therefore a conception of representation opposed to Hobbes’s and that can be qualified as liberal, according to which individuals bear rights in themselves, rights that exist prior to their representation, and they thus choose their representatives instead of being constituted by them. Here again, while not attempting to elucidate a history of this conception, it can be used as a starting point to construct a second ideal-type of representation, according to which representative claims are only valid on the condition that they rest on a composition of elements present in the represented itself. Several theories can then be included in this ideal-type: representation as composition can simultaneously justify the idea, as in Edmund Burke, that representatives are independent of their constituents and defend their interests without allowing themselves to be influenced by them, or, on the contrary, go hand in hand with the preference for an imperative mandate or various forms of control over the representative’s action. [28] Representation in this sense could also be founded on an abstract conception of the citizen, whose ideas alone are represented, or, on the contrary, it could take into consideration the social characteristics of the represented, opening the way to forms of descriptive representation, group representation, or a politics of presence. [29] However, aside from their very real differences, these theorizations of representation all recognize that those who are represented exist prior to their representation, that they are not the pure product of a representative claim, of a claim-making, but instead that all claim-making is organized according to elements that are present in the represented. [30] Representation is thus shown to be a composition, which can mean that the body of representatives is composed based on the body of those who are represented (their ideas, characteristics, etc.), or that the represented actively compose, for example through voting, the body of representatives. [31] Since the 1960s, studies in political theory as well as empirical studies, especially in the English-speaking world, have thus primarily involved discussing, categorizing, and measuring the ways in which representatives or the representative system enter into a relationship of composition with the wishes, ideas, and characteristics of the represented.

Representation as proposition

12We can thus see how the supposed constructivist turn, embodied by Michael Saward in particular, appeared as a rupture: it opened a breach in the field of representation in which the idea of representation as composition, closely tied to liberalism, had up to that point been dominant. However, as a result, Saward’s efforts were primarily viewed as forming part of a fairly specific oppositional stance to Anglo-American political theory. There are thus particular issues surrounding his reception in France, especially since the conception of political representation developed by Pierre Bourdieu makes Saward’s constructivism seem less polemical. The challenge then becomes how to take account of what, in Michael Saward’s suggestion concerning representative claims, goes beyond the mere repetition of the conception of representation as imposition. Viewed from this perspective, two elements stand out in the conceptualization proposed by Saward. On the one hand, for him, there is a separation between the claim-maker and the subject of representation: it is not the representative alone who institutes the represented, thereby giving rise to an indisputable imbalance, but also a relational configuration in which the one who claims that there is representation is not necessarily the one who claims to represent. On the other hand, the represented as such is far from merely passive: if the object of the representation is indeed constructed by the claim-maker, this always takes place before an audience and in relation to a referent, the social existence of which does not depend on representation. Therefore, there is some leeway to judge, transform, and contest representative claims – without necessarily needing to surrender oneself to other representatives. For Saward, representation does not come about through imposition or composition, but rather through a proposition: in front of an audience, the claim-maker proposes to establish a relation of representation between a subject, object, and referent, and the success or failure of this representative claim largely depends on the way in which those in whose name one is speaking respond to this proposition.

13It is then possible, based on the idea of representative claims, to bring together at least three very distinct trends in the analysis of representation, which nonetheless share a similar way of thinking about representation as a proposition. The first is the sociology of tests, concerning which a recent publication indicates that it draws mainly from two sources: “the anthropology of science and techniques developed by Michel Callon and Bruno Latour and the sociology of action regimes put forward by Luc Boltanski and Laurent Thévenot”. [32] While this publication does not develop the matter of representation, we can nonetheless find in Bruno Latour and Luc Boltanski two conceptions of representation that come together in the idea that representation is above all a proposition, where the representative does have a role in the formation of the represented, but in which those who are represented acquire, in the representation, the power to judge the representations proposed concerning them. In Bruno Latour, we find the model of the “circle of representation”: the representative causes the represented to exist by unifying it in political speech, but in a way that always betrays the multiplicity of the represented, which reappears in the various forms of obedience – that is, of interpretation and representation – to those who are represented. [33] For Luc Boltanski, it is certainly the idea of the “hermeneutic contradiction” that best exemplifies the treatment of representation as proposition: institutions, responsible for “stating the whatness of what is”, that is, for stabilizing representations of the real world, especially of social groups, only exist through spokespersons. [34] They do represent the institutions, but this ability to represent is always, at least potentially, subject to criticism by those who are subject to the power of the institution, precisely because equivalence between the institution and its spokespersons is never guaranteed – it is always a representative claim. For Bruno Latour and Luc Boltanksi, representatives may always have their representative claims tested by those who are represented.

14A second tradition describes a form of representation as proposition, that of deconstructionist feminism. This a school of thought which emerged in relation to the experiences of real feminist social movements, drawing especially on the challenge to the unity of women as the subject of feminism. At the beginning of the 1970s in the United States, the question was raised as to whether white women could speak for black women, and heterosexuals for homosexuals, contesting both the movements’ representatives and the representations of the group whom they claimed to represent. [35] The inability of (white) feminism to represent black women (and then Chicana women, post-colonial women, etc.) was at the heart of work done by black, [36] Chicana, [37] and post-colonial feminists, [38] and reflection on the co-construction of the dominations of race, gender, sexuality, class, etc., led to the emergence of the concept of intersectionality. [39] Initially proposed by jurist Kimberlé Crenshaw to examine the unequal treatment of white women and women of color facing gender violence, this concept is today widely used to try to account for the challenges involved in the representation of social groups, especially dominated ones. Criticism of feminist universalism led to an intense examination of the subject of feminist politics. Thus, Iris Marion Young precisely conceptualized the implications that this challenge to the unity of the subject represented by feminism had for the theory of representation. Using the concept of différance developed by Jacques Derrida, she proposed seeing the relation of representation not as a relation of identity, but as “a differentiated relationship between constituents”. The quality of this relationship depends on the ways in which representatives and those who are represented are connected, forming “a cycle of anticipation and recollection between constituents and representative, in which discourse and action at each moment ought to bear traces of the others”. [40] Any claim to represent an intrinsically diverse group is thus continuously shaped by the connections opened up by that diversity, and conceptualizing this in terms of the representative claim is certainly one approach to consider in order to avoid the fixation on represented identities, especially in a context where descriptive representation, through the use of quotas, is often presented as the solution to the weak representation of certain social groups. [41]

15Finally, the idea of representation as proposition is at the heart of certain attempts at reformulating the Marxist project in order to adapt it to situations where the subject of emancipation is no longer obviously the unified proletariat. Subaltern studies have certainly played a primary role in calling the Marxist narrative into question by denouncing the ethnocentrism of the claim of the Western industrial proletariat to represent workers; by showing the internal divisions in colonized worlds, hidden behind the claim by the elites of these worlds to represent the colonized; and finally by showing, on the basis of the linguistic turn initiated by Gayatri Spivak, that any discourse in the name of the subaltern, or even any discourse on the subaltern, contributes to rendering them invisible. [42] At the same time, other authors, heavily influenced by post-structuralism, have attempted to think about ways to recompose a subject of emancipation, taking into account the discursive nature of this subject – and therefore the relative contingency of its representation. In this vein, the studies by Chantal Mouffe and Ernesto Laclau in their book, Hegemony and Socialist Strategy, developed a conception of representation that can be linked to the idea of representation as proposition. The general idea is that the constitution of an emancipatory political subject involves the construction, by representatives, of a chain of equivalences between oppressions that can be affirmed as hegemonic in the representations of the social world. After a fashion, “political-hegemonic articulations retroactively create the interests they claim to represent”, [43] but they are always composed of demands that actually exist and that have been made equivalent, based on a contingent arrangement whose success depends largely on the represented themselves. A picture thus emerges of an agonistic democracy in which the popular subject is continuously reconstructed by representative claims that seek to mobilize it.

16Thus, even though Michael Saward’s theorization was primarily received in the Anglo-American world as breaking with the liberal conception of representation developed by Pitkin, it does not entirely adopt the forms of the constructivist idea of a representative monopolizing the power to institute the representative. As an alternative to representation as composition and imposition, Saward develops a reading of representation as proposition, in which a plurality of actors involved participate in the success or failure of representative claims – thus linking the sociology of tests, deconstructionist feminism, and certain forms of heterodox Marxism, approaches that share a certain sense of representation as proposition. It is in this regard, more than in terms of a supposed break with other conceptualizations of representation, that the “representative claims” approach proves interesting, both from a theoretical standpoint and as a way of enriching the empirical analysis of representation.

Representation as performance

18The question of the representative’s activities is at the core of substantive representation as defined by Pitkin, according to whom representing means above all acting in the interests of those who are represented. However, various studies on the profession or role of elected officials have revealed practices that allow these officials to “perform” their role as representative, practices that cannot be reduced to actions in the interest of the represented. [47] In this regard, there is an entire series of activities that the traditional approach to representation as composition is unable to grasp, even when there is a focus on elected officials. At first glance, the constructivist approach to representation may also seem inadequate, in that it seems to emphasize the discursive and symbolic aspect of the work of representation. However, representative claims are much more than simply discursive; they involve a range of practices, as clearly shown in the articles in this special issue by Julien Talpin concerning community organizing in Los Angeles, and by Marie-Hélène Sa Vilas Boas on women’s municipal conferences in Recife. This is particularly clear in the distinction Saward makes between the author of the representative claim and its subject, that is, the representative. Schematically, while the former “declares” the representation, it often falls to the latter to “perform” it. Obviously, the author and subject are not always clearly defined and separate entities, but the distinction is nonetheless useful at the analytical and empirical levels. The activity of the subject of representation can then be viewed as a performance, that is, a “stylized repetition of acts”, [48] which is thus never fully established or complete. From this perspective, the idea of “representation tests” proposed by Julien Talpin in this special issue provides a way of understanding the need of actors to constantly reiterate and validate their representation proposals, by implementing a series of activities and discourses intended to demonstrate their representativity.

19It is impossible to draw up an exhaustive list of the activities that constitute a representative claim. Indeed, representation is a process indexed to a specific context and cannot be reduced to a series of pre-defined activities. However, the articles included in this special issue, as well as the existing literature on interest groups or social movements, show that there are tests of representation that recur from one context to the next. In particular, it seems that when representation does not depend on election, it is essential that representatives pass the “test of numbers”, that is, that they visibly assemble those whom they represent. This is seen in collective mobilizations, in which it is a matter both of attesting to the existence of the group (we will return to this aspect) and its grouping behind its representative who can then legitimately speak for those who are represented. Thus, studies done on the mobilizations of those who are in some way deprived (for example, the unemployed or the inadequately housed) have shown how important it is for the various collectives to have the people they say they are defending participate in the various protests, such that their presence affects the aid that the collectives are prepared to provide to them. [49] In this issue, the test of numbers is closely analyzed in the articles by Marie-Hélène Sa Vilas Boas and Julien Talpin. Whether in the context of women’s conferences in Recife or community organizing in Los Angeles, a good representative is someone who can get a maximum number of people to participate in events. In this regard, the test of numbers sometimes seems to resemble elections, for example, when community organizing leaders are evaluated according to the “applause meter”. Thus, even though community organizing and women’s conferences are not part of the electoral system, and collectives of the unemployed or the inadequately housed, such as the AC [the French “Act together against unemployment” association] or the DAL [the French “Right to housing” association] explicitly reject this model, the aggregation of individuals remains at the heart of their representation work. Paradoxically, this question seems less crucial when representatives are more formally tied to the institutional representative system. Thus, in her article on the International Labor Organization (ILO), Marieke Louis emphasizes that while the number of members and votes in national union elections matter in designating worker representatives to the ILO, this criterion of numbers can be completely trumped by other elements of representativity. Similarly, when it is a matter of determining the methods of designating women representatives in the legislative assemblies of colonial India, the subject of Virginie Dutoya’s article in this issue, elections are not always viewed as the most appropriate procedure.

20However, the test of numbers is not merely a substitute for voting. As Marie-Hélène Sa Vilas Boas shows, it can also be seen as a way for female representatives to provide proof of their social anchoring in the group they claim to represent. Indeed, belonging to the group represented is an essential element for representatives, whether it involves regional anchoring, for example, for a local elected official, [50] or belonging to a social group. It is of course possible to affirm that a representative belongs to the group, but this belonging must also be performed. Thus, in her article, Virginie Dutoya indicates that Pakistani or Indian elected officials (or female activists for women’s rights) are often reproached for being unrepresentative of “real Indian women” or “real Pakistani women”, with their class habitus (materialized especially in their self-presentation, language, etc.) then being put forward as proof of their lack of representativity. In the case of community organizing, Julien Talpin shows that organizations are not only seeking leaders who have “good” traits, but that they are also aware of the performance of these representatives: their habitus, their style of dress, and the emotions they express and evoke. In this regard, the representative is not limited to embodying the socio-demographic traits of the group, but must also embody its aspirations and emotions. [51] Furthermore, Marie-Hélène Sa Vilas Boas points to a tension that exists for women’s delegates between the principles of identification and distinction. Depending on the context, these delegates will either present themselves as ordinary inhabitants of the neighborhood, or showcase their own abilities, such as engaging local elected officials, speaking the language of the administration, etc. This interplay between identification and distinction is also found in more traditional spheres of identification, in which representatives tend to modify their representative claims depending on the audience and representation contexts. [52] Thus, studies on the representation of farmers have shown how national leaders of the main agricultural union have to adapt their representation work, so as to appear “as farmers among farmers” or “spokespersons for a ‘profession of the future’ to other social milieux”, while at the same time joining the ranks of top-level management. [53]

21Finally, the goal of the representative’s performance is not only to establish his/her representativity, but also to establish what Michael Saward calls the object of representation. Thus, again with regard to agricultural representatives, Alexandre Hobeika analyzes the operation of a departmental branch of the Fédération nationale des syndicats d’exploitants agricoles. [54] He shows how important it is for local leaders to avoid standing out during union meetings, either by what they wear or their cultural practices; more generally, they need to perform a certain local sociability. What is important for these local leaders is being accepted as legitimate representatives, and – in a context of strong divisions within the federation – of creating consensus by reinforcing a farmer “group style” through their performance. In this regard, a representative claim always has two sides: it is both about the representativity of the subject of representation and the characteristics of the object of representation.

From instituted representation to instituting representation

22As we have seen, Michael Saward partially coincides with Pierre Bourdieu in his constructivist and instituting approach to representation, especially with the distinction he makes between the object of representative claims and their referent. This distinction is rarely taken up in Anglo-American studies on representative claims and yet is very heuristic, especially in the perspective of works on the construction of social groups. For Saward, the object of a representative claim is a symbolic construction that has a referent in the real world, but which does not emanate directly from this referent, and for which the definition of relevant traits constitutes one of the main challenges of the claim-making activity. In this regard, Saward’s approach reflects Pierre Bourdieu’s analysis of the work of spokespersons, who participate in instituting groups in whose name they speak by claiming to represent them. The work of both men encourages close study of the actual construction processes of social groups. For Saward, and more generally in approaches that consider representation as a proposition, particular importance is given to the dialogic nature of the construction of social groups, which involves numerous actors (not necessarily equally), and cannot be reduced to the relationship of the representative with those whom s/he represents.

23A distinction can be made, especially on the basis of the articles included in this special issue, between three linked processes in the construction of objects of representation: the affirmation of the existence of the object, the affirmation of its political legitimacy, and finally the affirmation of its unity. In certain cases, the existence of the object of representation seems to be fairly obvious, especially because it has a stable referent in the real world. This is, for example, the case of women who, in spite of studies showing the constructed nature of sex and gender, are always considered as a self-evident and natural category by the majority of political actors. As Virginie Dutoya demonstrates in her article, the existence of “women” as a social group is not a subject of debate in India or Pakistan; rather, it is the political legitimacy of the category as well as its unity that pose a problem. However, other social groups have less clearly defined profiles: this is especially the case of the groups that community organizing movements claim to represent, analyzed by Julien Talpin. The issue also arises when it comes to representing those who are “without” (papers, employment, housing), or those who are “excluded”. [55] In this regard, Sophie Maurer and Emmanuel Pierru have highlighted the importance for collectives attempting to mobilize the unemployed of “imposing the representation – in both senses of the word – of what they say, that is, the existence of the reference group as mobilized group”. [56] However, as the two authors emphasize, this construction is more than simply discursive and its “performative power assumes at a minimum that it is connected to the ‘reality’ it states: actually mobilizing the unemployed”. [57] In the same way, Julien Talpin emphasizes in this issue that community organizing can only be called representative of the “community” if it actually succeeds in assembling this community through mass events which affirm its unity.

24The task of unifying the object of representation is in fact crucially important. The social groups presented in the articles of this special issue – women, workers, or the “community” that community organizing organizations seek to mobilize – are far from homogeneous. This special issue reveals two strategies for unification: one strategy involves the negation or at least the political disqualification of divisions, and the other involves the group’s intersectional representation. In her article on the ILO, Marieke Louis reveals the significant effort expended by workers’ and employers’ representatives to present a united front and minimize national divisions. In India, the question of the unity of women as a political category is at the center of the debates over quotas for women, analyzed by Virginia Dutoya. In that country, the legitimacy of women’s demands for just representation in the political sphere is uncontested, but opponents of the quotas demand that this category be subdivided in order to account for caste and religion. Nonetheless, the major governing parties and women’s organizations argue for the indivisibility of women as a social group. By contrast, Julien Talpin’s article provides an example of the construction of an object of intersectional representation, in which community organizing organizations seek to represent unity while also taking account of racial, social, and religious differences within the community they are claiming to organize. This entails conscious institutional practices (paying attention to the social and racial identity of employees, the choice of sites for events, the selection of leaders who represent this diversity), as well as a discourse that seeks to showcase unity while also recognizing situated perspectives.

25In light of this last example, it seems that an analysis that simply underscores the violence of representation’s “symbolic takeover by force” [58] would necessarily be limited. There is no doubt that very often the work of representation involves the imposition of social and political identities, which contributes to the fact that some persons are rendered invisible. Thus, in the case of LGBT (lesbian, gay, bisexual, and transsexual) politics in India, various authors have denounced the imposition of categories modeled on European history. For them, it is not only an issue of whether these categories are relevant in the Indian context, but also of their impact on the attribution of international funding, most notably in the context of the fight against HIV/AIDS. [59] But at the same time, the debate has focused on the modalities of representation of sexual practices and identities and the risks of a queer or LGBT politics; [60] no one is fighting for a political invisibility strategy. As emphasized by Stuart Hall, access to representation, no matter how imperfect, can be liberating for those who do not have it, whether it is a matter of cultural or political representation. [61] Thus, in the case of the unemployed movement at the end of the 1990s, Sophie Maurer and Emmanuel Pierru show that the work to construct the “unemployed” as a social group was primarily carried out by activists who did not define themselves as unemployed, and sometimes were not unemployed, and that those most concerned hardly identified themselves as such. And yet, the movement had positive effects for the unemployed who participated, first in terms of objective rights obtained, and then in terms of its socializing and empowering aspects. [62]

26Furthermore, as each article in this special issue points out, representation is a malleable process, one that is never fully closed or the result of one person’s action. In her article, Virginie Dutoya shows that while the state plays a leading role in the construction of women as a political category, it does so most often together with feminist organizations. In this regard, it could be said that the state proposes a political representation of women, which is then negotiated and transformed by multiple actors. In the same way, Marieke Louis reveals the transformations that the conception of representation has undergone at the ILO, to answer the complaint made by different organizations decrying the fact that some workers (those in the informal sector), and even some regions of the world, are being rendered invisible. In political systems based on the notion of representation, being recognized as “representative” grants access to resources, both material and symbolic. Furthermore, this legitimizing function of representation does not end with the subjects of representation, but also extends to the authors of representative claims, whose legitimacy is strengthened by their capacity to arbitrate the representativity of individuals, organizations, and movements.

Representation, power, legitimation

27One problem with the constructivist approach to representation as imposition or proposition is the tendency to situate all claims to representation on the same level. Yet all claims do not get the same sort of recognition, whether that be from the groups that are said to be represented, the media, or even political authorities that may validate or invalidate certain demands. In this respect, while it is useful to understand representation in a decompartmentalized way, by accepting, for example, all representative claims made in the name of women by feminist organizations, elected officials, political parties, or administrations, [63] it is worth noting that all claims do not have the same chance of being accepted and that it is not possible to establish equivalencies between them. Indeed, representative claims are located within a sphere of power and thus participate in the constitution, transformation, and contestation of power relations.

28This is especially important in situations in which different actors say, or are told, that they are representative of the same social group. While there may be moments of cooperation between these actors, they are very often in competition, especially when representativity conditions access to resources. Thus, the representation of workers is at the heart of major national [64] or international disputes, as Marieke Louis’s article in this issue demonstrates. The notion of representative claims is then especially heuristic in order to understand the competition between social actors for access to the system of representation. This struggle takes place at different levels. Actors are in competition for the recognition of subjects of representation (or even of referents), recognition that can be expressed as votes during union representative elections, as a response to a call to demonstrate, or even as membership in an association. Very often, however, claim makers also seek institutional recognition – to the point sometimes that it supplants other requirements of representation such as numerical weight, as Pierre Rosanvallon shows concerning French unions. [65] Thus, when Yvan Bruneau analyzes the erosion of the FNSEA’s power of representation, he focuses especially on the acceptance of union pluralism by the public authorities (in this case, the ministry of agriculture). [66] In the same way, in the articles by Viginie Dutoya and Marieke Louis, competition for the recognition of the status of representative for women, workers, or employers occurs essentially at an institutional level, since the “real” women, workers, and employers are only very marginally involved. However, in these various examples, the institutional authorities recognizing (or not) the representative claims made by a particular actor (a union of female informal workers, a political party demanding quotas for women, etc.) are not outsiders to the process that led to the constitution of those claims. Thus, by finally accepting the request to join by the Self Employed Women’s Association, an Indian trade union, the ILO called into question the claim of “traditional” unions to represent all workers. But at the same time, it reinforced the institution’s overall legitimacy by affirming its ability to decide who can be said to be representative of workers.

29This situation allows us to understand again the distinction made by Michael Saward between the authors and subjects of representative claims. Indeed, this distinction is essential, as in the Bourdieusian perspective the focus is more on the “representative” (the subject of representation according to Michael Saward), not recognizing that this is not always the one who claims to be representative. The articles in this special issue analyze situations in which it proves useful to distinguish between the authors and subjects of representation, those who say there is representation, and those who are constructed as representatives. Indeed, whether we are considering Pakistani and Indian deputies, community organizing leaders, or even union delegates at the ILO, all participate in one (or several) representative claim(s) that exceed(s) them. Behind community organizing leaders is the matter of the representativity of their organization, behind a Pakistani deputy, there is the national representation of the parliament, etc. For example, Virginie Dutoya shows that the requirement for the political representation of women on the Indian subcontinent was generally led by colonial authorities. Prior to the appearance of a representative system, British authorities claimed to represent women. Then with the appearance of a representative system and women’s organizations after World War I, they positioned themselves as guarantors of women’s representation, and thus able to decide how women were to be represented – becoming no longer subjects, but authors of women’s representation. It then became a matter of ensuring the legitimacy of the colonial political system as the only system able to protect women (and by extension the most vulnerable) and to include them in decision-making processes.

30In this regard, to truly understand the “symbolic takeover by force” of representation, Bourdieu’s ventriloquist metaphor has to be more fully developed in order to focus on the one who “says” representation rather than the one who performs it. However, this would minimize the contribution of an approach to representation as proposition, which invites us instead to envision representation as a dialogic process. Indeed, the relationship between community organizing leaders and their organizations reveals, as does the relationship between unions and the ILO, that representative claims formulated in the name of the system and its subparts are closely interwoven and recursive, and the roles of authors and subjects are consistently interdependent, crossed, and even exchanged. Thus, in Julien Talpin’s article, when a community organization appoints a leader, they say s/he is the “representative” of the community or a specific group, but when s/he speaks in the name of that organization, the leader says that the organization is representative of his/her community. In this regard, while the leaders depend on the organization and can be replaced, they are not simply puppets, and the members of the community, the “represented”, have real power in the selection or marginalization process of the leaders, by not coming to meetings, by not applauding, or by expressing their discontent. The notion of the representative claim thus enables us to understand both the reality of the power relations and the disparity of power between the actors, while still allowing for the possibility that dominated actors can intervene in this process.

31* * *

32It thus appears that the theoretical framework proposed by Michael Saward allows for an analysis of the construction processes of political representation. This reveals the activity of several different actors in these processes, well beyond the face-to-face analysis offered by most approaches in terms of imposition or composition. By looking at representation as a proposition, we are able to place at the center of the analysis the dialogue – often unequal, conflictual, and limited – between representatives and represented, a dialogue which also, and sometimes mostly, involves the makers of representative claims and various audiences.

33The articles comprising this special issue, despite the diversity of the fields covered and the methods used, all take seriously the idea that representative claims are at the heart of the processes of constructing, negotiating, or invalidating relations of representation. This encounter between an approach from Anglo-American political theory and empirical studies for the most part stemming from Francophone political sociology proves, if proof were necessary, the need for a theoretical, methodological, and geographical decompartmentalization of political studies. Our hope is that the necessarily imperfect and incomplete nature of this essay will make it that much more stimulating for future encounters. [67]