COPYRIGHT NOTICE: THE FOLLOWING TEXT "Political Philosophy and Racial Injustice: From Normative to Critical Theory" IS FULLY PROTECTED BY COPYRIGHT LAW AND MAY ONLY BE QUOTED UNDER THE CORRESPONDING CONVENTIONS. © 2001 Thomas A. McCarthy
Philosophy and Racial Injustice:
In mainstream political philosophy the history of European racism, with its vast implications for the theory and practice of modern liberalism, has remained largely on the margins. This is nearly as astonishing as the theoretical marginality of gender until quite recently. I say “nearly,” because while gender relations have deeply structured every human society, racial relations, in the sense at issue here, have had major structural significance “only” for some five centuries. That is to say, they are contemporaneous with, and deeply implicated with, Western modernity from the first voyages of “discovery” to present-day neocolonialism. If one asked, in Rawlsian terms, which morally arbitrary facts about individuals and groups have had the greatest consequences for their legal and political standing in the modern world, gender and ascribed race would certainly be near the top of the list, along with class, though their comparative significance would vary from context to context. If the context were the global one of European expansionism since the fifteenth century, then racial classification would have a strong claim to being the most significant; for a central ingredient in the process by which more than 4/5 of the globe came to be under European and/or American rule before the start of World War I was the practice and theory of white supremacy.
The linkage of Western modernity and the rise of capitalism to the conquest and exploitation of the nonwestern world is a long-familiar story. Marx had already noted in the first volume of Capital, “The discovery of gold and silver in America, the extirpation, enslavement and entombment in mines of the aboriginal population, the beginning of the conquest and looting of the East Indies, the turning of Africa into a warren for the commercial hunting of black-skins, signalized the rosy dawn of the era of capitalist production.” But Marx and Marxism centered on class relations and generally treated race relations as derivative therefrom. An alternative account, running through W.E.B. Du Bois to the critical race theory of the present, has, by contrast, viewed racial categorization not simply as a dependent variable but as an irreducible structuring principle of social, cultural, and political relations in the modern world. On this view, the conquest and settlement of the Americas, the subjugation and extermination of indigenous peoples, and the massive expansion of the Atlantic slave trade in the early modern period were integrally bound up with the social construction of racial differences and racial hierarchies. During the course of this historical transformation, imperial ideologies of “Christians versus heathens” gradually gave way to those of “civilized Europeans versus uncivilized savages,” which were understood in biologized racial terms already in the eighteenth century. In the nineteenth century, the “age of empire,” the rapid expansion of colonialism in Asia, the Pacific, and Africa, fueled a further development of racial ideologies, particularly in conjunction with the rise of Darwinian biology and physical anthropology. In short, there was a constant interplay between colonialism and racism, between the establishment of imperial domination and the formation of racial ideologies.
In the view of this alternative tradition -- alternative not only to Marxism but also to liberalism -- these were not ideologies in the sense of mere epiphenomena of underlying social processes: they were social constructions that were real social facts with real effects on the ordering of social relations. This is the view expressed, for instance, by Omi and Winant in the claim that modernity saw the rise of “racial formations” at both national and global levels. Systems of racial categorization centered around visible body types had not only expressive but constitutive significance in modern society and politics. They not only justified preexisting practices of racial domination, they entered into and informed them. Stereotypical images of racial capacity and incapacity not only reflected institutional reality, they were essential to its very intelligibility and normativity. Similar appeals to “natural” racial differences were just as integral to US immigration policy in the nineteenth and twentieth centuries, to the formation of the American working class and labor movement, to “Jim Crow” and “separate but equal,” to eugenics and extermination camps. In short, “race has functioned as a marker of inclusion and exclusion, equality and inequality, freedom and unfreedom throughout the modern period, locally and globally. And despite the successes of the American Civil Rights Movement here and decolonization struggles abroad, there is widespread agreement in this alternative tradition that the legacy of institutionalized racism is still with us, that is, that local and global relations of wealth and power are still structured along racialized lines. In fact, the persistence of “race” as a significant ordering principle of social life, even after its political dismantling and theoretical deconstruction, is seen there to be one of the major problems of the age -- “the problem of the color-line.”
Since the 1960s, the centrality of race to the modern world in general and to the American experience in particular has finally permeated mainstream scholarship in most areas of the social sciences and the humanities. In mainstream political philosophy, however, the process has only just begun -- despite the evident fact that political discourses, practices, and institutions have been suffused with racism throughout the modern period and racial politics persist into the present as the legacy of centuries of oppression. As a result of this continued marginalization, the development of conceptual tools for analyzing the racialized dimensions of modern and contemporary politics has lagged, and the shift from legally institutionalized patterns of racial domination to domination anchored in lifeworld cultures and traditions, norms and values, socialization patterns and identity formations has remained largely untheorized in liberal political theory.
As a number of scholars have recently documented, however, theoretical marginalization is only part of the story. Most of the classical modern theorists were aware of and complicit in the emerging system of white supremacy. Thus Locke famously declared America to be a “vacant land” occupied only by nomadic savages still in the state of nature, and hence a land ripe for European expropriation, as no ownership-conferring labor had yet been mixed with it. Less famously, he was an original shareholder in the Royal African Company, which was chartered in 1672 to monopolize the English slave trade -- he increased his investment in 1674 and again in 1675 -- and a few years before (1669) he had helped author “The Fundamental Constitutions of Carolina,” which stated that “Every freeman of Carolina shall have absolute power and authority over his negro slaves, of what opinion or religion whatsoever.” Thus, the same Locke who declared in the opening line of his First Treatise that “Slavery is so vile and miserable an estate of man, and so directly opposite to the generous temper and courage of our nation, that ‘tis hardly to be conceived that an Englishman, much less a gentleman, should plead for it,” did in fact support it for the Africans being forcibly shipped to America to serve English gentlemen.
Our inclination might be to see to see this paradoxical combination as peculiar to him, or in any case untypical. But that is the problem: it is all too typical of Western political thought for the next two centuries and more. Even Kant, who developed what is arguably the philosophically purest version of European humanism, also developed what is arguably the most systematic theory of race and racial hierarchy prior to the 19th century. Kant himself was not only abreast of and indeed in some respects ahead of contemporary biological accounts of racial difference, he was also exceptionally well versed in the travel literature of his day. In connection with his lectures on physical geography and anthropology, which he delivered annually from 1756 to 1796, he immersed himself in the reports of explorers, settlers, missionaries, traders, and the like, which constituted a significant part of the “empirical” basis of comparative cultural studies in his day. And though he sometimes ridiculed the reliability of such reports, he nonetheless drew upon them in framing his views of non-European peoples. We could say, then, that his views mirrored the scope and nature of European contacts with non-Europeans in the early-modern period. And Kant being Kant, that mirror had a power and clarity second to none. The popular racism attendant upon “New World” conquest and enslavement found there its most highly resolved theoretical reflection prior to the 19th century, one that already displayed the chief characteristics of 19th-century racial “science”: racial differences were represented as biologically inherited determinants of cultural differences, particularly differences in intellectual and moral capacities. In consequence of what he thus took to be biologically grounded differences in talent and temperament, Kant conjectured that non-European peoples were incapable of autochthonously realizing their humanity, and in particular of attaining that perfectly just civil constitution which is the highest task that nature sets to mankind. Hence Europe would “probably legislate eventually for all other continents.” This was the plan of nature, and thus God’s plan, and thus humanity’s best hope. So here we have it already before the close of the eighteenth century: a theoretical rationale for global white supremacy, rooted in biology and featuring hereditary differences in ability and character, and replete with the civilizing mission of the white race favored by nature to be the pacemaker of cultural progress and to give the law to the rest of the world -- a rationale that had only to be further developed and adapted to meet the needs of the “floodtide of imperialism” in the nineteenth century.
There are, of course, important variations from context to context and author to author. But there is a also general pattern: non-European peoples are characterized as savage or uncivilized, as not possessed of fully developed rational capacities and incapable of fully rational agency, and thus declared to be in need of tutelage, not only for the good of those who command them but also for their own good, for the full development of their capacities. By the latter part of the 18th century, but especially in the 19th century, this scheme was filled in with developmental philosophies of history placing European civilization at the apex of social and cultural progress, and with allegedly scientific theories of racial difference and racial hierarchy. More recently, however, especially since World-War II, the widespread dismantling of colonial empires and of de jure discrimination in regard to citizenship rights has been accompanied by a change in the treatment of race in liberal political theory. There is no longer any attempt theoretically to justify or accommodate racial subordination within a putatively universalistic theory. Rather, the treatment of this persistent feature of the messy political reality we inhabit is consigned to the province of “nonideal theory.”
Part of the explanation behind this consignment is the tectonic shift in methodology that Richard Bernstein analyzed in his early study of The Restructuring of Social and Political Theory. There he noted the ambivalent attitude toward normative theory of the positivist approach to social and political inquiry that had gained ascendancy in the twentieth century. “On the one hand, there is an insistence on the categorial distinction between empirical and normative theory, but on the other hand, there is s widespread skepticism about the very possibility of normative theory.” That skepticism was based on a strict dichotomy between facts and values, the assignment of empirical science to the former domain and of normative theory to the latter, and the view that that there could be no rational determination of anything so subjective as “values.” Bernstein himself went on in that work to criticize this architectonic and to propose an approach to social and political theory that rejected the empirical/normative dichotomy in favor of a mode of inquiry that was at once empirical, interpretive, and critical -- that is, a mode of critical social theory. But just before he offered his diagnosis and proposal in the mid 1970s, a new venture in normative political theory had taken shape, one which by and large accepted the fact/value split, rejected the skepticism concerning rational discourse about values, and proposed another, more cooperative division of labor between empirical science and normative theory: the theory of justice of John Rawls, which has remained the dominant paradigm of normative political theorizing to this day.
In this paper, I want to (I) examine the division of labor between ideal and nonideal theory as it appears in Rawls’s thought, (II) identify some of the obstacles this paradigm presents to developing an adequate account of racial injustice, and (III) suggest that to overcome them we have to move in the direction of a critical theory of race.
As Susan Moller Okin and others have argued, Rawls effectively screened out gender and the gendered structure of the family from the purview of A Theory of Justice by designating the participants in the original position as “heads” or “representatives” of families. Behind the veil of ignorance, not only is one’s sex unknown to one, but participants seem to be unaware of the sex-gender systems that have deeply structured every society on record. Accordingly, the massively differential effect of the “morally arbitrary” fact of sexual difference does not become a central theoretical issue for justice as fairness. And although Rawls mentioned gender as a basic problem of our society in the introduction to Political Liberalism some 20 years later (PL, xxviii), it still remained untheorized therein. After a second round of criticism by Okin and other feminist theorists, Rawls briefly addressed the matter in his 1997 University of Chicago Law Review piece on “The Idea of Public Reason Revisited.” (CP, 575-615) There he endorses a principle of equal justice for women which requires that “wives have all the same basic rights, liberties, and opportunities as their husbands.” (CP, 597) He notes that the gendered division of labor in the family has been implicated historically in the denial of such equality, and holds that such a division might persist in a just society only if and when it were “fully voluntary” and arrangements were made to ensure that it did not undermine the equal liberties and opportunities of women. (CP, 600) It seems to follow -- and this is my main point -- that such arrangements would have to be given full consideration in the original position, and thus that the parties would have to be given access to the knowledge of “social theory and human psychology” (CP, 601) needed to deal rationally with them. In short, sex-gender would have to be theorized at the same level and in the same detail as other major axes of justice/injustice.
A similar question might be raised concerning cultural and institutional patterns of racial domination. Charles Mills has remarked on the surprising insignificance of racial discrimination as an explicit theme (rather than a tacit subtext) of A Theory of Justice, which appeared in a highly charged political atmosphere a few short years after African-Americans had finally won their centuries-long struggle for equal civil and political rights. One might add to this the absence of any sustained discussion of colonialism in a world convulsed with the dying gasps of European-American (formal) global rule. These, it seems, are features of the modern world about which parties in the original position are ignorant and are therefore unconcerned to address explicitly in laying out the basic structures of justice. But though the American Civil Rights Movement and the global decolonization struggles of the period did eliminate most forms of de jure inequality, many forms of de facto inequality remained in place --deeply entrenched in the beliefs and values, symbols and images, practices and institutions, structures and functionings of national and global society. Hence, while certain legalized forms of subordination like slavery and serfdom may now be, as Rawls says in Political Liberalism, “off the agenda” (PL,151, n.16)), the same cannot be said for racial relations generally, as he acknowledges in the introduction to that same work and as the persistent debates about affirmative action and other proposals for addressing the enduring legacy of legalized racism attest. Why, then, if “race” is admittedly still among “our most basic problems” (PL, xxviii), is it not theorized therein?
An important part of the answer, I want to suggest, has to do with the nature of “ideal theory.” Kantian in conception, ideal theory starts with “rationally autonomous agents” and allows them only so much information as is needed to achieve agreement on basic principles. This point is reached, according to Rawls, when the parties in the original position, represented now as symmetrically situated trustees of free and equal, rational and reasonable citizens, know all and only those “general” -- not particular or personal -- “facts” about society -- that is, laws, theories, and tendencies pertaining to politics, economics, psychology, and social organization -- required to design a just and feasible basic structure. Or, as Rawls also puts it in his most recent book, ideal theory seeks to construct a “realistic utopia.” (LP, 6) In this respect, it follows Rousseau in taking (a) “men as they are” and (b) “laws as they might be.” (LP, 7, 13)
(a) Rawls understands the former phrase to mean “persons’ moral and psychological natures and how that nature works within a framework of political and social institutions.” (LP, 7) To take men as they are in this sense means to “rely on the actual laws of nature and achieve the kind of stability these laws allow.” (LP, 12) Hence parties in the original position must have access to the relevant general knowledge about such laws. This has consistently been Rawls’s position. In TJ, while the veil of ignorance rules out knowledge of particular personal or social circumstances, the parties do know “the general facts about human society” pertaining to politics, economics, psychology, and social organization. Indeed, “there are no limitations on general information, that is, on general laws and theories, since conceptions of justice must be adapted to the characteristics of the systems of social cooperation which they are to regulate.” (TJ, 137-8) Some thirty years later, in The Law of Peoples, the same idea is expressed in similar terms: ideal theory is “realistic” when it depicts a social world that is “possible,” that is, one that is “achievable” in light of what we know about “the laws and tendencies of society.” (LP, 11) But there is added emphasis in the later work on “political culture” as a condition of feasibility. Thus in The Law of Peoples, civil society, the religious and moral traditions that support the basic structure of political and social institutions, the industriousness and cooperative talents of members, and their political virtues are all mentioned as “crucial” to the possibility of a society being well-ordered. (LP, 108) Accordingly, relevant “general” knowledge -- general facts, laws, theories, tendencies -- about political culture would have to be added to that concerning economic, politics, social organization, and psychology, if the parties are to design a realistic basic structure, one that is feasible as well as just. At the same time, however, this way of delimiting what is known behind the veil of ignorance entails that much of the “particular” knowledge required to understand and deal with racial injustice will not be available to parties in the original position -- knowledge, for instance, about “the particular circumstances of their own society,” or “to which generation they belong” (TJ, 137), or “the relative good or ill fortune of their generation.” (PL, 273)
(b) Rawls understands Rousseau’s phrase “laws as they might be” to mean “laws as they should or ought to be,” (LP, 7) and this too seems to entail that constructions of race have no place in ideal theory -- for “race” should not be a structuring principle of political and social relations. It is, in short, morally irrelevant. But this then threatens to render normative theory both unrealistic and unfair. To quote Charles Mills: “Failure to pay theoretical attention to this history [of racial subordination] will then just reproduce past domination, since the repercussions of white supremacy for the functioning of the state, the dominant interpretations of the Constitution, the racial distribution of wealth and opportunities, as well as white moral psychology...are not examined.” It is at this point that Rawls’s notion of “nonideal theory” comes into play. Taking the principles established by ideal theory as a guide, it approaches the “noncompliance” and “unfavorable conditions” of the real world in a spirit of reform: it asks how political ideals “might be achieved, or worked toward, usually in gradual steps. It looks for policies and courses of action that are morally permissible and politically possible as well as likely to be effective.” (LP, 89) Thus, though “ the specific conditions of our world at any time... do not determine the ideal conception, “ they “do affect the specific answers to questions of nonideal theory.” (LP, 90) Does this division of labor work?
We can understand this as a question about the relative strengths and weaknesses of alternative theoretical strategies, for political theory has been practiced in a great variety of ways, each with its peculiar advantages and disadvantages. One advantage of Rawls’s neoKantian strategy is precisely the “purity” of the conception of social justice it constructs: social arrangements are based on equal respect for free and equal moral persons. Morally irrelevant particularities are systematically excluded or, when unavoidable, compensated for, inasmuch as even at this high level of abstraction there are certain “impurities” that cannot be simply excluded from ideal theory if the envisioned “utopia” is to be at all “realistic.” Rawls already built some of these into A Theory of Justice --for instance, inequalities of birth and natural endowment -- and in Political Liberalism he in effect added the persistence of deep cultural and ideological differences to the list: ideal theory now has to accommodate, within its theory construction, the “fact of reasonable disagreement” about the meaning and value of human life. And Rawls seems now to concede that feminist arguments for adding sex and gender to the list are irresistible. At the very least, the “general fact” of the biological division of labor in the reproduction of the species has to be acknowledged by the parties in the original position and so accommodated in their design of a basic structure as to ensure women’s substantive freedom and equality. Our question now is whether the same may be said of race? For Rawls, I think, the answer has to be “no.” If “race,” in the sense at issue here, is at bottom a social, cultural, and political formation developed for purposes of subordinating certain groups to others, then it should simply be eliminated as a structuring element in a “well-ordered society.” In the ideal society, there would be a total absence of race in this sense . So ideal theory cannot be where it is theorized.
In itself, this need not be a problem. Political theorizing has always been carried on at various levels of abstraction. There is, in my view, nothing to be said against even the highest -- “Kantian” -- levels per se, providing that they are not taken to be exclusively valid for, or inclusively adequate to, their objects. Given that Rawls’s political theory is configured as a multilevel undertaking, with the intent of addressing relevant problems at one level or another, the question for us to consider is how his overall strategy, particularly the ideal/nonideal dichotomy, measures up against alternative strategies as a way of theorizing race.
To begin with, general “facts” about the social world are, as anyone who has followed the discussions in postpositivist philosophy of social science will be aware, hardly the uncontroversial matter that Rawls represents them to be. Facts are stated in languages, and so long as there is no single general theory on which a consensus has formed within and among the relevant communities of investigators in any of the major domains of social life, the languages, and hence the facts -- not to mention the “general laws” -- of the social “sciences” are up for debate. Unless Rawls wants to take a firm position on, say, the century-old understanding/explanation debates -- and build that position into his “freestanding” political liberalism -- he will have to leave open the possibility that social and political inquiry has an ineliminable interpretive dimension and thus that what the general facts about social life are cannot be settled from the standpoint of a neutral observer or a reflective equilibrator. If “realistic” political theory cannot be pursued without incorporating into it knowledge of the general characteristics of the social systems to which it is meant to apply, then political theory will have to get involved in just the sorts of interpretive-historical and social-theoretical disagreements which, in its self-understanding as normative theory, it hopes to avoid. And interpretive approaches to the human world typically place more and different weight on historical modes of inquiry than do positivist or empiricist approaches. Hermeneutic understanding is inherently historical: it aims to comprehend social phenomena as historical phenomena, often in narrative terms. But then Rawls strict separation of “general” from “particular” knowledge of society become problematic, if, as hermeneutically inclined social theorists maintain, general information about society always comes, even if often only tacitly, with an historical index.
If political theorists do not dispose of interpretation-free “facts” in the way that Rawls intends, neither do they have conflict-free “values” at their disposal. Rawls himself explicitly characterizes the political values that his conception of justice seeks to articulate as belonging to the public political culture of a particular historical society and not to some ideal realm beyond the world. But then it follows that, as such, they do not come with fixed, clear, uncontested meanings; rather, they have to be interpretively worked up from the variable, particular, often conflictual political contexts in which they figure. As a result, the basic terms of his political conception cannot but reflect and project the particular forms of life and situations of conflict from which they are prepared; and they must be understood and assessed in relation to them. Thus his guiding conceptions of “persons” with two “moral powers” as both “free and equal” and “rational and reasonable,” of “primary goods” as “specifying citizens needs” in a way that provides a “practicable basis of interpersonal comparison, ” and of the “stability” of a “well-ordered society” are laden with particular -- and contestable --interpretations and evaluations. It makes no sense to suppose that we could insulate their construction from the conflicts of interpretation and evaluation endemic to our public political culture, our constitutional tradition, our legal and political practices and institutions. Working them up theoretically via reflective equilibrium or rational reconstruction cannot remove the traces of their conflictual origins .
In A Theory of Justice, Rawls was clear that the method of reflective equilibrium could not simply articulate an already existing consensus on basic political values, for the reason, among others, that such values had always been and continued to be debated in the public political culture. In view of the admittedly deep divisions on many of the matters to which he addressed himself -- for example, the meanings and relative weights of liberty and equality -- he did not understand his method of reflective equilibrium to be the hermeneutically conservative operation interpreters and critics sometimes mistook it to be. He remarked, for instance, that the kind of wide reflective equilibrium proper to moral philosophy might bring about a “radical shift” in our sense of justice (TJ, 49). It involved, as he put it a few years later, asking “what principles people would acknowledge and accept the consequences of when they had the opportunity to consider other plausible conceptions and to assess their supporting grounds... [It] seeks the conception, or plurality of conceptions, that would survive the rational consideration of all feasible conceptions and reasonable arguments for them.” (“The Independence of Moral Theory”, CP, 289) But this might well put the theorist in the position of defending a comprehensive moral theory within the conflict of interpretations and reconstructions. And that is something which Rawls’s recent stress on “overlapping consensus” now explicitly disallows.
In Political Liberalism and The Law of peoples, the idea of the “reasonable” undergoes considerable dilution. The overriding concern of both these works is with the “feasibility” of liberal ideals and the “stability” of liberal institutions in the face of cultural and ideological pluralism. The irreducible plurality of basic views on the meaning and value of human life makes it necessary, Rawls now maintains, to construct a purely “political” conception of justice that “stands free” of “comprehensive doctrines” of any sort, including general philosophical views. This “strategy of avoidance” relocates the “reasonable” at some remove from the Kantian notion of reason, with its close connection to the idea of a critique that submits all claims to authority to the free examination of reason. The reasonable pluralism that we might expect to result from “the exercise of human reason under free institutions” (PL, 55-58) is, in Rawls’s construction of a “political” conception of justice, replaced in effect by the de facto pluralism of comprehensive doctrines that satisfy the much weaker requirements set by his revised notion of toleration. As a result, enlightenment and critique can play only a severely restricted role in normative theory of this sort, and that too makes it an unsuitable vehicle for theorizing racial injustice. In my concluding remarks I can do no more than gesture toward an alternative -- and in my view more promising -- theoretical strategy.
In interpreting the languages of political thought, normative theorists too often take abstract formulations at their word, as if what were left out of the saying were left out of the meaning; they neglect to attend to how key terms actually function in the multiplicity of contexts in which authors and their audiences put them to use, or to what in practice are regarded as conditions of satisfaction and acceptability for claims employing them. They tend also to disregard that general norms are always understood and justified with an eye to some range of standard situations and typical cases assumed to be appropriate, and that if that range shifts, then so too do the understandings and justifications of those norms, the conceptual interconnections and warranting reasons considered relevant to them. On the other hand, recognizing that ideals and principles of justice, however abstract their form, always come with contentful preunderstandings that derive from their locations not only in systems of thought but also in forms of life, does not in itself commit us to sheer localism. In the case in point of liberalism’s complicity with racial slavery, for instance, many of the ideas implicated in the justifications of slavery were also given more inclusive interpretations in the same cultural contexts as the dominant exclusionary versions highlighted in section I above. That is to say, those contexts also provided resources for arguments against slavery on religious or philosophical grounds, including arguments to the effect that the basic rights possessed by all human beings as such forbade it. One could say, then, that there were competing meanings -- networks of inferential connections, ranges of standard situations and typical cases -- which partly overlapped and partly diverged, but which were sufficiently interlinked to make disagreements real disagreements and not just incommensurable mutterings. And one might then understand the work of critics -- and critical theorists -- as an ongoing effort to reweave those connections and redefine those ranges so as to promote more genuinely inclusive versions. In doing so, they adopt the internal perspective of reflective participants and invoke the context-transcending validity claims of putatively universal ideals to argue that they have been betrayed, that existing formulations, though expressed in formally universal terms, are actually exclusionary.
On this view, the search for a genuinely inclusive theory of justice is a never ending, constantly renewed effort to rethink supposedly universal basic norms and reshape their practical and institutional embodiments to include what, in their limited historical forms, they unjustly exclude. What generally drives this effort are struggles for recognition by those whom the norms in their established versions fail to recognize. And the intellectual form it takes is the ongoing contestation of essentially contestable articulations of the universal demands of justice. Judith Butler puts the point this way: “the provisional and parochial versions of universality” encoded in law at any given time never exhaust “the possibilities of what might be meant by the universal.” Contestation by subjects excluded under existing definitions and conventions are crucial to “the continuing elaboration of the universal itself,” for “they seize the language of [the universal] and set into motion a ‘performative contradiction,’ claiming to be covered by [it] and thereby exposing the contradictory character” of conventional formulations. Butler here captures the important idea that the possibility of challenging putatively universal representations is inherent in those representations themselves, or more precisely, in their context-transcending claims, and that historically that possibility has been exploited to greatest effect by groups who, though not entitled under existing formulations of the universal, nevertheless appeal to it in formulating more inclusive conceptions of justice.
Viewed in this light, as a part --albeit a reflective part -- of historical processes of emancipation, normative theory is clearly not “freestanding” in any fundamental sense. And, as the shock-effects of Foucault’s genealogies have made clear, the familiar enlightenment metanarratives of universal principles discovered at the birth of modernity fail to acknowledge the impurity of the demands that have historically been made in the name of pure reason. Accordingly, there is a need for critical “histories of the present,” the aim of which is to alter our self-understandings by examining the actual genealogies of accepted ideas and principles of practical reason. This distinguishes critical approaches to moral and political theory from approaches like Rawls’s that seek to construct fundamental norms of justice from the “settled convictions” of our “public political culture” by way of “reflective equilibrium.” Critical histories make evident that the political values from which political liberalism seeks to construct a political conception of justice have always been and still are deeply contested, often fiercely, and usually in connection with matters of power, desire, and interest. And they make us aware that the quite varied, often conflicting ideas, principles, values, and norms that have been taken to express the demands of justice cannot adequately be comprehended or assessed without understanding that and how elements of the contexts and situations in which they were propounded entered into them.
It is not only this “context of origins” that contemporary normative theory leaves largely unexamined, but the “context of applications” as well. The distinction that Rawls and others draw between ideal and nonideal theory insulates political theorizing, at least initially, from the messiness of political reality. Subsequent forays into nonideal theory are all too often of limited value because of their loose, post hoc connection to empirical work. Specifically, discussions of race following this strategy usually end up as discussions of affirmative action in the broadest sense: since equal citizenship rights are now largely in place, the “unfavorable conditions” at issue are the substantive inequalities that are the enduring legacy of centuries of legalized oppression and discrimination. Of course, one then has to judge any proposed remedial measures, policies, and programs from a pragmatic as well as a moral point of view, for they are put forward as practical means to the desired end of eliminating or reducing those inequities. Hence the case for any concrete compensatory measures has to be made not just “in principle” but “all things considered,” that is, it has to take into account empirically likely consequences and side-effects, costs and benefits, comparative advantages and disadvantages vis-à-vis possible alternatives, political viability, long-term efficacy, and so on. Thus nonideal theorizing of this sort turns normative political theory back in the direction of the empirical social reality it began by abstracting and idealizing away from. But -- and this is my main point here -- there are no theoretical means at hand for bridging the gap between a color-blind ideal theory and a color-coded political reality, for the approach of ideal theory provides no theoretical mediation between the ideal and the real -- or rather, what mediation it does provide is usually only tacit and always drastically restricted.
On this last score, what Habermas argues in Between Facts and Norms to be true of legal theory holds ceteris paribus of normative political theory as well -- namely, that it always relies upon implicit background assumptions drawn from some preunderstanding of contemporary society’s structures, dynamics, potentials, and dangers. These implicit “images” or “models” of society tacitly enter into normative-theoretical constructions and often play a covert role in what appear to be purely normative disagreements. Deep differences in normative theory -- for instance, those separating classical from social-welfare liberals, or those dividing both from their socialist, radical feminist, or postcolonial critics -- often turn on disagreements about the “facts” being assumed, implicitly as well as explicitly, in regard to markets, classes, gender roles, global relations, and so forth. And as we know from the history of theory, significant shifts in thinking often come about as a result of challenges precisely to what have previously been taken for granted as the natural, unalterable facts of social life -- class-structured distributions of the social product, gendered divisions of labor, race-based hierarchies of social privilege, ethnocultural definitions of political membership, and the like. These considerations strongly suggest that such understandings, images, or models of society, which are always at work, though usually only tacitly, in normative theorizing, have to become an explicit theme if political theorists hope to avoid exalting intuitive preunderstandings of their social contexts into universal ideals. But political theory would then have somehow to combine intuitive knowledge from the perspective of the “insider” with counterintuitive knowledge from the perspective of the “outsider,” in the senses both of the observer and the excluded. It would have to join the constructive and reconstructive aims of normative theory to the interpretive, analytical, and explanatory aims of history and other empirically based human studies, and to the practical aims of social and cultural criticism. Given the existing institutionalization of research and scholarship, it would have to become interdisciplinary to the core. And this means that normative theory would have to become an interdependent -- not freestanding -- moment of a larger critical enterprise; that is, it would have to be pursued in a self-consciously interdisciplinary manner and remain theoretically responsive not only to the political struggles of the age but also to contemporary developments in historical, social, and cultural studies. At least it would have to do so if it hoped to have anything of interest to say about racial injustice.
 An earlier version of this paper was read at the Central Division Meeting of the American Philosophical Association in April of 2000. I am grateful to Lucius Outlaw for his comments on that occasion. For comments on a first draft, I am indebted to James Bohman, Felmon Davis, Maria Herrera, Richard Kraut, Christopher Zurn, and Robert Gooding-Williams, to whom I also owe many illuminating discussions of issues in race theory.
 Karl Marx, Capital (New York: Random House, 1906), p. 823.
 Michael Omi & Howard Winant, Racial Formation in the United states (New York: Routledge, 1994). They analyze race as a complex of social meanings that shape practices, identities, and institutions. This displaces the idea of race as a biologically fixed nature without reducing it to a mere fiction. Rather, it is a social construction that figures importantly in social, cultural, legal, political, economic, and psychological realities.
 As famously anticipated by W.E.B. Du Bois in 1903, The Souls of Black Folk, edited with and introduction by D.W. Blight & R. Gooding-Williams (Boston: Bedford Books, 1997), p. 45.
 And despite the best efforts of philosophers in the alternative tradition to bring this fact to the attention of mainstream political philosophy. In 1977-78, for instance, there was a double issue of The Philosophical Forum devoted to “Philosophy and the Black Experience.” Numerous other publications on political philosophy and race appeared in the 1970s and 1980s, from the hands of Bernard Boxill, Leonard Harris, Bill Lawson, Howard McGary, Lucius Outlaw, Laurence Thomas, and Cornel West, among others. Nonetheless, the agenda of mainstream, liberal, political theory has remained largely unchanged in this regard.
 Cited in Blackburn, The Making of New World Slavery, p. 275, n.92.
 John Locke, Two Treatises of Government, Peter Laslett, ed. (Cambridge: Cambridge University Press, 1988), p. 141.
 See Robert Bernasconi, “Who Invented the Concept of Race? Kant’s Role in the Enlightenment Construction of Race,” forthcoming in Race, R. Bernasconi, ed. (Oxford: Blackwell Publishers, 2001); and Emmanuel Chuckwudi Eze “The Color of Reason: the Idea of ‘Race’ in Kant’s Anthropology, “ In Eze, ed., Postcolonial African Philosophy (Oxford: Blackwell Publishers, 1997), pp. 103-140.
 I. Kant, “Idea for a Universal History with a Cosmopolitan Purpose,” in Kant: Political Writings, ed. by H. Reiss, tr. by H.B. Nisbet (Cambridge: Cambridge University Press, 1991), p. 52.
 However, Kant himself denied that there could be any right to chattel slavery and roundly condemned the European colonialism of his day. For the former, see The Metaphysics of Morals, tr. by M. Gregor (Cambridge: Cambridge University Press, 1996), p. 66; for the latter, see ibid., pp. 53, 121-2, and “Perpetual Peace,” in Kant; Political Writings, pp. 106-7.
 Exceptions to this pattern can be found, for instance, among the philosophes of the French Enlightenment. For a brief overview, see David Brion Davis, The Problem of Slavery in Western Culture (Oxford: Oxford University Press, 1988), chapter 13, “The Enlightenment as a Source of Antislavery Thought.”
 Philadelphia: University of Pennsylvania Press, 1978. See especially the section entitled “The Problems of Normative Theory,” pp. 45-51.
 Ibid., p. 45.
 A strong case for the inadequacy of mainstream political theory in regard to race and for the consequent need to develop a more historically and socioculturally informed critical approach is made by Charles W. Mills in The Racial Contract (Ithaca: Cornell University Press, 1997) and Blackness Invisible (Ithaca: Cornell University Press, 1998). I am indebted to Mills for opening up many of the paths of argument I pursue here, though the conclusions I arrive at sometimes diverge from his. An early call for a critical theory of race in the tradition of Frankfurt School critical social theory was issued by Lucius Outlaw in his contribution to David Theo Goldberg, ed., Anatomy of Racism (Minneapolis: University of Minnesota Press, 1990), pp. 58-82: “Toward a Critical Theory of ‘Race’.” The intentionally broad sense in which I use the term “critical theory” in this paper derives from but is not limited to that tradition. As the sketch of methodological desiderata in section III should make clear, it is meant to be sufficiently inclusive to accommodate productive strains of critical historical, social, and cultural studies that answer to different names. It might be worth noting explicitly that a critical theory of race in this sense would not be restricted to the topics or methodologies of the “critical race theory” that has taken shape within the field of critical legal studies. See Kimberle’ Crenshaw et al., eds., Critical Race Theory: the Key Writings that Formed the Movement (New York: New Press, 1995), and R. Delgado & J. Stefancic, eds., Critical Race Theory: the Cutting Edge (Philadelphia: Temple University Press, 2000).
 Susan Moller Okin, Justice, Gender, and the Family (New York: Basic Books, 1989). Rawls’s works will be cited by abbreviation in the text as follows: TJ, A Theory of Justice (Cambridge: Harvard University Press, 1971); PL, Political Liberalism (New York: Columbia University Press, 1993); CP, Collected Papers, Samuel Freeman, ed. (Cambridge: Harvard University Press, 1999); and LP, The Law of Peoples (Cambridge: Harvard University Press, 1999).
 See, for instance, Susan Moller Okin, “Political Liberalism, Justice, and Gender,” Ethics 105 (1994):23- 43.
 Mills, Blackness Visible, pp. 5, 152. This is not to deny that a deep concern with racial justice may well have been part of the motivation behind the argument of TJ, but only to point out that it is not systematically discussed therein -- though there are a few scattered remarks on race and slavery; see, e.g., pages 99, 158, 248, and 325, and the footnote to Martin Luther King (p. 364n.) in the extended discussion of civil disobedience in chapter VI. A number of theorists interested in questions of racial injustice have remarked and, in various ways, tried to make good on this lack in A Theory of Justice, among them Bernard Boxill, Blacks and Social Justice (Totowa, NJ: Rowman & Allanheld, 1984); David Theo Goldberg, Racist Culture (Oxford: Blackwell, 1993); and Howard McGary, Race and Social Justice (Oxford: Blackwell, 1999). Though some of the themes I pursue in this section and the next are commonplaces of this genre of Rawls critique, my hope is that the particular line of argument I develop will repay revisiting them.
 Rawls described it this way in “Kantian Constructivism in Moral Theory,” CP, 336. Despite his shift to a conception of political theory as “freestanding,” he continues to acknowledge the Kantian provenance of certain key ideas. Thus the earlier characterization in TJ of the veil of ignorance as “implicit, I think, in Kant’s ethics” (TJ, 140-1) recurs in PL, where he characterizes his preference for a “thick” veil of ignorance -- in which “the parties are to be understood so far as possible solely as moral persons and in abstraction from contingencies” -- as “a Kantian view.” (PL, 273) A footnote to this passage points us back to the distinction between a thicker and a thinner veil of ignorance which was elaborated in “Kantian Constructivism in Moral Theory” (CP, 335-6) and which is credited to Joshua Rabinowitz in both places. There we are told that “a Kantian doctrine aims at the thickest possible veil of ignorance,” initially allowing the parties no information at all and then adding “just enough so that they can make a rational agreement,” thus ensuring that the first principles of justice are those of “rationally autonomous agents” represented solely as “free and equal moral persons” with no more information than is required for agreement. He contrasts this with a “Humean” approach that would initially allow the parties full information and then rule out just enough “to achieve impartiality in the sense of the elimination of threat advantage.” This veil of ignorance is thinner than the Kantian in that the parties still have access to information about the configuration of the society for which they are designing principles of justice. Even if both approaches led to the same principles, Rawls writes, “the thicker veil of ignorance would still be preferable, since these principles are then connected more closely to the conception of free and equal moral persons,” whereas the alternative approach “obscure[s] how intimately the principles are tied to the conception of the person.”
 Blackness Visible, p. 108.
 On the other hand, the evidence seems plain that ethnocentric modes of thought and action have been “general facts” or “general tendencies” of human society throughout recorded history. See Hannaford, Race, and Davis, Slavery, op. cit. Whether there are any “general laws” at work here is hard to say, particularly as no stable meaning can be given to that notion in political theory, as I shall argue below.
 As to whether it might be “conserved” in some other sense, see the differing views on Du Bois’s “The Conservation of Races” expressed by Anthony Appiah, “Illusions of Race,” chapter two of In My Father’s House (New York: Oxford University Press, 1992); Lucius Outlaw, “‘Conserve’ Races?”, in W.E.B. Du Bois on Race and Culture, eds. B. Bell, E. Grosholz, & J. Stewart (New York: Routledge, 1996), pp. 15-37; and Robert Gooding-Williams, “Outlaw, Appiah, and Du Bois’s ‘The Conservation of Races’,” in the same volume, pp. 39-56.
 Rawls does not use the terminology of “levels” of theory, but of “parts” (ideal and nonideal) and of “stages” (the four-stage sequence). I shall be focusing on the former, but a full treatment of the issues I raise would require examining the latter as well. My conjecture is that similar considerations would apply, as the basic structure of justice is already designed at the first stage, but I cannot argue this here. See Habermas’s remarks on the four-stage sequence in “Reconciliation and the Public Use of Reason: Remarks on John Rawls’s Political Liberalism,” in The Journal of Philosophy, XCII (1995):109-131, pp. 118 and 128; and Rawls’s “Reply to Habermas,” in the same volume, pp. 132-180, here pp. 151-3.
 For an elaboration and defense of this point, see Richard Bernstein, Restructuring, and, in a different but related context, Between Objectivism and Relativism (Philadelphia: University of Pennsylvania Press, 1983).
 The resultant strains can, I think, be glimpsed in the tension, in his “Reply to Habermas,” between his characterization of discourse in civil society as an “omnilogue” in which citizens directly debate the relative merits of competing conceptions of justice (p. 140) and his account a few pages later of public justification as a kind of indirect “overlapping consensus”: “Public justification happens when all the reasonable members of political society carry out a justification of the shared political conception by embedding it in their several reasonable comprehensive [doctrines]...[C]itizens do not look into the content of others’ doctrines...Rather, they take into account and give some weight to only the fact -- the existence -- of the reasonable overlapping consensus itself.” (pp. 143-4)
 For the details of this argument, see my “Kantian Constructivism and Reconstructivism: Rawls and Habermas in Dialogue,” Ethics 105 (1994):44-63, and “On the Idea of a Reasonable Law of Peoples,” in J. Bohman & M. Lutz-Bachmann, eds., Perpetual Peace (Cambridge, MA: The MIT Press, 1997), pp. 201-217.
 I hope to flesh out this alternative in future publications; the discussion here remains at the level of abstract methodological requirements.
 This is essentially the strategy Charles Mills recommends in The Racial Contract.
See Axel Honneth, The Struggle for Recognition, tr. J. Anderson (Cambridge, MA: The MIT Press, 1996).
 Judith Butler, Excitable Speech (New York: Routledge, 1997), p. 89. Her position is interestingly elaborated but not essentially changed in J. Butler, E. Laclau, & S. Zizek, Contingency, Hegemony, Universality (London: Verso, 2000).
 See Richard Bernstein, “Philosophy, History, and Critique,” in Bernstein, The New Constellation (Cambridge, MA: The MIT Press, 1992), pp. 15-30.
 Bernstein makes similar points in regard to Rorty’s appropriation of Rawls in The New Constellation, pp. 238-249.
 This is the general approach to racial injustice taken by liberal theorists such as Bernard Boxill, Ronald Dworkin, Gertrude Ezorsky, Thomas Nagel, and Judith Jarvis Thompson, among others. A recent and rich elaboration of it is Amy Gutmann’s “Responding to Racial Injustice,” in K. Anthony Appiah & A. Gutmann, Color Consciousness (Princeton: Princeton University Press, 1996), pp. 106-178.
 J. Habermas, Between Facts and Norms, tr. W. Rehg (Cambridge, MA: The MIT Press, 1996), chapter 9, “Paradigms of Law.”
 I do not mean to imply here that “counterintuitive” knowledge could not also come from critical “insiders,” as is argued by Michael Walzer, for instance, in The Company of Critics (New York: Basic Books, 1988).
 To be sure, critical race theorists from Du Bois onwards have typically worked in interdisciplinary ways. But their work has largely been ignored by mainstream political theory. So this paper could also be construed as an argument for changing the canon.
COPYRIGHT NOTICE: THE PRECEDING TEXT "Political Philosophy and Racial Injustice: From Normative to Critical Theory" IS FULLY PROTECTED BY COPYRIGHT LAW AND MAY ONLY BE QUOTED UNDER THE CORRESPONDING CONVENTIONS. © 2001 Thomas A. McCarthy