Jump to ContentJump to Main Navigation
Women, Culture, and DevelopmentA Study of Human Capabilities$

Martha C. Nussbaum and Jonathan Glover

Print publication date: 1995

Print ISBN-13: 9780198289647

Published to Oxford Scholarship Online: November 2003

DOI: 10.1093/0198289642.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 20 June 2021

Why not a Feminist Theory of Justice?

Why not a Feminist Theory of Justice?

(p.298) Why not a Feminist Theory of Justice?
Women, Culture, and Development

Ruth Anna Putnam

Oxford University Press

Abstract and Keywords

Putnam first considers and critiques Rawls's theory of justice from a feminist perspective, then endeavours to address the problem of how to preserve the universality of justice without repressing the multiplicity of voices of distress by employing both a Rawlsian stance and a pragmatist approach. Finally, Putnam defends Rawls's conception of justice against Sen's concern regarding its limited scope and Glover's worry of its culturally imperialist appropriation in the Third World. Putnam's conclusion is that a universalist theory is necessary and that any such theory ought to draw from Rawls's work.

Keywords:   feminist critique of Rawls, justice, liberalism, Rawls, universalism

Why not a feminist theory of justice? How are we to hear this question? And in what sort of context does this suggestion arise? Is it as if we were in some store, shopping around for a suitable theory of justice, and someone suggested a feminist theory, as in another context someone might suggest a bottle of Pinot Noir? Or is it as if one were choosing a career and someone suggested becoming a philosopher? The latter comes closer to the right sort of tone. Choosing a career is a serious matter, much hangs on it. Choosing a career is a commitment, afterwards one is committed to behave in certain ways and not in others. Similarly choosing a theory of justice, to the realization of which one will then be committed, is a serious matter: it will shape one's subsequent conduct. Again, when one chooses a career one chooses to bring about certain things, to build a career—careers do not lie on shelves, ready‐made, waiting to be bought. Similarly, choosing a theory of justice is choosing to bring about certain things; even if the perfect conception of justice already existed, one would have to work out how to apply it to the society in which one happens to live. Just as one's prior conception of a career may be modified as one prepares for and then pursues it, so one's conception of justice may be modified as it is used to combat the injustices one encounters.

One is prompted to seek a theory of justice when one's more or less inarticulate sense of justice is outraged by some feature or features of a society.1 If the outrageous features include a large number of systematic injustices to women, then one might well suggest seeking a feminist theory of justice. Thus when Drèze and Sen point out that millions of women are ‘missing’ in South and West Asia, in China, and in North Africa, one is moved to agree with Susan Moller Okin that we confront ‘a major “justice crisis” in contemporary society arising from issues of gender’ (Drèze and Sen, 1989: 50ff; Okin, 1989: 7). One may then seek a feminist theory of justice as a basis from which to launch one's protest against the gender injustices that explain these horrific statistics.

Suppose now that Pinot Noir has been suggested, or a career in philosophy, and the response is, ‘No, that will never do’. Then one asks, ‘Why not?’ So, one might ask with a different emphasis, ‘Why not a feminist (p.299) theory of justice?’ if someone were to reject the suggestion that a feminist theory of justice would enable us to deal with some of the glaring injustices of our society that the existing theories neglect. Obviously, a theory of justice which would establish and entrench women as first‐class citizens and relegate men to second‐class status would be no better than a theory which attempts to justify the oppression of women. What is wanted, as Okin notes correctly, is a humanist theory, a theory which treats all human beings as equal from the moral point of view, or at least all those human beings who are in Seyla Benhabib's phrase ‘capable of speech and action’ (Benhabib, 1989: 6). So, I shall understand the phrase ‘feminist theory of justice’ in a weaker sense. A feminist theory of justice focuses attention on those issues of justice which it claims to be (a) of particular concern to women as women,2 and (b) generally ignored by major contemporary theories of justice. Okin's book Justice, Gender, and the Family may be regarded as a feminist theory of justice in this sense in spite of the fact that she seeks a humanist theory. Hers is an important voice that we need to hear on our way to a fully human theory of justice.

My present engagement with theories of justice is prompted by Chen's concluding remark (Chen, in this volume, p. 55) that we need a concept of justice that takes seriously the problems of women who must provide for themselves and their children but face enormous traditional barriers to gainful employment. Glover, responding to Chen, wonders whether our saying that these women suffer from injustice is just an expression of our Western bias, whether we can defend that judgement against the objections of communitarians and relativists. He concludes that more needs to be said in defence of moral objectivity (Glover, in this volume).

Chen's and Glover's remarks suggest that we have no adequate theory of justice. Against that, one may be inclined to say that the Rawlsian conception of justice as fairness is the sort of thing Chen seeks and that Rawls's theory provides an adequate account of what makes his conception objective, or at least objective enough for our purposes. What is needed is a basis on which women whose very survival is at stake can stand and demand access to gainful employment as a matter of right, not as a matter of mercy for which they would have to beg. Such a basis would also enable one to justify the intervention of various relief organizations such as BRAC in traditional village life.

Any appeal to Rawls's theory for these purposes must, however, defeat or circumvent two objections. First, as already mentioned, it has been shown, e.g., by Okin, that Rawls's theory fails to take adequate notice of the problems of women; this failure, she argues, is built deeply into the structure of the theory. While I agree with the first part of Okin's claim, I shall oppose her suggestion that women must be represented as women in the (p.300) Rawlsian original position. Secondly, Sen believes that Rawls, in his more recent writings, has restricted the applicability of his theory to societies that are already democratic in the sense that they practice tolerance of a plurality of conceptions of the good life (Sen, in this volume, pp. 265–6; see also 1989). This condition fails to be met in many of the Third World countries with which we are concerned in this volume. I am inclined to dispute this reading of Rawls's more recent papers; in any case, when we appropriate Rawls's theory for our purposes, we are not obliged to accept his restrictions. Although the priority of liberty is a key feature of Rawls's theory, and although freedom of conscience (read: tolerance for a plurality of conceptions of the good) seems to be the most important of the liberties, one can nevertheless protest against any violation of either principle even while one objects also against violations of these most fundamental requirements of justice.

My strategy will be the following. After laying out abstractly what I take to be major threats to a universalistic theory of justice (Part 1), I shall examine what I take to be the most important, most influential contemporary liberal theory—that due to John Rawls—from a feminist perspective. As I do this, benefiting from the work of leading feminists, I note also that criticisms analogous to those raised by feminists can be raised by a variety of other more or less disadvantaged groups (Part 2), and that since a concrete human being tends to belong to the intersection of various groups, no one who speaks as a member of one of these groups can quite succeed to speak for that concrete person (Part 3). The problem we face is how to preserve the universality of justice while paying heed to the multiplicity of voices of distress (Part 4). I try to find some answers to this problem, first within the framework of Rawls's ‘four stage sequence’ (Part 5) and then in a more general pragmatist approach (Part 6). Finally, I address Sen's concern that the scope of Rawls's conception of justice is too limited and Glover's worry that extending its scope to Third World countries may be mere Western arrogance. In the end (Part 7) I claim that we need a universalistic theory; it is implicit in my discussion that such a theory will be in some sense ‘Rawlsian’.

1 The Issues

Before I begin, and before I become bogged down in more or less specific details, I want to state the underlying philosophical issues in general terms. What I shall call the liberal conception of justice presents itself as universal; it is the Enlightenment successor to the conceptions of justice found in Judaism and its daughter religions, and these too present themselves as universal, as do socialist conceptions. Justice is universal in a double sense. On the one hand, justice grants rights to and imposes obligations on every (p.301) human being (with appropriate caveats for natural incapacities); on the other hand, it limits the claims that any human being may press in the name of justice. Disagreements about justice tend to be disagreements about these limits. For example, if justice demands adequate nutrition for all, then those in authority have an obligation to enable healthy adults to earn that nutrition for themselves and their children (as well as to provide it to those unable to work). This obligation would not be fully discharged if the government, as in Bangladesh, sets up food‐for‐work sites but does not allow women to work. On the other hand, if justice does not demand adequate nutrition for all, then upper caste Hindu widows are entirely at the mercy of their male relatives; they cannot demand as a matter of right to be allowed to seek gainful employment outside the homestead.

The notion of justice at issue in this paper is political justice, and in our world that means for the most part justice within national boundaries,3 it means, for the most part, that we as individuals are obliged to demand that our governments enact just policies and to support such policies when they are enacted. However, given that we live in societies which are far from just, many of us encounter opportunities to rectify to some small extent existing injustices or to refuse to benefit from them. When we act in this way it is justice not charity which we pursue, it is justice not charity which our beneficiaries receive. Their dignity as they demand justice is and remains intact, and justice requires that we acknowledge this. In contrast, one often, not always, begs for charity and receives it humbly or even with shame. However, justice is not confined within national boundaries, both nations and the citizens of different nations may deal with one another justly or unjustly. In particular, when natural disasters cause famines, food aid across national boundaries should not be regarded as charity but as what is owed to those stricken, by their more fortunate neighbours.

This view of the universality of justice has come under attack in recent decades. Because a conception of justice limits the claims that may be pressed, the following objection arises: how can one be bound by a conception of justice, in particular how can one's legitimate claims be limited by such a conception, if one has had no voice in formulating it? In the pursuit of universality, justice speaks with a single voice, but that voice, so the challenge maintains, is the voice of the dominant group in society. It seems inevitable that the resulting conception of justice will be biased, will fail to provide adequate grounds for the claims of the less favoured.4 This (p.302) challenge may be understood as a challenge to widen the scope of justice. Thus when women in India and Bangladesh attempt to break the bounds of occupational purdah, they seek to be included among those who have the right to seek gainful employment outside the home. Similarly, the struggle against apartheid in South Africa seeks to extend various rights and privileges of citizenship to previously excluded groups. Needless to say, I support wholeheartedly these and other attempts to widen the scope of justice. However, and this is what I intend to show, in the writings of some feminists, attempts to widen the scope of justice have opened a line of reasoning which threatens to fragment justice by emphasizing the diverse particularities of those to whom justice applies.

Because a conception of justice imposes obligations, the view that justice, and in particular a liberal conception of justice, is universal, has been charged with incoherence. How can the view be binding on members of another tradition, especially if that tradition does not recognize the fundamental moral equality of human beings, or if its conception of the person is different from that which is implicit in the liberal conception of justice? I take it that this objection, unlike the earlier one, seeks to narrow the scope of the justice; a particular conception of justice is said to cover not all human beings but only those who belong to the tradition in which that conception originates. Thus, this objection, too, if it could be sustained, would result in a fragmentation of justice.

Both attempts to widen and attempts to narrow the scope of justice threaten fragmentation. I shall concentrate for the most part on showing how the former threat develops. I shall then suggest how one can pay attention to the claims of particular groups within our society without sacrificing universality. In conclusion, more briefly, I shall examine relativist and communitarian attempts to narrow the scope of justice.

2 Rawls through Women's Eyes

I shall begin by showing how a feminist critique of a liberal conception of justice threatens to undermine the possibility of a universal conception of justice. Although my argument is limited to a liberal conception of justice and to its liberal critics, similar arguments could be presented not only for other interpretations of liberal justice and their critics but for socialist conceptions of justice and their socialist feminist critics as well.

We must begin where we are, that is, in a world in which vast numbers of people live at the edge of starvation and/or under dictatorships of various forms.

We must begin where we are, that is, in a world in which women suffer more than men from economic deprivation and in which women are (p.303) struggling to overcome the special handicaps that were traditionally imposed upon them.

We must begin where we are (or where, at any rate, this writer is), that is, in a society in which many people have become conscious of the injustices suffered by women, people of colour, the aged, children, the disabled, gays and lesbians, pregnant women, the ill, the poor; the list seems to grow each year. Of course, the categories overlap. Women, for example, are found in all of them, as are members of oppressed racial, ethnic or religious groups. The aged are often ill or disabled, and so on.

We must begin where we are, that is, two decades after the publication of John Rawls's A Theory of Justice (1971). One cannot, at least in the English‐speaking world, think about justice without taking one's position relative to that work. I take it that the Rawlsian conception of justice as fairness articulates the liberal conception of justice, and that the liberal conception of justice is the dominant conception in post‐Enlightenment industrial democracies.

We must begin where we are, that is, in the presence of a mature feminist movement.

We must begin where we are, that is, in a world in which women of colour and lower class women point out that the dominant feminist critique fails to speak to their situation, keeps them as invisible as they have always been.

I shall elaborate on these points by recalling, briefly, Rawls's theory of justice as fairness. As I go along in this exposition, I shall interject objections and concerns, many of these are feminist, all of them are particularist in the sense that they are prompted by injustices suffered by a particular group of people and do not seek to transcend that particular perspective.

I shall follow the interpretation which Rawls has given to his own work in his 1985 article ‘Justice as Fairness: Political not Metaphysical’. We must distinguish between the conception of justice, which is embodied in the two, by now, famous principles of justice, and the theory, which clarifies and justifies the principles and explains how they are to be applied to the basic social and political institutions of a society.

The two principles are formulated as follows:

  1. 1. Each person has an equal right to a fully adequate scheme of equal basic rights and liberties which scheme is compatible with a similar scheme for all.

  2. 2. Social and economic inequalities are to satisfy two conditions: first, they must be attached to offices and positions open to all under conditions of fair equality of opportunity; and second, they must be to the greatest benefit of the least advantaged members of society. (Rawls, 1985: 227)

Rawls has pointed out, and presumably does not mean to rescind this, that realization of the second part of the second principle must be consistent (p.304) with a ‘just savings principle’, a restriction which becomes increasingly relevant to practical political and economic decisions as we become more and more aware of ecological issues (Rawls, 1971: 292).

Justice as fairness is offered ‘as a political conception of justice for a democratic society, it tries to draw solely upon basic intuitive ideas that are imbedded in the political institutions of a constitutional democratic regime and the public traditions of their interpretation’ (Rawls, 1985: 225). The most basic of these ideas is that of ‘society as a fair system of cooperation between free and equal persons’ (ibid.: 231), where ‘a person is someone who can be a citizen, that is, a fully cooperating member of society over a complete life’ (ibid.: 233). While this does not rule out occasional illness or accidents, for which provisions must be made, Rawls adds rather startlingly ‘for our purposes here I leave aside permanent physical disabilities or mental disorders so severe as to prevent persons from being normal and fully cooperating members of society in the usual sense’ (ibid.: 234; my italics). This is both disturbing and puzzling. A person was defined as someone capable of being a fully co‐operating member of society; are we then to conclude that a permanently disabled human being is not a person, or at any rate not a citizen? If not, how fully co‐operating must one be before one will be regarded as a citizen? Surely one wants to remember here that the extent to which a disabled person can participate in political and economic life depends not only on the nature of the disability but on the provisions made by the society to enable handicapped people to participate in public life. If handicapped people are not citizens before such provisions are made, how will they make their voices heard? And if they are not regarded as persons, who will even speak for them? It will not do to answer that we have spokespersons even for non‐humans, e.g., for whales, for the claims made on behalf of whales are not justice claims while those made by and for handicapped people are. Of course, the text of A Theory of Justice, especially the interpretation of equal opportunity in the light of the difference principle, has always seemed to me to support the kind of efforts that have recently culminated in passage of the Americans with Disabilities Act (see Rawls, 1971: 17). Perhaps we are to read these discussions as a reminder that a just society attempts to enable all its citizens to participate in the ‘skillful and devoted exercise of social duties [lest they] be deprived of one of the main forms of human good’ (ibid.: 84).

Unfortunately, the second point in Rawls's more explicit account of what is meant by the claim that citizens of a democratic society regard themselves as free persons, increases one's disquiet. To say that citizens of a democratic society consider themselves as free is to say that they ‘regard themselves as self‐originating sources of valid claims’ (Rawls, 1985: 242). What does this rather puzzling statement mean? It seems to mean that people recognize moral rights and duties that are not rooted in the political conception of justice. In other words, our (various) moral conceptions in so (p.305) far as they are compatible with the requirements of political justice give rise to moral claims and obligations. Rawls contrasts this view of ourselves and each other as free and equal persons with that prevailing in a slave‐holding society. In such a society, ‘slaves are human beings who are not counted as sources of claims . . . they are not publicly recognized as persons at all. Thus, the contrast with a political conception which allows slavery makes clear why conceiving of citizens as free persons in virtue of their moral powers and their having a conception of the good, goes with a particular political conception of the person’ (Rawls, 1985: 243; emphasis mine). This will not do. The framers of the US Constitution would have agreed wholeheartedly with Rawls that citizens are free persons in the sense explained, but they would also have maintained that only white men of a certain social standing are citizens. All other human beings (men and women of African descent, white women, men without property), they would have claimed, fail to have the moral powers to the requisite degree or lack other capacities necessary to be fully co‐operating members of society. In other words, the egalitarian tendencies of justice as fairness are undermined by Rawls's use of the word ‘citizen’. Monarchies, to be sure, have no citizens, they have subjects. But in many historical societies, including our own in the not too distant past, a notion of citizenship limited to some human beings has prevailed. These citizens conceived of themselves and each other as free and equal persons, they may even have claimed to live under a democratic regime; that was scant comfort to those who were completely or partially excluded from the rights of citizenship.

Obviously, Rawls means to say something quite different. Justice as fairness is based on the assumption of universal citizenship in a modern democracy. Only when all women and all men of whatever colour or economic status are considered to be free and equal persons does it even make sense to say, as he does in A Theory of Justice, that inequalities based on sex, race or other fixed natural characteristics are justified only if they are to the advantage of the less favoured, and that this is in fact ‘seldom, if ever’ the case (Rawls, 1971: 99). How then can we explain this careless and potentially dangerous (to anyone who is not a white middle or upper class male) substitution of the word ‘citizen’ for the word ‘human being’? Of course, those who have always enjoyed citizen status in so‐called democratic societies might well fail to remember that for most of our history neither ‘human being’ nor ‘human being capable of speech and action’ have been co‐extensive with ‘citizen’.

Let us complete Rawls's account of the respects in which citizens regard themselves as free persons. ‘Citizens are free,’ we are told, ‘in that they conceive of themselves and of one another as having the moral power to have a conception of the good’ (Rawls, 1985: 240) and ‘as capable of taking responsibility for their ends’ (ibid.: 243). Here it is worth noting that one's conception of the good will include ‘a view of our relation to the world—religious, (p.306) philosophical, or moral—by reference to which the value and significance of our ends and attachments are understood’ (ibid.: 234). In a democratic society citizens may change their conception of the good in these as well as in other respects without loss of any basic political or economic rights. Now it is, of course, true that over the course of the last 300 years religious toleration over a wider and wider range of faiths, and the lack of any faith, have become a common feature of democratic societies. But we must remember that toleration has grown slowly, and that a certain amount of religious and moral intolerance and even outright oppression still plays a role in our public life. One welcomes, therefore, the fact that Rawls notes that meaningful freedom of conscience requires not merely permission to worship as one pleases, if one pleases, but absence of political or economic inequalities based solely on a difference in faith (and I would add: sexual preference). One wonders, however, why Rawls devotes several sections in A Theory of Justice to liberty of conscience and toleration, but only one paragraph to problems of sexism and racism. To be sure, in that paragraph he points out that ‘from the standpoint of persons similarly situated in an initial situation which is fair, the principles of explicit racist doctrines are not only unjust. They are irrational’ (Rawls, 1971: 149). It is clear that he would say the same about sexist doctrines. Unfortunately, the irrationality of racism and sexism is far from obvious to many persons, including many Americans. Why does he not devote more space to these urgent matters? Rawls might respond by saying that further arguments for racial and sexual political and socio‐economic equality are implicit in the general argument for the two principles. He may add that he chose to argue explicitly for freedom of conscience because (a) he wished to give one sustained argument under the rubric of the first principle, and (b) he chose this particular topic because freedom of conscience is built into the historical foundations of the liberal conception of justice, while opposition to the exclusion from political life of non‐white males and all women is a more recent development. But is that not all the more reason to develop the liberal argument against racism and sexism? Finally, Rawls might say that the argument against racism and sexism belongs to partial compliance theory, but so does his own argument concerning conscientious objection. To be sure, conscientious objection was a burning issue during the years just prior to the publication of A Theory of Justice, but so were a host of issues related to overcoming the vestiges of slavery and issues raised by feminists. Once again one cannot fail to notice that in a country which does not draft women, conscientious objection to service in the armed forces is a particularly male issue (see also Okin, 1989: 95–6).

A person's conception of the good includes various attachments and loyalties to persons and groups of persons, and justice requires that one may change one's conception of the good within the limits of justice without suffering political or economic sanctions. Unfortunately, divorce (surely a (p.307) change of attachment, hence a change in one's conception of the good) has often economically disastrous consequences for women and their children. Because women tend to earn less than men, because women tend to be the parent whose career is ‘put on hold’ while the children are young, and because of current divorce legislation, the economic situation of women tends to deteriorate drastically after divorce (Okin, 1989: ch. 7). Feminists note that Rawls and other major theories of justice fail to pay attention to this problem. Once again one cannot help being aware of the fact that one hears a male voice.

3 Substitutionalism

The aim of A Theory of Justice was to develop and justify a conception of political justice which would specify fair terms of social co‐operation for free and equal persons. To do so, Rawls introduced the idea of the ‘original position’, i.e., of ‘a point of view, removed from and not distorted by the particular features and circumstances of the all‐encompassing background framework, from which a fair agreement between free and equal persons can be reached’ (Rawls, 1985: 235). As is well known, the parties in the original position, whose task it is to agree on principles of justice, were to be behind a veil of ignorance, they were to be ignorant of all particulars about themselves and their society, though they were to have general knowledge. Rawls refers to the original position as ‘a device of representation’ (1985: 236 and elsewhere); it represents what we are to think about and what we are to leave out of consideration when we seek principles of justice to guide our political evaluation of the basic institutions of our society. Such disinterested reflection is at least one kind of political philosophizing. Thus, we may also think of the original position as the forum in which political philosophy is done, provided we do not limit access to professional philosophers.

Criticisms of Rawls's theory of justice are often formulated as criticisms of the original position.5 Feminist critics have noted that the original position is profoundly monological and that the one voice heard is that of a man. Women, they point out, are both literally and figuratively excluded from the original position. Women are figuratively excluded because the parties are described as heads of households in order to guarantee that each care for some members of the next generation. Although women certainly care for their children as much as do men, and although we have in our society a large number of female‐headed households, many of which belong (p.308) to the economically least advantaged group, the phrase ‘head of household’, these critics say, when it is not qualified, refers to a man.

What difference would it make if women were included in the original position? Okin writes: ‘If Rawls were to assume throughout the construction of his theory that all human beings are participants in what goes on behind the veil of ignorance, he would have no option but to require that the family, as a major social institution affecting the life chances of individuals, be constructed in accordance with the two principles of justice’ (Okin, 1989: 97). If Okin means here that the parties in the original position are to be ignorant of their sex, I am in wholehearted agreement with her. But I believe that Rawls never meant to say that the parties knew themselves to be men, i.e., he never meant to say that when we ponder questions of justice we should consider only questions of interest to men. Whether the principles of justice are to be applied to the family is, in any case, not decided in the original position itself. What counts among basic structures to which a conception of political justice is to be applied is part of the general knowledge that the parties retain even behind the veil of ignorance.6 Rawls envisages that the principles will be applied in a four‐stage sequence: the Original Position, a constitutional convention, a legislative assembly, and finally, the courts. If the principles will be applied to the family, this will happen at the second and perhaps only at the third step in the four stage sequence, i.e., during a constitutional convention and/or at the legislative stage. Finally, if as Okin's remark suggests, the principles of justice will be the same whether or not the parties in the original position know themselves to be men, or are ignorant of their sex, or include both men and women who know their sex, then I fail to understand the point of the demand that women be included in the original position.

I want to make quite clear what I am saying and what I am not saying. Of course, everyone should be included in the original position, and of course, justice should prevail in the family as well as in the larger society. But just what does Okin want when she wants the family to be constructed according to the principles of justice? How much of this constructing is to be done by legislation? Legislation, either by the state or by religious authorities with the ability to enforce their will, has always regulated marriage, remarriage, and divorce. Legislation has also regulated what happens within a marriage, sometimes to a greater and sometimes to a lesser degree, injustices may be due to such regulations as often as they are due to the lack thereof. It is precisely such regulations, when they prevent women from being equally valuable members of the family by restricting their activities to the homestead, that seem to explain the ‘missing women’ who are a major concern of this symposium. Thus, while it may be more or less clear to all of us what (p.309) justice in the family demands (or at least what it prohibits), I doubt that any single remedy will do the trick. When women speak out on these matters, their voices seem to be pointing in two directions, wanting marriage and the family to be both more and less private (more and less subject to legislation) than they have been. In fact, there is no ambivalence and no contradiction. Women object to the fact that their personal choices are limited by political (or religious) choices in which they did not participate, since legislators (or religious authorities) are, for the most part, male. Sometimes the remedy for this situation is to liberalize restrictive legislation, at other times it requires the introduction of protective legislation, and sometimes what is needed is not legislation but education. None of this is, I believe, contentious.

The question I wish to raise is whether women are to be represented as women, and hence men as men, or whether the original position is to be sex‐blind. It is, I hope, evident that my objections below are directed only against the first alternative. I fully agree with Okin when she says that the concept of the original position in which the parties are ignorant of their sex as well as of other particulars ‘is a powerful concept for challenging the gender structure’ (Okin, 1989: 109).

That women are literally excluded from the original position is seen when we recall that it may be taken to represent the forum in which we engage in political reflection, in particular in reflection concerning matters of justice. When we look at the development of liberalism, in particular of the conception of justice as fairness, and at the ongoing discussion in political philosophy, the continued adjusting of the reflective equilibrium, we notice that women are under‐represented in these discussions. There is, to be sure, a considerable feminist literature, but for the most part feminist voices fail to be heard and appropriated by male philosophers. Feminists have, therefore, demanded that women be included in the original position, or in the philosophical reflections which it represents. Once again, I cite Okin. ‘In a gender‐structured society there is such a thing as the distinct standpoint of women . . . The notion of the standpoint of women, while not without its own problems, suggests that a fully human moral or political theory can be developed only with the full participation of both sexes’ (Okin, 1989: 1067). Here it seems to be clear that women are to be represented as women and are to speak as women, for they are to be present in approximately equal numbers with men and in positions of comparable authority. Women so represented would insist that the principles of justice be applied to the family. For: ‘the theory as it stands contains an internal paradox. Because of his assumptions about gender, [Rawls] has not applied the principles of justice to the realm of human nurturance, a realm that is essential to the achievement and maintenance of justice’ (ibid.: 108). I am inclined to think that the charge of ‘internal paradox’ is far too strong since it depends on certain empirical developmental theories, in particular that of Chodorow (p.310) (ibid.: 106, 131–2 and n. 58). These theories suggest to Okin that men and women raised in gender‐structured families, hence experiencing different psychological and moral development, cannot agree on principles of justice even when they are ‘behind the veil of ignorance’. The very fact that Okin and I are inclined to accept the Rawlsian principles, or something very much like them, suggests to me that empirical theories cannot be extrapolated to highly abstract theoretical settings. This is not meant to be understood as either a criticism or an endorsement of developmental psychology, feminist or otherwise; I am unqualified to deliver either.

Obviously, women should not be silenced or ignored in our actual philosophical reflections, nor should they be excluded from equal participation in our actual political lives when the principles of justice are translated into effective legislation. However, the demand that women be included in the hypothetical original position as women (as opposed to being included as persons‐who‐do‐not‐know‐their‐sex) seems to me misleading, even misguided. The demand that the parties in the original position know their sex and that both sexes be equally represented, leads, it seems to me, directly to the demand that the parties know their race, religion, sexual preference, state of health, etc. Translated into actual philosophizing and practical politics, this would mean that there would be only persons pleading for ‘special interests’ but no one attempting to take a more or less impersonal stance; it would mean only particularist moral and political philosophy but no universalistic views, only lobbyists but no legislators, only prosecutors and defence attorneys but no judges. In short, a small tear in the veil of ignorance would lead, I believe, to it being rent altogether. What follows is my argument for this apprehension.

The single voice which is heard in the original position is not the voice of MAN. It is the voice of a white upper or middle class adult male who is heterosexual and, if he is not a practising Christian, he comes at least from a Christian background; he is also in perfect health and not old.7 Thus, if women are to be included in the original position, lower class men and men who are not white need also be included, as well as Jews and Pagans, gay men, old men and disabled men, and no doubt others who are so invisible that we have not yet begun to think about them. Why is it that some feminists, though not all, have failed to notice how many men are not heard? I believe that the main reason is that these feminists happen to be women for whom sexism is the dominant, often the only, form of oppression that they experience, i.e, they are white professional women. For these women, it is easy, consciously or unconsciously, to fail to notice and even to perpetuate the implicit class, race, etc. bias of the dominant theories of justice. They recognize that human experience has been identified with male experience, they do not recognize that a particular kind of male (p.311) experience, that of upper or middle class males of European descent has been substituted for male experience. As a result, when they insist on the representation of women, they fail to ask which women are to be represented. In fact, they substitute the experience of white middle and upper class women (mainly professional women) for that of all women.8

I do not here refer to the fact that not all women are feminists, nor to differences among feminists concerning the just society; these matters are acknowledged. I refer to the fact that some feminists substitute their own point of view for that of all women.9 Seyla Benhabib has suggested that we must replace the ‘substitutionalist universalism’ that characterizes modern universalistic political theories by an ‘interactive universalism’. Instead of substituting the experiences of ‘white male adults who are propertied or professional’ for the experiences of all human beings, interactive universalism ‘acknowledges the plurality of modes of being human . . . [and] aims at developing moral attitudes and encouraging political transformations that can yield a point of view acceptable to all’ (Benhabib, 1987: 81). I am suggesting that some feminist theories of justice have offered us a ‘substitutionalist feminism’ that needs to be replaced by an ‘interactive feminism’.

How true is the charge of substitutionalism? Let us consider first, briefly, the charge of substitutionalism as raised against Rawls. I believe enough has been said to substantiate the claim that we hear the voice of an able‐bodied man. Perhaps being gay or lesbian falls under one's conception of the good; still just as Okin insists that more needs to be said about the difference that divorce makes in the lives of men and women, so more needs to be said about the difference that sexual preference makes in the lives of straight and gay people. Just as failure to do the former constitutes substitutionalism, so does the latter. Those of us who are at or near retirement age are painfully aware of the fact that the parties in the Rawlsian original position will return to a life of gainful employment; they are not old and we do not know whether they gave any thought to the aged. (The just savings principle does not prescribe saving for this generation's old age, it saves for the benefit of future generations.) Finally, I suggested that we hear the voice of someone who comes from a Christian background; that is not quite accurate. We hear the voice of someone for whom tolerance, not only of religion properly so called but of a quite wide range of conceptions of a life worth living, is of the utmost importance. However, for many religious persons, including (p.312) the Muslim women in Bangladesh and the Hindu widows in India, the problem of main concern is not whether the government will tolerate their religious practices but whether their religious practices will allow them to survive at all. As an anonymous wit has said with respect to a different country, the problem is not freedom of religion but freedom from religion. For the overwhelming majority of Christian sects, though not for all, that is simply not a problem, and it is not a problem recognized by Rawls.10

For the charge of ‘substitutionalist feminism’ I offer these pieces of evidence. First, in the introduction to Feminism as Critique, Benhabib and Cornell write: ‘Third world women have challenged precisely the assumption that there is a generalizable, identifiable and collectively shared experience of womanhood. To be Black and to be a woman, is to be a Black woman, a woman whose identity is constituted differently from that of white women’, a remark which suggests that they too are aware of the dangers of substitutionalist feminism. Unfortunately, this remark introduces the question, ‘How can feminist theory base itself upon the uniqueness of the female experience without reifying thereby one single definition of femaleness as the paradigmatic one—without succumbing, that is, to an essentialist discourse on gender?’ (Benhabib and Cornell, 1987: 13) None of the authors who contribute to their book ever return to the issue raised by Third World feminists, a debate internal to white feminism has been substituted for that issue.

Secondly, having mentioned that ‘feminists have been criticized for developing theories of gender that do not take sufficient account of differences among women, especially race, class, religion, and ethnicity,’ Okin writes: ‘While such critiques should always inform our research and improve our arguments, it would be a mistake to allow them to detract our attention from gender itself as a factor of significance. Many injustices are experienced by women as women, whatever the differences among them and whatever other injustices they also suffer from’ (Okin, 1989: 6–7).11 Of course, there are injustices that are suffered by women as women: women are raped, women are battered, women are abused, medical research has focused on men to the detriment of women, etc. But when medical research (p.313) does focus on women, it substitutes white women. ‘Currently, only reproductive research involves women subjects in large numbers. However, the majority, if not all, of the participants are white women . . . it results in the development of products with little to no information on how they affect obese women and women with high blood pressure, how they interact with hypertension medications, or impact other cardiovascular problems which disproportionately affect women of colour, particularly African American women.’ (Sojourner, 1991)

Another example. For Okin, rape is essentially an act of violence by a man against a woman, it becomes an issue of political injustice only when the courts refuse to recognize and prosecute marital rape.12 In contrast, for bell hooks in Ain't I a Woman: black women and feminism, rape of a black woman by a white man is an act of racism as well as sexism, indeed it is a quintessential racist act, a devaluation not simply of a woman but of black womanhood. Of course, it is unfair to compare Okin's book and hooks's on this issue, since their aims are quite different. Okin's book is devoted to issues of justice, gender, and the family, while hooks describes her book as ‘an examination of the impact of sexism on the black woman during slavery, the devaluation of black womanhood, black male sexism, racism within the recent feminist movement, and the black woman's involvement with feminism’ (hooks, 1981: 13). Hooks's point is that we will not succeed in eradicating sexism unless we eradicate racism and classism at the same time. In contrast, Okin concentrates primarily on the ways in which the gender structure of the family, and the parallel gender structure of the employment market, make women and their children economically vulnerable, though she recognizes that as a result of being economically vulnerable, married or partnered women are also more or less defenceless against physical and psychological abuse. Nevertheless, I believe that when Okin discusses the vulnerability of women and their children, she focuses primarily on the situation of white women, and indeed of privileged white women.13

I do not mean to suggest that Okin ignores the existence of non‐white, in particular of black, women. She notices that, unlike most white women, black women have always worked, that the educational and career choices women make in the expectation of getting married are particularly disadvantageous to women who do not get married, in particular ‘poor urban black women, whose actual chances of marrying and being economically (p.314) supported by a man are small’ (Okin, 1989: 142), that the divorce rate is much higher for black women than for whites, and that court‐ordered child support is, on the average, highest for white women and least for black. But the bulk of Okin's discussion deals with women who, during their married or at least their child‐rearing years, do not work (hence not poor black women) and with the impact of divorce on such women, with the special problems of married women raising children while attempting to have professional careers (hence not poor women), with equal parenting (which assumes the presence of the father in the home or at least his living in near proximity) and with the need for quality day‐care. Only the last of these seems to me to be of equal importance to women across the socio‐economic spectrum.14 However, the need for day‐care seems to be peculiar to urban societies, or perhaps only to urban societies in which the extended family has disintegrated.

Okin is surely correct in believing that for many American women the chief obstacles to economic security lie in the gender structure of the market place and the family, and that these are deeply inter‐twined. In contrast for the poor women of Bangladesh and upper caste Hindu widows studied by Chen, the chief barrier is the occupational purdah imposed by their respective religions and enforced by both religious and secular authorities.

When one speaks of substitutionalism, one does not charge bad faith. No doubt Rawls is sincere in believing that his argument for the principles of justice is an argument that anyone behind the veil of ignorance would be able to accept, or to put it another way, the argument is addressed to and believed to be rationally acceptable by any reader. And in spite of the difficulties already raised and the modifications to be suggested, I find the conception of justice as fairness quite compelling; I would not devote so much attention to it were it otherwise. The medical researchers who use only white women in their research do not think that non‐white women do not matter; they fail to ask the question whether there are medically significant differences between white and non‐white women. Feminists whose arguments revolve primarily about the experiences and needs of women like themselves but speak about women as women, do not deny the existence of women who are more or less different from them; they fail to appreciate the magnitude and significance of the difference. I suspect that none of us can avoid substitutionalism, all of us implicitly substitute our own perspective for a more general one. Indeed, in pointing out the dangers of (p.315) substitutionalism, I speak from a perspective which magnifies the differences between women; feminists might retort that I substitute my own experiences for those of women for whom the similarities dominate. Nevertheless, recognizing the possible universality of substitutionalism, I plead nevertheless that we should attempt to overcome it.

Let us suppose then that, in the spirit of Benhabib, we try to replace substitutionalist feminism by an interactive feminism, that we listen to the voices of women of colour and women of a different class, and that we appropriate what we hear. What will we have learned? We will have learned that for women who have always had to earn a living (in the fields, in factories, in stores and offices, doing piece‐work at home, or working as domestics in other women's homes), the demand that women be allowed to leave the domestic sphere is not a demand for liberation. On the other hand, we will have learned that for women who are by religious law confined to the home, being able to go out and earn a living is literally a matter of physical survival. We will have learned that the phrase ‘women and minorities’ denies the existence of minority women since everyone understands it as ‘white women and minority men’. We will have learned to care more about women's medical needs and about safety on the streets and less about chances for tenure or making partner in a law firm. We will have learned that women worry about the safety of their sons as well as their daughters. We will come to understand that it is our responsibility as feminists, as white professional women, to learn from all these other women.

4 The Dilemma

Elizabeth Spelman, who suggests learning from women of colour and poor women as a strategy for overcoming feminist racism and classism, writes in her book Inessential Woman: ‘Modern feminism is faced with a dilemma: will throwing out the bath water of white middle‐class privilege involve throwing out the baby of feminism? . . . If we can't isolate gender from race or class, if we can't talk about the oppression women face as women, or about the experience of women as women, isn't feminism left without a foundation, without a specific focus?’ (Spelman, 1988: 171–2).

I should like to rephrase these questions. Modern theories of justice, she should have written, are faced with a dilemma: will throwing out the bath water of white, middle‐class male privilege throw out the baby of universalism? If they cannot appeal to a conception of justice which is based on the idea that all human beings capable of speech and action are equal from the political point of view, that what is justice for one is justice for each, and that nothing is justice unless each can come to see that it is just, then those who are excluded, exploited, or otherwise oppressed have no (p.316) Archimedean point on which to stand and from which to move the world. I am not so politically naïve as to think that mere moral appeals, even appeals to political morality, will carry the day, but I am also not so politically cynical as to think that appeals to justice carry no weight at all. We need, therefore, an adequate theory of justice.

The principles of an adequate theory of justice are principles to which the disadvantaged can refer when they demand that justice be done; they are principles which will enable us to settle divisive disputes because they are acknowledged by all parties to the dispute; they are, in short, principles to which we can all agree. The argument just reviewed began with feminists demanding that women be included in the original position, that is, that the monological reflection which leads to Rawls's principles should be replaced by a dialogue in which women are able to remind men of their particular problems. If women are not included in the reflection which leads to the adoption of principles of justice, it was suggested, the resulting principles will not enable women to press their claims, hence women will not be able to agree to these principles. But, so the argument continued, there are many kinds of women and many kinds of men whose concerns will not be expressed by white, middle‐class men and women. Must we then introduce the voices of men and women who are poor, or persons of colour, or disabled, or not Christians, or old, or ill, or homosexuals? Must they all participate in the conversation which is to lead to the formulation of a conception of justice?

Rawls has introduced the notion of an overlapping consensus in order to explain how an agreement concerning justice may be reached in a liberal society in which persons embrace widely divergent comprehensive moral/ religious doctrines (Rawls, 1987). That appeal to an overlapping consensus fails to respond, however, to the concerns of this paper. It is one thing to find that as part of one's comprehensive conception of the good one subscribes to a conception of political justice that one holds in common with other members of one's society; it is quite another thing to be confronted with a conception of justice that determines which of one's claims are legitimate, when one of one's claim is that one has been ‘unjustly’ excluded from the deliberations that led to the adoption of this conception. Thus I ask, yet again, whether the original position must include representatives of all disadvantaged groups. An affirmative answer would entail a radical reconception of the original position: the veil of ignorance would be rent. The representatives of the disadvantaged groups would know to which group they belong, though they would still be ignorant of the details of their individual lives, they would be, to use Benhabib's phrase, neither completely generalized nor fully concrete others (Benhabib, 1987).

It seems to me clear that this new kind of ‘original position’ will not do for three reasons. First, we shall continue to be haunted by the spectre of (p.317) substitutionalism. Secondly, we shall be unable to come to an agreement on any conception of justice. Thirdly, it is impossible to perform the suggested thought experiment. Consider the first point. The argument just reviewed made us aware of the fact that when persons speak as representatives of a certain group, e.g., ‘women as women’, they may in fact speak only for a subset of that group, e.g., white, middle‐class women. There is then no reason to think that the conception of justice to which all these ‘representatives’ could agree would not be again a ‘substitutionalist’ conception of justice. This concern may also be expressed by pointing out that there would be no guarantee that all groups that deserve to be heard would be heard. Perhaps orthodox Jewish women would be ‘represented’ by Jewish men speaking ‘as Jews’, or by women speaking ‘as women’. No doubt, there are groups whom no one as yet even misrepresents; their situation is comparable to that of homosexuals as recently as twenty years ago, that is, their voices would be inaudible. In fact, every human being belongs to more than one ‘group’. One is not merely a philosopher, one is also male or female, white or a person of colour, married or single, childless or with children, hetero‐ or homosexual, aged or young, physically challenged or not, belonging to some faith community or other or to none, and one may or may not subscribe to unpopular political ideals. These features and yet others make one the concrete unique individual that one is, and every theory of justice, whether it claims to be universal or admits frankly that it is particularist, will substitute the experience of some more or less generalized other, for one's own concreteness. Should we then, each of us, in the name of universal moral equality, construct our very own conception of justice? Surely, this would defeat the point of a conception of justice. Unless a conception of justice is shared, or has at least the potential of being shared, by both the advantaged and the disadvantaged, the oppressor and the oppressed, it does not provide a basis from which the latter can with dignity press their claims against the former. A conception of justice tailormade for some individual's particular circumstances cannot be such a shared conception. Yet where would one stop once one has stepped on the slippery slope and suggested looking for a feminist theory of justice and then notices that a feminist theory would be open to objections just like the objections that lead some feminists to demand that they be included in the original position?

What has just been said must not be misunderstood as a general condemnation of groups, or of persons who speak as representatives of certain groups. It is only at the first stage in Rawls's four‐stage sequence, that is, when one searches for a conception of justice, that permitting some groups to be explicitly represented will open the door to total fragmentation. The lesson to be learned from the above discussion is that everyone (not merely ‘heads of households’) needs to be present in the original position, i.e., all of us need to reflect upon and participate in the conversation concerning (p.318) principles of justice, but not only must we make an effort to discount what concerns each of us as a concrete individual, we must also refrain from acting as spokespersons for some group to which we belong; the veil of ignorance must remain whole.

I now turn to the second point. Here we ignore the worry about substitutionalism. Everyone participates in the reflections and conversations which are to lead to the adoption of principles of justice that are to enable us to settle our most divisive disputes. Once again, one would be expected to ignore or discount one's particular situation, but know to which groups one belongs and consider the kinds of claims one would wish to be able to press as a member of one of these groups. For example, as an American working woman one would wish to tailor the principles of justice so that they would permit demands for equal pay for equal work, but as a Bangladeshi woman working at a food‐for‐work site, one would wish to tailor the principles so that they would permit enough food for oneself and one's family for a day's work. As an American Jew, I would seek principles which would permit me to obey the rules of my religion; as a Hindu widow, I would seek principles which would enable me to transgress the rules of my religion. In each case it would be the government that would be expected to act as guarantor. As an American man, on the other hand, I might worry that equal pay for equal work would lead to a lowering of my income; as a Hindu man, I might worry that allowing Hindu women to work outside the home would undermine my authority. In other words, the conflicts which we hope to resolve by an appeal to mutually agreed‐upon principles of justice would be brought into the original position. Knowing to which group (or groups) one belongs, one would be able to estimate how the adoption of one set of principles or another would affect one's life prospects. I doubt that under such conditions one would be able to come to an agreement on anything except the most general and vacuous principles (see also Rawls, 1991: 141). Fragmentation may be avoided but only at the cost of having a conception of justice that is hardly better than no conception at all.

Finally, this is my third point, we must remember that the original position is a hypothetical position, a device of representation. In so far as it represents thinking of which every one of us is capable, it is obvious that no one can think, simultaneously, or even in succession, as a woman and as a man, as Jew and as Muslim, as one who has AIDS and as one who does not, etc. etc. What one can do, according to Rawls, is to attempt to ignore any considerations that depend on one's being a woman or being a man, being a Jew or being a Muslim, etc.

I have also suggested that we may think of the original position as representing philosophical discussions. Different persons are then quite able to think and articulate the views from different perspectives. However, my first point suggests that in so far as they do that, none of these persons (p.319) can claim to offer a universal conception of justice, nor can they be sure to speak for all members of the group they claim to represent; my second point suggests that any philosophical agreement achieved by persons who speak and think only from limited perspectives would be vacuous.

Once again I need to guard against being misunderstood. I do not claim that there is no place for group representation; in spite of the dangers of substitutionalism, the realities of political life are such that only organized groups of the disadvantaged can hope to be successful in the struggle for justice. Again, in spite of the dangers of substitutionalism, the complexities of philosophical, political, and moral discourse are such that we need to hear frankly particularist voices lest we ignore what William James called ‘the cries of the wounded’. Indeed, I want to insist on heeding what we have learned from persons in other countries, from feminists, from persons of colour, from members of various other disadvantaged groups. Our dilemma is precisely how to preserve the universality of justice while giving due considerations to the variety of voices of distress.

5 Some Answers

What we learn from the cries of the wounded modifies what Rawls calls our ‘considered judgments’, and since they stand in reflective equilibrium with our conception of justice, it may be the case that our conception of justice needs to be modified as well. Thus I do not claim that principles of justice once adopted will stand in perpetuity. I do not know whether this represents a serious disagreement with Rawls. Rawls insists that when one reflects on principles of justice, one must not only discount one's particular circumstances but one must also assume that the principles one adopts are adopted once and for all (Rawls, 1971: 147, 176). He wants to rule out the possibility of saying to oneself: ‘I'll try to live by these principles, but if it is either too hard or too disadvantageous, I'll reconsider.’ Someone with that attitude betrays a lack of seriousness, a lack of a sense of justice; of course, I quite agree with that. My point is, rather, that what we learn from the complaints of the victims of injustice may be that our conception of justice itself needs to be modified because it fails to acknowledge injustices which we nevertheless recognize to be such. However, it may also be the case that what we learn is that our realization of our conception of justice in legislation and practice has been woefully incomplete, i.e., that our society falls very short indeed of being even ‘near just’. I am inclined to think that both are the case.

Consider the first alternative. The principles of justice themselves need revision; they fail to provide a basis from which certain disadvantaged groups may press claims that we recognize to be just. To use an example of Sen's: ‘a person may have more income and more nutritional intake, but (p.320) less freedom to lead a well‐nourished existence because of a higher metabolic rate, greater vulnerability to parasitic diseases, larger body size, or pregnancy’ (Sen, 1989: 25). If Sen is right, Rawls's theory fails to accommodate the claims of say, pregnant women. Yet the existence in the USA of the WIC programme (providing food supplements for pregnant and nursing women) shows that we already recognize these claims to be just. On reentering the original position, we would retain the knowledge that the principles must leave room for these kinds of claim, but we would once again not know (i.e., discount) who we are, to what group we belong. This would make it possible, once again, to reach agreement on a substantive conception of justice. I believe that this represents the sort of reflecting and revising one would in fact attempt if one were dissatisfied with the liberal conception of justice or its Rawlsian articulation, or indeed some other, e.g., a socialist conception, because it fails to be in reflective equilibrium with one's considered judgements. The revised conception would again claim to be universal, to both legitimize and limit the kinds of claims persons can press in the name of justice. One example of such a ‘revision’ is Rawls's own reformulation of his first principle in light of an objection offered by H. L. A. Hart (1973).15 Another example is Sen's shift, in response to the problem mentioned above, from an index of socially primary goods to an index of capabilities. This move responds to the concerns of a variety of ‘less favored’ groups, including pregnant women, and the women at the food‐for‐work sites in Bangladesh. Unlike the search for particularist theories of justice, Sen's proposal retains the universality of principles of justice while taking due account of the diversity of human beings. It is, as he says, a shift in the information available to the parties in the original position, i.e., a shift in what we are to think about when attempting to articulate our conception of justice (Sen, 1989).16 As far as I know, Sen has never spelled out how the principles of justice would be reformulated if one substituted capabilities for primary social goods. It is not clear to me, for example, whether the Rawlsian distinction between political and civil liberties on the one hand and opportunities, income, and wealth on the other would be entirely obliterated, or to what extent something like it might be retained. In particular, considering the plight of women whose religion condemns them to such economic inferiority that their very lives are at stake, one would like to know how a capabilities approach might combine the ability to practise one's religion with the ability to receive adequate medical care and nutrition, so that, for examples, the ability of mullahs and village elders to limit the mobility of women (p.321) would be curtailed in order to increase the ability of women (and female children) to survive.

Whether or not the principles of justice need modification or have been modified, the construction of a conception of justice is not an idle exercise, rather it is to enable us to overcome deep and pervasive conflicts within democratic culture. To do that we apply the principles of justice first to the basic institutions of our society, then in the confines of these institutions to legislation, and ultimately to individual decisions of judges and administrators. We do so either from the point of view of equal citizenship or from the point of view of a representative of a least advantaged group; here what we have learned from listening to a variety of voices is of the utmost importance. Rawls assumes that the least advantaged group is economically disadvantaged, anyone whose average income and wealth is equal to or less than that of an unskilled labourer, or anyone whose income and wealth is less than half of the median, belongs to it (Rawls, 1971: 98). To be sure, Rawls is aware of the fact that one may be disadvantaged in other ways. As noted above, he recognizes that inequalities may be based on such fixed natural characteristics as one's sex, race or ethnicity. If so, these inequalities determine least favoured positions and in a just society must be, and must be able to be shown to be, to the advantage of those least favoured. But, Rawls continues, ‘these inequalities are seldom, if ever, to the advantage of the less favored, and therefore in a just society the smaller number of relevant positions should ordinarily suffice’ (ibid.: 99). This will not do. Feminists and civil rights advocates deserve credit for pointing this out. Ours is not a just society, and there are deep political disagreements concerning both the nature of the injustices and the permissible remedies. By making explicit what he takes to be our implicit convictions about justice, Rawls hopes to provide a basis for resolving these differences at least sufficiently to continue to function as a community. But then we must ask not merely who will remain least favoured in a just society but who is least favoured now. Indeed, even when we take up the point of view of the economically least favoured, it behoves us to ask why they are in that position. It is one thing to be an unskilled labourer because one lacked, for whatever tragic personal reasons, the drive, the opportunity, or the ability to become anything else, it is another to be an unskilled labourer because the colour of one's skin barred one from acquiring skills. In the former case justice requires that the economic distance between the labourer and those in a better position be to the former's advantage, in the latter case, in addition and more importantly, racist barriers to training and education must be removed. It is one thing to be a widow who finds herself at a disadvantage in the labour market when she is forced to return to it after a lengthy voluntary absence, it is another to be a widow who is prevented by caste rules from entering the labour market. The kind of changes that Okin advocates to protect women of the former kind do not (p.322) require changes in the religious convictions of anyone in the United States, the kind of changes required to rescue Metha Bai do. It is one thing to know that one's life is made difficult by institutions that one recognizes to be unjust, it is another to be required to break rules that one still accepts as proper. Of course, I am not saying that Rawls would fail to see that Metha Bai suffers an injustice that is quite distinct from that of the American widow: Metha Bai's freedom of movement is severely restricted. Even if Metha Bai were a rich widow she would still be a victim of this injustice. But that is to say that the position of Hindu women is a position of disadvantage, a position from which the institutions and laws of India need to be scrutinized.

Feminists point out that the point of view of women (all women) is the perspective of a least favoured group; hence, they argue, we must look at the basic structures from this perspective. Such scrutiny reveals unequal citizenship as well as unequal economic status and unequal bases of self‐respect; none of these inequalities are to the advantage of women. Not only are women under‐represented in our legislatures and the judiciary; even in situations of direct participatory democracy (town meetings, for example) studies have shown that the most authoritative participants turn out to be white, middle‐class men (Young, 1989). We must note, however, that these privileged men silence, or out‐talk and out‐vote, not only women but people of colour, poor people, and the elderly of either sex. Political impotence, in turn, undermines self‐respect, and has effects on the distribution of public moneys, thus leading to unequal access to public services. While married women whose husbands belong to the politically influential group are protected against the last mentioned effect of unequal citizenship, it is in general true that a vicious cycle keeps the poor powerless and the powerless poor.

Most women are economically disadvantaged because of complementary features of the family and the labour market. Because in the traditional family women carry the major burden of home‐making and child rearing, they cannot compete in the labour market on equal terms with men who can devote themselves wholeheartedly to their work. This inequality is exacerbated when we compare single mothers with married men. Because women are shunted into less prestigious and lower paid jobs, their careers are more easily sacrificed to the needs of their children, than are those of the fathers. Although Rawls (1971) mentioned the family as one of the basic institutions of society and described how children will acquire a sense of justice through growing up in a just family, and although he recognized that the stability of the well‐ordered society depends on its citizens having a sense of justice, he never discussed the applications of the two principles of justice to the family and does not confront the implications on the worlds of work and politics of a restructuring of the family (Okin, 1989: ch. 5).

(p.323) If the principles of justice are applied to the family, one is forced to conclude that men and women should share equally in providing for the family's economic, physical, and emotional needs. This requires that not only in the family but also in the worlds of work and education and on every level, the distinction between women's jobs and men's jobs will have to be eradicated, on every level the issue of equal pay for equal worth will have to be raised. Everywhere and in every way barriers to women's flourishing must come down. Both women's economic well‐being, and that of their children, and women's self‐respect are at stake (ibid.: 1989, chs. 7 and 8). I shall not pursue this issue further, the arguments are familiar to all of us.

Most of what has just been said concerning women is easily repeated if we replace ‘men’ and ‘women’ everywhere by racial or ethnic phrases. We all know that the job of railroad porter was a ‘black job’, we all know that the job of grocer in certain parts of New York is now ‘Korean’ and used to be ‘Jewish’. Medical schools were not only bastions of male privilege, they were and are bastions of white privilege as well. School administrators are predominantly male, they are also predominantly white. And in all these cases also the self‐respect and the economic well‐being of the less favoured and their children is diminished. Precisely because it seems that the veil of ignorance must be retained lest we lack any conception of justice at all, it is important that a large variety of relevant ‘least advantaged’ positions be recognized.17 From all these positions arguments analogous to those by Okin and other feminists can be offered, from all these positions claims need to be pressed—pressed as claims of justice not as appeals to charity. It seems to me that the main result of an ‘interactive universalism’ would consist precisely in enhancing one's ability to recognize a multiplicity of ‘least advantaged’ positions and an ability to hear and heed the complaints raised from these perspectives.

6 A Second Look

For the last several pages I have assumed that a universal conception of justice is at hand to which all the less favoured groups may appeal. Doing that, I have ignored the earlier worry that the universal conception of justice is the result of reflections which substitute the experiences of a privileged group of human beings for the experiences of all of us, that it may, (p.324) therefore, ‘in effect burden or exclude anyone who does not share the characteristics of privileged, white, Christian, able‐bodied, heterosexual, adult men’ (Minow and Spelman, 1990: 1601) by whom and for whom the conception was developed. Moreover, in speaking of group perspectives, especially of the perspective of women, I have ignored the burden of the second part of my argument, namely that those perspectives may substitute the experiences of more privileged (less oppressed) members of the group for the experience of all its members. The loud trumpet of universalism and the lesser trumpets of various communalisms threaten to drown out the cries of oppressed individuals; yet these individuals need to be able to appeal to the universalism of justice, and they need to be able to join their feeble voices into one mighty cry. With these observations, I have returned to the problem raised in the beginning; however, I shall now formulate it somewhat differently. The principles of justice, whatever they may be, must be universal if they are to reflect the liberal conception of justice, for the fundamental intuition behind that conception is the moral equality of all human beings capable of speech and action. The principles must be acceptable to all as a basis from which they may press their claims and as grounds of the obligations which they assume, otherwise they fail to provide a resource for settling disputes concerning the proper allocation of burdens and benefits. The task of articulating such principles appears impossible; whatever we do, so it seems, we shall either offer principles so vague and so general that they are useless, or we shall come up with principles that ride roughshod over the legitimate interests of some. It is time to make a new beginning.

I do not consider principles of justice to be eternal truths, nor, as already indicated, do I consider it helpful to regard a conception of justice, once adopted, to be the last word. Instead, I take principles of justice to be hypotheses concerning what social arrangements will best encourage human flourishing. As hypotheses based on rather extensive human experience, we may think of these principles as analogous to laws of science, provided we take laws of science themselves as hypotheses that aid us in our understanding and control of nature. Thinking of science in this light we recall that the formula which we now know as the ideal gas law was some centuries ago an empirical generalization based on relatively crude observations of actual gases. As better instruments were constructed, better data were obtained, and more accurate formulae were developed containing additional variables that refer to the properties of particular gases. In the nineteenth century the theory of gases as swarms of molecules was developed, and in that theory Boyle's original formula reappears as the ideal gas law. We need to remember also that the better instruments which enabled scientists to discover both the more accurate empirical formulae and the molecular theory were developed with the aid of theories based on data obtained with the older, cruder instruments. In every sense, science pulls (p.325) itself up by its own bootstraps.18 At every stage of scientific progress there are laws which are taken to be universal and which are applied to particular circumstances. We do not ask how we can retain the universality of the laws while paying due attention to the particular circumstances in which we apply them. Knowing that a law was discovered in one sort of setting may cause us to be cautious when we apply it to another sort of setting, but one of the aims of scientific investigation is to discover laws that are universal, that can be applied in all circumstances to which that sort of law is relevant. Learning to be a chemist is learning both what the laws of chemistry are and how to apply them; the particular problem at hand will determine what to ignore and what to heed in a given situation and which laws are relevant to it.

Using this account of science as a model, let us consider again the problem of the fundamental principles of justice. Boyle's law of gases was based on relatively crude observations, the earliest versions of the liberal conception of justice (say the version implicit in the pre‐Jacksonian understanding of the US Constitution as it was at that time) was clearly a conception by and for middle and upper class, white males. Yet that conception turned out to be broad enough so that those who subscribed to it were able to hear the demands of poor white men for the right to vote and hold public office. Once Jacksonian democracy was firmly established, the conception of justice, or at any rate the way in which it was interpreted had changed. Using the Rawlsian notion of social primary goods, we might say that participation in the political process was added under the rubric of liberty. Thus when slavery had been abolished, the enlarged conception of justice enabled the newly freed African–American men to demand the right to vote and hold public office. Women, too, began to demand all the rights and privileges of citizenship.19 In fact, the conception of justice implicit in the institutions and practices of liberal democracies have developed over time, and will, presumably, continue to develop. Just as laws of science that hold only approximately because they are based on crude data enable us to discover laws that hold more precisely, so a conception of justice that is partial because based on substitution (crude data) enables us to advance to a more comprehensive conception. Like science, democracy must pull itself up by its own bootstraps.

Given, then, an admittedly imperfect universal conception of justice, how will it be applied to a particular social problem? Having emphasized the (p.326) diversity of human beings, one wonders whether one will be overwhelmed by details. It is not sufficient to have a universal conception of justice, it is also necessary to deal with problems of justice politically, that is, not as individual problems but as problems of a certain kind. Just as scientists learn which circumstances to heed and which to ignore, so we must learn to know when it is appropriate to think of, say, women as women, and when it is necessary to think specifically of Japanese Buddhist women, when it is important to think of the homeless as homeless and when it is necessary to think specifically of homeless persons with AIDS, etc. Political genius consists at least in part in recognizing possible alliances where none were suspected, just as scientific genius consists in recognizing a common factor in apparently disparate phenomena. Each social problem is, for some of us, our problem. Each problem is also, at least at first sight, for most of us not our problem. The very first task any of us face when we are the victims of injustice is to see that we are victims of injustice and not simply of the arbitrary cruelty of some individual, i.e., we must see that ours is a political problem, not merely a personal one. Our next task is to make others see that there is a problem. The technique called ‘blaming the victim’ is used to prevent us from taking that very first step, and when it fails at that, it is used to prevent others from acknowledging that our problem calls for a political solution.

In science we need to identify the laws or theory relevant to solving the problem at hand; we work with the best theory that we already have, though in the course of solving our problem we may also change the theory. Likewise, in dealing with a political problem we work with the best theory that we already have, and in the course of applying the theory to concrete problems, we learn how it must be modified. I have discussed this above; now I want to emphasize that the origin of a scientific theory does not matter, what matters is how well it fares in experimentation and application, whether experience satisfies or frustrates the expectations to which the theory gave rise. The origin of the principles of justice in a tradition dominated by white Christian bourgeois men does not matter, what matters is how well it enables the rest of us to press our claims successfully, without either stepping on each other's toes or simply changing places with our erstwhile oppressors.

The view I have attempted to sketch in this section is a version of what Richard Rorty (1991: 68) calls ‘Dewey's experimentalism in moral theory’, its ‘whole point’ is, as he writes, ‘that you need to keep running back and forth between principles and the results of applying principles.’20 Without the universality of justice there are only particular cruelties and particular pains, without attention to the particular cruelties and the particular pains, theories of justice are utopian fantasies, at best comforting, at worst causes (p.327) of great suffering.21 With a universal but modifiable conception of justice and attention to particulars, victims of injustice may press their claims with dignity. There is no doubt that our conception of justice has widened over the last several centuries and will continue to widen; the process of widening I have envisaged in this section will not lead to fragmentation: on the contrary, our principles have become and will continue to become more inclusive.

7 Universalism Revisited

I now turn to the claim that the liberal conception of justice is a conception of justice only for modern constitutional democracies. Sen (1989: 14) seems to read Rawls's insistence that justice as fairness is a political conception of justice in this way when he writes: ‘The definitional exclusion contained in Rawls's “political conception” limits the scope of the concept of justice drastically and abruptly, and it would often make it hard to identify political rights and wrongs that a theory of justice should address.’ While agreeing with Sen's sentiment, I want to dispute his reading of ‘Justice as Fairness: Political not Metaphysical’.

When Rawls (1985: 224) speaks of a political conception of justice he means both that it is a conception of justice ‘worked out for a specific kind of subject, namely, for political, social, and economic institutions’, and that it should be ‘so far as possible, independent of controversial philosophical and religious doctrines’ (ibid.: 223) because ‘as a practical political matter no general moral conception can provide a publicly recognized basis for a conception of justice in a modern democratic state’ (ibid.: 225). According to Sen (1989: 15–16) ‘on grounds of the absence of toleration a whole lot of comprehensive doctrines may be ruled out of court (indeed, in some cases none may remain), and yet there may be very perspicuous problems of justice and injustice in disputes between different sides. To be without a theory that can deal with such problems (when the different sides are intolerant), and to see the disputes as lying outside the purview of the so‐called political conception of justice, would appear to be oddly limiting.’ Sen goes on to point out that even an intolerant political party may point out the injustice of the policies of an intolerant government that fails to provide for the victims of famine, and that they may do so using Rawlsian arguments. Of course, this last point of Sen's is correct, and Rawls may well accept it.

What Rawls (1985: 225) points out repeatedly is that ‘since justice as fairness is intended as a political conception of justice for a democratic society, it tries to draw solely upon basic intuitive ideas that are embedded (p.328) in the political institutions of a constitutional democratic regime and the public traditions of their interpretation. Justice as fairness is political in part because it starts from within a certain political tradition.’ We need not read this as claiming that justice as fairness may be used as a basis for social critique only of democratic societies. Indeed, who would make such a claim? Surely one cannot say from within the liberal tradition that its conception of justice applies only to democratic societies. From within the liberal tradition one does say that its conception of justice can be realized only under a democratic regime; from within the liberal tradition one acknowledges that the conception of justice serves as a publicly recognized regulative ideal only in democratic societies. However, these acknowledgements are in themselves criticisms of other societies; one says in effect that non‐democratic societies are eo ipso unjust. One cannot say from within the liberal tradition that one cannot criticize non‐liberal regimes precisely because universalism is built deeply within the foundations of the liberal conception of justice. Every human being capable of speech and action is entitled to live in a society which realizes the liberal conception of justice. (Contrast this with the Jewish prohibition of eating shellfish which explicitly applies only to Jews.) Again, there is nothing incoherent in demands for liberty when they are raised by revolutionaries against an intolerant regime, though such demands may strike an outsider as hypocritical if that outsider believes that the revolutionaries would establish a dictatorship.

But, perhaps, from outside the liberal tradition one can say that its conception of justice cannot serve as a basis of critique within a non‐liberal society. Lest this claim seem immediately absurd—surely Germans could have and should have objected on liberal grounds against National Socialism—a non‐liberal society must here be understood as a society without a tradition of toleration. This is the sort of objection which communitarians and moral relativists raise against any claim to universalism. With respect to the liberal political conception of justice, the claim is either that a liberal critique will be unintelligible in such a society or at the least that it will have no motivational force. This is not the place to examine the communitarian critique of Rawls or more generally the claims of moral relativism. It suffices to say that (1a) Sen is surely right in holding that the second principle of justice may serve as a basis for a critique of economic arrangements even when neither the regime nor its critics accept the first principle. Such critiques are intelligible and have motivational force, they have been known to lead to reform or revolution. (1b) When Amnesty International criticizes regimes of all stripes for human rights violations, i.e., violations of Rawls's first principle, its protests are not pointless. While Amnesty cannot take credit for any revolutions (nor would it wish to), it has been and continues to be able to save some lives. Clearly, it is able to make its objections intelligible and to find ways to give them motivational force.

(p.329) (2) When Rawls (1985: 225) says that he has articulated the political conception of justice implicit in the democratic tradition, he is not saying that it cannot be applied to other societies; he says, rather, that whether it can be so extended is a separate question which he wants to avoid prejudging. I take it that Sen's point is that we need a political conception of justice that can be applied to all societies, and I am saying the liberal conception of justice is one such conception. I do not say that the liberal conception is the only such conception, nor do I say that A Theory of Justice or ‘Justice as Fairness: Political not Metaphysical’ provide the best articulation of that conception. I am saying that the liberal conception of justice is the best we have, I am also saying that we must treat it as provisional, being ever ready to deepen and broaden our understanding of it.

(3) One wonders where this position outside the liberal tradition is located. The communitarian and relativist critics of liberalism are in fact not products of traditional societies, they are themselves heirs of the liberal tradition. Perhaps this is why, in the end, I find it difficult to make their position intelligible, but that is the subject of another paper.

If it were the case, however, as communitarians and relativists claim, that justice judgements make sense, or have a truth value, or have motivational force only relative to ‘social meanings’ (Walzer, 1983), ‘an adequate moral system’ (Wong, 1984), or relative to an agreement (Harman, 1975), then justice would be fragmented. It would be nonsensical to judge the social meanings of one society in terms of the social meanings of another, to judge one adequate moral system from the perspective of another, or one agreement in terms of another. It would then be nonsensical or pointless or a mere expression of one's frustration to say that the practices of another society based on different ‘understandings of justice and rationality’ (MacIntyre, 1988) are unjust. While a particular action or even a particular practice can be judged from within the comprehensive understanding to which both agent and critic belong, and Walzer (1987) has argued that such criticism can be far‐reaching, there would be no point of view from which it would make sense to judge an understanding of justice and rationality itself. But to object to the fact that, e.g., women in India receive less adequate nutritional and health care than men (Sen, 1990) is neither pointless nor nonsensical. It is, in fact, a first step toward understanding a situation which needs correcting. It would, of course, be insufferable arrogance to straight away condemn Indian fathers and husbands without understanding the complex reasons for this phenomenon; it is that complex situation which needs modifying.

Narrowing the range of applicability of the liberal conception of justice leads once again to fragmentation. The liberal conception of justice would then be reconceived as the ‘Western’ conception, an alternative to, say, the Indian conception (but India is, in fact, a constitutional democracy!) and so on. But why would it stop there. Surely there are significant differences in (p.330) the traditions of different European countries (e.g., Catholic versus Protestant) or of the diverse communities (Muslim, Hindu, Sikh) in India etc. Once again, as earlier when considering a feminist theory of justice, I can see no natural stopping point once we step on the slippery slope of demanding communally anchored theories of justice. Moreover, even internal criticism will be called into question. Is it not always open to the guardians of the old order to claim that their critics have stepped outside the tradition? Women, in particular, are endangered by moves toward communal conceptions of justice since the traditions of most communities have denied women's moral equality; so are ‘minorities’ (racial, ethnic, religious, etc.) for most communal traditions refuse to grant them an equal place. I conclude that demands for feminist theories of justice or for limiting the range of the liberal conception of justice to the so‐called ‘West’ (is Australia west?) is not to the advantage of the least advantaged. Let us not open, in the name of toleration, the doors to intolerance, but let us, in the name of universalism, be prepared to learn from the cries of the oppressed, whoever and wherever they are.22


Bibliography references:

Benhabib, Seyla (1987). ‘The Generalized and the Concrete Other: The Kohlberg‐Gilligan Controversy and Feminist Theory’ in S. Benhabib and D. Cornell, eds. Feminism and Critique. Minneapolis, Minn.: University of Minnesota Press.

—— (1989). ‘In the Shadow of Aristotle and Hegel: Communicative Ethics and Current Controversies in Practical Philosophy’, The Philosophical Forum, XXI, 77–95.

Benhabib, Seyla and Cornell, Drucilla (eds.) (1987). Feminism and Critique. Minneapolis, Minn.: University of Minnesota Press.

Drèze, Jean and Sen, Amartya (1989). Hunger and Public Action. Oxford: Clarendon Press.

Harman, Gilbert (1975). ‘Moral Relativism Defended’, Philosophical Review, 84, 3–22.

Hart, H. L. A. (1973). ‘Rawls on Liberty and Its Priority’, University of Chicago Law Review, 90, 534–55.

Hooks, Bell (1981). Ain't I a Woman: black women and feminism. Boston: South End Press.

Lorde, Audre (1984). Sister Outsider. Freedom, Calif.: The Crossing Press.

(p.331) MacIntyre, Alasdair (1988). Whose Justice? Which Rationality? Notre Dame, Ind.: University of Notre Dame Press.

Minow, Martha and Spelman, Elizabeth V. (1990). ‘In Context’, Southern California Law Review, 63, 1597–652.

Okin, Susan, Moller (1989). Justice, Gender and the Family. New York: Basic Books.

Rawls, John (1971). A Theory of Justice. Cambridge, Mass.: Harvard University Press.

—— (1985). ‘Justice as Fairness: Political not Metaphysical’, Philosophy and Public Affairs, 14, 223–51.

—— (1987). ‘The Idea of an Overlapping Consensus’, Oxford Journal of Legal Studies, 7, 1–25.

—— (1988). ‘The Priority of Right and Ideas of the Good’, Philosophy and Public Affairs, 17, 251–76.

Rorty, Richard (1991). Objectivity, Relativism, and Truth. Cambridge: Cambridge University Press.

Sen, Amartya (1989). ‘The Territory of Justice’, Discussion Paper Number 1425, Harvard Institute of Economic Research.

—— (1990). ‘More than 100 Million Women are Missing’, The New York Review of Books 20 Dec.

Sojourner, Sabrina (1991). ‘Race, Class and Reproductive Freedom’, National Now Times, 23, 8.

Spelman, Elizabeth V. (1988). Inessential Woman: Problems of Exclusion in Feminist Thought. Boston: Beacon Press.

Walzer, Michael (1983). Spheres of Justice. New York: Basic Books.

—— (1987). Interpretation and Social Criticism. Cambridge, Mass.: Harvard University Press.

Williams, Bernard (1978). Descartes: The Project of Pure Enquiry. Harmondsworth: Penguin.

Wong, David B. (1984). Moral Relativity. Berkeley, Calif.: University of California Press.

Young, Iris Marion (1989). ‘Polity and Group Difference: A Critique of the Ideal of Universal Citizenship’, Ethics, 99, 250–74.


(1) I just said that one is prompted to seek a theory of justice because one is outraged by injustices. That suggests that the citizens of a Rawlsian well‐ordered society do not seek a theory of justice; one might suppose that they take themselves to have found the perfect theory of justice already. I don't want to be committed to that view. I think of a well‐ordered society as a human, worldly society, and any such society can be further improved. As our institutions become more just, we shall, I hope, become more sensitive to injustice.

(2) Elizabeth Spelman writes, ‘I have come to think even of the phrase “as a woman” as the Trojan horse of feminist ethno‐centrism’ (Spelman (1988: x)).

(3) Justice within national boundaries is, however, not confined to citizens. Issues of who is entitled to the full range of rights and obligations of citizenship are among the most fundamental of justice conflicts. Moreover, even non‐citizens are able to press some claims in the name of justice.

(4) There is also the reverse challenge, the claim that the voice of justice is the voice of the poor and the powerless, resentment and envy masquerading as moral indignation. I don't accept that characterization of justice although I readily agree that the poor and the powerless are victims of injustice.

(5) Some critics have denied that rational choosers so situated would choose the Rawlsian principles, others have suggested modification in what is known to the parties, i.e., in what we are to consider and not consider when we think about justice. Finally, there are critics who reject the very idea of an ‘original position behind a veil of ignorance’.

(6) One might, of course, wonder whether that general knowledge is not already biased in favour of whatever groups are dominant in society or in the academy where that general knowledge is developed and disseminated.

(7) Here I have been guided by characterizations offered by Seyla Benhabib (1987: 81) and Audre Lorde (1984: 116).

(8) I borrow the language of substitution from Seyla Benhabib who recognizes the particular nature of the male voice that is heard in our supposedly universalistic theories of justice. The charge that the woman's voice we hear is also a particular kind of voice is documented by bell hooks, Audre Lorde, Elizabeth Spelman, and others.

(9) Of course, one can only speak from one's own point of view, but one can, indeed when deliberating about justice one must, attempt to discount one's personal preoccupations, and one can refrain from claiming to speak for a greater collectivity than one is capable of representing.

(10) Rawls does discuss ‘intolerant sects’ (1971: 35), and to what extent they are to be tolerated in the just society. He does not discuss whether a just society that tolerates such a sect has any obligations to protect the sect's members from the sect's hierarchy.

(11) Further evidence is gleaned from this: in Sister Outsider, Audre Lorde prints a letter which she wrote to Mary Daly. In this letter Lorde points out that in her book Gyn/Ecology Daly has made ‘the assumption that the herstory and myth of white women is the legitimate and sole herstory and myth of all women to call upon for power and background, and that nonwhite women and our herstories are noteworthy only as decorations or examples of female victimization . . . This dismissal,’ she continues, ‘does not essentially differ from the specialized devaluations that make Black women prey, for instance, to the murders even now happening in your own city.’ (69) This letter, though written in anger, is a plea for a feminist movement that will recognize racial and ethnic diversity as well as what women have in common; it acknowledges Daly's contributions as well as noting her failings. Lorde published this originally private letter because Daly never responded to it.

(12) What has just been said is true as far as the book goes; however, I am sure Okin understands that while the courts recognize and prosecute extra‐marital rape, it is difficult for rape victims to establish the fact that they have been raped. That difficulty is due to the gender structure of our society against which Okin's argument is directed.

(13) Of course, both Okin and hooks focus on the situation of women in the United States; they do not claim to speak for women in other countries, in particular, they do not claim to speak for women in the developing countries. What I have to say in the next few pages is thus only indirectly related to the topic of this conference.

(14) Day‐care providers, one cannot help feeling, take the place occupied by domestics in the utopias of an earlier generation of feminists. Of course, we need quality day‐care, of course, government subsidies should enable the children of poorer parents to receive the same high quality care as those of the better off, and, of course, only day‐care will keep single parents off welfare. But it is hard to believe that day‐care will ever become a high prestige job, or a very highly paid one, hence it is troubling that it should play such a central role in the more just future society envisaged by Okin.

(15) Rawls replaced the phrase ‘the most extensive total system of equal basic liberties’ by ‘a fully adequate scheme of equal basic liberties’.

(16) This is not the first time Sen made this suggestion. It has been rejected by Rawls, e.g. in Rawls, 1988. This is not the place to recount or to attempt to adjudicate this dispute. I do, however, lean toward Sen's position.

(17) I owe this point to a personal communication from Robin Avery. She brought to my attention the situations of (a) homeless, intravenous drug users who need stable housing situations if they are to kick the habit, and (b) of HIV‐infected persons whose symptoms do not match the criteria for AIDS established by the Centers for Disease Control and who are, therefore, not deemed disabled and denied benefits. These persons seem to be women, the poor, children, and intravenous drug users.

(18) This description of scientific progress should be acceptable both to realists who would, however, want to add that the successive descriptions approach more and more closely to what Bernard Williams (1978) calls the absolute conception' of the world, and to those who reject that notion and are satisfied with a more modest understanding of scientific progress.

(19) It is worth remembering that some white women failed to acknowledge the justice of the African–Americans' demand, and some African–American men failed to acknowledge the justice of the women's demand.

(20) A similar approach is found in Minow and Spelman, 1990.

(21) I take it that both Christianity and Marxism are, or were in their origins, theories of justice.

(22) I am indebted to Robin Avery; to members of the audience when an earlier version of this paper was read at the New School for Social Research, and above all to my colleagues Ann Congleton, Alison McIntyre, and Ken Winkler who subjected an earlier draft to searching criticism and helped to clarify my thinking. I am also indebted to my commentator Susan Okin and to other members of the conference on Human Capabilities: Women, Men and Equality, at which the penultimate version of this paper was read. None of these are responsible for the errors which remain.