Jump to ContentJump to Main Navigation
The Quality of Life$

Martha Nussbaum and Amartya Sen

Print publication date: 1993

Print ISBN-13: 9780198287971

Published to Oxford Scholarship Online: November 2003

DOI: 10.1093/0198287976.001.0001

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: 05 August 2021

Women and the Quality of Life: Two Norms or One?

Women and the Quality of Life: Two Norms or One?

(p.279) Women and the Quality of Life: Two Norms or One?
The Quality of Life

Julia Annas

Oxford University Press

Abstract and Keywords

Questions about the quality of life are questions about the lives of individuals in society. But, apart from the complication that individuals come already grouped into families,1 there is the crucial complication that individuals come divided into two sexes. The biological sex differences between men and women bring with them, in all known societies, enormous cultural divisions. To a greater or lesser extent, the shape of men's lives is in all societies different from that of women's. And not only do social institutions and attitudes every‐where divide up kinds of activity between the sexes, but women and men see their own lives, from the inside, very differently....


Questions about the quality of life are questions about the lives of individuals in society. But, apart from the complication that individuals come already grouped into families,1 there is the crucial complication that individuals come divided into two sexes. The biological sex differences between men and women bring with them, in all known societies, enormous cultural divisions. To a greater or lesser extent, the shape of men's lives is in all societies different from that of women's. And not only do social institutions and attitudes every‐where divide up kinds of activity between the sexes, but women and men see their own lives, from the inside, very differently.

Even casual reflection on the details of everyday life shows that these differences are deep and wide. It is, I think, fair to say that there is no society in which there is not, in some degree, sexual division of activity; and so, for any individual, the fact of being a man or a woman to some extent determines both what that individual's options are for undertaking various kinds of activity and how that individual sees his or her own life. There are everywhere two actual norms for human life: in no society is it indifferent to the shape of your life and what you can make of it, whether you are a man or a woman. Your sex may close some options to you entirely, or merely make them more difficult; it always makes a difference to what your options are over your life as a whole.2

Societies all lie on a spectrum from what I shall call traditional societies at (p.280) one end to what I shall call liberal societies at the other. A society is traditional if the fact of having two norms for the lives of men and women produces a strongly enforced actual division of activities and ways of living; it is liberal if this division of actual activities is weakly enforced. This gives us a way of contrasting societies which cuts across other differences in values, and makes it possible to say that women in otherwise very different societies face the same problem, without overdramatizing what the problem is.3

We are used to living with the two norms, and often cease to notice them. But they generate deep problems when we ask about the quality of individuals' lives.

Let us make the point more vividly by setting up a situation where we are on the outside. Consider a society which is more traditional than ours, say one in which the man is the family's main provider and the woman has total charge of bringing up the children. In this society men are educated to make as much as they can of their talents and to achieve as much as they can; but women are not, because they do not use their talents outside the domestic area. We do not need to suppose that there are actual laws forbidding women to hold certain jobs; custom will suffice, for example by making parents reluctant to ‘waste’ resources on educating daughters.4 Nor do we have to imagine that the women are not educated at all; suppose them to be educated to the level needed for them to be companions to the appropriate men.5 But now we know that educational opportunities in this society are distributed unfairly; regardless of ability, men will take more than they should.

There are two obvious retorts, of course. The first is that in this society women do not want education; they want to be home‐makers. And even if some of them do want it, it will still be said that in this society women have less need for these opportunities, for they lack the opportunity to put them to use. If it is unthinkable for a woman to have a career, or a least one equal to her husband's, then opportunities equal to his will be wasted on her. So the opportunities are in fact, given the two actual norms, being distributed in accordance with a principle of need.

But it is clear that we cannot justify the distribution by pointing out that it answers to a pattern of need created by the differences between the two norms. Rather, it is because we can clearly appreciate that the distribution is unjust that we come to criticize the two norms and the different patterns of (p.281) need that they create. It is, of course, easier to appreciate the injustices that the norms of a society give rise to if we do not have to live with the norms ourselves. It is much harder to stand back from, and criticize, the injustices arising from the differentiated norms that are part of our own society's current ways of thinking.6 None the less, as soon as we do position ourselves with regard to a more traditional society it is obvious that injustice results from the existence of two norms.

This is not just an ‘intuition’, if that means a judgement which we happen to make but might not have, and cannot defend. Rather we have here a systematic pattern of judgements; whatever the adequacy of our view of our own society, when faced by a society more traditional than ours, we conclude that the two norms for men and women produce injustice. This holds, I think, whoever ‘we’ are; even a traditional society will find fault with an even more traditional society in this area. The Greeks in Aristotle's time formed a very traditional society; yet Aristotle records the feeling that they had progressed in not buying and selling women for wives as their ancestors had done.7 So we do not just happen to feel that these women are being wronged; this is a judgement that people systematically come to when they see things from a less traditional viewpoint than the situation being considered. It is an empirical fact that people tend to judge in this way; but it is also an important and, as we shall see, revealing fact. Of course I am not claiming that it will be easy to isolate people's judgements about the position of women from their judgements about other matters in the more traditional society (about inheritance, for example). And of course societies can consciously revert to more traditional ways. (Though it is a moot point how successfully one can go back to these ways, as opposed to imposing them on people.) But this does not affect the point that, when we look from the outside at a society which is more traditional than our own, we systematically perceive injustice in the ways in which the two norms impose different kinds of life on men and women.

Can we locate more precisely the nature and source of the injustice? A popular approach is that the injustice arises from the frustration of well‐being in some form. If we pass over obviously inadequate accounts of this in terms (p.282) of pleasure of happiness,8 the most promising approach is in terms of the frustration of desires. But attempts to locate the source of the injustice in the frustration of women's desires are doomed to failure for familiar reasons. For people's desires can be in large part formed by the circumstances and options that they perceive as being open to them. (The point holds across different conceptions of what a desire is.) In societies in which the options open to them are fewer than those open to men, it has always been a common adaptive strategy for women to adjust their desires to what they can realistically expect. So examining the actual desires of women may lead us to the conclusion that women on the whole get what they want. But it is clear that desires which owe their nature even in part to the agent's reduced circumstances cannot adequately adjudicate questions of justice. This approach would lead us to conclude that it is women in the most traditional societies—those where in every area of living activity is divided by sex in the most marked and rigid way—who are most satisfied. For women in these societies, where alternatives are hardest even to conceive, are most likely to have adjusted their desires successfully to achieve a measure of content with their lot.9 But it cannot, of course, be right that the happiest women should turn out to be those whose horizons are so limited that they cannot even conceive of alternatives. And even apart from this issue, we have only to look at the way that women's desires have expanded with the expansion of the alternatives open to them to see that women's actual desires cannot settle this issue for us.

We might appeal to ‘informed’ desires: the women in the traditional society, with their domestic futures, do not desire education (at least, not the kind their husbands get). But, it could be argued, their desires are superficial, resting on an unreflective view of their circumstances. If their desires were to be fully informed—if they were the desires which they would form given full knowledge of all aspects of their situation—then they would form desires for the kind of education their husbands have.

Often the appeal to informed desires is fruitful, but in this case it is hard to see how it could be adequate. Giving these women information in the ordinary sense cannot produce what we need. The more facts the women acquire, not only about men's and women's abilities but about their situation, the explanation for it, the historical background, the patterns of needs it generates, and so on, the less likely they are to develop the relevant desire. For in this situation, and in many involving differentiated activity between the sexes, the more you know about the situation, the more hopeless it seems (p.283) to develop the relevant desire; the more clearly you see why that desire cannot be fulfilled, given the existing two norms, the more motivation you have to adjust to things the way they are to avoid frustration. Perhaps, then, ‘informed’ means more than just having more information: if these women understood their situation fully they would find in themselves desires for education like their husbands have. But any notion of understanding that could produce this result would be an unsatisfactorily expansive one. The women would have to grasp that they had no reason not to desire what they were being unjustly denied; and that is rather a lot to demand of a notion of understanding. No concept of informed desire can on this issue avoid the Scylla of inadequacy and the Charybdis of begging the question.

We might appeal to considerations more abstract than happiness and the fulfilment of desires. We might just point to unequal possession of goods on the part of men and women in a given society. Suppose that we aggregate the amounts of a certain good possessed by men and by women, and compare. And suppose, not too implausibly, that the men have more money, education, etc.; surely this is what gives rise to an unjust situation, unless we can show that the inequality rests on some basis other than the sheer fact of difference between the sexes.

But apart from familiar objections to this kind of procedure, there are special reasons why it will not work in this case, so that we do not need to work out sophisticated versions of it. It provides no answer to the person who claims that the inequality does answer to unequal needs, given the two norms. And it could well give the wrong answers, given the complication that class always cuts across sex. In nearly every society there will be a small number of élite women, wives and daughters of the socially most prominent and powerful men, who have large amounts of the desirable goods that most women are lacking in. So when we are looking at the amount of a certain good possessed by women in total, the results will be most misleading about the position of all but a few women.10 Moreover, in many cases the great wealth possessed by a few élite women in a society makes no impact on the ability even of these women to live in a way different from that imposed on them by the two norms. Until fairly recently, for example, rich women could spend money to influence votes and politicians, but could not themselves vote or hold political office. No approach in terms of what men and women have can get to the heart of the problem, which is what men and women do.

(p.284) Many prefer at this point to find the source of the injustice that we discern in terms of the violation of rights, or principles of equality, or the wrongness of denying autonomy to people in imposing sex roles on them. I do not think that these approaches have been conspicuously successful. In this area there is no consensus as to what the basic rights are that we would appeal to, nor as to the principles of equality that we would have to use, nor as to the justifiability of appealing to autonomy when this would involve overriding people's actual considered judgements. Further, it is hard to feel that these approaches go deep enough. Why is it important for men and women to have equal rights and not to have distinct norms for living their lives imposed on them? Some further answer is needed, and I suspect that it will have to be in terms of something like the importance of their capacities for living a flourishing human life.11 However, nothing in this paper hinges on these other approaches having been ruled out, and I offer no arguments against them here. What I would like to do instead is simply to defend the most intuitive way that we actually think about these matters, and to clear away some misconceptions about it. Defending it fully against all alternatives, and working out its relation to them, particularly to the rights approach, is clearly a much larger task.

We grasp that what is done in the more traditional society is unjust. What could be the basis for this, if not the frustration of desires or the violation of rights? The key seems to lie, as hinted above, not in what people want, or in what they possess, but in what they do. In the traditional society the women's activities, and hence the ways they shape and give meaning to their lives, are limited by the different norms of life and activity for the two sexes, in a way that bears no relation to, and can frustrate, their abilities, and the ways they could act and shape their lives if the norms were not distinct. Men and women can benefit equally from education, can learn skills and develop their intellectual capacities equally; and this does not depend on whether they desire to do these things.

The appeal seems to be to a general and indeterminate shared notion of human nature — human nature, that is, as shared by men and women. And, of course, though we are most likely to come to the thought of it by noticing injustices to women, the point applies equally to men. In the case of either sex, we discern injustice when relevantly similar groups of people are treated differently. In this case the thought that these groups of people are in fact relevantly similar rests, I have claimed, not on thoughts about similarity of desires or of rights, but on similarity in the way they are, their human nature.

Serious objections have been made to such an appeal to human nature, but before turning to them I want to expand on the idea and make it clearer what is involved. We see that what is done about, for example, education in the traditional society is unjust because what people make of their lives is restricted (p.285) by norms that do not answer to differences in their natures, differences of actual ability. But is this not far too general a claim, and one that will end up being far too strong? Many social norms can, or could, rest on differences that are conventional, answering to no natural differences of ability. Such norms can be mutually advantageous, and it seems no objection to them that they answer to no natural differentiation. So the objection to sex roles, and the distinct norms they spring from, on the grounds that they do not answer to natural distinctions between the sexes, would seem to imply, absurdly, objections to any conventionally based social norms, whether or not they are mutually advantageous and acceptable on other grounds. If not, there must be some demonstrable differences as regards sex roles and distinct norms of life for the sexes which makes it plausible that they create injustice in ways that other conventionally based norms may not.

There is such a difference, I think, but we find it not by argument but by considering history. In the course of recorded history people have objected to distinct sexually based norms for life in ways that they have not objected to other norms which do not seem to be based on differences in people's natures. We are back with the point that people do systematically, from a certain perspective, discern injustice in the operation of sex roles. There could be worlds in which enforced sexual division of activity did not produce injustice, but ours is not one of them, as history shows us.12

In appealing to the way that people have objected to sex roles, are we not back to some version of the subjective approach in terms of people's desires? No. The objections have taken the form of claims that the imposition of sex roles is unjust, and I have suggested that behind this lies the claim that they do not answer to real differences in the natures of men and women. In fact, I think that if we take seriously the claim that this is the basis of the objections, we can better understand why it is that our judgements about the injustice of sex roles are typically made from the perspective of a less traditional society than the one being criticized.

In any society reasons will be given for having distinct norms for the lives and activity of men and women. (In a very unreflective society the members will perhaps feel no need to give these reasons to one another, and so they will remain implicit, discernible only by a more reflective outsider.) And these reasons typically come down to various assertions of relevant differences (p.286) between men's and women's natures. Thus it has been held, for example, that women may justly be kept from participation in public life because they are more self‐centred and less capable of impartial thought than men. In any given society these reasons may well be accepted. It is only from the perspective of a less traditional society, in which men and women are no longer restricted by this division of activity, that people can clearly see that the reasons given for the restriction were not good reasons but mere rationalizations. The backward‐looking nature of our discernment of the injustice of sex roles is grounded in this point: it is those no longer bound by a sexual division of activities who can see that the reasons offered for it when it was in force were not good ones; for they are the ones who can see that there is in fact no such natural division of abilities as the one the division of activities was supposed to rest on.13

Aristotle, as I mentioned, congratulated his society on being less traditional than their ancestors, and non‐Greeks, on the position of women. We do not have to assume that he was aware of the level of frustration of women's desires, or that he had even a faint grasp of the principle of equality of rights for women and men. Rather, living in a society where women were not treated as property, he could see that the reasons offered in the past and by non‐Greeks for treating women as property were not good ones, and did not answer to a relevant natural similarity. It is easy to see how this could coexist with a massive failure to see what was wrong with the reasons offered by Aristotle himself for having two norms of life for women and men; nor is it hard to explain this failure, since Aristotle was living within the norms in question. And this case illustrates what I have claimed is a systematic pattern. It is in societies more traditional than our own that we discern the injustice of sex roles; this is the perspective from which we can see that there is no natural basis for these roles.

The backward‐looking perspective on this issue can throw light on two other points also. One is the striking non‐reversibility of attitudes to the division of sex roles. For while societies can become more reactionary, they cannot reinstate past attitudes to sex‐divided norms. This is most plausibly explained as the straightforward cognitive difficulty in coming to regard as a reason something that you have seen to be a mere rationalization.14 The other is the acceleration of liberalization on this issue in societies where discussion and questioning of the reasons behind various socially marked divisions is prominent. Societies which are traditional in the sense I introduced of strongly enforcing sex‐divided norms are also likely to be traditional in the more common (p.287) sense of discouraging public reflection on the basis of those norms. And societies that are liberal in the sense of enforcing sex‐divided norms only weakly are likely to be liberal also in encouraging public rational reflection on just this kind of issue.15


We should now face the principled objections to the idea that it is in terms of human nature that we understand, and have to reflect on, the injustice of sex‐divided norms.

Why might one be reluctant to adopt this approach? One reason might be a reluctance to accept objectivism in ethics, at least until all else fails. Now would not be the time to defend objectivism in ethics in a general way; but it should be pointed out that it is really very difficult to avoid in this case. For, as I have noted, we cannot doubt that having two norms for the lives of men and women produces injustices; this is clear as soon as we turn to a society more traditional than our own. And we cannot understand why that is so as long as we do not go deeper than the desires people have. I have argued that this is so because the desires in question are infected by the norms in question. But there seems in this case to be another reason also. For once we come to the conclusion about injustice there are large implications. The injustices in question are not minor ones; if we take them seriously then we are committed to serious criticism of the social institutions and attitudes that help to create and sustain them. How could we be entitled to do that on the basis of people's desires? However information‐enriched, how could desires ever have that authority? A large‐scale criticism a of fundamental and pervasive fact of social life can only be solidly based if it is founded on the way people actually are, and makes a claim that social life is failing to answer to the way people actually are.16

An interesting objection enters at this point. We might reject informed desire accounts without accepting that notion of human nature.17 For we might accept that objectively there are important goods for people, but be unwilling to accept the further, more restrictive notion of human nature. This objection, I think, comes from seeing the ethical employment of human (p.288) nature as being in a tradition that is committed to things like essentialist thinking, which can independently be seen to be dubious or controversial; it can thus seem best to steer clear of this tradition altogether. I have not taken this option because I believe that the human nature tradition is more complex than often thought, and that when properly understood it is not open to familiar objections. While this project can obviously not be undertaken here, I believe there to be good reasons for not resting content with notions less robust than that of human nature.

Someone might accept what has been said so far and still not proceed further because of a feeling of futility in the face of the problem. Every known society has upheld different norms for the lives of men and women. No more basic social fact can be imagined. Further, it has entered in many ways into every less basic social fact. What makes us think that we have any chance of going behind the lives of men and women as they have developed and asking what they could be like if human activities were not divided by sex? This objector will claim that it is theoretically futile to try to establish what people are ‘really’ like here: even if we could in some areas, we surely cannot in this one, where people's conception of their lives is moulded by the basic division of life activities produced by different norms for the sexes. And this is matched by the practical futility of trying to do anything about the situation. We might hope to rectify injustices of a more tractable nature. But when we are considering injustices on this scale, it is romantic to think that we might make any difference.

It is this feeling of futility which explains, I think, why so many people ignore or acquiesce in these injustices in their own society. But the difficulty is surely exaggerated on both fronts. Recognizing the practical difficulties here should lead us neither to complacency nor to an unrealistic overestimation of what we personally can achieve. The claims of theoretical difficulty are more serious, however.

One familiar objection is that there is no such thing as the notion of human nature that we think we are appealing to. There is no bedrock of natural fact about humans, facts true of anyone regardless of culturally produced perspective. For our own culture affects the way we consider ourselves and others so deeply that it is useless to try to produce an objective theory of human nature; all we succeed in doing is to project our own parochial concerns on to a wider screen. It is worth noting that this type of objection has been rather popular with feminist writers, who, far from basing any feminist arguments on an appeal to human nature, tend to regard it as discredited. Indeed it is easy to produce examples of appeals to human nature in past writers, particularly political philosophers, which are rather discreditable. Women have been presented as defective in terms of theories of human nature which are manifestly biased towards men. (Rousseau is perhaps the most flagrantly un‐self‐critical example of this.) The response to this by many feminists has been to relativize one's theory of human nature to one's political theory. Thus past male political (p.289) thinkers have a theory of human nature to support their theories, feminists have their own, and so forth. In a recent book Alison Jaggar openly sets out different theories (liberal feminism, radical feminism, and so on), each with its associated theory of human nature. You choose between the theories on several grounds, of which the plausibility of the associated theory of human nature is one. But it is assumed throughout that there are no facts about human nature to which one could appeal to decide between theories.18

But as a response to past theories of human nature this is surely faulty. Of course Aristotle, and some more recent thinkers, have put forward ‘factual’ claims about the inferiority of women that manifestly spring from other preconceptions, and what they say about human nature to back this up is a transparent rationalization. But then they are wrong; they are doing something badly and in a biased way which might be done better and in a less biased way. To the extent that we disagree, as we do, with people like Aristotle and Rousseau on what women are like, the remedy is to improve on what they do, not to retreat to the position that we can do no better.19 Feminists have been unwise to accept relativism so readily on this issue, and to withdraw from discussion about human nature rather than continuing it.20 Consistent relativism here, for example, weakens the feminist case for social reform as well as blunting criticisms of past thinkers.

Secondly, I have argued that we need the notion of human nature here, to take the issue seriously. Just as we cannot stay at the level of people's desires, because these can be infected by the norms in question, so we cannot stay at the level of a theory of human nature which answers to and depends on a particular political theory or outlook. If I accept a theory of human nature but recognize that this depends on my position as, say, a liberal feminist, then I am accepting that my position on human nature is but one of many alternatives among which I choose. My choice depends on a number of factors, of course, and can be made on rational grounds and not on whim, but it is still a matter of choice. But then we are not talking about human nature, but about something else; human nature is the way we are, and we cannot (p.290) choose the way we are.21 And on this issue what we need to appeal to is nothing less than the way we are. When we judge that there is injustice in the society more traditional than our own, we are judging that the social arrangements are out of kilter with the way these people are—not that they are out of kilter with what these people want, or would want given certain information, nor that they are out of kilter with the way that we (liberal feminists, for example) judge people to be. The crucial thought is that there is something wrong, and that this wrongness does not depend on what these people happen to want, or think, or on what alternative theories we have decided to hold about them. We do not, of course, always have to have this thought when judging some state of affairs to be unjust, but in this case it is hard to avoid.

We need to appeal, then, to something which could decide between various moral and political theories, not something which is itself a product of the one we happen to hold. It does not follow, of course, that what we are appealing to is an account of humans which is ‘value‐free’ or ‘value‐neutral’, or in the domain of the ‘hard’ sciences. One reason why feminists have been eager to relativize theories of human nature to specific moral and political theories is that they have conceived the alternative to be a theory which is ‘purely factual’ and scientific. And there are major problems with this view.

We know, just from looking at the attempts, that it is very difficult to produce a scientific, value‐neutral view of the lives of our own species in a way that is independent of cultural factors. Biology can tell us about an individual's heart or liver, but we find no clear way to resolve the dispute as to the social and sexual patterns of the species homo sapiens, independently of cultural values, nor to improve its terms. We are primates, but primates have developed completely opposed social patterns of life, so biology seems to leave the dispute as to whether we are naturally monogamous, promiscuous, etc., no further on than what we can learn from an Updike novel. Of course there are facts here; but it is grotesque to think that we can appeal to the present state of what we know about these facts to learn about our present encultured state.22

What we need, then, is a notion of human nature which tells us about ourselves—not just biology but something on which ethics can build. Human nature, as I understand it, is not ‘value‐free’; we find out about it empirically, but by examining human life and the values it contains, not by doing science. But neither is it already moralized; it gives us a basis for our moral judgements, but it is not itself a moral notion.23 This conception of human nature (p.291) has recently become more respectable in ethics; it is no longer automatically ruled out as a confusion of ‘fact’ and ‘value’. But putting it to serious work is another matter, and several objections are commonly still thought to stand against this. It has commonly been argued that we have no reason to think that there could be such a conception unless we are committed to a teleological view of the world and the functions of humans within it; and that if such a view were correct it would preclude our ethical options in an unacceptable way. I have argued elsewhere that these objections are unsuccessful.24 Appeal to human nature in ethics has nothing to do with teleology (it turns up in ancient ethics in conjunction with a variety of views on teleology, including its complete denial). Nor is it dependent on any background theory of a metaphysical kind. As we have seen, it emerges in a quite straightforward way from ethical reflection. It is simply an appeal to the way people are, grounded in a refusal to take the level of people's expressed desires as the deepest we can go.


But even if we accept this there remains a serious problem: we may think that if the appeal to human nature is not to be uselessly vague, an appeal to a merely edifying ideal, then it must deliver to us a fairly specific pattern for the course of a human life. But if we think this, then in the present context of the roles of the sexes we run into a striking problem.

A judgement about the injustice done to women in the traditional society turned out to be a case of a more general pattern: when we criticize the two actual norms in a given society, it is on the grounds that they ignore the fact that men and women are actually similar in the relevant respect. And if this is grounded in human nature, it looks as though what we are doing is appealing beyond the two actual norms to a single ideal norm, an ‘androgynous’ or unisex one. And since there are no obvious limits set in advance to the ways in which more traditional societies than our own give us material for criticism, we seem to get an argument applicable in principle to all sex roles. If we continue with this argument we reach the conclusion that all sex roles are to be rejected (p.292) because they falsify and distort the nature that we share. There is a single ideal of human life as far as our nature goes; all actual division of activity by sex is the creation of social norms.

It is not hard to find statements of this view in feminist writers, for example: ‘male/female roles are neither inevitable results of “natural” biological differences between the sexes, nor socially desirable ways of socializing children in contemporary societies.’25 From Mill on, these writers are ready to admit that we have little idea in advance what the details of this unisex ideal will turn out to be, but they have no doubt that sexual differentiation of activity is imposed on our nature, not part of it.

But more recent writers, also feminist, have queried this whole way of proceeding. Thus Elizabeth Wolgast claims that not all aspects of sex roles and imposed on us: a unisex society, free of sex roles, finds no resonance in our nature, she says, and talks of the respect we owe to our nature to recognize essential differences based on but not limited to biological differences.26 We should note that this is a dispute among people all of whom accept the appeal to objective human nature as a basis for criticism of the two actual norms in our society. But they give different answers to the question, What is the ideal form of life that answers just to human nature? Some feminists argue that this ideal is unisex: sex roles are learned and could be learned differently. Others argue that we should have two ideal norms and not one; that the biological differences between men and women bring with them differences in instinct, need, and way of seeing one's life, and that to deny this is to flout biology. We are one species among others; in all other species biological differences bring with them different social patterns. Further, distinct social patterns may well give rise to different kinds of virtues in men and in women; (p.293) if women's role as mothers leads them, for example, to excel in caring and nurturing qualities, then why not value this alongside the different qualities that men have? A single specific ideal for both sexes disregards what is characteristically different and valuable about women's perspectives.27

This debate about our nature looks depressingly insoluble. Both sides can recognize the biological evidence, for what it is worth; but they give it a different significance. Both can recognize that as things are the sex roles in our society (and even more in more traditional societies) render many people's lives unfairly stunted; but one side will claim that we should get rid of sex roles altogether, the other that we should adjust them. It looks as though it is hard to find any consideration which will be decisive as to whether human nature provides us with one ideal norm or two for human lives.28 We find thoughtful opinion divided between the single unisex ideal norm and two ideal norms. They have very different implications for practice. A unisex ideal norm would base no differential treatment merely on difference of sex; activities and roles would be divided as little as possible by sex. A proponent of two ideal norms would condone and strengthen some sex roles, for example, mothers caring for children, and would insist on differential treatment for women on the grounds of different needs. Human nature seems compatible with two conflicting ideals for humans; but if we choose on other grounds between the conflicting ideals we seem to have lost the whole advantage of appealing to human nature in the first place.

The problem here comes from expecting the notion of human nature itself to provide us with a specific ideal for human life. I have argued elsewhere29 that we misconceive the role of human nature in ethics if we expect it to produce a specific ideal form of life. Human nature does not impose on us a specific set of things to do and way of living; rather it functions in a more unspecific and negative way, as a constraint on proposed forms of life and ethical rules. In the case of this issue the point comes out with especial clarity. We appeal to human nature deeper than the level of actual desires; but if we expect this to lead us to a specific ideal we find that all the unresolved problems about the importance of biology for our conception of human nature (p.294) reappear within the notion of human nature itself. Are we essentially a two‐sexed species, as Wolgast claims? Or is this a fact about us that does not have to have specific moral consequences? As long as we look for a specific ideal, we shall find the situation dear to the hearts of ancient sceptics: there is dispute, it is unresolvable, and the person who thinks about the matter tends to end up in suspension of judgement. (And since she has to act, she will act in accordance with her own view, but having no rational confidence in that view.)

But why should we expect our notion of human nature to be capable of delivering to us a specific norm or two specific norms for human life? We know that we cannot derive a single specific way of life or set of virtues from the notion of human nature alone. The difficulties here come from the common problem of trying to overwork the notion of human nature, to get from it what could not possibly be in it. And in the present case it is particularly obvious that we do not know what precise conclusion we should draw for human life from primate social behaviour and other facts of biology; how could the intervening notion of human nature magically solve the problems for us?

We should reject the two actual social norms that divide activities by sex without aiming at any specific ideal norm at all. We know enough about human nature to know that even if we had a complete account of it, it would not determine in detail what kinds of life we should lead. So the fact that we must appeal to it here is not undercut by the fact that we know depressingly little about it. We can work towards an ideal norm without being committed in advance to the form it will take. The actual argument I have concentrated on involves rejecting actual discrimination about education on the grounds that men and women are similar in the relevant respect. The best course would be to proceed entirely in this way, talking always about particular abilities. An advantage of a piecemeal approach like this is that it relates the claims we do make about human nature fairly directly to particular claims about the injustices in actual society. This is just the backward‐looking perspective that I have stressed. Thus, it is because we can see the injustice of discriminating in educational opportunities on the basis of sex (at least in more traditional societies than our own) that we can see that human nature supports equal opportunities for education for men and women, for there is no relevant basis for denying it. This claim is not based on a prior commitment to a specific ideal unisex way of life for men and women, and so is not disturbed by claims that might be made in other areas to the effect that men and women have distinct needs.

Thus we expand our knowledge of human nature and our natural needs and wants on the basis of our recognition of what goes wrong when society imposes two actual norms for male and female lives. This is our only route to discovering human nature, and our only hope of disentangling, to the extent that we can, very general facts about our nature from their specific, culturally differentiated appearances. Why should we think that there is a quicker and (p.295) better route to this knowledge, that we can know now either that there is an ideal, natural, unisex form of life, with all sexual division of roles and activity on the culturally imposed side, or that biology pervades culture and gives us two ineliminably different norms of life for men and women? Indeed, progress in serious consideration of particular capabilities in men and women has often been blocked by prior commitment to one or other of these pictures. I have argued that our judgements about injustice where sex roles are concerned are systematically backward‐looking: we are often confused or uncertain about our own society, but able to discern more clearly the wrongs in a society more traditional than our own. And there is a good reason for this: it is those who no longer have to live with a particular sex‐linked restriction who can rationally reject the claim that it was needed to answer to a natural difference.


This stress on the backward‐looking perspective eliminates familiar problems, but does it then give us too exiguous a basis to make judgements about injustice in our own society?30 If we can clearly discern the injustice only of practices that we no longer suffer from, will we not always be one stage behind, unable rationally to criticize the injustice of practices that we do still suffer from? This is an important and difficult issue; in fleeing premature dogmatizing we do not want to end up excessively sceptical. Our best route to avoiding both lies, I think, in stressing the piecemeal nature of what we gain from the backward‐looking perspective. Rather than having a complete perspective on gender justice at a given time, we have a variety of views, at different levels and of different institutions. Our grasp of the injustice of a sex‐divided norm that we live with is built up from the insights we have as to the injustice of various other norms we no longer live with, together with our usual capacities to make inferences, generalizations, and so on. And it is, I think, bound to be less secure than our grasp of past injustice. But it need not lead us to an over‐sceptical view if we give due account to the different sources and levels of past injustice and our ability to reason about these.

The appeal to human nature has become discredited through its association with authoritarian traditions, and with claims to produce specific patterns of human life that give us answers to difficult questions right at the start. I have argued that once we clear misconceptions away, the appeal appears more modest, and also gives us the basis for progress and advance. This is a distinct advantage over alternative approaches in terms of rights, for example—though I emphasize that I have not here argued against these approaches.31 We start from the unspecific notion that we have of human nature, arrived at negatively (p.296) through our recognition of various injustices arising as a result of imposed sex roles. We make it more specific as we consider the various capabilities and the kinds of life that they can coexist in. For a consideration of history gives us reason to think that the problem of ideal unisex versus sex‐divided norms is unlikely to be solved in just the terms we now conceive it in. It is more likely to dissolve into many smaller and more tractable problems about the various capacities of men and women. And progress here will, I suspect, for a long time reflect the fact that our best route to discovering human nature is reflection on the mistakes and rationalizations of those who got it wrong.


Bibliography references:

Elster, J. (1982). ‘Sour Grapes’, in A. Sen and B. Williams (eds.), Utilitarianism and Beyond. Cambridge: Cambridge University Press.

Ferguson, Ann (1977). ‘Androgyny as an Ideal for Human Development’, in Vetterling‐Braggin et al. (1987).

Gill, Christopher (1990). ‘The Human Being as an Ethical Norm’, in C. Gill (ed.), The Person and the Human Mind. Oxford: Oxford University Press.

Grimshaw, Jean. (1986). Feminist Philosophers. Brighton: Harvester Press, ch. 3.

Jaggar, Alison (1980). Feminist Politics and Human Nature. Totowa, NJ: Rowman and Allanheld.

Midgley, Mary (1979). Beast and Man. Brighton: Harvester Press.

——and Hughes, Judith (1983). Women's Choices. London: Weidenfeld and Nicholson.

Nussbaum, Martha (forthcoming). ‘Aristotle on Human Nature and the Foundations of Ethics’, in a volume on the philosophy of Bernard Williams, ed. J. Altham and R. Harrison. Cambridge: Cambridge University Press.

Rainone, Francine, and Moulton, Janice (1982). ‘Sex Roles and the Sexual Division of Labor’, in Vetterling‐Braggin (1982).

Sen, A. (1984). ‘Economics and the Family’, in Resources, Values and Development. Oxford: Blackwell.

Trebilcot, Joyce (1982). ‘Two Forms of Androgynism’, in Vetterling‐Braggin (1982).

Vetterling‐Braggin (ed.) (1982). ‘Femininity’, ‘Masculinity’ and ‘Androgyny’. Totowa, NJ: Rowman and Littlefield.

——et al. (eds.) (1977). Feminism and Philosophy. Totowa, NJ: Littlefield, Adams.

Warren, M.A. (1982). ‘Is Androgyny the Answer to Sexual Stereotyping?’, in Vetterling‐Braggin (1982).

Williams, Bernard (1985). Ethics and the Limits of Philosophy. London: Fontana.

Wolgast, Elizabeth (1980). Equality and the Rights of Women. Ithaca, NY: Cornell University Press.


I am very grateful for helpful comments to Ann Levey, David Owen, Patricia Scaltsas, and Holly Smith. Allen Buchanan enormously improved the early stages of this paper with comments and discussion. The discussion at WIDER was wide‐ranging and helpful; I would like to thank especially my commentator, Margarita Valdés, and also Christine Korsgaard, Hilary Putnam, Paul Seabright, Tim Scanlon, and Michael Walzer. I would also like to thank Martha Nussbaum for useful comments at more than one stage, and for her role in getting me to explore the actual usefulness as well as the historical roles of the notion of human nature.

(1) Cf. Elizabeth Wolgast, Equality and the Rights of Women (Ithaca, NY: Cornell University Press, 1980), ch. 6, esp. p. 143; if society is viewed straightforwardly as a collection of individuals, children come out as anomalies. Cf. A. Sen, ‘Economics and the Family’, in Resources, Values and Development, (Oxford: Blackwell, 1984), 369–85.

(2) There are of course other divisions that we find salient in everyday life which tend to cut across sexual division of activity: wealth and poverty, for example. (I am grateful to Paul Seabright for pressing this point.) But present and past experience shows that sexually divided norms of activity do present themselves to people as particularly salient even before explicit reflection on the topic, perhaps even more vividly because of the importance of the distinctions that cut across them.

(3) For ‘traditional’ and ‘liberal’ see below. ‘Traditional’ carries no implications about economic or cultural development; thus America in the 1950s is on my account a highly traditional society.

(4) We can suppose that there are some women who do get educated, and aim for a career; but they have to choose between the man's role (career) and the woman's role (domesticity). This does not affect the roles: the fact that a few women can become ‘honorary men’ has no effect on the existence of two norms. (In most societies class cuts across sex, and there are some élite women who are as advantaged as most men, but this makes no impact on the different roles and kinds of life for the two sexes. See below.)

(5) A situation which was far from dead when I was an undergraduate at Oxford. The recent past of one's own society will often provide a good example of a traditional society.

(6) Thus we should not be too surprised or shocked that it has generally been found easier, by both women and men, to acquiesce in injustices arising from sexually differentiated norms, than to focus on the injustices and think about the norms. It is, after all, often perfectly acceptable to acquiesce in one's society's ways of thinking; moral theories are suspect which put us at odds with them right at the start. Serious moral thinking about the sexes is very difficult just because of this fact, that it requires us not just to reflect on but to stand aside from some everyday habits of thought.

(7) Politics 1268b38–41. The passage is discussed in Martha Nussbaum's paper in this volume, ‘Non‐Relative Virtues: An Aristotelian Approach’. At Politics 1252a34–b6 Aristotle also criticizes non‐Greeks for treating women like slaves, not recognizing the natural difference between them. He regards his own society as less traditional than other societies, and than its own past.

(8) Happiness, that is, on a modern subjective view of what it is, which has prevailed in most ethical discussions: whether someone is happy is settled by what they think, happiness can be momentary, temporary, etc. I think that the notion of happiness as a satisfactory and well‐lived life, an objective notion well worked out in ancient ethics, has more footing in our ethical thinking than most ethical discussion allows, but I shall not pursue the point here.

(9) See J. Elster, ‘Sour Grapes’, in A. Sen and B. Williams (eds.), Utilitarianism and Beyond (Cambridge: Cambridge University Press, 1982).

(10) Two frivolous examples: Aggregate the personal wealth of British women, and add that of the Queen. Or: aggregate the personal political power of British women, and add that of Mrs Thatcher. The totals will be impressive, but not a good guide to the wealth and power of British women in general. (Of course this is only a crude approach. The Queen and Mrs Thatcher raise the average for women, but we would realistically focus on something more fine‐grained, e.g. average of personal property, or average of property held by women compared with that for men. The more realistic our approach, the less likely we are to reach results that are misleading with respect to the position of women in a society. This point thus has limited force when compared with the following one.)

(11) To avoid misunderstanding: of course there are many areas where appeal to rights or autonomy need involve nothing like this; I am simply claiming that on this issue it seems to.

(12) Several writers have seen the point that there must be something special about sex roles and norms for them to create injustice when other conventionally based roles and norms need not. But they tend at this point to appeal straight away to some theory to explain what it is about sex roles in particular that produces injustice: e.g. ‘the problem is not that sex roles restrict freedom but that sex roles and the sexual division of labour are used in patriarchal societies to oppress women’ (Francine Rainone and Janice Moulton, ‘Sex Roles and the Sexual Division of Labor’, in M. Vetterling‐Braggin (ed.), ‘Femininity’, ‘Masculinity’ and ‘Androgyny’ (Totowa, NJ: Rowman and Littlefield, 1982), 222). I want to avoid this approach and to pay more attention to the question of what thoughts we actually have, what we are in fact appealing to, when we perceive injustice flowing from sexually divided norms but not from other norms.

(13) This phenomenon is often clearest on a small scale. It was very obvious, for example, in Oxford during the period when single‐sex colleges were becoming coeducational, and colleges stood to one another as more to less traditional societies.

(14) Past attitudes can of course be recreated indirectly: if the government restricts opportunities for discussion and public reflection, destroys evidence of the liberal past, etc., for long enough people may be returned to their previous cognitive state.

(15) One could develop this point to show why liberal societies can justifiably claim to be advancing our conception of what is natural, even though the traditionalist may claim that changes in a liberal direction are unnatural. Public rational reflection, rather than traditional acquiescence, is more likely to produce an understanding of, and an explanation for, both what is rejected and what is accepted about human nature. I am grateful to Michael Walzer for stressing the point that a theory of human nature must be explanatory and not just critical of what it replaces.

(16) It could, of course, be based on moral concern, for example at the violation of rights. But, as suggested above, I think that in this case this is not what is fundamental; such concern rests ultimately on beliefs about what people's natures are.

(17) This point was raised by Tim Scanlon and by several others; I have learned much on this issue from Scanlon's paper in this volume and the comments by James Griffin.

(18) Alison Jaggar, Feminist Politics and Human Nature (Totowa, NJ: Rowman and Allanheld, 1983). Cf. p. 170: ‘The values that identify each group of feminists are connected conceptually with their characteristic view of human nature.’ According to Jaggar an individual's political values and conception of human nature hang together in a holistic way.

(19) Jaggar (Feminist Politics, 22) says that feminist critiques of male political philosophy ‘will necessitate a reconstruction not only of political philosophy, but of all the human sciences and perhaps of the physical sciences as well’. But talk of reconstruction suggests an engagement with, rather than a repudiation of, the past tradition, and this is incompatible with a thoroughgoing relativism. Jean Grimshaw, (Feminist Philosophers (Brighton: Harvester Press, 1986), ch. 3) makes some pertinent criticisms on this point.

(20) Some feminists argue at this point that the notion of objectivity in any science is a product of male thinking; women must do their own science and cannot appeal to any of the sciences in our tradition as a neutral point. This way of thinking puts women outside any current or past debate and leaves them to a discussion among themselves not bound by any current methodological rules. It is then not clear why any current thinking should take this seriously.

(21) More accurately: we cannot choose the total way we are, although this certainly includes aspects about which we can do nothing and aspects which we can develop or manipulate.

(22) Though this is a pseudo‐exercise many like to do; cf. the popularity of reductive forms of sociobiology. Since these are usually crudely unaware of the sexist values they are reading into biology, they have unfairly discredited serious attempts to take biology seriously in ethics. See M. Midgley, Beast and Man (Brighton: Harvester Press, 1979).

(23) Hilary Putnam, in the closing remarks of his paper in this volume, remarks that human nature can figure as very different kinds of notion. As I use it, it is an empirical notion: what we find out about ourselves and refine in an ongoing way as we learn more about various forms of social life. It is to be distinguished from human nature as a conceptual notion: what is shared and obvious and can be appealed to in a distinctively human form of life—and also from human nature as a moral notion: a normative idea playing a determinate role in an ethical theory.

(24) I have examined the appeal to human nature in ancient ethics, and modern understandings (and misunderstandings) of this, in ‘Naturalism in Greek Ethics’, Proceedings of the Boston Area Colloquium in Ancient Philosophy, 4 (1988). Cf. also Martha Nussbaum, ‘Aristotle on Human Nature and the Foundations of Ethics’ (forthcoming), and ‘Non‐Relative Virtues’ (in this volume); also the work of Christopher Gill in ‘The Human Being as an Ethical Norm’, in C. Gill (ed.), The Person and the Human Mind (Oxford: Oxford University Press, 1990). From different directions we have converged upon a criticism of certain assumptions, articulated most powerfully in Bernard Williams's Ethics and the Limits of Philosophy (London: Fontana, 1985), ch. 3.

(25) Ann Ferguson, ‘Androgyny as an Ideal for Human Development’, in M. Vetterling‐Braggin et al. (eds.), Feminism and Philosophy (Totowa, NJ: Littlefield, Adams, 1977), 45. Cf. also p. 56, where Ferguson claims that, compared with any ‘static universal theory of what the “natural relationship” of man to woman should be’, ‘it seems more plausible to assume that human nature is plastic and moldable, and that women and men develop their sexual identities, their sense of self, and their motivations through responding to the social expectations placed upon them’. On pp. 61–2 she claims, ‘There is good evidence that human babies are bisexual, and only learn a specific male or female identity by imitating and identifying with adult models . . . on this analysis, if the sexual division of labor were destroyed, the mechanism that trains boys and girls to develop heterosexual sexual identities would also be destroyed.’ On ‘androgyny’ as an ideal, see further Joyce Trebilcot, ‘Two Forms of Androgynism’, in Vetterling‐Braggin et al., Feminism and Philosophy, 70–8, and in Vetterling‐Braggin, ‘Femininity’, 161–9; also M. A. Warren, ‘Is Androgyny the Answer to Sexual Stereotyping?’ in Vetterling‐Braggin, ‘Femininity’, 170–86.

(26) Wolgast, Equality, 108: ‘an arrangement [where the roles of men and women will be indistinguishable] may not find answering resonance in our nature—on the contrary, our nature may constitute an obstacle’, and p. 110: ‘when we come to deal with sex roles and their justice, we should try to understand the respect we owe our nature.’ Cf. Mary Midgley and Judith Hughes, Women's Choices (London: Weidenfeld and Nicholson, 1983), 185, where they talk of ‘The extraordinary assumption that everything physical (the whole ‘genetic and morphological structure’) could be different and yet everything mental could remain the same’. Midgley and Hughes are, however, cautious in inferring from pervasiveness of sexual difference to legitimacy of sexual roles.

(27) I am indebted to Martha Nussbaum for stressing the importance of this point in the argument.

(28) We might try to find such a consideration in a specific case where it makes a difference to one's decision which side one takes. A central role has been taken in this debate by the Supreme Court case Geduldig v. Aiello (cf. Wolgast, Equality, 88–90; Midgley and Hughes, Women's Choices, 160–1), in which the court decided that there was no wrongful discrimination against women in an insurance programme that did not cover pregnancy; it simply excluded pregnant persons, and would equally exclude pregnant men, were there any. Supporters of the two‐ideal‐norms view conclude that this is what happens if one goes for a unisex ideal. Since only women get pregnant, they have different needs which should be allowed for in a distinct ideal for female life. Even this is hardly decisive, though; the other side could say that in this case what has gone wrong is that the ideal in question is not unisex at all, but just the traditional old male one; allowing that women have special needs, e.g. in pregnancy, is just the kind of thing needed to make the ideal unisex, rather than traditionally male.

(29) In the article referred to in n. 24 above.

(30) As is lucidly argued by Margarita Valdés in her reply to this paper.

(31) Onora O'Neill's lucid and forceful paper in this volume presents an excellent example of a different, more Kantian approach.