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: 21 June 2021

Human Capabilities, Female Human Beings

Human Capabilities, Female Human Beings

(p.61) Human Capabilities, Female Human Beings
Women, Culture, and Development

Martha C. Nussbaum (Contributor Webpage)

Oxford University Press

Abstract and Keywords

Nussbaum argues that the best way to tackle the issue of women's equality in developing countries is to begin with a conception of the human being and human functioning. By seeing certain capabilities and functions as more central to human life than others, Nussbaum anticipates feminist and cultural relativistic challenges and defends her particular brand of universalism and essentialism. This paper provides a self‐consciously normative conception of what it is to be human and to live a good life, as well as a proposal for a culturally sensitive development ethic.

Keywords:   central human capabilities, central human functioning, common humanity, development policy, inclusiveness, methodology, pluralism, universalism

Human beings are not by nature kings, or nobles, or courtiers, or rich. All are born naked and poor. All are subject to the miseries of life, to frustrations, to ills, to needs, to pains of every kind. Finally, all are condemned to death. That is what is really the human being; that is what no mortal can avoid. Begin, then, by studying what is the most inseparable from human nature, that which most constitutes humanness.

Jean‐Jacques Rousseau, Emile, Book IV

Women, a majority of the world's population, receive only a small share of developmental opportunities. They are often excluded from education or from the better jobs, from political systems or from adequate health care . . . In the countries for which relevant data are available, the female human development index is only 60% that of males.

Human Development Report 1993, United Nations Development Program

Were our state a pure democracy there would still be excluded from our deliberations women, who, to prevent depravation of morals and ambiguity of issue, should not mix promiscuously in gatherings of men.

Thomas Jefferson

Being a woman is not yet a way of being a human being.

Catharine MacKinnon

1 Feminism and Common Humanity1

Begin with the human being: with the capacities and needs that join all humans, across barriers of gender and class and race and nation. To a person concerned with the equality and dignity of women, this advice should appear in one way promising. For it instructs us to focus on what all human beings share, rather than on the privileges and achievements of a (p.62) dominant group, and on needs and basic functions, rather than power or status. Women have rarely been kings, or nobles, or courtiers, or rich. They have, on the other hand, frequently been poor and sick and dead.

But this starting point will be regarded with scepticism by many contemporary feminists. For it is all too obvious that throughout the history of political thought, both Western and non‐Western, such allegedly‐unbiased general concepts have served in various ways to bolster male privilege and to marginalize women. Human beings are not born kings, or nobles, or courtiers. They are, or so it seems,2 born male and female. The nakedness on which Rousseau places such emphasis reveals a difference that is taken by Rousseau himself to imply profound differences in capability and social role. His remarks about human nature are the prelude to his account of Emile's education. Sophie, Emile's female companion, will be said to have a different ‘nature’ and a different education. Whether, as here, women are held to be bearers of a different ‘nature’ from unmarked ‘human nature’, or whether they are simply said to be degenerate and substandard exemplars of the same ‘nature’, the result is usually the same: a judgement of female inferiority, which can then be used to justify and stabilize oppression.3

I shall argue nonetheless that we should in fact begin with a conception of the human being and human functioning in thinking about women's equality in developing countries. This notion can be abused. It can be developed in a gender‐biased way. It can be unjustly and prejudicially applied. It can be developed in ways that neglect relevant differences among women of different nationalities, classes, and races. But I shall argue that, articulated in a certain way (and I shall be emphatically distinguishing my approach from others that use an idea of ‘human nature’) it is our best starting point for reflection. It is our best route to stating correctly what is wrong with the situations that confronted Saleha Begum and Metha Bai,4 the best basis for claims of justice on their behalf, and on behalf of the huge (p.63) numbers of women in the world who are currently being deprived of their full ‘human development’.

I note that the concept of the human being has already been central to much of the best feminist and internationalist thinking. Consider, for example, J. S. Mill's remarks on ‘human improvement’ in The Subjection of Women; Amartya Sen's use of a notion of ‘human capability’ to confront gender‐based inequalities; the Sen‐inspired use of a notion of ‘human development’ in the UN Report to describe and criticize gender‐based inequalities; Susan Moller Okin's proposal for a ‘humanist justice’ in her recent major work of feminist political theory; Catharine MacKinnon's graphic description of women's current situation, quoted as my epigraph; and, of course, the role that various accounts of ‘human rights’, or even ‘The Rights of Man’ have played in claiming justice for women.5 Much the same can be said more generally, I think, about internationalist thought.6 To cite just one example, I take my proposal to be the feminist analogue of the proposal recently made by Ghanaian philosopher Kwame Anthony Appiah when he wrote, ‘We will only solve our problems if we see them as human problems arising out of a special situation, and we shall not solve them if we see them as African problems, generated by our being somehow unlike others.’7

My proposal is frankly universalist and ‘essentialist’. That is, it asks us to focus on what is common to all, rather than on differences (although, as we shall see, it does not neglect these), and to see some capabilities and functions as more central, more at the core of human life, than others. Its primary opponents on the contemporary scene will be ‘anti‐essentialists’ of various types, thinkers who urge us to begin not with sameness but with difference—both between women and men and across groups of women—and to seek norms defined relatively to a local context and locally held beliefs.8 This opposition takes many forms, and I shall be responding to several distinct objections that opponents may bring against my universalist (p.64) proposal. But I can begin to motivate my enterprise by telling several true stories of conversations that have taken place at WIDER, in which the relativist position9 seemed to have alarming implications for women's lives. I have in some cases conflated two separate conversations into one; otherwise things happened as I describe them.10

1. At a conference on ‘Value and Technology’, an American economist who has long been a left‐wing critic of neoclassical economics delivers a paper urging the preservation of traditional ways of life in a rural area of India, now under threat of contamination from Western development projects. As evidence of the excellence of this rural way of life, he points to the fact that, whereas we Westerners experience a sharp split between the values that prevail in the workplace and the values that prevail in the home, here, by contrast, there exists what the economist calls ‘the embedded way of life’; the same values obtaining in both places. His example: just as in the home a menstruating woman is thought to pollute the kitchen and therefore may not enter it, so too in the workplace a menstruating woman is taken to pollute the loom and may not enter the room where looms are kept. Amartya Sen objects that this example is repellant, rather than admirable: surely such practices both degrade the women in question and inhibit their freedom. The first economist's collaborator, an elegant French anthropologist (who would, I suspect, object violently to a purity check at the seminar room door), replies to Sen. Doesn't he realize that there is, in these matters, no privileged place to stand? This, after all, has been shown by both Derrida and Foucault. Doesn't he know that he is neglecting the (p.65) otherness of Indian ideas by bringing his Western essentialist values into the picture?11

2. The same French anthropologist now delivers her paper. She expresses regret that the introduction of smallpox vaccination to India by the British eradicated the cult of Sittala Devi, the goddess to whom one used to pray in order to avert smallpox. Here, she says, is another example of Western neglect of difference. Someone (it might have been me) objects that it is surely better to be healthy rather than ill, to live rather than to die. The answer comes back: Western essentialist medicine conceives of things in terms of binary oppositions: life is opposed to death, health to disease.12 But if we cast away this binary way of thinking, we will begin to comprehend the otherness of Indian traditions.

At this point Eric Hobsbawm, who has been listening to the proceedings in increasingly uneasy silence, rises to deliver a blistering indictment of the traditionalism and relativism that prevail in this group. He lists historical examples of ways in which appeals to tradition have been used to support oppression and violence.13 His final example is that of National Socialism in Germany. In the confusion that ensues, most of the relativist social scientists—above all those from far away, who do not know who Hobsbawm is—demand that he be asked to leave the room. The radical American economist, disconcerted by this apparent tension between his relativism and his affiliation with the left, convinces them, with difficulty, to let Hobsbawm remain.

(p.66) 3. We shift now to another conference two years later, a philosophical conference organized by Amartya Sen and me.14 Sen makes it clear that he holds the perhaps unsophisticated view that life is opposed to death in a very binary way, and that such binary oppositions can and should be used in development analysis. His paper15 contains much universalist talk of human functioning and capability; he begins to speak of freedom of choice as a basic human good. At this point he is interrupted by the radical economist of my first story, who insists that contemporary anthropology has shown that non‐Western people are not especially attached to freedom of choice. His example: a new book on Japan has shown that Japanese males, when they get home from work, do not wish to choose what to eat for dinner, what to wear, etc. They wish all these choices to be taken out of their hands by their wives. A heated exchange follows about what this example really shows. I leave it to your imaginations to reconstruct it. In the end, the confidence of the radical economist is unshaken: Sen and I are both victims of bad universalist thinking, who fail to respect ‘difference’.16

Here we see the relativist position whose influence in development studies motivated the work that has led to the present volume. The phenomenon is an odd one. For we see here highly‐intelligent people, people deeply committed to the good of women and men in developing countries, people who think of themselves as progressive and feminist and anti‐racist, people who correctly argue that the concept of development is an evaluative concept requiring normative argument17—effectively eschewing normative argument and taking up positions that converge, as Hobsbawn correctly saw, with the positions of reaction, oppression, and sexism. Under the banner of their fashionable opposition to ‘essentialism’ march ancient religious taboos, the luxury of the pampered husband, educational deprivation, unequal health care, and premature death. (And in my own universalist Aristotelian way, I say it at the outset, I do hold that death is opposed to life in the most binary way imaginable, and freedom to slavery, (p.67) and hunger to adequate nutrition, and ignorance to knowledge. Nor do I believe that it is only, or even primarily, in Western thinking that such oppositions are, and should be, important.)

The relativist challenge to a universal notion of the human being and human functioning is not always accompanied by clear and explicit philosophical arguments. This is especially true in the material from development studies to which I have referred, where the philosophical debate concerning relativism in ethics and in science is not confronted, and universalism is simply denounced as the legacy of Western conceptions of ‘episteme’18 that are alleged to be in league with imperialism and oppression.19 The idea behind this volume as a whole was that to sort out various strands in the philosophical debate on these questions would be of the first importance in making further progress on women's issues; and the papers by Alcoff, Benhabib, Glover, and Hilary Putnam carry out various aspects of this anti‐relativist project. Here, then, I shall simply set out rather schematically and briefly, for the purposes of my own argument, several objections to the use of a universal notion of human functioning in development analysis to which I shall later respond.

2 The Assault on Universalism

Many critics of universalism in ethics are really critics of metaphysical realism who assume that realism is a necessary basis for universalism. I shall (p.68) argue that this assumption is false. By metaphysical realism I mean the view (commonly held in both Western and non‐Western philosophical traditions) that there is some determinate way the world is, apart from the interpretive workings of the cognitive faculties of living beings. Far from requiring technical metaphysics for its articulation, this is a very natural way to view things, and is in fact a very common daily‐life view, in both Western and non‐Western traditions. We did not make the stars, the earth, the trees: they are what they are there outside of us, waiting to be known. And our activities of knowing do not change what they are.

On such a view, the way the human being essentially and universally is will be part of the independent furniture of the universe, something that can in principle be seen and studied independently of any experience of human life and human history. Frequently it is held that a god or gods have this sort of knowledge, and perhaps some wise humans also. This knowledge is usually understood to have normative force. The heavenly account of who we are constrains what we may legitimately seek to be.20 It is this conception of inquiry into the nature of the human that the Marglins are attacking in their critique of what they call Western episteme. They clearly believe it to be a necessary prop to any ethical universalism.

The common objection to this sort of realism is that such extra‐historical and extra‐experiential metaphysical truths are not in fact available. Sometimes this is put sceptically: the independent structure may still be there, but we cannot reliably grasp it. More often, today, doubt is cast on the coherence of the whole realist idea that there is some one determinate structure to the way things are, independent of all human interpretation. This is the objection that non‐philosophers tend to associate with Jacques Derrida's assault on the ‘metaphysics of presence’,21 which he takes to have dominated the entirety of the Western philosophical tradition, and with Richard Rorty's closely related assault on the idea that the knowing mind is, at its best, a ‘mirror of nature’.22 But it actually has a far longer and more complicated history, even within Western philosophy, beginning at least as early as Kant's assault on transcendent metaphysics, and perhaps far earlier, in some of Aristotle's criticisms of Platonism.23 A similar debate was long familiar in classical Indian philosophy, and no doubt it has figured in (p.69) other philosophical traditions as well.24 Contemporary arguments about realism are many and complex, involving, frequently, technical issues in the philosophy of science and the philosophy of language.

The debate about realism appears to be far from over. The central issues continue to be debated with vigour and subtlety, and a wide range of views is currently on the table. On the other hand, the attack on realism has been sufficiently deep and sufficiently sustained that it would appear strategically wise for an ethical and political view that seeks broad support not to rely on the truth of metaphysical realism, if it can defend itself in some other way. If, then, all universalist and humanist conceptions in ethics are required to regard the universal conception of the human being as part of the independent furniture of the world, unmediated by human self‐interpretation and human history, such conceptions do appear to be in some difficulty, and there may well be good reasons to try to do without them.

But universalism does not require such support.25 For universal ideas of the human do arise within history and from human experience, and they can ground themselves in experience. Indeed, if, as the critics of realism allege, we are always dealing with our own interpretations anyhow, they must acknowledge that universal conceptions of the human are prominent and pervasive among such interpretations, hardly to be relegated to the dustbin of metaphysical history along with rare and recondite philosophical entities such as the Platonic forms. As Aristotle so simply puts it, ‘One may observe in one's travels to distant countries the feelings of recognition and affiliation that link every human being to every other human being.’26 Or, as Kwame Anthony Appiah eloquently tells the story of his bicultural childhood, a child who visits one set of grandparents in Ghana and another in rural England, who has a Lebanese uncle and who later, as an adult, has nieces and nephews from more than seven different nations, comes to notice not unbridgeable alien ‘otherness’, but a great deal of human (p.70) commonality, and comes to see the world as a ‘network of points of affinity’.27 Pursuing those affinities, one may accept the conclusions of the critics of realism, while still believing that a universal conception of the human being is both available to ethics and a valuable starting point. I shall be proposing a version of such an account, attempting to identify a group of especially central and basic human functions that ground these affinities.

But such an experiential and historical universalism28 is still vulnerable to some, if not all, of the objections standardly brought against universalism. I therefore need to introduce those objections, and later to test my account against them.

2.1. Neglect of Historical and Cultural Differences

The opposition charges that any attempt to pick out some elements of human life as more fundamental than others, even without appeal to a transhistorical reality, is bound to be insufficiently respectful of actual historical and cultural differences. People, it is claimed, understand human life and humanness in widely different ways: and any attempt to produce a list of the most fundamental properties and functions of human beings is bound to enshrine certain understandings of the human and to demote others. Usually, the objector continues, this takes the form of enshrining the understanding of a dominant group at the expense of minority understandings. This type of objection is frequently made by feminists, and can claim support from many historical examples, in which the human has indeed been defined by focusing on the characteristics of males, as manifested in the definer's culture.

(p.71) It is far from clear what this objection shows. In particular it is far from clear that it supports the idea that we ought to base our ethical norms, instead, on the current preferences and the self‐conceptions of people who are living what the objector herself claims to be lives of deprivation and oppression.29 But it does show at least that the project of choosing one picture of the human over another is fraught with difficulty, political as well as philosophical.

2.2. Neglect of Autonomy

A different objection is presented by liberal opponents of universalism; my relativist opponents, the Marglins, endorse it as well. (Many such objectors, though not, I believe, the Marglins, are themselves willing to give a universal account of the human in at least some ways, holding freedom of choice to be everywhere of central importance.) The objection is that by determining in advance what elements of human life have most importance, the universalist project fails to respect the right of people to choose a plan of life according to their own lights, determining what is central and what is not.30 This way of proceeding is ‘imperialistic’. Such evaluative choices must be left to each citizen. For this reason, politics must refuse itself a determinate theory of the human being and the human good.

2.3. Prejudicial Application

If we operate with a determinate conception of the human being that is meant to have some normative moral and political force, we must also, in applying it, ask which beings we shall take to fall under the concept. And here the objector notes that, all too easily—even if the conception itself is equitably and comprehensively designed—the powerless can be excluded. Aristotle himself, it is pointed out, held that women and slaves were not full‐fledged human beings; and since his politics were based on his view of human functioning, the failure of these beings (in his view) to exhibit the desired mode of functioning contributed to their political exclusion and oppression.

It is, once again, hard to know what this objection is supposed to show. In particular, it is hard to know how, if at all, it is supposed to show that we (p.72) would be better off without such determinate universal concepts. For it could be plausibly argued that it would have been even easier to exclude women and slaves on a whim if one did not have such a concept to contend with. Indeed, this is what I shall be arguing.31 On the other hand, it does show that we need to think not only about getting the concept right but also about getting the right beings admitted under the concept.

Each of these objections has some merit. Many universal conceptions of the human being have been insular in an arrogant way, and neglectful of differences among cultures and ways of life. Some have been neglectful of choice and autonomy. And many have been prejudicially applied. But none of this shows that all such conceptions must fail in one or more of these ways. But at this point I need to advance a definite example of such a conception, in order both to display its merits and to argue that it can in fact answer these charges.

3 A Conception of the Human Being: The Central Human Capabilities

Here, then, is a sketch for an account of the most important functions and capabilities of the human being, in terms of which human life is defined. The basic idea is that we ask ourselves, ‘What are the characteristic activities32 of the human being? What does the human being do, characteristically, as such—and not, say, as a member of a particular group, or a particular local community?’ To put it another way, what are the forms of activity, of doing and being, that constitute the human form of life and distinguish it from other actual or imaginable forms of life, such as the lives of animals and plants, or, on the other hand, of immortal gods as imagined in myths and legends (which frequently have precisely the function of delimiting the human)?33

We can get at this question better if we approach it via two somewhat more concrete questions that we often really ask ourselves. First is a question about personal continuity. We ask ourselves what changes or transitions are compatible with the continued existence of that being as a member of the human kind, and what are not. (Since continued species (p.73) identity seems to be at least necessary for continued personal identity, this is also a question about the necessary conditions for continuing as one and the same individual.) Some functions can fail to be present without threatening our sense that we still have a human being on our hands; the absence of others seems to signal the end of a human life. This question is asked regularly, when we attempt to make medical definitions of death in a situation in which some of the functions of life persist, or to decide, for others or (thinking ahead) for ourselves, whether a certain level of illness or impairment means the end of the life of the being in question.34

The other question is a question about kind inclusion. We recognize other humans as human across many differences of time and place, of custom and appearance. Kwame Anthony Appiah writes about the experience of seeing his heterogenous nieces and nephews playing together, and the term ‘the human future’ naturally occurs to him.35 Much though we may love our dogs and cats, we recognize such scenes as crucially different from scenes of a child playing with a dog or cat. On what do we base these recognitions? We often tell ourselves stories, on the other hand, about anthropomorphic creatures who do not get classified as human, on account of some feature of their form of life and functioning. On what do we base these exclusions? In short, what do we believe must be there, if we are going to acknowledge that a given life is human?36

This inquiry proceeds by examining a wide variety of self‐interpretations of human beings in many times and places. Especially valuable are myths and stories that situate the human being in some way in the universe, between the ‘beasts’ on the one hand and the ‘gods’ on the other; stories that ask what it is to live as a being with certain abilities that set it apart from the rest of the world of nature and with, on the other hand, certain limits that derive from membership in the world of nature. The idea is that people in many different societies share a general outline of such a conception. This is not surprising, since they do recognize one another as members of the same species,37 marry one another, have children together, and so forth—and indeed do tell one another such stories, without much difficulty (p.74) of translation. This convergence gives us some reason for optimism, that if we proceed in this way, using our imaginations, we will have in the end a theory that is not the mere projection of local preferences, but is fully international and a basis for cross‐cultural attunement.

Several important methodological points must now be emphasized:

  1. 1. The procedure through which this account of the human is derived is neither ahistorical nor a priori. It is an attempt to set down a very general record of broadly shared experiences of human beings within history. A related point can be made about the results of the inquiry: they do not claim to be ahistorical or a priori truth, but, rather, an especially deep and continuous sort of experiential and historical truth.

  2. 2. On the other hand, the guiding questions of the inquiry direct it to cross national and temporal boundaries, looking for features that ground recognitions of humanness across these boundaries. Thus we can expect that its results will embody what is continuous rather than rapidly changing, international rather than local.

  3. 3. The account is neither a biological account nor a metaphysical account. (For these reasons I have avoided using the term ‘human nature’, which is usually associated with attempts to describe the human being either from the point of view of an allegedly value‐free science or from the point of view of normative, often theological, metaphysics.) The inquiry pays attention to biology, but as it figures in and shapes human experience. It is an evaluative and, in a broad sense, ethical inquiry. It asks us to evaluate components of lives, asking which ones are so important that we would not call a life human without them. The result of this inquiry is, then, not a list of value‐neutral facts, but a normative conception.38

  4. 4. The account is meant to be both tentative and open‐ended. We allow explicitly for the possibility that we will learn from our encounters with other human societies to recognize things about ourselves that we had not seen before, or even to change in certain ways, according more importance to something we had thought more peripheral. (We may also shift to reach a political consensus.)

  5. 5. The account is not intended to deny that the items it enumerates are to some extent differently constructed by different societies. It claims only that in these areas there is considerable continuity and overlap, sufficient to ground a working political consensus.39

  6. 6. Although the account appeals to consensus in this way, it should be understood that the consensus is acceptable only if it is reached by (p.75) reasonable procedures, where the notion of reasonableness has normative content.40 In this way it is different from consensus as mere overlap.41

  7. 7. The list is heterogeneous: for it contains both limits against which we press and capabilities through which we aspire. This is not surprising, since we began from the intuitive idea of a creature who is both capable and needy.

  8. 8. The concept ‘human being’, as this view understands it, is in one way like the concept ‘person’ as used elsewhere in moral philosophy: that is, it is a normative ethical concept. On the other hand, because of its link with an empirical study of a species‐specific form of life, and with what is most central in such a form of life, it may prove more difficult to withhold from certain beings in an arbitrary way (see Section 7 below). This may commend it to feminists: for the label ‘person’ has frequently been withheld from women, without substantial argument.42

Here then, as a first approximation, is a story about what seems to be part of any life we will count as a human life:

(p.76) 3.1. Level One of the Conception of the Human Being: The Shape of the Human Form of Life

3.1.1. Mortality

All human beings face death and, after a certain age, know that they face it. This fact shapes more or less every other element of human life. Moreover, all human beings have an aversion to death. Although in many circumstances death will be preferred to the available alternatives, the death of a loved one, or the prospect of one's own death, is an occasion for grief and/or fear. If we encountered an immortal anthropomorphic being, or a mortal being who showed no aversion to death and no tendency at all to avoid death, we would judge, in both of these cases, that the form of life was so different from our own that the being could not be acknowledged as human.

3.1.2. The Human Body

We live all our lives in bodies of a certain sort, whose possibilities and vulnerabilities do not as such belong to one human society rather than another. These bodies, similar far more than dissimilar (given the enormous range of possibilities) are our homes, so to speak, opening certain options and denying others, giving us certain needs and also certain possibilities for excellence. The fact that any given human being might have lived anywhere and belonged to any culture is a great part of what grounds our mutual recognitions; this fact, in turn, has a great deal to do with the general humanness of the body, its great distinctness from other bodies. The experience of the body is culturally shaped, to be sure; the importance we ascribe to its various functions is also culturally shaped. But the body itself, not culturally variant in its nutritional and other related requirements, sets limits on what can be experienced and valued, ensuring a great deal of overlap.

There is much disagreement, of course, about how much of human experience is rooted in the body. Here religion and metaphysics enter the picture in a non‐trivial way. Therefore, in keeping with the non‐metaphysical character of the list, I shall include at this point only those features that would be agreed to be bodily even by determined dualists. The more controversial features, such as thinking, perceiving, and emotion, I shall discuss separately, taking no stand on the question of dualism.

  1. 1. Hunger and thirst: the need for food and drink. All human beings need food and drink in order to live; all have comparable, though varying, nutritional requirements. Being in one culture rather than another does not (p.77) make one metabolize food differently. Furthermore, all human beings have appetites that are indices of need. Appetitive experience is to some extent culturally shaped; but we are not surprised to discover much similarity and overlap. Moreover, human beings in general do not wish to be hungry or thirsty (though of course they might choose to fast for some reason). If we discovered someone who really did not experience hunger and thirst at all, or, experiencing them, really did not care about eating and drinking, we would judge that this creature was (in Aristotle's words) ‘far from being a human being’.

  2. 2. Need for shelter. A recurrent theme in myths of humanness is the nakedness of the human being, its relative unprotectedness in the animal world, its susceptibility to heat, cold, and the ravages of the elements. Stories that explore the difference between our needs and those of furry or scaly or otherwise protected creatures remind us how far our life is constituted by the need to find protection through clothing and housing.

  3. 3. Sexual Desire. Though less urgent as a need than the needs for food, drink, and shelter (in the sense that one can live without its satisfaction) sexual need and desire are features of more or less every human life, at least beyond a certain age. It is, and has all along been, a most important basis for the recognition of others different from ourselves as human beings.

  4. 4. Mobility. Human beings are, as the old definition goes, featherless bipeds—that is, creatures whose form of life is in part constituted by the ability to move from place to place in a certain characteristic way, not only through the aid of tools that they have made, but with their very own bodies. Human beings like moving about, and dislike being deprived of mobility. An anthropomorphic being who, without disability, chose never to move from birth to death would be hard to view as human.

3.1.3. Capacity for Pleasure and Pain

Experiences of pain and pleasure are common to all human life (though, once again, both their expression and, to some extent, the experience itself may be culturally shaped). Moreover, the aversion to pain as a fundamental evil is a primitive and, it appears, unlearned part of being a human animal. A society whose members altogether lacked that aversion would surely be judged to be beyond the bounds of humanness.

3.1.4. Cognitive Capability: Perceiving, Imagining, Thinking

All human beings have sense‐perception, the ability to imagine, and the ability to think, making distinctions and ‘reaching out for understanding’.43 And these abilities are regarded as of central importance. It is an open question what sorts of accidents or impediments to individuals in these (p.78) areas will be sufficient for us to judge that the life in question is not really human any longer. But it is safe to say that if we imagine a group of beings whose members totally lack sense‐perception, or totally lack imagination, or totally lack reasoning and thinking, we are not in any of these cases imagining a group of human beings, no matter what they look like.

3.1.5. Early Infant Development

All human beings begin as hungry babies, aware of their own helplessness, experiencing their alternating closeness to and distance from that, and those, on whom they depend. This common structure to early life44—which is clearly shaped in many different ways by different social arrangements—gives rise to a great deal of overlapping experience that is central in the formation of desires, and of complex emotions such as grief, love, and anger. This, in turn, is a major source of our ability to recognize ourselves in the emotional experiences of those whose lives are very different in other respects from our own. If we encountered a group of apparent humans and then discovered that they never had been babies and had never, in consequence, had those experiences of extreme dependency, need, and affection, we would, I think, have to conclude that their form of life was sufficiently different from our own that they could not be considered part of the same kind.

3.1.6. Practical Reason

All human beings participate (or try to) in the planning and managing of their own lives, asking and answering questions about what is good and how one should live. Moreover, they wish to enact their thought in their lives—to be able to choose and evaluate, and to function accordingly. This general capability has many concrete forms, and is related in complex ways to the other capabilities, emotional, imaginative, and intellectual. But a being who altogether lacks this would not be likely to be regarded as fully human, in any society.

3.1.7. Affiliation With Other Human Beings

All human beings recognize and feel some sense of affiliation and concern for other human beings. Moreover, we value the form of life that is constituted by these recognitions and affiliations. We live with and in relation to others, and regard a life not lived in affiliation with others to be a life not worth the living. (Here I would really wish, with Aristotle, to spell things out further. We define ourselves in terms of at least two types of affiliation: intimate family and/or personal relations, and social or civic relations.)

(p.79) 3.1.8. Relatedness to Other Species and to Nature

Human beings recognize that they are not the only living things in their world: that they are animals living alongside other animals, and also alongside plants, in a universe that, as a complex interlocking order, both supports and limits them. We are dependent upon that order in countless ways; and we also sense that we owe that order some respect and concern, however much we may differ about exactly what we owe, to whom, and on what basis. Again, a creature who treated animals exactly like stones and could not be brought to see any difference would probably be regarded as too strange to be human. So too would a creature who did not in any way respond to the natural world.

3.1.9. Humour and Play

Human life, wherever it is lived, makes room for recreation and laughter. The forms play takes are enormously varied—and yet we recognize other humans, across cultural barriers, as the animals who laugh. Laughter and play are frequently among the deepest and also the first modes of our mutual recognition. Inability to play or laugh is taken, correctly, as a sign of deep disturbance in a child; if it proves permanent we will doubt whether the child is capable of leading a fully human life. An entire society that lacked this ability would seem to us both terribly strange and terribly frightening.

3.1.10. Separateness

However much we live with and for others, we are, each of us, ‘one in number’,45 proceeding on a separate path through the world from birth to death. Each person feels only his or her own pain and not anyone else's. Each person dies without entailing logically the death of anyone else. When one person walks across the room, no other person follows automatically. When we count the number of human beings in a room, we have no difficulty figuring out where one begins and the other ends. These obvious facts need stating, since they might have been otherwise. We should bear them in mind when we hear talk about the absence of individualism in certain societies. Even the most intense forms of human interaction, for example sexual experience, are experiences of responsiveness, not of fusion. If fusion is made the goal, the result is bound to be disappointment.

3.1.11. Strong Separateness

Because of separateness, each human life has, so to speak, its own peculiar context and surroundings—objects, places, a history, particular friendships, locations, sexual ties—that are not exactly the same as those of anyone else, and in terms of which the person to some extent identifies herself. Though societies vary a great deal in the degree (p.80) and type of strong separateness that they permit and foster, there is no life yet known that really does (as Plato wished) fail to use the words ‘mine’ and ‘not mine’ in some personal and non‐shared way. What I use, live in, respond to, I use, live in, respond to from my own separate existence. And on the whole, human beings recognize one another as beings who wish to have at least some separateness of context, a little space to move around in, some special items to use or love.

This is a working list. It is put out to generate debate. It has done so and will continue to do so, and it will be revised accordingly.

As I have said, the list is composed ot two different sorts of items; limits and capabilities. As far as capabilities go, to call them parts of humanness is to make a very basic sort of evaluation. It is to say that a life without this item would be too lacking, too impoverished, to be human at all. Obviously, then, it could not be a good human life. So this list of capabilities is a ground‐floor or minimal conception of the good. (In the sense that it does not fully determine the choice of a way of life, but simply regulates the parameters of what can be chosen, it plays, however, the role traditionally played in liberal political theory by a conception of the right.)46

With the limits, things are more complicated. In selecting the limits for attention, we have, once again, made a basic sort of evaluation, saying that these things are so important that life would not be human without them. But what we have said is that human life, in its general form, consists of the awareness of these limits plus a struggle against them. Humans do not wish to be hungry, to feel pain, to die. (Separateness is highly complex, both a limit and a capability. Much the same is true of many of the limits implied by the shape and the capacities of the body.) On the other hand, we cannot assume that the correct evaluative conclusion to draw is that we should try as hard as possible to get rid of the limit altogether. It is characteristic of human life to prefer recurrent hunger plus eating to a life with neither hunger nor eating; to prefer sexual desire and its satisfaction to a life with neither desire nor satisfaction. Even where death is concerned, the desire for immortality, which many human beings certainly have, is a peculiar desire: for it is not clear that the wish to lose one's finitude completely is a desire that one can coherently entertain for oneself or for someone one loves. It seems to be a wish for a transition to a way of life so wholly different, with such different values and ends, that it seems that the identity of the individual will not be preserved. So the evaluative conclusion, in mapping out a ground‐floor conception of the good (saying what functioning is necessary for a life to be human) will have to be expressed with much caution, clearly, in terms of what would be a humanly good way of countering the limitation.

(p.81) 4 The Two Thresholds

Things now get very complicated. For we want to describe two distinct thresholds: a threshold of capability to function beneath which a life will be so impoverished that it will not be human at all; and a somewhat higher threshold, beneath which those characteristic functions are available in such a reduced way that, though we may judge the form of life a human one, we will not think it a good human life. The latter threshold is the one that will eventually concern us when we turn to public policy: for we don't want societies to make their citizens capable of the bare minimum. My view holds, with Aristotle, that a good political arrangement is one ‘in accordance with which anyone whatsoever might do well and live a flourishing life’.47

These are clearly, in many areas, two distinct thresholds, requiring distinct levels of resource and opportunity. One may be alive without being well nourished. As Marx observed, one may be able to use one's senses without being able to use them in a fully human way. And yet there is need for caution here. For in many cases the move from human life to good human life is supplied by the citizen's own powers of choice and self‐definition, in such a way that once society places them above the first threshold, moving above the second is more or less up to them. This is especially likely to be so, I think, in areas such as affiliation and practical reasoning, where in many cases once social institutions permit a child to cross the first threshold its own choices will be central in raising it above the second. (This is not always so, however: for certain social conditions, for example certain mindless forms of labour or, we may add, traditional hierarchical gender relations, may impede the flourishing of affiliation and practical reason, while not stamping it out entirely.) On the other hand, it is clear that where bodily health and nutrition, for example, are concerned, there is a considerable difference between the two thresholds, and a difference that is standardly made by resources over which individuals do not have full control. It would then be the concern of quality‐of‐life assessment to ask whether all citizens are capable, not just of the bare minimum, but of good life in these areas. Clearly there is a continuum here. Nor will it in practise be at all easy to say where the upper threshold, especially, should be located.

I shall not say much about the first threshold, but shall illustrate it by a few examples. What is an existence that is so impoverished that it cannot properly be called a human life? Here we should count, I believe, many forms of existence that take place at the end of a human life—all those in which the being that survives has irretrievably lost sensation and consciousness (in what is called a ‘permanent vegetative condition’); and also, I (p.82) would hold, some that fall short of this, but in which the capacity to recognize loved ones, to think and to reason, has irreversibly decayed beyond a certain point. I would include the extreme absence of ability to engage in practical reasoning that is often the outcome of the notorious frontal lobotomy. I would also include an absence of mobility so severe that it makes speech, as well as movement from place to place, impossible.

It follows from this that certain severely damaged infants are not human ever, even if born from two human parents: again, those with global and total sensory incapacity and/or no consciousness or thought; also, I think, those with no ability at all to recognize or relate to others. (This of course tells us nothing about what we owe them morally, it just separates that question from moral questions about human beings.)48

Again, we notice the evaluative character of these threshold judgements. The fact that a person who has lost her arms cannot play a piano does not make us judge that she no longer lives a human life; had she lost the capacity to think and remember, or to form affectionate relationships, it would have been a different matter.

Many such disasters are not to be blamed on social arrangements, and in those cases the first threshold has no political implications. But many are, where bad nutrition and health care enter in. The role of society is even more evident if we think of a more controversial group of first‐threshold cases, in which the non‐human outcome was environmentally caused: the rare cases of children who have grown up outside a human community, or in a severely dysfunctional home, and utterly lack language and reason, or lack social abilities in an extreme and irreversible way. We can focus the political question more productively, however, if we now turn from the question of mere human life to the question of good life, the level we would really like to see a human being attain.

Here, as the next level of the conception of the human being, I shall now specify certain basic functional capabilities at which societies should aim for their citizens, and which quality of life measurements should measure. In other words, this will be an account of the second threshold—although in some areas it may coincide, for the reasons I have given, with the first: once one is capable of human functioning in this area one is also capable, with some further effort and care, of good functioning. I introduce this list as a list of capabilities to function, rather than of actual functionings, since I (p.83) shall argue that capability, not actual functioning, should be the goal of public policy.

4.1. Level 2 of the Conception of the Human Being: Basic Human Functional Capabilities

  1. 1. Being able to live to the end of a human life of normal length,49 not dying prematurely, or before one's life is so reduced as to be not worth living.

  2. 2. Being able to have good health; to be adequately nourished;50 to have adequate shelter;51 having opportunities for sexual satisfaction, and for choice in matters of reproduction;52 being able to move from place to place.

  3. 3. Being able to avoid unnecessary and non‐beneficial pain, so far as possible, and to have pleasurable experiences.

  4. 4. Being able to use the senses; being able to imagine, to think, and to reason—and to do these things in a way informed and cultivated by an (p.84) adequate education, including, but by no means limited to, literacy and basic mathematical and scientific training.53 Being able to use imagination and thought in connection with experiencing and producing spiritually enriching materials and events of one's own choice; religious, literary, musical, and so forth. I believe that the protection of this capability requires not only the provision of education, but also legal guarantees of freedom of expression with respect to both political and artistic speech, and of freedom of religious exercise.

  5. 5. Being able to have attachments to things and persons outside ourselves; to love those who love and care for us, to grieve at their absence; in general, to love, to grieve, to experience longing and gratitude.54 Supporting this capability means supporting forms of human association that can be shown to be crucial in their development.55

  6. 6. Being able to form a conception of the good and to engage in critical reflection about the planning of one's own life. This includes, today, being able to seek employment outside the home and to participate in political life.

  7. 7. Being able to live for and to others, to recognize and show concern for other human beings, to engage in various forms of social interaction; to be able to imagine the situation of another and to have compassion for that situation; to have the capability for both justice and friendship. Protecting this capability means, once again, protecting institutions that constitute such forms of affiliation, and also protecting the freedoms of assembly and political speech.

  8. 8. Being able to live with concern for and in relation to animals, plants, and the world of nature.

  9. 9. Being able to laugh, to play, to enjoy recreational activities.

  10. (p.85) 10. Being able to live one's own life and nobody else's. This means having certain guarantees of non‐interference with certain choices that are especially personal and definitive of selfhood, such as choices regarding marriage, childbearing, sexual expression, speech, and employment.

  11. 10a. Being able to live one's own life in one's own surroundings and context. This means guarantees of freedom of association and of freedom from unwarranted search and seizure; it also means a certain sort of guarantee of the integrity of personal property, though this guarantee may be limited in various ways by the demands of social equality, and is always up for negotiation in connection with the interpretation of the other capabilities, since personal property, unlike personal liberty, is a tool of human functioning rather than an end in itself.

My claim is that a life that lacks any one of these capabilities, no matter what else it has, will fall short of being a good human life. So it would be reasonable to take these things as a focus for concern, in assessing the quality of life in a country and asking about the role of public policy in meeting human needs. The list is certainly general—and this is deliberate, in order to leave room for plural specification and also for further negotiation. But I claim that it does, rather like a set of constitutional guarantees, offer real guidance in the ongoing historical process of further refinement and specification, and far more accurate guidance than that offered by the focus on utility, or even on resources.

A few comments are in order about the relationship of this version of the list to other versions I have published previously. First, taking some lessons from the Human Development Report, it is considerably more specific about matters such as education and work, so as to give the development theorist something concrete to measure. Secondly, it is far more explicitly concerned with guarantees of personal liberty of expression, reproductive choice, and religion.56 This was not only called for in general, but called forth by the attempt to articulate the specific requisites of equal female capability.57 Thirdly, in accordance with its commitment to the distinction between ends and means, it understands ‘property rights’ as instrumental to other human capabilities,58 and therefore to a certain extent, as up for negotiation in general social planning.

The list is, emphatically, a list of separate components. We cannot satisfy the need for one of them by giving a larger amount of another. All are of central importance and all are distinct in quality. This limits the trade‐offs (p.86) that it will be reasonable to make, and thus limits the applicability of quantitative cost‐benefit analysis. At the same time, the items on the list are related to one another in many complex ways. For example our characteristic mode of nutrition, unlike that of sponges, requires moving from here to there. And we do whatever we do as separate beings, tracing distinct paths through space and time. Notice that reproductive choices involve both sexual capability and issues of separateness, and bind the two together in a deep and complex way.

A further comment is in order, concerning the relationship of this threshold list to an account of human equality. A commitment to bringing all human beings across a certain threshold of capability to choose represents a certain sort of commitment to equality: for the view treats all persons as equal bearers of human claims, no matter where they are starting from in terms of circumstances, special talents, wealth, gender, or race. On the other hand, I have said nothing so far about how one should regard inequalities that persist once the threshold level has been attained for all persons. To some extent I feel this would be premature, since the threshold level has so rarely been attained for the complete capability set. On the other hand, one can imagine a situation—perhaps it could be that of the USA or Japan, given certain large changes in health support here, or educational distribution there, that would meet threshold conditions and still exhibit inequalities of attainment between the genders or the races. We have two choices here: either to argue that this situation actually contains capability failure after all; or to grant that the capability view needs to be supplemented by an independent theory of equality. I am not yet certain what I want to say about this, but I am inclined to the first alternative, since I think that gender inequality of the sort one sees in a prosperous nation does none the less push the subordinated racial or gender group beneath an acceptable threshold of autonomy, dignity, and emotional well being. Indeed, subordination is itself a kind of capability failure, a failure to attain complete personhood. So I am inclined to say that, properly fleshed out, the second threshold would be incompatible with systematic subordination of one group to another.

5 The Role of the Conception in Development Policy

My claim is that we urgently need a conception of the human being and human functioning in public policy. If we try to do without this sort of guidance when we ask how goods, resources, and opportunities should be distributed, we reject guidance that is, I think, superior to that offered by any of the other guides currently available.

I shall focus here on the area of most concern to our project: the assessment of the quality of life in a developing country, with special (p.87) attention to the lives of women. For the time being, I shall take the nation state as my basic unit, and the question I shall ask is, ‘How is the nation doing, with respect to the quality of life of its citizens?’ In other words, I shall be asking the sort of question asked by the UN Human Development Report. I shall not propose a general theory about how the needs revealed by such an assessment should be met: whether by centralized government planning, for example, or through a system of incentives, and whether through direct subsidies or through the provision of opportunities for employment. Nor shall I ask what responsibilities richer nations have to poorer nations, in ensuring that the needs of all human beings are met the world over. That is an urgent question, and it must at a later date be confronted. For now, however, I shall focus on the correct understanding of the goal, where each separate nation is concerned.

The basic claim I wish to make—concurring with Amartya Sen—is that the central goal of public planning should be the capabilities of citizens to perform various important functions. The questions that should be asked when assessing quality of life in a country are (and of course this is a central part of assessing the quality of its political arrangements) ‘How well have the people of the country been enabled to perform the central human functions?’ and, ‘Have they been put in a position of mere human subsistence with respect to the functions, or have they been enabled to live well?’ In other words, we ask where the people are, with respect to the second list. And we focus on getting as many people as possible above the second threshold, with respect to the interlocking set of capabilities enumerated by that list.59 Naturally, the determination of whether certain individuals and groups are across the threshold is only as precise a matter as the determination of the threshold; and I have left things deliberately somewhat open‐ended at this point, in keeping with the procedures of the Human Development Report, believing that the best way to work toward a more precise determination is to allow the community of nations to hammer it out after an extended comparative inquiry, of the sort the report makes possible. Again, we will have to answer various questions about the costs we are willing to pay to get all citizens above the threshold, as opposed to leaving a small number below and allowing the rest a considerably above‐threshold life quality. Here my claim is that capability‐equality, in the sense of moving all above the threshold, should be taken as the central goal. As with Rawls's Difference Principle, so here: inequalities in distribution above the threshold should be (p.88) tolerated only if they move more people across it;60 once all are across, societies are to a great extent free to choose the other goals that they wish to pursue.

The basic intuition from which the capability approach starts, in the political arena, is that human capabilities exert a moral claim that they should be developed. Human beings are creatures such that, provided with the right educational and material support, they can become fully capable of the major human functions, can cross the first and second thresholds. That is, they are creatures with certain lower‐level capabilities (which I have elsewhere called ‘basic capabilities’)61 to perform the functions in question. When these capabilities are deprived of the nourishment that would transform them into the high‐level capabilities that figure on my list, they are fruitless, cut off, in some way but a shadow of themselves. They are like actors who never get to go on the stage, or a musical score that is never performed. Their very being makes forward reference to functioning. Thus if functioning never arrives on the scene they are hardly even what they are. This may sound like a metaphysical idea, and in a sense it is (in that it is an idea discussed in Aristotle's Metaphysics). But that does not mean that it is not a basic and pervasive empirical idea, an idea that underwrites many of our daily practises and judgements in many times and places. I claim that just as we hold that a child who dies before getting to maturity has died especially tragically—for her activities of growth and preparation for adult activity now have lost their point—so too with capability and functioning more generally: we believe that certain basic and central human endowments have a claim to be assisted in developing, and exert that claim on others, and especially, as Aristotle saw, on government. We shall see the work this consideration can do in arguments for women's equality. I think it is the underlying basis, in the Western philosophical tradition, for many notions of human rights. I suggest, then, that in thinking of political planning we begin from this notion, thinking of the basic capabilities of human beings as needs for functioning, which give rise to correlated political duties.

(p.89) There is, then, an empirical basis for the determination that a certain being is one of the ones to which our normative conception and its associated duties applies. It is the gap between potential humanness and its full realization that exerts a moral claim. If the worker described by Marx as not capable of a truly human use of his senses62 had really been a non‐human animal, the fact that he was given a form of life suited to such an animal would not be a tragedy. If women were really turtles, the fact that being a woman is not yet a way of being a human being would not be, as it is, an outrage. There is, of course, enormous potential for abuse in determining who has these basic capabilities. The history of IQ testing is just one chapter in an inglorious saga of prejudiced capability‐testing that goes back at least to the Noble Lie of Plato's Republic. Therefore we should, I think, proceed as if every offspring of two human parents has the basic capabilities, unless and until long experience with the individual has convinced us that damage to that individual's condition is so great that it could never in any way arrive at the higher capability level.

The political and economic application of this approach is evident in a variety of areas. Amartya Sen has developed a number of its concrete implications in the areas of welfare and development economics, and has focused particularly on its application to the assessment of women's quality of life.63 With his advice, the UN Human Development Reports have begun to gather information and to rank nations in accordance with the type of plural‐valued capability‐focused measuring the approach suggests. In a closely related study, Iftekhar Hossein has used the approach to give an account of poverty as capability failure.64 Independently, a very similar approach has been developed by Finnish and Swedish social scientists, above all Erik Allardt and Robert Erikson.65 Wishing to develop ways of gathering information about how their people are doing that would be more sensitive and informationally complete than polls based on ideas of utility, they worked out lists of the basic human capabilities for functioning, and then examined the performance of various groups in the population—above all women and minorities—in these terms, thus anticipating the procedures of the Human Development Report, which devotes a great deal of attention to gender differences, urban‐rural differences, and so forth.

(p.90) The ‘capabilities approach’ has clear advantages over other current approaches to quality of life assessment. Assessment that uses GNP per capita as its sole measure fails to concern itself with the distribution of resources, and thus can give high marks to countries with enormous inequalities. Nor does this approach examine other human goods that are not reliably correlated with the presence of resources: infant mortality, for example, or access to education, or the quality of racial and gender relations, or the presence or absence of political freedoms. The Human Development Report for 1993 informs us, for example, that the United Arab Emirates has Real GNP per capita of $16,753—tenth highest in the world, higher, for example, than Norway or Australia—while overall, in the aggregation of all the indicators of life quality, it ranks only sixty‐seventh in the world (out of 173 nations measured). Its adult literacy rate is 55%, far lower than any of the 66 countries generally ahead of it, and also than many generally below it. (Both Norway and Australia have adult literacy of 99%.) The maternal mortality rate of 130 per 100,000 live births is comparatively high. The proportion of women progressing beyond secondary education is very low, and only 6% of the labour force is female (as opposed, for example, to 42% in Seychelles, 35% in Brazil, 43% in China, 47% in Viet Nam, 26% in India, and 20% in Nigeria). In fact, in all the world only Algeria (4%) has a lower proportion of females in the labour force, only Iraq (6%) ties it, and only Qatar (7%), Saudi Arabia (7%), Libya (9%), Jordan (10%), Pakistan (11%), Bangladesh (7%), and Afghanistan (8%) come close. Evidence links female wage‐earning outside the home strongly to female health care and life‐expectancy.66 And in fact, we find that the ratio of females to males in the United Arab Emirates is the amazing 48:100, lowest in all the world. If this is discounted as employment related, we may pursue the other countries in our low external employment comparison class. The ratio of females to males in nations in which there is no reason to suppose sexual discrimination in nourishment and health care is, Sen has shown, about 106:100 in Europe and North America—or, if we focus only on the developing world, taking sub‐Saharan Africa as our ‘norm’, 102:100. In Qatar it is 60:100, in Saudi Arabia 84, in Libya 91, in Jordan 95, in Pakistan 92, in Bangladesh 94, in Afghanistan 94.

These are some of the numbers that we start noticing if we focus on capabilities and functioning, rather than simply on GNP. They are essential to the understanding of how women are doing. In fact, they are the numbers from which Sen's graphic statistics regarding ‘missing women’ emerge. (The number of ‘missing women’ is the number of extra women who would be in a given country if that country had the same sex ratio as sub‐Saharan Africa.) They strongly support Martha Chen's argument that the right to work is a right basic to the lives of women not only in itself, but (p.91) for its impact on other basic capabilities and functionings. Saleha Begum's employment led to better nutritional and health status for herself and, indeed, her children and family. Metha Bai may soon become one of the statistics from which the number of missing women is made.

Would other available approaches have done the job as well? The common approach that measures quality of life in terms of utility—polling people concerning the satisfaction of their preferences—would have missed the obvious fact that desires and subjective preferences are not always reliable indicators of what a person really needs. Preferences, as Amartya Sen's work has repeatedly shown, are highly malleable.67 The rich and pampered easily become accustomed to their luxury, and view with pain and frustration a life in which they are treated just like everyone else. Males are a special case of this: we do not need to go abroad to know that males frequently resent a situation in which they are asked to share child care and domestic responsibilities on an equal basis.68 The poor and deprived frequently adjust their expectations and aspirations to the low level of life they have known. Thus they may not demand more education, better health care. Like the women described in Sen's account of health surveys in India, they may not even know what it is to feel healthy.69 Like the rural Bangladeshi women so vividly described in Martha Chen's A Quiet Revolution,70 they may not even know what it means to have the advantages of education. We may imagine that many women in the countries I have mentioned would not fight, as Seleha Begum did, for participation in the workforce; nor would they be aware of the high correlation between work outside the home and other advantages. As Sen argues, they may have fully internalized the ideas behind the traditional system of discrimination, and may view their deprivation as ‘natural’. Thus if we rely on utility as our measure of life quality, we most often will get results that support the status quo and oppose radical change.71

(p.92) If these criticisms apply to approaches that focus on utility in general, they apply all the more pointedly to the sort of local‐tradition relativism espoused by the Marglins, in which the measure of quality of life will be the satisfaction of a certain group of preferences, namely the traditional ones of a given culture. Indeed, it is illuminating to consider how close, in its renunciation of critical normative argument, the Marglin approach is to the prevailing economic approaches of which it presents itself as a radical critique. A preference‐based approach that gives priority to the preferences of traditional culture is likely to be especially subversive of the quality of life of women who have been on the whole badly treated by prevailing traditional norms. And one can see this clearly in the Marglins' own examples. For menstruation taboos impose severe restrictions on women's power to form a plan of life and to execute the plan they have chosen. They are members of the same family of traditional attitudes about women and the workplace that made it difficult for Saleha Begum to support herself and her family, that make it impossible for Metha Bai to sustain the basic functions of life. And the Japanese husband who allegedly renounces freedom of choice actually enhances it, in the ways that matter, by asking the woman to look after the boring details of life. One can sympathize with many of the Marglins' goals—respect for diversity, desire to preserve aspects of traditional life that appear to be rich in spiritual and artistic value—without agreeing that extreme relativism of the sort they endorse is the best way to pursue these concerns.

As for liberal approaches that aim at equality in the distribution of certain basic resources, these have related problems, since these, too, refuse to take a stand on the ends to which the resources are means.72 Wealth and income are not good in their own right; they are good only insofar as they promote human functioning. Secondly, human beings have widely varying needs for resources, and any adequate definition of who is ‘better off’ and ‘worse off’ must reflect that fact.73 Women who have traditionally not been educated, for example, may well require more of the relevant resources to attain the same capability level: that is why, in the case (p.93) discussed by Martha Chen, the Bangladesh Rural Advancement Committee created a special female literacy programme, rather than a programme that distributed equal resources to all. Thirdly, by defining being ‘well‐off’ in terms of possessions alone, the liberal fails to go deep enough in imagining the impediments to functioning that are actually present in many lives—in their conditions of labour or exclusion from labour, for example, in their frequently unequal family responsibilities, in the obstacles to self‐realization imposed by traditional norms and values.74 The stories of Saleha Begum and Metha Bai are vivid examples of such unequal obstacles. No right‐to‐work effort, and no expenditure of resources in that connection, were necessary in order to make men capable of working in the fields in Bangladesh. No male of Metha Bai's caste would have to overcome threats of physical violence in order to go out of the house to work for life‐sustaining food.

6 Answering the Objections: Human Functioning and Pluralism

I have commended the human‐function view by contrast to its rivals on the development scene. But I must now try to show how it can answer the objections I described earlier.

Concerning neglect of historical and cultural difference, I can begin by insisting that this normative conception of human capability and functioning is general, and in a sense vague, for precisely this reason. The list claims to have identified in a very general way components that are fundamental to any human life. But it allows in its very design for the possibility of multiple specifications of each of the components. This is so in several different ways. First, the constitutive circumstances of human life, while broadly shared, are themselves realized in different forms in different societies. The fear of death, the love of play, relationships of friendship and affiliation with others, even the experience of the bodily appetites never turn up in simply the vague and general form in which we have introduced them here, but always in some specific and historically rich cultural realization, which can profoundly shape not only the conceptions used by the citizens in these areas, but also their experiences themselves. Nonetheless, we do have in these areas of our common humanity sufficient overlap to sustain a general conversation, focusing on our common problems and prospects. (p.94) And sometimes the common conversation will permit us to criticize some conceptions of the grounding experiences themselves, as at odds with other things human beings want to do and to be.

When we are choosing a conception of good functioning with respect to these circumstances, we can expect an even greater degree of plurality to become evident. Here the approach wants to retain plurality in two significantly different ways: what I may call the way of plural specification, and what I may call the way of local specification.

Plural specification means what its name implies. Public policy, while using a determinate conception of the good at a high level of generality, leaves a great deal of latitude for citizens to specify each of the components more concretely, and with much variety, in accordance with local traditions, or individual tastes. Many concrete forms of life, in many different places and circumstances, display functioning in accordance with all the major capabilities.

As for local specification: good public reasoning, I believe and have argued, is always done, when well done, with a rich sensitivity to the concrete context, to the characters of the agents and their social situation. This means that in addition to the pluralism I have just described, the Aristotelian needs to consider a different sort of plural specification of the good. For sometimes what is a good way of promoting education in one part of the world will be completely ineffectual in another. Forms of affiliation that flourish in one community may prove impossible to sustain in another. In such cases, the Aristotelian must aim at some concrete specification of the general list that suits, and develops out of, the local conditions. This will always most reasonably be done in a participatory dialogue75 with those who are most deeply immersed in those conditions. For though Aristotelianism does not hesitate to criticize tradition where tradition perpetrates injustice or oppression, it also does not believe in saying anything at all without rich and full information, gathered not so much from detached study as from the voices of those who live the ways of life in question. Martha Chen's work, both here and in her book, gives an excellent example of how such sensitivity to the local may be combined with a conviction that the central values on the list are worth pursuing even when tradition has not endorsed them.

The liberal charges the capability approach with neglect of autonomy, arguing that any such determinate conception removes from the citizens the chance to make their own choices about the good life. This is a complicated issue: three points can be stressed. First, the list is a list of capabilities, not a list of actual functions, precisely because the conception is designed to leave room for choice. Government is not directed to push citizens into acting in certain valued ways; instead, it is directed to make sure that all (p.95) human beings have the necessary resources and conditions for acting in those ways. It leaves the choice up to them. A person with plenty of food can always choose to fast. A person who has been given the capability for sexual expression can always choose celibacy. The person who has access to subsidized education can always decide to do something else instead. By making opportunities available, government enhances, and does not remove, choice.76 It will not always be easy to say at what point someone is really capable of making a choice, especially in areas where there are severe traditional obstacles to functioning. Sometimes our best strategy may well be to look at actual functioning and infer negative capability (tentatively) from its absence.77 But the conceptual distinction remains very important.

Secondly, this respect for choice is built deeply into the list itself, in the architectonic role it gives to practical reasoning. One of the most central capabilities promoted by the conception will be the capability of choice itself.78 We should note that the major liberal view in this area (that of John Rawls) agrees with our approach in just this area. For Rawls insists that satisfactions that are not the outgrowth of one's very own choices have no moral worth; and he conceives of the two moral powers (analogous to our practical reasoning), and of sociability (corresponding to our affiliation) as built into the definition of the parties in the original position, and thus as necessary constraints on any outcome they will select.79

Finally, the capability view insists that choice is not pure spontaneity, flourishing independent of material and social conditions. If one cares about autonomy, then one must care about the rest of the form of life that supports it, and the material conditions that enable one to live that form of life. Thus the approach claims that its own comprehensive concern with flourishing across all areas of life is a better way of promoting choice than is the liberal's narrower concern with spontaneity alone, which sometimes tolerates situations in which individuals are in other ways cut off from the fully human use of their faculties.

I turn now to the objection about application; it raises especially delicate questions where women are concerned.

(p.96) 7 Who Gets Included? Women as Human Beings

In a now well‐known remark, which I cite here as an epigraph, the feminist lawyer Catharine MacKinnon claimed that ‘being a woman is not yet a way of being a human being.’80 This means, I think, that most traditional ways of categorizing and valuing women have not accorded them full membership in the human species, as that species is generally defined. MacKinnon is no doubt thinking in particular of the frequent denials to women of the rational nature that is taken to be a central part of what it is to be human. It is sobering to remind oneself that quite a few leading philosophers, including Aristotle and Rousseau, the ‘fathers’ (certainly not mothers) of my idea, did deny women full membership in human functioning as they understood that notion. If this is so, one might well ask, of what use is it really to identify a set of central human capabilities? For the basic (lower‐level) capacity to develop these can always be denied to women, even by those who grant their centrality. Does this problem show that the human function idea is either hopelessly in league with patriarchy or, at best, impotent as a tool for justice?

I believe that it does not. For if we examine the history of these denials we see, I believe, the great power of the conception of the human as a source of moral claims. Acknowledging the other person as a member of the very same kind would have generated a sense of affiliation and a set of moral and educational duties. That is why, to those bent on shoring up their own power, the stratagem of splitting the other off from one's own species seems so urgent and so seductive. But to deny humanness to beings with whom one lives in conversation and interaction is a fragile sort of self‐deceptive stratagem, vulnerable to sustained and consistent reflection, and also to experiences that cut through self‐deceptive rationalization.81 Any moral conception can be withheld, out of ambition or hatred or shame. But the conception of the human being, spelled out, as here, in a roughly determinate way, in terms of circumstances of life and functions in these circumstances, seems much harder to withhold than other conceptions that have been made the basis for ethics—‘rational being’, for example, or (as I have suggested) ‘person’.

To illustrate this point, I now turn to the earliest argument known to me in the Western philosophical tradition that uses a conception of the human being for feminist ends. It is not the first feminist argument in the Western tradition: for Plato's Republic precedes (and influences) it.82 But Plato's (p.97) argument in favour of equal education for women is heavily qualified by his élitism with respect to all functions for all human beings; thus it is able to generate only élitist conclusions for males and females alike. Platonic justice is not the ‘humanist justice’ of Susan Okin's powerful phrase. The argument I have in mind is, instead, the first argument of the Roman Stoic thinker Musonius Rufus in his brief treatise, ‘That Women Too Should Do Philosophy’, written in the first century AD.83 This argument is all the more interesting in that it, in effect, uses Aristotelian concepts to correct Aristotle's mistake about women—showing, I think, that an Aristotelian who is both internally consistent and honest about the evidence cannot avoid the egalitarian normative conclusion that women, as much as men, should receive a higher education (for that is in effect what is meant by doing philosophy).84

The argument has a tacit premise. It is that—at least with respect to certain central functions of the human being—the presence in a creature of a basic (untrained, lower‐level) capability to perform the functions in question, given suitable support and education, exerts a claim on society that those capabilities should be developed to the point at which the person is fully capable of choosing the functions in question. This premise needed no argument in the philosophical culture of Greco‐Roman antiquity, since that moral claim is more or less taken to be implicit in the notion of capability itself. I have tried to give it intuitive support in the argument of this paper.

The argument itself now follows with a truly radical simplicity. Its second premise consists of an appeal to the experience of the imaginary recalcitrant male interlocutor. Women, he is asked to concede on the basis of experience, do in fact have the basic capabilities to perform a wide variety of the most important human functions. They have the five senses. They have the same number of bodily parts, implying similar functional possibilities in that sphere. They have the ability to think and reason, just as males do. And, finally, they have responsiveness to ethical distinctions, making (whether well or badly) distinctions between the good and the bad. Some time is then spent establishing a third premise: that ‘higher education’ of (p.98) the sort offered by the Stoic ideal of liberal education, is necessary for the full development of the perceptual, intellectual, and moral capabilities. Conclusion: women, like men, should have this education.

The puzzle, for us, is the second premise. Why does the interlocutor accept it? We see from the surrounding material that the interlocutor is a husband who interacts with his wife in a number of areas of life that are explicitly enumerated: planning and managing a household (where she is the one who manages most of the daily business); having and raising children (where he observes, or imagines, her in labour, enduring risk and pain for the sake of the family and, later, caring for and educating the child); having sexual relations with him, and refusing to have sex with others; having a real friendship with him, based on common contemporary ideas of ‘sharing life together’;85 deciding how to treat the people around her; being fair, for example, to the household staff; and, finally, confronting all the dangers and the moral ambiguities of the politics of first century AD Rome—refusing to capitulate, he says, to the unjust demands of a tyrant. In all of these operations of life, the argument seems to be, he tacitly acknowledges, in fact strongly relies upon, his wife's capability to engage in practical reasoning and ethical distinction making. Indeed, he is depicted as someone who would like these things done well—for he wants his wife not to reason badly when political life gets tough, or to treat the servants with cruelty, or to botch the education of the children. So in his daily life he acknowledges her humanity, her possession of the basic (lower‐level) capabilities for fully human functioning. How, then, Musonius reasonably asks him, can he consistently deny her what would be necessary in order to develop and fulfil that humanity?

This, I believe, is an impressively radical argument. And it led to (or reflected) a social situation that marked a high point for women in the Western tradition for thousands of years since and to come.86 We do not need to show that the views of Musonius on women were perfect in all respects; in many ways they were not. But his argument shows, I believe, the power of a universal conception of the human being in claims of justice for women. For the interlocutor might have refused to acknowledge that his wife was a ‘person’: it was to some extent up to him to define that rather refined and elusive concept. He could not fail to acknowledge that she was a human being, with the basic capability for the functions in question. For he had acknowledged that already, in his daily life.

(p.99) 8 Women and Men: Two Norms or One?

But should there be a single norm of human functioning? It has often been argued, in both non‐Western and Western traditions, that there should be two different standards of human functioning and capability, corresponding to the different ‘natures’ of the male and the female. Usually these overlap in the areas of bodily health, mobility, and perception, but differ sharply in the areas of practical reason and affiliation. Most commonly, citizenship, public activity, and full practical autonomy are assigned to males, care for home and family to females. We must now confront the claims of this position.

Those who recognize separate spheres of functioning for males and females have taken up two importantly‐different positions, which we need to be careful to distinguish. The first, which I shall call Position A, assigns to both males and females the same general normative list of functions, but suggests that males and females should exercise these functions in different spheres of life. The second, which I shall call Position B, insists that the list of functions, even at a high level of generality, should be different. (It is B rather than A that is usually associated with the claim that males and females have different ‘natures’.)

Position A is compatible with a serious interest in equality and in gender justice. For what it says, after all, is that males and females have the same basic needs for capability development and should get what they need. It is determined to ensure that both get to the higher (developed) level of capability with respect to all the central functions. It simply holds that this can (and perhaps should) be done in separate spheres. It is a kind of gender‐based local specification. A is, after all, the position of Musonius, who holds that the major functions of affiliation and practical reason may be exercised by the woman in the management of the home and by the man in the public sphere.87 It evidently seems to him convenient, given women's childbearing role, that the customary divisions of duties should not be overturned, and he believes that all the major capabilities can flourish in either sphere. Is this any more problematic than to say that human functioning in India can, and even should, take a different concrete form from functioning in England?

The difficulty is, however, that once we have recognized the extent to which gender divisions have been socially constructed in morally arbitrary (p.100) and injurious ways, and once we insist, instead, on using common humanity as our moral and political basis, it is difficult to see what good arguments there are for Position A, which just happens to maintain in place divisions that have often proven oppressive to women. What could such arguments be?

I have mentioned biological differences. But how much separation of function is really suggested by women's childbearing, especially today? Even in the fourth century BC, Plato was able to see that the situation of males and females is not very different from the situation of male and female hunting dogs: the female needs a period of rest for childbearing and nursing, but this in no way requires, or even suggests, a life‐long differentiation of functions. Advances in the control of reproduction are making this less and less plausible. And it should be evident to all that the disability imposed by childbearing for the member of the labour force is to a large extent constructed, above all by the absence of support for child care, both from the public sphere and from employers. Other bodily differences that have standardly been mentioned—for example, differences in bodily strength that have often been held to imply a differentiation of functions—are increasingly being found to be based on bad scientific argument,88 and are also less and less plausible as bases for functional differentiation. Military functions, for example, depend less and less upon bodily strength and more and more on education. The recognition of this by the US Congress in its recent equalization of military roles simply grants what should long ago have been obvious.

One might also point to contingent social facts. Societies are already divided along gender lines. So if we are going to move to a situation in which women will be capable of exercising all the major functions, it will be prudent to develop the resources of that gender‐divided structure, seeking greater independence and fulfilment for women within it, rather than trying to break it up. This, I think, is what is really going on in Musonius. As a Greek‐speaking philosopher in Nero's Rome, he hasn't the ghost of a chance of making institutional changes of the sort recommended in Stoic views of the ideal city, in which males and females were to be fully equal citizens with no distinction of spheres and even no distinction of clothing!89 He does have a hope of convincing individual husbands to allow their wives access to education, so he does what he can. Much the same is true in Martha Chen's A Quiet Revolution. Neither Chen nor her colleagues proposed to jettison all gender divisions within the village. Instead, they found ‘female jobs’ that were somewhat more dignified and important than the old jobs, jobs that looked continuous with traditional female work but were outside the home and brought in wages.

(p.101) Frequently this is a prudent strategy in bringing about real social change. As Martha Chen shows, the ‘revolution’ in women's quality of life never would have taken place but for the caution of the women, who at each stage gave the men of the village reason to believe that the transformations were not overwhelmingly threatening and were good for the well‐being of the entire group. On the other hand, such pragmatic decisions in the face of recalcitrant realities do not tell us how things ought to be. To hold that a gender‐divided two‐spheres result is an acceptable specification of the norm is deeply problematic. For very often the traditionally female norm is socially devalued, and the traditionally male functions powerfully connected with important advantages. In Musonius' Rome, a husband can be both a citizen and a household manager; a wife does not have the choice to be a citizen. In Metha Bai's contemporary India, the confinement of women to the domestic sphere cuts them off from the choice to earn a living, a powerful determinant of overall capability status. In short, ‘separate but equal’ assignments usually serve the ends of a dominant group and perpetuate the oppression of the powerless.90

This point needs particular attention in thinking about divisions of labour within the family. It seems perfectly reasonable that in any household there should be a division of labour, even a long‐standing one, with some members gaining greater skills at one task, some at another. It would already be great progress, vis‐à‐vis the current state of things in all known countries, if domestic duties were equally divided by time and effort. But even in that utopian situation, assignment of tasks along traditional gender‐divided lines may be suspect, on account of its possible association with lack of respect and self‐respect. If all and only girls are taught to cook, for example, this does not seem to be a morally neutral case of functional specialization (like teaching one child the piano, another the clarinet); for it reinforces stereotypes that are associated, historically, with the denial to women of citizenship and autonomy.

I conclude that there are no good arguments for position A, and that even the prudent use of A in promoting gradual social change should be viewed (p.102) with caution, and with a constant awareness of more genuinely equal norms.

I turn now to Position B, which has been influentially defended by many philosophers, including Rousseau and some of his contemporary followers.91 This position may be criticized in a number of different ways. First, we should insist that, insofar as it rests on the claim that there are two different sets of basic capabilities, this claim has not been borne out by any responsible scientific evidence. As Anne Fausto Sterling's Myths of Gender repeatedly shows, experiments that allegedly show strong gender divisions in basic (untrained) abilities are full of scientific flaws; these flaws removed, the case for such differences is altogether inconclusive.

Secondly, we should note that even what is claimed without substantiation in this body of scientific material usually does not amount to a difference in what I have been calling the central basic capabilities. What is alleged is usually a differential statistical distribution of some specific capacity for a high level of excellence, not for crossing the threshold, and excellence in some very narrowly defined function (say, geometrical ability), rather than in one of our large‐scale capabilities such as the capability to perform practical reasoning (which may, recall, be done in a number of different ways, in accordance with the particular tastes and abilities of the individual). So: even if the claim were true it would not be a claim about capabilities in our capacious sense; nor, since it is a statistical claim, would it have any implications for the ways in which individuals should be treated. So the political consequences of such gender differences in our scheme of things, even had they been established, would be nil.

Finally, we must also note that it is in principle next to impossible, right now, to do the sort of research that would be required if such differences were ever to be convincingly established. For it has been shown that right now, from birth on, babies of the two sexes are differently treated by parents and other adults, in accordance with the perception of their external genitalia. They are handled differently, spoken to differently, given different toys. Their emotions are labelled differently—thus a crying infant tends to be labelled ‘angry’ if the observer believes it to be a boy, and ‘frightened’ if the observer believes it to be a girl.92 This means that in the present gender‐divided state of things we cannot get beneath culture reliably enough to get the necessary evidence about basic capabilities. I think this supports the conclusion I defended earlier: the potential for error and abuse in capability testing is so great that we should proceed as if every individual has the basic capabilities.

(p.103) But we can also criticize Position B in a different way. For I believe that it can also be shown that the differentiated conceptions of male and female functioning characteristically put forward by B are internally inadequate, and fail to give us viable norms of human flourishing.93

What do we usually find, in the versions of B that our philosophical tradition bequeaths to us? (Rousseau's view is an instructive example.) We have, on the one hand, males who are ‘autonomous’, capable of practical reasoning, independent and self‐sufficient, allegedly good at political deliberation. These males are brought up not to develop strong emotions of love and feelings of deep need that are associated with the awareness of one's own lack of self‐sufficiency. For this reason they are not well equipped to care for the needs of their family members, or, perhaps, even to notice those needs. On the other hand, we have females such as Rousseau's Sophie, brought up to lack autonomy and self‐respect, ill equipped to rely on her own practical reasoning, dependent on males, focused on pleasing others, good at caring for others. Is either of these viable as a complete life for a human being?

It would seem not. The internal tensions in Rousseau's account are a good place to begin seeing this; they have been well described by Susan Okin and Jane Roland Martin. Rousseau, in Emile, places tremendous emphasis on compassion as a basic social motivation. He understands compassion to require fellow feeling, and a keen responsiveness to the sufferings of others. And yet, in preparing Emile for autonomous citizenship, in the end he shortchanges these emotional functions, allocating caring and responsiveness to the female sphere alone. It appears likely that Emile will be not only an incomplete person but also a defective citizen, even by the standards of citizenship recognized by Rousseau himself.

With Sophie, things again go badly. Taught to care for others, but not taught that her life is her own to plan, she lives under the sway of external influences and lacks self‐government. As Rousseau himself shows, in his fascinating narrative of the end of her life, she comes to a bad end through her lack of judgement. Moreover—as Musonius already argued to his Roman husband, defending equal functioning—she proves to be a bad partner and deficient in love. For love, as we come to see, requires judgement and constancy if it is to be truly deep and truly perceptive. So each of them fails to live a complete human life; and each fails, too, to exemplify fully and well the very functions for which they were being trained, since those functions require support from other functions for which they were not trained. The text leads its thoughtful reader to the conclusion that the (p.104) capabilities that have traditionally marked the separate male and female spheres are not separable from one another without a grave functional loss. They support and educate one another. So society cannot strive for completeness by simply adding one sphere to the other. It must strive to develop in each and every person the full range of human capabilities.

This more inclusive notion of human functioning admits tragic conflict. For it insists on the separate value and the irreplaceable importance of a rich plurality of functions. And the world does not always guarantee that individuals will not be faced with painful choices among these functions, in which, in order to pursue one of them well they must neglect other (and thus, in many cases, subvert the one as well). But this shows once again, I believe, the tremendous importance of keeping some such list of the central functions before us as we assess the quality of life in the countries of the world and strive to raise it. For many such tragedies—like many cases of simple capability failure—result from unjust and unreflective social arrangements. One can imagine, and try to construct, a society in which the tragic choices that faced Emile and Sophie would not be necessary, in which both males and females could learn both to love and to reason.

Being a woman is indeed not yet a way of being a human being. Women in much of the world lack support for the most central human functions, and this denial of support is frequently caused by their being women. But women, unlike rocks and plants and even dogs and horses, are human beings, have the potential to become capable of these human functions, given sufficient nutrition, education, and other support. That is why their unequal failure in capability is a problem of justice. It is up to us to solve this problem. I claim that a conception of human functioning gives us valuable assistance as we undertake this task.94


(1) The argument of this paper is closely related to that of several other papers of mine, to which I shall refer frequently in what follows: ‘Nature, Function, and Capability’, Oxford Studies in Ancient Philosophy, suppl. vol. 1 (1988), 145–84; ‘Non‐Relative Virtues: An Aristotelian Approach’, Midwest Studies in Philosophy 13 (1988), 32–53, and, in an expanded version, in M. Nussbaum and A. Sen (eds.), The Quality of Life (Oxford: Clarendon Press, 1993), 242–76; ‘Aristotelian Social Democracy’, in R. B. Douglass, G. Mara, and H. Richardson (eds.), Liberalism and the Good (New York: Routledge, 1990), 203–52; ‘Aristotle on Human Nature and the Foundations of Ethics’, in World, Mind, and Ethics: Essays on the Philosophy of Bernard Williams, R. Harrison and J. Altham (eds.) (Cambridge: Cambridge University Press, 1995); ‘Human Functioning and Social Justice: In Defense of Aristotelian Essentialism’, Political Theory 20 (1992), 202–46.

(2) By this I mean that the difference in external genitalia figures in social life as it is interpreted by human cultures; thus we are never dealing simply with facts given at birth, but always with what has been made of them (see below, section 8 for discussion of the role of culture in biological claims about male/female differences). Thus, even the common distinction between ‘gender’, a cultural concept, and ‘sex’, the allegedly pure biological concept, is inadequate to capture the depth of cultural interpretation in presenting even the biological ‘facts’ to human beings, from the very start of a child's life. See Anne Fausto‐Sterling, Myths of Gender (2nd edn., New York: Basic Books, 1992). I have discussed these issues further in ‘Constructing Love, Desire, and Care’, forthcoming in D. Estlund and M. Nussbaum (eds.), Laws and Nature: Shaping Sex, Preference, and Family (Oxford University Press).

(3) For a historical argument along these lines from the history of Western scientific thought, see Thomas Laqueur, Making Sex (Berkeley and Los Angeles: University of California Press, 1989). The papers in this volume by Amartya Sen, Xiaorong Li, and Roop Rekha Verma show that the use of ideas of nature to convey a false sense of appropriateness, ‘justifying’ unjust practices, is by no means confined to the Western tradition.

(4) See Martha Chen's paper in this volume.

(5) J. S. Mill, The Subjection of Women (Indianapolis: Bobbs Merrill, 1988); Amartya Sen, ‘Gender and Cooperative Conflicts’, in I. Tinker (ed.), Persistent Inequalities (New York: Oxford University Press, 1990); ‘Gender Inequality and Theories of Justice’ in this volume and ‘More Than a Million Women are Missing’, New York Review of Books: Human Development Report 1993, for the United Nations Development Programme (UNDP) (New York and Oxford: Oxford University Press, 1993); Susan Moller Okin, Justice, Gender, and the Family (New York: Basic Books, 1989), see my review of Okin, ‘Justice for Women’, New York Review of Books October 1992;Catharine MacKinnon, remark cited by Richard Rorty in ‘Feminism and Pragmatism’, Michigan Quarterly Review 30 (1989), 263. MacKinnon has since acknowledged the remark.

(6) For a compelling argument linking feminism and internationalism, see Onora O'Neill, ‘Justice, Gender, and International Boundaries’, in M. Nussbaum and A. Sen (eds.), The Quality of Life, 303–23.

(7) Kwame Anthony Appiah, In My Father's House: Africa in the Philosophy of Culture (New York and Oxford: Oxford University Press, 1992), 136.

(8) On the other hand, it is closely related to Kantian approaches using the universal notion of personhood. See, for example, Onora O'Neill, ‘Justice, Gender, and International Boundaries’, with my commentary (324–35). In the present volume, see the papers of Onora O'Neill, Ruth Anna Putnam, and Roop Rekha Verma. Below I shall be making some criticisms of the concept of ‘person’ in feminist argument, and related criticisms of liberal Kantian approaches (on which see also ASD and my review of Okin). But these differences are subtle and take place against a background of substantial agreement. See also David Crocker, ‘Functioning and Capability: The Foundation of Sen's and Nussbaum's Development Ethics’, Political Theory 20 (1992), 584 ff.

(9) By relativism, I mean the view that the only available criterion of adjudication is some local group or individual. Thus relativism, as I understand it, is a genus of which the brand of reliance on individuals' subjective preferences frequently endorsed in neoclassical economics is one species. (Economists, of course, are relativist only about value, not about what they construe as the domain of scientific ‘fact’.) This affinity will later be relevant to my comments on the Marglin project. My opponents also frequently employ the term ‘postmodernist’ to characterize their position: this is a vaguer term, associated in a very general way with the repudiation of both metaphysical realism (to be defined below) and universalism.

(10) Much of the material described in these examples is now published in Dominating Knowledge: Development, Culture, and Resistance, F. A. Marglin and S. A. Marglin (eds.) (Oxford: Clarendon Press, 1990). The issue of ‘embeddedness’ and menstruation taboos is discussed in S. A. Marglin, ‘Losing Touch: The Cultural Conditions of Worker Accommodation and Resistance’, 217–82, and related issues are discussed in S. A. Marglin, ‘Toward the Decolonization of the Mind’, 1–28. On Sittala Devi, see F. A. Marglin, ‘Smallpox in Two Systems of Knowledge’, 102–44; and for related arguments see Ashis Nandy and Shiv Visvanathan, ‘Modern Medicine and Its Non‐Modern Critics’, 144–84.

(11) For Sen's own account of the plurality and internal diversity of Indian values, one that strongly emphasizes the presence of a rationalist and critical strand in Indian traditions, see M. Nussbaum and A. Sen, ‘Internal Criticism and Indian Relativist Traditions’, in M. Krausz (ed.), Relativism (Notre Dame, Ind.: Notre Dame University Press, 1989)—a paper originally presented at the same WIDER conference and refused publication by the Marglins in its proceedings; and ‘India and the West’, The New Republic, 7 June 1993.

(12) S. A. Marglin, in ‘Toward the Decolonization’, 22–3, suggests that binary thinking is peculiarly Western. But such oppositions are pervasive in all traditions with which I have any acquaintance: in the Upanishads, for example (see the epigraph to ‘Human Functioning’), in Confucian thought (see, again, the epigraph to ‘Human Functioning’), in lbo thought (see, for many examples, Chinua Achebe's Things Fall Apart (London: William Heinemann, 1958)). Critics of such oppositions have not explained how one can speak coherently without bouncing off one thing against another. I believe that Aristotle was right to hold that to say anything at all one must rule out something, at the very least the contradictory of what one puts forward. The arguments of Nietzsche, which are frequently put forward as if they undermine all binary oppositions, actually make far more subtle and concrete points about the origins of certain oppositions, and the interests served by them.

(13) See E. Hobsbawm and T. Ranger (eds.), The Invention of Tradition (Cambridge: Cambridge University Press, 1983). In his New Republic piece, Sen makes a similar argument about contemporary India: the Western construction of India as mystical and ‘other’ serves the purposes of the fundamentalist BJP, who are busy refashioning history to serve the ends of their own political power. An eloquent critique of the whole notion of the ‘other’, and of the associated ‘nativism’, where Africa is concerned, can be found in Appiah (above n. 7), especially in the essays ‘The Postcolonial and the Postmodern’, 137–57 and ‘Topologies of Nativism’, 47–72.

(14) The proceedings of this conference are now published as Nussbaum and Sen (eds.), The Quality of Life (n. 1 above).

(15) ‘Capability and Well‐Being,’ in Nussbaum and Sen, 30–53.

(16) Marglin has since published this point in ‘Toward the Decolonization’, His reference is to Takeo Doi, The Anatomy of Dependence (Tokyo: Kedansho, 1971). On women and men in Japan, See Human Development Report 1993, 26: ‘Japan, despite some of the world's highest levels of human development, still has marked inequalities in achievement between men and women. The 1993 human development index puts Japan first. But when the HDI is adjusted for gender disparity, Japan slips to number 17 . . . Women's average earnings are only 51% those of men, and women are largely excluded from decision‐making positions . . . Their representation is even lower in the political sphere . . . In legal rights in general, Japan's patrilineal society is only gradually changing to offer women greater recognition and independence. Japan now has political and non‐governmental organizations pressing for change . . . ’ The question of freedom of choice is thus on the agenda in Japan in a large way, precisely on account of the sort of unequal functioning vividly illustrated in Marglin's example, where menial functions are performed by women, in order that men may be free to perform their managerial and political functions.

(17) See S. A. Marglin, ‘Toward the Decolonization’.

(18) See S. A. Marglin, ‘Losing Touch’. I put the term in quotes to indicate that I am alluding to Marglin's use of the term, not to the concept as I understand it.

(19) See S. A. Marglin, ‘Toward the Decolonization’ and ‘Losing Touch’. Similar claims are common in feminist argument. For example, in The Feminist Theory of the State (Cambridge, Mass.: Harvard University Press, 1989), Catharine MacKinnon argues that ‘objectivity’ as traditionally conceived in the Western epistemological tradition is causally linked to the objectification and abuse of women. This line of argument is effectively criticized in Louise M. Antony, ‘Quine as Feminist: The Radical Import of Naturalized Epistemology’, in L. M. Antony and C. Witt (eds.), A Mind of One's Own: Feminist Essays on Reason and Objectivity (Boulder, Colo.: Westview Press, 1992), 185–225. See also the detailed examination of MacKinnon's argument in the same volume by Sally Haslanger, in ‘On Being Objective and Being Objectified’, 85–125. MacKinnon's fundamental contributions in the areas of sexual harassment and pornography do not depend on this analysis, and are actually undermined by it. The core of her thought actually reveals a strong commitment to a type of ethical universalism, as my epigraph indicates. See, in the Antony volume, the persuasive analysis by Liz Rappaport, ‘Generalizing Gender: Reason and Essence in the Legal Thought of Catharine MacKinnon’, 127–43. Alcoff's contribution in the present volume continues the debate about feminism and reason; and see also L. Alcoff and E. Potter (eds.), Feminist Epistemologies (New York: Routledge 1993). For a healthy scepticism about the role of ‘anti‐essentialism’ within feminism, see Seyla Benhabib, ‘Feminism and the Question of Postmodernism’, in Situating the Self: Gender, Community, and Postmodernism in Contemporary Ethics (New York: Routledge, 1992), 203–42; Sabina Lovibond, ‘Feminism and Postmodernism’, New Left Review, 178 (November–December 1989), 5–28; Val Moghadam, ‘Against Eurocentrism and Nativism’, Socialism and Democracy, fall/winter (1989: 81–104); Moghadam, Gender, Development, and Policy: Toward Equity and Empowerment, UNU/WIDER Research for Action series, November (1990).

(20) For an account of this sort of normative argument, see Alasdair Maclntyre, After Virtue (Notre Dame, Ind.: Notre Dame University Press, 1989).

(21) J. Derrida, Of Grammatology, trans. G. Spivak (Baltimore: The Johns Hopkins University Press, 1976). The term is meant to suggest the idea that reality is simply ‘there’ and that knowledge consists in being ‘present’ to it, without any interfering barrier or mediation.

(22) R. Rorty, Philosophy and the Mirror of Nature (Princeton, NJ: Princeton University Press, 1979).

(23) See, for example, G. E. L. Owen, ‘Tithenai ta Phainomena’, in Logic, Science, and Dialectic (London: Duckworth, 1986), and M. Nussbaum, The Fragility of Goodness: Luck and Ethics in Greek Tragedy and Philosophy (Cambridge: Cambridge University Press, 1986). See also Hilary Putnam, Aristotle After Wittgenstein, Lindlay Lecture, University of Kansas, 1991.

(24) See the illuminating discussion in B. K. Matilal, Perception (Oxford: Clarendon Press, 1985). It is worth noting that this fundamental work is not cited anywhere in Marglin and Marglin, although Matilal was present at the conference and delivered a paper critical of the Marglins' characterization of Indian traditions. This paper was dropped from the volume. Matilal also described the implications of the realism debate for Indian ethical thought: see ‘Ethical Relativism and the Confrontation of Cultures’, in Krausz (ed.), Relativism (Notre Dame, Ind.: Notre Dame University Press, 1989), 339–62.

(25) There is a longer version of my criticism of contemporary attacks on universalism in ‘Human Functioning’. See also ‘Skepticism About Practical Reason in Literature and the Law’, Harvard Law Review 107 (1994), 714–44. In both of these papers I study the surprising convergence between ‘left’ and ‘right’ in the critique of normative argument, the ‘postmodern’ positions of many thinkers on the left proving, often, difficult to distinguish from claims about the arbitrariness of evaluation in neoclassical economics. In Barbara Herrnstein Smith's Contingencies of Value (Durham: Duke University Press, 1988), we even see a fusion of the two positions, a postmodernism concluding that, in the absence of transcendent standards, we should understand value judgements as attempts to maximize expected utility.

(26) Aristotle, Nicomachean Ethics VIII.I, 1155a 21–2. I discuss this passage in ‘Aristotle on Human Nature’ and ‘Non‐Relative Virtues’.

(27) K. A. Appiah, In My Father's House, pp. vii–viii: ‘If my sisters and I were “children of two worlds”, no one bothered to tell us this; we lived in one world, in two “extended” families divided by several thousand miles and an allegedly insuperable cultural distance that never, so far as I can recall, puzzled or perplexed us much.’ Appiah's argument does not in any sense neglect distinctive features of concrete histories; indeed, one of its purposes is to demonstrate how varied, when concretely seen, histories really are. But his argument, like mine, seeks a subtle balance between perception of the particular and recognition of the common. In his essay ‘The Postcolonial and the Postmodern’ (137–57), Appiah shows that it is all too often the focus on ‘otherness’ that produces a lack of concrete engagement with individual lives. Speaking of the sculpture ‘Yoruba Man with Bicycle’ that appears on the cover of the book, Appiah comments: ‘The Man with a Bicycle is produced by someone who does not care that the bicycle is the white man's invention—it is not there to be Other to the Yoruba Self; it is there because someone cared for its solidity; it is there because it will take us further than our feet will take us . . . ’ (157).

(28) In this category, as closely related to my own view, I would place the ‘internal–realist’ conception of Hilary Putnam articulated in Reason, Truth, and History (Cambridge: Cambridge University Press, 1981), The Many Faces of Realism (La Salle: Open Court Publishing, 1987), and Realism With a Human Face (Cambridge, Mass.: Harvard University Press, 1990); and also the views of Charles Taylor, for example in Sources of the Self: The Making of Modern Identity (Cambridge, Mass.: Harvard University Press, 1989), and ‘Explanation and Practical Reason’, in Nussbaum and Sen (eds.), The Quality of Life, 208–31.

(29) In this sense I am thoroughly in agreement with Susan Okin's reply to the charge of ‘substitutionalism’ that has been made against her book, and in agreement with both Okin and Ruth Anna Putnam that it is a mistake to conceive of the moral point of view as constituted by the actual voices of all disadvantaged parties, see Okin's and Putnam's papers in this volume. See my further comments below, Section 5.

(30) Can the Marglins consistently make this objection while holding that freedom of choice is just a parochial Western value? It would appear not; on the other hand, F. A. Marglin (here differing, I believe, from S. A. Marglin) also held in oral remarks delivered at the 1986 conference that logical consistency is simply a parochial Western value.

(31) The politics of the history of Western philosophy have been interpreted this way, with much plausibility though perhaps insufficient historical argumentation, by Noam Chomsky, in Cartesian Linguistics (New York: Harper & Row, 1966). Chomsky argues that Cartesian rationalism, with its insistence on innate essences, was politically more progressive, more hostile to slavery and imperialism, than empiricism, with its insistence that people were just what experience had made of them. My analysis of Stoic feminist argument (below Section 7) bears this out.

(32) The use of this term does not imply that the functions all involve doing something especially ‘active’. (See here Sen, ‘Capability and Well‐Being’, in The Quality of Life, 30–53.) In Aristotelian terms, and in mine, being healthy, reflecting, being pleased, are all ‘activities’.

(33) For further discussion of this point, and examples, see ‘Aristotle on Human Nature’.

(34) Ibid. discusses the treatment of this point in contemporary medical ethics. Could one cease to be one's individual self without ceasing to be human? This is ruled out, I think, in Aristotle's conception, but is possible in some other metaphysical conceptions. But the sort of case that would most forcefully raise this possibility is not the sort involving illness or impairment, but instead the sort involving personality or memory change; and I shall not attempt to deal with such cases here.

(35) Appiah, In My Father's House, p. viii.

(36) In ‘Aristotle on Human Nature’, there is a more extended account of this procedure and how it justifies.

(37) This of course is not incompatible with calling certain groups non‐human or subhuman for political purposes. But such denials are usually either transparent propaganda or forms of self‐deception, which can be unmasked by critical argument. See below for a case involving women; and for an extensive analysis of the psychology of such self‐deception, and its unmasking, see Raoul Hilberg, The Destruction of the European Jews, abridged edition (New York: Holmes & Meier, 1985), 274–93.

(38) In order to make this clear, I speak of it as a conception of the good, at a very minimal and general level. The phrase I have elsewhere used is ‘the thick vague theory of the good’. The term ‘thick’ contrasts this account, in its comprehensiveness, with Rawls's ‘thin’ theory of the good, which is designed to avoid even partial comprehensiveness.

(39) On this see especially ‘Non‐Relative Virtues’.

(40) I have discussed my own views about practical rationality elsewhere, particularly in ‘The Discernment of Perception’, in Love's Knowledge (New York: Oxford University Press, 1990). A related account, which I admire and to a large extent agree with, is given by Henry Richardson in Practical Deliberation About Final Ends (Cambridge University Press, 1994). Richardson's account is closely related, as well, to the pragmatist conception supported by Hilary Putnam in his paper in this volume.

Should the conception of reasonableness be defined with reference to democratic procedures, as Seyla Benhabib has recommended? I see the attractions of this proposal, but I have not followed it. First of all, it seems to me that democratic procedures as they actually are do not always embody reasonableness; so to describe what makes a democratic procedure reasonable we will have to have a notion of the reasonable that is to at least some extent independent of the notion of democracy. Secondly, to build democracy into the ground level of the conception of the human from the start prevents us from raising later on the question of what political arrangement will best secure to citizens the list of human capabilities, in a wide variety of circumstances. It may turn out that the answer will always be ‘democracy’. But even then, I think it will rarely be just democracy (ancient Athenian or New England town‐meeting style). No modern democratic state is a pure democracy, and it should at this point remain an open question as to what role should be played by relatively undemocratic institutions such as the US Supreme Court in promoting the capabilities of citizens.

(41) For Rawls's use of a notion of consensus, see Rawls, ‘The Idea of an Overlapping Consensus’, Oxford Journal of Legal Studies, 7 (1987), and now Political Liberalism (New York: Columbia University Press, 1993). Rawls's notion of consensus appears ambiguous between the two notions I identify here. See, on this, the exchange between Joshua Cohen and Jean Hampton in The Idea of Democracy (New York: Oxford University Press, 1992). Cohen argues that Rawls needs, and can consistently defend, the weaker ‘overlap’ reading; Hampton argues that, whatever Rawls intends, the plausibility of his argument rests on his opting for the normative reading. I concur with Hampton.

(42) To cite only a few recent examples with serious practical consequences: in the United Sates in the 1890s, the Supreme Court, denying a Virginia woman's appeal against a law forbidding women to practise law, judged that it was up to the state Supreme Court ‘to determine whether the word “person’ ” in the statute on which the woman based her appeal ‘is confined to males’. (In re Lockwood, 154 US 116, discussed in Okin, Women, 251 and n. 10, and see Sunstein's paper in this volume.) In Massachusetts in 1932, women were denied eligibility for jury service, although the law stated that ‘every person qualified to vote’ was eligible. The state Supreme Court wrote: ‘No intention to include women can be deduced from the omission of the word “male’ ” (Commonwealth v. Welosky, 276 Mass. 398, cert. denied, 284 US 684 (1932)), discussed in Okin, Women, 251 and n. 11. Such readings no doubt reflect faithfully enough the views that the Founders had about the term ‘person’ when they used it in the Constitution: see my Jefferson epigraph. Although this construal of the term does not prevail today in American law, its legacy is with us in countless more informal ways.

(43) Aristotle, Metaphysics I.I.

(44) I discuss this issue in much more detail in Lecture 3 of my 1993 Gifford Lectures, University of Edinburgh (forthcoming, Cambridge University Press).

(45) Aristotle, ubiquitously in the accounts of substance.

(46) On these issues, see further in ‘Aristotelian Social Democracy’.

(47) Aristotle, Politics VII.I: see ‘Nature, Function, and Capability’.

(48) It may support what James Rachels calls ‘moral individualism’ (Created From Animals (Oxford and New York: Oxford University Press, 1990)), in which our moral obligations flow from the endowments of the individual creature with whom we are dealing, rather than from its species, and our goal should be to promote—or at least not to impede—the form of flourishing of which the being is basically capable. On this view such an infant should get the same treatment that we would give to an animal of similar endowment. But we may also decide to give the fact that it is an offspring of humans some moral weight; nothing I have said here rules that out.

(49) Although ‘normal length’ is clearly relative to current human possibilities, and may need, for practical purposes, to be to some extent relativized to local conditions, it seems important to think of it—at least at a given time in history—in universal and comparative terms, as the Human Development Report does, to give rise to complaint in a country that has done well with some indicators of life quality, but badly on life expectancy. And although some degree of relativity may be put down to the differential genetic possibilities of different groups (the ‘missing women’ statistics, for example, allow that on the average women live somewhat longer than men), it is also important not to conclude prematurely that inequalities between groups—for example, the growing inequalities in life expectancy between blacks and whites in the USA—are simply genetic variation, not connected with social injustice.

(50) The precise specification of these health rights is not easy, but the work currently being done on them in drafting new constitutions in South Africa and Eastern Europe gives reason for hope that the combination of a general specification of such a right with a tradition of judicial interpretation will yield something practicable. It should be noticed that I speak of health, not just health care: and health itself interacts in complex ways with housing, with education, with dignity. Both health and nutrition are controversial as to whether the relevant level should be specified universally, or relatively to the local community and its traditions: for example, is low height associated with nutritional practices to be thought of as ‘stunting’, or as felicitous adaptation to circumstances of scarcity? For an excellent summary of this debate, see S. R. Osmani (ed.), Nutrition and Poverty, WIDER series (Oxford: Clarendon Press, 1990), especially the following papers: on the relativist side, T. N. Srinivasan, ‘Undernutrition: Concepts, Measurements, and Policy Implications’, 97–120; on the universalist side, C. Gopalan, ‘Undernutrition: Measurement and Implications’, 17–48; for a compelling adjudication of the debate, coming out on the universalist side, see Osmani, ‘On Some Controversies in the Measurement of Undernutrition’, 121–61.

(51) There is a growing literature on the importance of shelter for health: e.g. that the provision of adequate housing is the single largest determinant of health status for HIV infected persons. Housing rights are increasingly coming to be constitutionalized, at least in a negative form—giving squatters grounds for appeal, for example, against a landlord who would bulldoze their shanties. On this as a constitutional right, see proposed Articles 11, 12, and 17 of the South African Constitution, in a draft put forward by the ANC committee, advisor Albie Sachs, where this is given as an example of a justiciable housing right.

(52) I shall not elaborate here on what I think promoting this capability requires, since there is a future volume in the WIDER series devoted to this topic: J. Glover, M. Nussbaum, and C. Sunstein (eds.), Women, Equality, and Reproduction.

(53) A good example of an education right that I would support is given in the ANC South African Constitution draft, Article 11: ‘Education shall be free and compulsory up to the age of sixteen, and provision shall be made for facilitating access to secondary, vocational and tertiary education on an equal basis for all. Education shall be directed towards the development of the human personality and a sense of personal dignity, and shall aim at strengthening respect for human rights and fundamental freedoms and promoting understanding, tolerance and friendship amongst South Africans and between nations.’ The public (or otherwise need‐blind) provision of higher education will have to be relative to local possibilities, but it is at least clear that the USA lags far behind most other countries of comparable wealth in this area.

(54) On the emotions as basic human capabilities, see, in addition to my other chapter in this volume, my 1993 Gifford Lectures, Upheavals of Thought: A Theory of the Emotions (forthcoming, Cambridge University Press). My omission of anger from this list of basic emotional capabilities reveals an ambivalence about its role that I discuss at length, both in Gifford Lectures 3 and 10, and in The Therapy of Desire: Theory and Practice in Hellenistic Ethics (Princeton, NJ: Princeton University Press, 1994) chs. 7, 11, and 12. See also ‘Equity and Mercy’, Philosophy and Public Affairs, spring 1993.

(55) In my 1993 Gifford Lectures, I spell out what I think this entails where ‘the family’ is concerned. On the whole, I am in agreement with Susan Okin that some form of intimate family love is of crucial importance in child development, but that this need not be the traditional Western nuclear family. I also agree with Okin that the important educational role of the family makes it all the more crucial that the family should be an institution characterized by justice, as well as love. See Okin, Justice, Gender, and the Family.

(56) ‘Aristotelian Social Democracy’ said that a list of such liberties needed to be added to the Aristotelian scheme, but it did not include them in the account of capabilities itself. These issues are further developed in a future WIDER volume on reproductive rights and women's capabilities, based on the papers given at our 1993 conference, and edited by Jonathan Glover and Martha Nussbaum.

(57) For reproductive choice as an equality issue, see Sunstein's paper in this volume, and also his ‘Gender, Reproduction, and Law’, forthcoming in Glover and Nussbaum.

(58) On this see also ‘Aristotelian Social Democracy’.

(59) With Sen, I hold that the capability set should be treated as an interlocking whole: for my comments on his arguments, see ‘Nature, Function, and Capability’. Tensions will frequently arise among members of the list, and I shall comment on some of those below. But it should be clear by now that the architectonic role of practical reasoning imposes strict limits on the sort of curb on personal autonomy that will be tolerated for the sake of increased nutritional well‐being, etc.

(60) Chris Bobonich ‘Internal Realism, Human Nature, and Distributive Justice: A Response to Martha Nussbaum’, Modern Philology, May 1993 supplement, 74–92, worries that this will impose enormous sacrifices. But I think that this is because he has not imagined things in detail, nor thought about my claim that once people have what they basically need, they can get all sorts of other good things through their own efforts. If I have enough food to be well‐nourished, more food will just rot on the shelf or make me fat. If my basic health needs are met, it seems right that I should not be able to claim expensive unnecessary luxuries (say, cosmetic surgery) at the public expense so long as even one person in my country is without support for basic needs. And so forth. One must take seriously the Aristotelian idea, which is basic to both Sen's and my programmes, that resources are just tools for functioning and have a limit given by what is needed for that functioning. Above that limit, they are just a heap of stuff, of no value in themselves.

(61) See ‘Nature, Function, and Capability’, with reference to Aristotle; and below, Section 9.

(62) Marx, Economic and Philosophical Manuscripts of 1844, discussed in ‘Nature, Function, and Capability’ and ‘Aristotle on Human Nature’.

(63) See especially Sen's paper in this volume; also ‘More than a Million Women are Missing’, New York Review of Books 37 (1990) 61–6.

(64) Iftekhar Hossein, ‘Poverty as Capability Failure’, Ph.D. Dissertation in Economics, Helsinki University, 1990.

(65) See Allardt, ‘Having, Loving, Being: An Alternative to the Swedish Model of Welfare Research’, and Erikson, ‘Descriptions of Inequality: The Swedish Approach to Welfare Research’, in Nussbaum and Sen, The Quality of Life, 88–94 and 67–84.

(66) See Sen, ‘More than a Million Women’.

(67) See also Jon Elster, Sour Grapes (Cambridge: Cambridge University Press, 1983); Cass R. Sunstein, ‘Preferences and Politics’, Philosophy and Public Affairs, 20 (1991), 3–34.

(68) Päivi Setälä, Professor of Women's Studies at the University of Helsinki, informs me that recent studies show that even in Finland, only 40% of the housework is done by males. This, in the second nation in the world (after New Zealand, in 1906) to give females the vote, a nation as committed to sex equality as any in the world. We can assume that the situation is causally related to male preferences.

(69) On the disparity between externally observed health status and self‐reports of satisfaction about health, see Sen, Commodities and Capabilities (Amsterdam: North‐Holland, 1985).

(70) Martha Chen, A Quiet Revolution: Women in Transition in Rural Bangladesh (Cambridge, Mass.: Schenkman, 1983). I describe this account of a rural women's literacy project, and its large‐scale impact on women's quality of life, in ‘Non‐Relative Virtues’, ‘Aristotelian Social Democracy’, and ‘Human Functioning and Social Justice.’

(71) This is a criticism of economic utilitarianism, not of sophisticated philosophical forms of utilitarianism that build in means to filter or correct preferences. Nonetheless, the human‐functioning approach would still object to the role played by the commensurability of values in utilitarianism, and to the related suggestion that for any two distinct ends we can, without loss of what is relevant for choice, imagine trade‐offs in purely quantitative terms. Furthermore, most forms of utilitarianism are committed to aggregating utilities across lives, and thus to neglecting separateness, which I have defended as fundamental. I have addressed some of these questions elsewhere, for example, in ‘The Discernment of Perception’ in Love's Knowledge, and in ‘The Literary Imagination in Public Life’, New Literary History (fall 1993). Sen's work has addressed them in greater detail. I therefore leave them to one side for the purposes of the present inquiry.

(72) For a detailed consideration of these approaches, see ‘Aristotelian Social Democracy’, ‘Human Functioning’, with references to related arguments of Sen. ‘Aristotelian Social Democracy’ contains a detailed account of the relationship between Rawls's resourcism and my project, which is a particularly subtle one. Rawls is willing to take a stand on certain items: thus liberty and the social conditions of self‐respect figure on his list of ‘primary goods’, as well as wealth and income. On the other hand, he has repeatedly denied that his index of primary goods could, or should, be replaced by an index of functionings as in the Human Development Report.

(73) This is the central point repeatedly made by Sen against Rawls; for an overview, see ‘Capability and Well‐Being’ in The Quality of Life, with references.

(74) In Rawls's liberalism the problem is even more acute, since the parties who are either well or not well off are ‘heads of households’, usually taken to be male, who are alleged to deliberate on behalf of the interests of their family members. But women cannot in fact rely on the altruism of males to guarantee their economic security, or even survival. In addition to Sen's work on this issue, see Susan Moller Okin, Justice, Gender, and the Family. In my review of Okin, I offer this as a reason for Okin to be more critical of resource‐based liberalism than she is.

(75) Martha Chen and her fellow development workers, in the project described in A Quiet Revolution, were indebted in their practice to Paolo Freire's notion of ‘participatory dialogue’.

(76) Sen has stressed this throughout his writing on the topic. For an overview, see ‘Capability and Well‐Being’.

(77) This is the strategy used by Erikson's Swedish team, when studying inequalities in political participation: see ‘Descriptions of Inequality’. The point was well made by Bernard Williams in his response to Sen's Tanner Lectures: see Williams, ‘The Standard of Living: Interests and Capabilities’, in G. Hawthorn (ed.), The Standard of Living (Cambridge: Cambridge University Press, 1987). To give just one example of the issue, we will need to ask to what extent laws regulating abortion, sodomy laws, the absence of civil rights laws, etc., restrict the capability for sexual expression of women and homosexuals in a given society. The gay American military officer who chooses celibacy for fear of losing his job has not, in the relevant sense, been given a capability of choosing.

(78) See also Sen, Commodities and Capabilities.

(79) The relevant textual references are gathered and discussed in ‘Aristotelian Social Democracy’.

(80) The remark was cited by Richard Rorty in ‘Feminist and Pragmatism’, Michigan Quarterly Review, 30 (1989), 231; it has since been confirmed and repeated by MacKinnon herself.

(81) See n. 37 above on Raoul Hilberg's account, in The Destruction of the European Jews, of the Nazi device of categorizing Jews as animals or inanimate objects, and the vulnerability of that stratagem to ‘breakthroughs’, in which the mechanisms of denial were caught off guard.

(82) The most comprehensive and incisive account of Plato's arguments about women is now in Stephen Halliwell, Plato: Republic Book V (Warminster: Aris and Phillips, 1992), Introduction and commentary to the relevant passages. See also Okin, Women in Western Political Thought.

(83) For Musonius' collected works see the edition by O. Hense (Leipzig: Teubner Library, 1905). Other works with radical conclusions for women's issues include ‘Should Boys and Girls Have the Same Education?’ (answering yes to that question); ‘Should One Raise all the Children Who are Born?’ (arguing against infanticide, a particular threat to female offspring); ‘On the Goal of Marriage’ (arguing against the sexual double standard and in favour of equal sexual fidelity for both sexes; arguing as well against the common view that female slaves were available for sexual use).

(84) Stoics are of course highly critical of much that passes for higher education, holding that the traditional ‘liberal studies’ are not ‘liberal’ in the right way, that is, do not truly ‘free’ the mind to take charge of its own reasoning. See Seneca, Moral Epistle 88.

(85) See Musonius, ‘On the Goal of Marriage’. Similar conceptions are defended by Seneca and Plutarch. On this shift in thinking about the marital relationship, see the useful discussion in Foucault, History of Sexuality, vol. III, trans. R. Hurley (New York: Pantheon, 1985).

(86) On the way in which Christianity disrupted the emerging feminist consensus, see G. E. M. de Ste. Croix, The Class Struggle in the Ancient Greek World (London: Duckworth, 1987).

(87) See the last section of ‘That Women Too’, where he answers the male interlocutor's imaginary objection that educated women will spend too much time sitting around and talking, and neglect their practical duties, by telling him that the very same issue arises for him: he too has practical duties that may seem less interesting than talking about ideas, and he too should make sure that he doesn't neglect them. It is, I think, because Musonius has a pretty low view of the worth of male public life that he can easily view that sphere as equivalent and equal to the female sphere.

(88) See Anne Fausto‐Sterling, Myths of Gender.

(89) For the evidence, see Malcolm Schofield, The Stoic Idea of the City (Cambridge: Cambridge University Press, 1992).

(90) Is the Nigerian situation depicted in Nzegwu's paper an exception? We can agree with her that the traditional system in which women controlled certain vital agricultural functions and men others, was somewhat better, in capability terms, than the system of confinement to the domestic sphere imposed by British colonialism, without being altogether sure that the traditional system was morally acceptable. This would depend on a closer scrutiny of the whole system of functionings and capabilities, as affected by gender divisions. I am no expert in Ibo culture, clearly; but the traditional Ibo families depicted in Chinua Achebe's novels, for example, do not seem to me to manifest full gender equality in capability. Okonkwo (in Things Fall Apart) can decide to beat his wife; she cannot choose to beat him in return, or even to stop him, in all but the most egregious of cases. Okonkwo can choose to take another wife; no wife of his can choose another husband. The reason why Okonkwo keeps wishing that Ezinma had been a boy rather than a girl is that he perceives that, being a girl, she is debarred from many functions for which she seems well suited. His fear of being seen as a ‘woman’ is, by contrast, a fear of capability failure.

(91) On Rousseau, see Okin, Women, and Jane Roland Martin, Reclaiming a Conversation (New Haven: Yale University Press, 1985). On some related contemporary arguments, for example those of Allan Bloom, see Okin, Justice, ch. 1.

(92) On all this, see Fausto‐Sterling.

(93) Here I am in agreement with the general line of argument in Okin, Women, and Martin, Reclaiming, and with the related arguments in Nancy Chodorow's The Reproduction of Mothering, which I discuss in my other chapter.

(94) I am grateful to all the members of our meeting for valuable comments, and especially to Amartya Sen for valuable discussions and to David Crocker, Jonathan Glover, Cass Sunstein, and Susan Wolf for helpful written comments. I am also grateful to Chris Bobonich, David Estlund, and Henry Richardson for comments on related earlier work.