Feminist logic operates according to its own special rules, which allow its practitioners to condemn in others methods of argument used by themselves. Mr and Mrs Belsey and Marcia Yudkin take me to task for using stereotypes, but seem very ready to use stereotypes themselves, and, since I reject the doctrines of female egalitarianism, cast me in the role of male chauvinist pig. 1 But these are not exhaustive alternatives. To poke fun at the follies of feminism is not to commend a brutal machismo. Indeed, it is one of my complaints against feminist logic that its crude either-or approach encourages those who can see that women are not in all respects the same as men to conclude that they are therefore inferior, and to be treated as sex-objects rather than persons. The age of chivalry may be dead but it is still at least logically possible to think of women as persons, sharing a common humanity with men although different from them in certain, often important, respects. Mr and Mrs Belsey find it `quite unclear whether' I suppose that the distinction between the two stereotypes of the peer-group and the pairgroup is exhaustive. 2 Of course I do not. In another paper in this journal I explore others relevant to the general issue of egalitarianism. 3 The reasons for concentrating on the peer-group and pair-group are that they are pure cases and we can understand their logic better. The family, the neighbourhood and the institution embody differences of age, authority power or position which may appear adventitious and obscure the logic of the situation. Feminists are fond of attributing features of actual societies to outmoded patriarchal practices of male dominance and female subordination. It is very much in point, therefore, to argue that even if we were not born of women, were never politically or economically dependent on others, and never found ourselves under pressure to conform to the established mores of an already existing neighbourhood, we should still have reason to form peer-groups and pair-groups which are based, respectively, on the principles of sameness and of difference. Trudy Govier allows that pair-grouping leads to differentiation within roles, but protests [p.112] that `individual role differentiation within pairs is one thing, general differentiation of roles, based upon sex, quite another'. 4 Not quite. Once there are some differences, there are great advantages in partners having reciprocal roles, and therefore in roles going together in two groups, and in persons who expect to occupy roles of one group choosing partners who expect to occupy roles of the other. Games-theoretical considerations alone will generate firm, though modifiable, conventions about who does what, and the absence of sexually assigned roles constitutes a serious problem for gay relationships. 5 Stereotypes are illuminating. They enable us to notice, organize and explain social phenomena. Of course there are other stereotypes. But neither that fact nor the fact that stereotypes never exhaust the particularity of individual people and their relationships renders them useless, though it should make us cautious in using them. Not every man buys in the cheapest or sells in the dearest market: but economics becomes more intelligible when we follow out the behaviour of economic man. True, some men and some women do not have the instincts I ascribe to them; some find a ménage a trois tolerable, and Platonic friendships abound: nevertheless, the fidelity expected of a lover is very different from the fidelity expected of a friend, and jealousy is a phenomenon difficult to understand without some such account as I have offered. The use of stereotypes can distort, as can any limited scheme of classification. But classification is none the less necessary. Marcia Yudkin finds it `dehumanizing to be seen first, foremost and essentially as a woman', but seems to be confused about the meaning of these terms. It would be dehumanizing to be seen only as female---or for that matter only as male, only as human, only as American, only as a philosopher. But to say that she is essentially a woman is not to say that she is essentially nothing else. Nor do I maintain that her being a woman should `be permitted to colour everything else about me, no matter what the context' and give trials and examinations as examples to the contrary. 6 These, however, are somewhat untypical cases in which we are at pains to deem irrelevant factors which in other contexts we naturally regard as relevant. The concepts of justice and of equality of opportunity do not admit of indefinite extension.7 In most contexts we are under no obligation to disregard anybody's characteristics---indeed to regard someone as being neither male nor female  would be far more dehumanizing than to recognize her femininity along with her many other personal characteristics. Mr and Mrs Belsey, Marcia Yudkin and Trudy Govier none the less perform a useful function in articulating an alternative account of the proper relations between the sexes. There is, in their view, no profound difference between men and women. No differentiation in emotional expectations or in social or legal responsibilities or rights is justified. Homosexual `marriages' are just as acceptable as heterosexual ones. There is no reason why marriages should be permanent. Women ought to be prepared, emotionally and economically, for the end of their marriages. And the labour market should be structured with an eye to those who enter it rather than to those who would rather not. The sexless society is conceivable, and may obtain in the kingdom of heaven:8 but terrestrially there is not much future in it. So long as man is mortal and has any care for the future, it is rational for him to be concerned for posterity in general and his children in particular. Although Plato and some kibbutzniks tried to abolish the family, it is at least arguable that children need their mothers not only for nine months before their birth but for many years after. Society may legitimately have an interest in encouraging women to stay at home, and hence in structuring our economic and social institutions so as not to pressure them into entering the labour market. It is a bad thing if women feel obliged to go out to work in order to keep up the mortgage payments, but it inevitably will happen if women earn the same as men. Although we can go a small way towards mitigating the costs of parenthood by means of family allowances, taxpayers are always likely to be more reluctant to part with their money for the children of other men than fathers are to part with theirs for their own. The pay differentials between male and female workers are a recognition of this. Marcia Yudkin finds them shocking. But justice is not equality, nor absolutely determinate; political, social and economic institutions can legitimately be shaped by social needs and ideals. How far and under what conditions the expectations of single women may be legitimately abridged for the sake of their married sisters is a question that needs much more discussion than I was able to give in my first paper. Here I only maintain, as against the advocates of the sexless society, that a moral philosopher can reasonably regard potential motherhood and the needs of children as factors properly relevant to social institutions. I am castigated for my depraved nature due to my animal ancestry, and for holding that men are stronger than women, that there are differences of disposition and need between the sexes, that women are vulnerable, and that the raising of a family is for most men and women their major creative activity. Of course, these alleged facts about human nature are not  ineluctable truths of physics.9 And again, of course, if all my ancestors had been sexless, I would not have the nasty, brutish and lupine nature I do have. But moral philosophy is not greatly advanced by wishing men unsexed. The sexless society is spectral in being bloodless. Our actual society, composed of real men and women, may be less admirable. But, and this is my second claim about moral philosophy, facts are relevant, although not conclusive. Moral philosophers ought to take into account what human beings are like and what instincts they have and wherein their happiness lies. It is relevant to my behaviour in an airport or on a bus that I am big and strong and can hump suitcases, whereas women are often dizzy or in pain and have difficulty with their luggage. It does not follow, and I never suggested, that all human instincts are to be hallowed. According to the latest arguments of the sociobiologists, the fact that men are bigger than women points to there having been a period in our evolution when we were polygamous, which would support and explain the common belief that men often suffer from a roving eye. It does not follow that I ought to be polygamous and am entitled on account of my size and instincts to help myself to any Bathsheba who catches my fancy. But the fact of my being likely to be in a position of power and having such instincts would be good reason for having some institution or etiquette which differentiated between the sexes and protected the secretary`s bottom from being pinched by her boss, or the undergraduette from being graded according to the favours she shows to her tutor. Mr and Mrs Belsey and Trudy Govier tax me with supposing that there are two and only two ways of resolving disputes and discounting the possibility of disagreements being resolved by calm and rational discussion. But I do not. I fully allow that many disputes are resolved by  discussion and compromise, and that it is much the best that they should be. But it cannot be relied on always to happen, and therefore it cannot be a decision-procedure. For a merely cohabiting couple it constitutes a tolerable basis for living together, since if it fails to yield agreement they can go their separate ways, but it does not provide a final fall-back for two people who are committed to each other for keeps. As in all communities, some other rule is required, which in a community of two cannot be a majority rule.10 So far as the argument goes, it is not essential, as I was at pains to point out, that the men have the final say. Matriarchy is a possible basis for marriage, and might conceivably, as Trudy Govier argues, work out better. But in that case she will have to acknowledge, what at present she claims not to know about, the conceptual link between decisions and responsibility.11 Mr and Mrs Belsey and Trudy Govier object to my acknowledging permanence in marriage as an ideal. I make no apology. It is rooted not only in the needs of children, important though they are, but in the nature of love, which demands not just togetherness for so long as neither party tires of the other, but a commitment which transcends time and once made can never be unmade. One needs to be a poet rather than a philosopher to spell out the logic of love, but even a non-Poet can point out that the shifting liaisons of the sexless society, however intimate and intense while they last, are only transient and lack the absolute unconditionality of total commitment. They cannot express the whole of a lover's devotion. It is impossible to enter into absolute, unconditional commitments on a temporary basis--- Take the charm `for ever' from them, and they crumble into dust.