I begin, as I shall end, with fictions.

In a well-known tale, The Sandman, Hoffmann has a student, Nathaniel, fall in love with a beautiful doll, Olympia, whom he has spied upon as she sits at a window across the street from his lodgings. We are meant to suppose that Nathaniel mistakes an automaton for a human being (and so a person). The mistake is the result of an elaborate but obscure deception on the part of the doll's designer, Professor Spalanzani. Nathaniel is disabused quite by accident when he over-hears a quarrel between Spalanzani, who made Olympia's clockwork, and the sinister Coppelius, who contributed the eyes (real eyes, it seems).

The subject of this symposium is sometimes introduced by asking whether machines could think. This way of introducing it may be misleading, for it may seem as if it were merely about a particular activity, called ‘thinking’. The question would then seem to have the same character as ‘Can machines make a noise?’. But thinking is not something that can be treated in isolation from other personal qualities. What we need to consider is whether, or to what extent, a machine could participate in the whole complex of qualities, activities, attitudes, thoughts, feelings and moral relationships that we regard as essential to being a person—whether, in this sense, machines could be persons.

I want to argue for the importance of the notion human being in ethics. Part I of the paper presents two different sorts of argument against treating that notion as important in ethics.

A. Here is an example of the first sort of argument.

What makes us human beings is that we have certain properties, but these properties, making us members of a certain biological species, have no moral relevance. If, on the other hand, we define being human in terms which are not tied to biological classification, if (for example) we treat as the properties which make us human the capacities for reasoning or for self-consciousness, then indeed those capacities may be morally relevant, but if they are morally significant at all, they are significant whether they are the properties of a being who is a member of our species or not. And so it would be better to use a word like ‘person’ to mean a being that has these properties, to bring out the fact that not all human beings have them and that non-human beings conceivably might have them.

I wish from my Heart, I could avoid concluding, that since Morality, according to your Opinion as well as mine, is determin'd merely by Sentiment, it regards only human Nature & human Life. … If Morality were determin'd by Reason, that is the same to all rational Beings: But nothing but Experience can assure us, that the Sentiments are the same. What Experience have we with regard to superior Beings? How can we ascribe to them any Sentiments at all? They have implanted those Sentiments in us for the Conduct of Life like our bodily Sensations, which they possess not themselves. (Letter from Hume to Hutcheson: March 16, 1740)

When we ask whether morality ‘regards only human nature and human life’ we might be concerned with one of two kinds of question. We may be asking what kinds of being could share our moral point of view. What is the potential scope of the community of moral agents and assessors? Is the moral point of view essentially a human point of view? Could we, or should we, adopt a wider standpoint (say that of rational agency as such) which would leave room for a significant moral dialogue with non-human moral agents, if there are any? Alternatively, we could be asking whether only human beings should be the objects of moral concern or whether we should widen the circle of concern to include other kinds of being. We may make the same point using a helpful device of Cora Diamond's. We may be concerned with what should replace x or with what should replace y in the formula: We xs assess together conduct and character in so far as it affects ourselves and our fellow ys.

David McNaughton argues that in footnote 48 of my paper I provide justification for including the severely retarded among those whom we think of as with us in being human. Let me fill in the background to that note.

Where two people differ about the application of some term to a particular object, to what can they appeal to settle their disagreement? In my paper I suggested that the disputants can legitimately draw attention to various features of the contested case to try to show the continuity, or lack of it, between this case and other instances that fall under the concept. In so doing they are offering reasons to justify their judgement in this particular case.

Are persons substances or modes? (The terminology may seem archaic, but the issue is a live one.) Two currently dominant views may be characterized as giving the following rival answers to this question. According to the first view, persons are just biological substances. According to the second, persons are psychological modes of substances which, as far as human beings are concerned, happen to be biological substances, but which could in principle be non-biological. There is, however, also a third possible answer, and this is that persons are psychological substances. Such a view is inevitably associated with the name of Descartes, and this helps to explain its current unpopularity, since substantial dualism of his sort is now widely rejected as ‘unscientific’. But one may, as I hope to show, espouse the view that persons are psychological substances without endorsing Cartesianism. This is because one may reject certain features of Descartes's conception of substance. Consequently, one may also espouse a version of substantial dualism which is distinctly non-Cartesian. One may hold that a person, being a psychological substance, is an entity distinct from the biological substance that is (in the human case) his or her body, and yet still be prepared to ascribe corporeal characteristics to this psychological substance. By this account, a human person is to be thought of neither as a non-corporeal mental substance (a Cartesian mind), nor as the product of a mysterious ‘union’ between such a substance and a physical, biological substance (a Cartesian animal body). This is not to deny that the mind—body problem is a serious and difficult one, but it is to imply that there is a version of substantial dualism which does not involve regarding the ‘mind’ as a distinct substance in its own right.

My topic is personal identity, or rather, our identity. There is general, but not, of course, unanimous, agreement that it is wrong to give an account of what is involved in, and essential to, our persistence over time which requires the existence of immaterial entities, but, it seems to me, there is no consensus about how, within, what might be called this naturalistic framework, we should best procede. This lack of consensus, no doubt, reflects the difficulty, which must strike anyone who has considered the issue, of achieving, just in one's own thinking, a reflective equilibrium. The theory of personal identity, I feel, provides a curious contrast. On the one side, it seems highly important to know what sort of thing we are, but, on the other, it is hard to find any answer which has a ‘solid’ feel.

The central fact about the problem of personal identity is that it is a problem posed by an apparent dichotomy: the dichotomy between the objective, third-person viewpoint on the one hand and the subjective perspective provided by the first-person viewpoint on the other. Everyone understands that the mind/body problem is precisely the problem of what to do about another apparent dichotomy, the duality comprising states of consciousness on the one hand and physical states of the body on the other. By contrast, contemporary discussions of the problem of personal identity generally display little or no recognition of the divide which to my mind is at the heart of the problem. As a consequence, there has been a relentlessly third-personal approach to the issue, and the consequent proposal of solutions which stand no chance at all of working. I think the idea that the problem is to be clarified by an appeal to the idea of a human being is the latest manifestation of this mistaken approach. I am thinking in particular of the claim that what ought to govern our thinking on this issue is the fact that human beings constitute a natural kind, and that standard members of this kind can be said to have some sort of essence. Related to this is the idea that ‘person’, while not itself a natural kind term, is not a notion which can be framed in entire independence of this natural kind.

Most of us, at one time or another, will have been struck by a thought that we might wish to express in the following words: ‘I could have been born in a different time and place, my position in life and all my personal characteristics could have been completely different from what they are; how amazing then that it should have fallen to my lot to live my life, the only life I shall ever live, as this particular individual rather than any other.’ This thought need not derive from a sense that there is anything unusual about one's life; what it expresses, rather, may be the sense that there is something gratuitous or contingent about one's being any particular individual at all. This sense of contingency might be connected with a feeling of gratitude, perhaps of responsibility towards others less fortunate in life; or it might be bound up with envy, or pride, or self-pity, etc.

Suppose that someone writes an argument on a blackboard which leads to the conclusion that he may, at that time, be dreaming. He goes over it, considers its validity, the truth of its premises, its assumptions and so on, and then to his dismay, he judges that he is compelled to conclude that he may be dreaming. He goes over the argument repeatedly and carefully, but finds the conclusion ‘inescapable’. If reviewing the argument on the blackboard may be taken as an analogue of reviewing thoughts before one's mind, then his condition seems like the condition which Descartes describes at the beginning of the Second Meditation: ‘The meditation of yesterday filled my mind with so many doubts that it is no longer in my power to forget them. And yet I do not see in what manner I can resolve them; and just as if I had all of a sudden fallen into very deep water, I am so disconcerted that I can neither make certain of setting my feet on the bottom, nor can I swim and support myself on the surface.’

In a footnote to the preface to the second edition of his Critique of Pure Reason (1787) Kant remarked that ‘it still remains a scandal to philosophy and to human reason in general that the existence [Dasein] of things outside us … must be accepted on faith, and that if anyone thinks good to doubt their existence, we are unable to counter his doubts by any satisfactory proof’ (B XL). In Being and Time (1927) Heidegger remarks, somewhat less famously, that the scandal of philosophy, far from being the continuing absence of philosophically satisfactory proof of the existence of the world outside human subjectivity, is rather-the-very idea that such proof need be sought at all: ‘If Dasein is understood correctly, it defies such proofs, because, in its being, it already is what subsequent proofs deem necessary to demonstrate for it’ (BT, 205).

The renaissance of philosophy of mind within the analytical tradition owes a great deal to the intellectual midwifery of Ryle and Wittgenstein. It is ironic, therefore, that the current state of the subject should be one in which scientific and Cartesian models of mentality are so widely entertained. Clearly few if any of those who find depth, and truth, in the Wittgensteinian approach are likely to be sympathetic to much of what is most favoured in contemporary analytic philosophical psychology. Finding themselves in a minority, they might well look elsewhere for support, hoping to establish the idea that opposition to scientific and Cartesian ways of thinking is by no means philosophically eccentric. Perhaps this partly explains the increasing British and North American interest in ‘continental’ thought, particularly as it bears (as most of it does) on the nature of human beings. Husserl, Heidegger, Merleau-Ponty and Sartre are obvious enough subjects for such attention.

Cartesian accounts of the mental make it axiomatic that consciousness is transparent: what I feel, I know I feel, however many errors I may make about its cause. ‘I’ names a simple, unextended, irreducible substance, created ex nihilo or eternally existent, and only associated with the complete, extended, dissoluble substance or pretend-substance that is ‘my’ body by divine fiat. Good moderns take it for granted that ‘we’ now realize how shifting, foggy and deconstructible are the boundaries of the self; ‘we’ know that our own motives, feelings and intentions constantly escape us; ‘I’ names only the current speaker, or the momentarily dominant self among many fluid identities.

The answer to the title question which I want to defend in this paper is ‘none’. That is: I doubt strongly that the notion of ‘a self’ has any use whatsoever as part of an explanans for the explanandum ‘person’.

Put another way: I shall argue that the question itself is misguided, pointing the inquirer in quite the wrong direction by suggesting that the term ‘self’ points to something which can sustain a philosophically interesting or important degree of reification.

The question about ‘identity’ under consideration in this paper is different from the one discussed in some of the other papers—for instance by Geoffrey Madell and Lars Herzberg. That question arises from the fact that human beings change in appearance and behaviour in the course of their life. By and large we have no trouble in recognizing them but we may wonder what it is that remains the same in them or about them so that we recognize them, address them by the same name, respond to them as to someone we know. What is it for a person to be the same person through these changes? By virtue of what do we call them by the same name?