home | philosophy index | bibliography | comment

Part II: Identity

by David Foss

March 18, 1994

Williams and Parfit

What is this thing personal identity. It seems that, while a great deal of attention is being cast upon the concept of identity, the underlying notion of personality is, at best, being taken for granted. Williams nicely argues that any attempt to “externally” equivocate across bodies in terms of a single “person” cannot distinguish between numerical identity and mere similitude. From the “outside,” any claim as to the identity of [embodied] person X and [embodied] person Y, where X and Y are not spatio-temporally contiguous (in bodily terms), can appeal to no evidence which will not more readily support a thesis of “clairvoyance” (as strange as asserting that X is the same exact person as Y), coincidence, or confusion. From the “inside,” there seems even less of a reason to suppose that a person could wonder, “Am I the person I remember myself to be?” without risk of contradiction, confusion, or merely asking “Do I really remember doing X or witnessing Y?” Perhaps the most helpful notion developed in the course of his discussion, Williams points out that persons, identified roughly in terms of personalities, cannot be person-tokens, but must remain person-types. It is the particular embodiment, or so Williams claims, that makes a peculiar person-type a particular person-token.

In Parfit we see, most obviously, a confusion as to the content of person-tokens. It seems here that the principle content of personality is considered a set of memories, intentions, desires, etc., such that a change of body is no more dramatic than, say, changing the water in the vat (or changing vats) for our classic “brain in the vat.” Williams asserted the rather strong (albeit believable) claim that personality is significantly tied to the peculiar embodiment of a mind. There is a certain symbiosis between brain and body, such that personality is as much a sort of organic congress of body, thoughts, desires, relations with others, and so on, such that the “mere” change of body destroys whatever continuity we had been using to ensure personal integrity. Admittedly, Parfit disagrees with Williams far less than he seems to believe. But a central mistake still seems to be made in terms of the substantive individuating conditions, or characterization, of the person, such that persons are thought to hold a merely convenient parasitic relationship with the body they currently inhabit. There can be no evidence, from the “outside” or the “inside,” which could support such a view. Worse, such a view considers friends, loved ones, political affiliations, etc., as utterly extrinsic to a person’s “identity” (understood in terms of lived character).

March 31, 1994

Second Assignment: Immaterial Duplicity

Is bodily identity necessary for personal identity? (That is, is it the case that necessarily, if x is the same person as y then x has the same body as y?)


Does it even make sense to ask of oneself, “Am I the same person I was yesterday.” Surely, if I am asking whether I was me yesterday, this is trivially true (or absurd, for the case in which I did not exist). So I cannot be speaking about an identity relation between, or making an identity claim of, two “I”’s which are both me (without strangely finding within me two me-parts; one corresponding to the “I” of yesterday, and the other to the “I” of now, but both being the self same I now). From the “inside,” the question of identity hinges on the logical truth x=x, and can only be a question for someone who considers it possible that x=x is false. The question of identity only arises for objects viewed from the “outside”. Even in the case of myself, I can only intelligibly wonder (that is, consistently consider the falsity of) whether I am the same person I was yesterday so long as I suspend the belief that the person yesterday who highly resembled me today, and whose experiences I seem to remember, was me; supposing instead that I am possibly mistaken, and that the person I believe to have been me yesterday, bears no more special relations to me than any other person. I am forced to abandon the internal perspective in such cases, for the simple reason that the question only arises when I do. Taking refuge in a strictly internal perspective only seeks to solve the “problem” by denying it logical validity, and thereby loosing the question.

But what are these perspectives I’ve called “internal” and “external?” So far, the internal perspective need merely involve a commitment to identity ex hypothesi. That is, what characterizes an internal perspective (in the relevant sense here) is that it take for granted the identity of each side of an identity claim (x is the same P as y) such that the identity claim becomes a logical truth. In the case of personal identity, this internal perspective takes the form of “consciousness” or “psychological continuity.” It is attractive to appeal to such notions when considering the identity of oneself over time. It seems there is something special about one’s relation to oneself, such that the question of identity over time is both nontrivial (in content and consequence) and definite. There is something terribly dissatisfying about an answer to the question of personal identity which locates me ambiguously in the future or the past. I wonder if I was that person I see in the picture. It is not enough to know, as the metaphysician will tell me, that I may be identical with that person from some perspectives but not from others. “Either I was there, or I was not,” I will protest, “How could it be both?”

Intuitively, when I ask, “Was I there?” I am asking something like, “If I could remember being there, would such a memory be true [veridical]?” or “Was it this same stream of consciousness which passed through that place, time, and body which now experiences and feels things as I?” These would be strictly internal questions if they sought resolution from entirely within the subject; wherein these questions could all be reduced to one of two forms, “Was I me then?” or “Am I me now?” They are external to the extent that they ask, “Was that me there [or then]?” or “Is this me now?” where the “that” and the “this” are ostensive acts pointing in some manner away from the subject. This pointing away, or differentiating in order to open the possibility of non-identity, can only be done with respect to particulars. And particulars, it seems, need to be embodied. For this is what we normally mean by a particular: a thing; a physical entity; a spatiotemporal instance.

Persons are particulars. Identity claims can only be made of particulars (To say two blue objects have the same color, is just to say that the particular blue-shade of object A is the same as the particular blue-shade of object B. It is, in other words, to ask about two identifiably distinct objects). Bernard Williams (in Personal Identity and Individuation, pp.155-165) claims that psychological criteria, as applied “from the outside” can only characterize type-persons, or capture universals. Physical or bodily criteria, on the other hand, as applied from the outside, clearly characterize token-persons, or refer to particular people. Therefore, a bodily criterion for personal identity is indispensable. Whatever else might be said of persons and the relation or property of identity, it is true that [personal] identity concerns persons, is intelligibly asked of persons only from the outside, and persons are from the outside only clearly distinguishable (or identifiable) by bodily criteria.

There are then two inferences which force the necessity of a bodily criterion of personal identity to the fore. The first is that identity claims only make sense with respect to particulars (as viewed in part or in whole, from the outside). The second is that persons are only particular either, when they are viewed from the inside according to psychological criteria, or when they are viewed from the outside according to bodily criteria. Because psychological criteria are only able to individuate a person from the inside, it can never even pose the question of identity with respect to that same person. To the extent that it can (I wonder whether I had the thought that p) it must implicitly appeal to a spatiotemporal distinction (I wonder whether there was a person s.t. that person is spatially and/or temporally distinct from me, thought that p, and is [in some manner] identical with me). Once the question is about some object x, and its properties or relations with respect to me, then purely psychological, or strictly internal, criteria are no longer able to bridge the gap. It is no longer an internal question. And once external, the question cannot be framed or answered (or so it is claimed) without appeal to bodily identity.

This final point needs clarification, and at least some hand-waiving at a proof. But we should first distinguish between treating personal identity as bodily identity and the appeal to bodily identity as providing some necessary criteria for personal identity. Obviously, in the normal course of events, when we in fact wonder about the identity of ourselves or others through time, we let bodies (our own and others’) provide the answers. The reason, simply enough, is that we are unable to know anything about psychological continuity with respect to others, or even past selves, without inferring it from discernible activity of their bodies. This is plainly an epistemic restriction, and one which may have nothing to do with the actual content of the relation (or property) we are seeking to affirm or deny. In this sense, then, bodily identity is (most of the time) taken as a necessary and sufficient condition for personal identity. This is simply to reduce the question of personal identity, to the question of bodily identity, for the sake of pragmatic constraints upon our cognitive powers as the sort of rational beast that we are. The criterion we in fact use, then, is what Thomas Reid disdainfully called “similitude.” So a claim that bodily identity (of a special sort of physical body) is personal identity, would seem to entail that similitude is identity when it comes to persons. And this seems to return us to the strange predicament in which there may be no truth about whether I am the same person I was yesterday (read in its most non-trivial sense).

There are two ways this may be avoided. First it might be objected that, in fact, while we can only use similitude to make a guess about bodily identity, it does not follow that similitude is bodily identity any more than it would in the case of personal identity. Second, it could be said that while it may not be sufficient for a person to remain the same person by their body holding some similitude relation with a previous body, it may nevertheless be necessary. I think that both of these objections can and need to be made in response to Reid. The use of a bodily criterion, no more than the use of a psychological criterion or an immaterial criterion, does not entail similitude as the guarantor of identity. “Same body” is no more “identity by similitude” than “same psyche.” In both cases external identity attributions are epistemically limited to measures of similitude, and in neither case are we committed to the view that such similitude is identity.

Still, troubles with bodily identity might be communicated directly to personal identity if personal identity were simply being equated with bodily identity. Here the distinction between bodily identity as necessary and sufficient for personal identity (in which case personal identity simply is bodily identity), and bodily identity as necessary but not sufficient for personal identity (in which case various bodily criteria must be met in order to satisfy personal identity) is critical. Any objection which seeks to undermine the necessity of the satisfaction of bodily criteria for personal identity, on the ground that bodily identity is inadequate for the task or is itself incoherent, must show that a bodily criterion actually blocks important aspects of the identity claim. It is not enough to say that a body analysis of personal identity can not capture the whole picture. It must be shown that a body analysis prevents a coherent identity claim by the introduction of paradoxes no other criteria may be adequate to heal, or that a body analysis is irrelevant.

The issue of relevance is an important one. But we should first respond to the effort to show that bodily criteria actually corrupt or prevent a proper analysis of personal identity. There are, admittedly, clear problems with respect to our intuitions about bodily identity for bodies of things other than persons. And it is rather unnerving to suppose that these problems might become intractable when applied to ourselves. Bodies are divisible, and I, it would seem, am not. We normally speak of a table being the same table in spite of it now missing, say, a leg. So with me also. If I loose my leg in an accident, I would certainly consider myself to be the same person (minus a leg). But what about that particular table, or my particular body prior to, and then after, the accident. Or more dramatically, consider that over the course of the past seven years almost every molecule in my body has been replaced, and that I now have a totally different set of “parts” from those I had just seven years ago. This sort of identity is what Reid calls “imperfect.” Reid goes further, and claims that such imperfect identity is only called identity for convenience. It is, curiously, not “real” identity. Real (or perfect) identity, we are asked to suppose, is a relation between a thing and itself. So long as there are (real) parts, there is no (real) identity. If bodies have (real) temporal parts, then they are non-identical through time; and if persons require bodily identity in order to maintain identity, then I am no more the person I was before I lost my leg than the table is the self-same table with and without its leg.

But, Reid complains, “A part of a person is a manifest absurdity.” (Of Identity and On Mr. Locke’s Theory of Personal Identity, p.125) This seems a puzzling claim. We certainly do talk as if there were parts of a person. Someone who has been severely brain damaged is said to have lost a part of herself, whether it be memory, substantial aspects of her character, or what not. And we still understand what is being said. To say that such an idea is manifestly absurd implies that it would make no sense to talk of such changes as real, concrete, and terrible. A close friend who has been in a car accident, and as a result has suffered serious brain damage, will be the same person only in a limited or partial sense. A part of the person has been lost. At least, this is the way in which we normally describe the situation. Is this pure nonsense? I think not.

What Reid finds troubling about the “softness” of bodily identity is that it threatens the conceptual clarity and unity of the I. What I have suggested is that we have good reasons for supposing (from actual occurrences) that this clarity and unity is itself a convenient fiction. Bodily identity, however, does present a different sort of problem, which may be more troubling here. In what ways does bodily change effect personal change? Clearly, some bodily changes do not effect my identity (in the logical sense): I cut my fingernails, and I am still the same person I was before I cut them. Clearly, however, some bodily changes do effect my identity: My brain damaged friend being a perfect case. Admittedly, those cases in which a clear non-identity exists between a person now, and “who they were” (whatever this means) five or ten years ago, depends as much upon bodily non-identity (of a special sort) as upon psychological non-continuity. After the accident, it is not merely a manner of speaking which makes it the case that my friend is no longer numerically the same person. Memories are lost, deep character traits are suddenly changed, the brain and body have suffered substantial trauma, and physical and cognitive structures have been significantly altered or lost.

Unfortunately, a case of this sort shows that whatever problems are seemingly generated by bodily criteria (eg. identity as an imperfect relation) were there all along. Psychological continuity cannot help us any more that bodily continuity. And the appeal to some non-psychological and non-physical substrate to straighten out the problem succumbs to the problems identified by Locke: that it can contribute nothing to personal identity unless it include psychological criteria. Otherwise, the substrate alone cannot distinguish between clearly distinct persons (because it is not particular). If it is psychologically particular, then it no longer remains a criterion for perfect identity between my friend before the accident and my friend after the accident.

This should be sufficient to show that bodily criteria do not introduce problems into questions of personal identity which we might avoid by appeal to exclusively non-bodily criteria. Still, it might be easier to show that bodily criteria are not needed to explain personal identity. Unfortunately, whether we appeal to some notion of veridical memory, causally linked temporal parts (ala Armstrong, Identity Through Time, pp.147-154), some immaterial substance, or some sort of brain criterion, we need the criterion to establish a relation between two spatiotemporally distinct objects. The concept of veridical memory, in order to pose any threat to bodily criteria, needs to be strictly non-physical while at the same time “external” to the judgment (or the judge). And this does not seem conceptually possible: the concept of memory entails bodily continuity. Likewise, any brain criterion for personal identity is going to be a bodily criterion. Can there be a concept of same brain that is not dependent upon bodily continuity? Similarly, any notion of causal connectedness intended to unify persons (or bodies in general) across time, such that bodily continuity were now unnecessary, would have to be built upon a non-continuity notion of causality. Perhaps I am simply not smart enough, but I cannot even begin to understand what causality could mean if it did not require spatiotemporal continuity. It seems that in order not to be a bodily criterion itself, the notion of causality would have to not rely upon a physicalistic core.

The proposal of an immaterial substance, or a substance unnameable and unknowable, which does the work of making an identity claim true or false, as I have said, seems blatantly inadequate unless it incorporates individuating psychological or physiological traits. And these would, then, be the actual criteria for personal identity. Perhaps the correct account of personal identity will ultimately look a great deal Aristotelian. In order that a person be particular, in order that a person be the sort of thing for which spatiotemporal identity is neither merely a logical truth nor an fiction, a person needs to be embodied (as a particular character). This is not an epistemic constraint upon the occurrence of particulars, but a conceptual one. In so far as the conceptual structure of the question accurately maps the real structure of the problem (which I take it necessarily to do; otherwise the “problem” will not have even been asked about), the conceptual constraints upon the question (i.e., the necessity of embodied particularity) entail real constraints upon the problem.

Proseminar: Metaphysics
PHIL-767-01, Georgetown University
Spring 1994
(© David Foss, 1994)

Send comments to hok007@shlobin-foss.net
Last modified January 30, 2000

home | philosophy index | bibliography | comment