Disembodied Minds and Personal Identity

Abstract

Discussion of the human soul has bulked large in the literature of philosophy and religion. I defend the possibility of disembodied Cartesian minds by examining the criticisms of three philosophers who argue that there are serious difficulties about any attempt to account for the identity of such Cartesian minds through time. I argue that their criticisms of the possibility of disembodied minds are damaging but not fatal. I hold that the central issue behind their criticisms of Cartesian minds is whether any nonphysical mental criterion can be formulated for the identity of such entities. Even though no such criterion can be given, disembodied minds that persist through time remain logical possibilities. Inlhree distinguished philosophers-Peter Strawson, Terence Penelhum, and Derek Parfit-have given arguments against the existence of disembodied Cartesian minds based on considerations about personal identity. I shall rebut their arguments. Although I see no convincing reason to believe there are disembodied Cartesian minds, I think they are distinct logical possibilities_ I will defend this possibility against these arguments because I do not think the arguments hold any water. P.F. Strawson gives a very typical criticism of Cartesian dualism and disembodied minds. He says that in order to be able to reidentify individual items of any kind, we must first be able to identify them. In order to identify any given item, we must have a way of individuating items of that kind; we must know when we have one such item rather than two. Cartesian dualists, however, hold that the concept of a mind is genuinely independent of talk about a person, a human being or a man. For a Cartesian,the concept of a mind is not dependent on the concept of a person the way talk of surfaces is dependent on the concept of a material object. It is essential to Cartesianism "that the application of the notions of identity and numerability of souls (consciousness) should not be determined by their application to persons."1 We know how to apply the concepts of identity and difference to individual human beings, but the Cartesian cannot rely on this since, "the concept of the identity of a soul or consciousness over time is not derivative from, dependent upon, the concept of the iden416 THOMAS W. SMYTHE tity of person over time".2 The Cartesian must either admit that the concepts of identity and difference of minds are derivative from the concepts of identity and difference of human beings, or supply us with an independently intelligible account of the individuation and identity of Cartesian minds. The former amounts to giving up Cartesian dualism. The latter cannot be done since no mental criteria for personal identity are sufficient of themselves. Therefore, Cartesian dualism and talk about disembodied minds is mistaken. I shall consider Strawson's criticism as it pertains only to the real possibility of disembodied minds. The main point is that we lack any way of identifying and individuating disembodied minds. There would be no way to pick out some other disembodied mind A rather than B because there is no observation we could make to show we have one mind A rather than B. The knowledge we could in principle have of the identity of any disembodied mind would have to depend on our ability to identify and reidentify such entities, and the identification of a disembodied mind requires that we be able to pick out or individuate such an entity. But there does not seem to be any way of individuating Cartesian minds which would allow us to identify them for they do not occupy space. One might say that two distinct disembodied minds are distinct because they have different mental histories, but this will not help us pick out one mind from another. Since the identity of disembodied minds over time depends on being able to identify them, there is no reason to think any account of the identity of disembodied minds may be given. The criticism is not just dependent on the impossibility of having any physical means of identification for disembodied minds. A disembodied mind might appear to us in the guise of a fire or vary its appearance in different mediums such as a voice, a beam of light, and so on. The difficulty is one of knowing that it is one disembodied mind A rather then B which appears to us this way (since the Cartesian mind and its properties itself never appear to us even when embodied), and knowing it is the same mind which appears to us now one way and now another. If we assumed that a disembodied Cartesian mind can act on, or initiate change in, the physical world, then we still may be able to formulate a criterion for the identity of minds. However, if we fail we cannot conclude that the account of the identity of disembodied minds is impossible in principle. We may still be able to know of the identity of such minds in some other way. Such a way of knowing the identity of disembodied minds can be made intelligible by considering the notion of a "criterion" for personal identity construed in terms of necessary evidence. It is reasonable to believe that memory claims, when sincere and confident, count as a criterion for personal identity because it is inconceivable that a memory claim made by a person should be irrelevant as evidence for his identity, or that his memory claims would not count as evidence for his identity. It is necessarily true that a person's sincere and confident memory reports made about his past history count as some evidence for his past identity. This way of telling whether memory claims are criteria for personal identity involves asking ourselves whether we can imagine a case where sincere and confident memory statements could ever fail to carry some presumption that the person who utters such statements is identical with a certain past DISEMBODIED MINDS AND PERSONAL IDENTITY 417 self. This method of deciding whether memory statements count as necessary evidence for personal identity, which has become well known in the literature of personal identity, can be applied to disembodied minds. In this sense of "criterion" it seems that a disembodied mind could speak to us through a medium, or out of blue, and make a memory claim to be a certain person. Suppose we hear a voice that says or asserts that he is a certain person we knew in the past; the person thereby identifies himself as being a certain person we knew in the past and previously identified. It would seem, off hand, that his claim or assertion that he is identical with a certain person we previously identified could not fail to be evidence that he is the person. Certainly we would be inclined to place as much weight on the memory statements of disembodied minds as we would on memory statements made by embodied persons. Sydney Shoemaker has argued that memory is a criterion for personal identity in a more complicated and careful way. He gives some arguments for the necessary truth that sincere and confident memory statements are generally true. If so, inferences of the form, "He claims to remember doing x, so he probably did x", are noninductive and memory claims are criteria for personal identity? If all of what Shoemaker says in support of this is adequate (and I am inclined to think there are some serious problems here), one might be able to argue, in a similar way, that it is a necessary truth that sincere and confident memory claims made by a disembodied mind are usually true. Whether Shoemaker is correct and whether or not it is a conceptual truth that honest and confident assertions by people that they are identical with people previously identified are generally correct, I think enough has been said to dispel the view that it is impossible to give an account of the identity of another disembodied mind. Thus, Strawson is not justified in concluding that there is no reason to think any account can be given of the identity of an individual consciousness or a disembodied Cartesian mind. Strawson is open to just this criticism since he says we can conceive of disembodiment in a secondary way, derivative from our existence and identity as a person with physical properties. He seems to think that personal identity, once established, has a kind of inertia that can keep it intact even after physical properties have dropped away. Thus my criticism is at least consistent with what Strawson himself says. Let us now turn to what is apparently a more serious difficulty for disembodied minds based on considerations about our knowledge of personal identity. Terence Penelhum has an even more basic attack on the notion of disembodied minds. It is based on difficulties in saying we have the same disembodied mind at different times.s His criticism is not based on the problem of identification and individuation of Cartesian minds, but more directly on considerations necessary for speaking of the identity of any item through time. The cogent part of Penelhum's attack stems from his application of Shoemaker's arguments that there is a sense in which bodily identity is the most fundamental criterion for personal identity. The most basic consideration is the way in which memory depends on local bodily identity in order to be used as a criterion for personal identity. In order to recognize people's personal memory claims, we must be able to identify and recognize them in a bodily way over a long enough period of time to organize and understand their memory 418 THOMAS W. SMYTHE claims. This requires some bodily way of telling that we are conversing with the same person which is independent of memory. In addition, for disembodied minds, we will also need bodily identity to show that a person is using words like "remember" correctly and as the most fundamental way of distinguishing correct and incorrect memory claims. This is because there can be no reidentification of any physicalistic realization of memories in disembodied minds. The effect of this is to make the memory criter