It is well known that, in debates concerning ‘personal identity’, philosophers are apt to invoke fantastical thought experiments in order to illustrate some point or other for which they wish to argue. One such thought experiment, or genre of thought experiments, bears some similarities to what many millions of people around the world and throughout human history have believed to be possible and to have actually happened from time to time. The genre in question is that in which it is imagined that someone claims to remember having undergone certain experiences that she could not possibly have undergone if we assume that her existence as a person began only in, or at the beginning of, her current biological lifetime. In short, it involves imagining that someone claims, sincerely, to remember experiences from a ‘previous life’ and to be the person who underwent those experiences – to be, in other words, a reincarnation of the person who lived that previous life and who has now been ‘reborn’ in this one.

Probably the best-known use of a thought experiment belonging to the genre just outlined is that of Bernard Williams in his paper ‘Personal Identity and Individuation’, which was first published in 1957.Footnote ¹ Williams used his version of the thought experiment as part of an argument for the conclusion that bodily continuity is a necessary condition for personal identity. Hence he drew from it the conclusion that a scenario of the kind that the reader is invited to entertain in the thought experiment could not occur. The strength of this ‘could not’ was held by Williams to be logical, in the sense that to suppose that it could occur would be to misdescribe the scenario; it would be to wrongly apply the concept of identity when, at most, only the concept of ‘exact similarity’ is applicable. Williams’ paper made a stimulating contribution to the philosophical literature on personal identity and sent forth ripples that reverberate through into contemporary discussions. Among the philosophers who were impressed by Williams’ argument was the young D. Z. Phillips, who endorses it in the context of a very brief consideration of reincarnation in his book Death and Immortality. Phillips’ endorsement of the argument involves using Williams’ thought experiment in order specifically to discredit the belief in reincarnation whereas Williams himself had used it for the more general purpose of discrediting denials that bodily continuity is necessary for personal identity to be retained. But these two uses are very closely connected, and entirely mutually compatible.

In the present paper, I will take Williams’ thought experiment, and Phillips’ use of it, as a starting point for reflecting upon beliefs in reincarnation and the intelligibility thereof. Having outlined the features of the thought experiment that are most salient for Phillips’ purposes, and having noted Phillips’ favourable response to Williams’ conclusion, I will then consider two alternative ways of responding. One of these is exemplified in a thoughtful paper by David Cockburn, wherein Cockburn examines the ethical implications of regarding someone, specifically a child, as a reincarnation of a deceased individual, such as a sibling who died before the present child was born. The second way of responding to the sorts of issues raised in Williams’ thought experiment that I will consider involves paying attention to the work of those who have studied the traditions and cultures wherein beliefs in reincarnation or rebirth have a central place. Of particular significance for one of the key strands in Williams’ argument is a belief – held by some Amerindian and Inuit people, among others – in a phenomenon that anthropologists have dubbed ‘multiple simultaneous reincarnation.’ Taking this latter form of belief seriously – that is, seeking to improve one's understanding of it, and acknowledging that it does in fact have a place in certain people's lives – will, in my view, help to deepen philosophical inquiries into reincarnation and personal identity. I will argue that, in contrast with the somewhat shallow appreciation displayed by both Williams and Phillips concerning the possible senses that a belief in reincarnation may have for someone, both of the ways of responding to Williams’ thought experiment that I shall be considering in sections 2 and 3 have the potential to enrich our appreciation of those senses.

1. Allegedly fatal objections to a belief in reincarnation

In the opening chapter of his little book Death and Immortality D. Z. Phillips, as he admits, ‘simply note[s] without a great deal of discussion some logical objections to various suggestions about the possibility of survival after death.’Footnote ² According to Phillips, ‘The notions of the survival of non-material bodies, disembodied spirits or new bodies, after death, all seem open to fatal logical objections.’Footnote ³ Among the notions that Phillips takes to be (or ‘seem’) ‘open to fatal logical objections’ is that of reincarnation, which he briefly characterizes as ‘the departed soul liv[ing] again in a new body.’Footnote ⁴

Phillips’ reason for holding that beliefs in reincarnation are logically suspect is his acceptance of the argument offered by Bernard Williams for the conclusion that bodily identity is a necessary condition for personal identity. The thought experiment, which (for Phillips’ purposes) forms the crucial part in the argument, requires us to suppose that someone named Charles claims, upon awaking one day, ‘to remember witnessing certain events and doing certain actions which earlier he did not claim to remember; and that under questioning he could not remember witnessing other events and doing other actions which earlier he did remember.’Footnote ⁵ The question in which Williams is interested is whether this scenario would give us grounds for regarding the man who has woken up as a different person from the one who earlier went to sleep in the same bed. Williams subsequently elaborates the thought experiment slightly by inviting us to suppose that, after investigating the events and actions that Charles is now claiming to remember, ‘our enquiry has turned out in the most favourable possible way, and that all the events he claims to have witnessed and all the actions he claims to have done point unanimously to the life-history of some one person in the past – for instance, Guy Fawkes.’Footnote ⁶ In such circumstances, Williams suggests, the temptation to say that Charles is now Guy Fawkes (or to say something similar to this) is certainly ‘very strong’, and it ‘is difficult to insist that we couldn't say that Charles (or sometime Charles) had become Guy Fawkes’.Footnote ⁷

Williams gives us a clue as to the force of ‘couldn't’ in the sentence just quoted when he subsequently writes that declaring Charles to be Guy Fawkes ‘is certainly what the newspapers would say if they heard of it’.Footnote ⁸ It thus appears that he is suggesting that, since some people (such as newspaper journalists) certainly would say that Charles is now Fawkes, it would be difficult and implausible to maintain that no one could say this. But in view of where Williams’ argument ends up, he would have to say that, regardless of what some people would say, they ought not to say it, because it is logically incorrect. From what Phillips says, I take it that he would concur with Williams’ judgement on this matter.

In order to ward off the temptation to refer to Charles as Fawkes, Williams embellishes his thought experiment in a way that generates what he calls his ‘reduplication argument’.Footnote ⁹ This involves supposing that, simultaneous to the changes that Charles has undergone, someone else, such as his brother Robert, has undergone ‘the same changes’.Footnote ¹⁰ In this new scenario, Williams contends, it would be ‘absurd’ to maintain that Charles is Fawkes. It would be absurd because there is no more reason for regarding Charles as Fawkes than for regarding Robert as Fawkes; and it would be absurd to regard them both as Fawkes because, firstly, ‘if they were, Guy Fawkes would be in two places at once, which is absurd’; and secondly, ‘if they were both identical with Guy Fawkes, they would be identical with each other, which is also absurd.’Footnote ¹¹ On Williams’ view, from the absurdity of regarding either Charles or Robert as Fawkes when both Charles and Robert exist contemporaneously, it follows that it would also be absurd to regard either of them as Fawkes even if only one of them existed at a particular time. This inference follows on the basis of a principle which Derek Parfit, in his interpretation of Williams’ argument, states in the following terms: ‘Whether a future person will be me must depend only on the intrinsic features of the relation between us. It cannot depend on what happens to other people.’Footnote ¹²

In his précis of Williams’ argument, Phillips expresses approval of Williams’ contention that the distinction between a relation of ‘identity’ (if we wish to call numerical identity a ‘relation’) and a relation of ‘exact similarity’ can operate only in the case of bodies, and not of memories. Williams maintains that, to say of Charles (or Robert) that he has memories that are identical with – i.e., they are the very same memories as – those of Fawkes, in the absence of continuity between the respective bodies of Fawkes and Charles (or Fawkes and Robert), fails to mark a contrast with saying that the memories are merely exactly similar. This is because the only means of distinguishing a relation of identity from one of exact similarity in the memory case, is to trace a continuity of body between the person who claims to remember having done or experienced a given action or event and the person who actually did or experienced the action or event in question. ‘What is wrong with the supposed criterion of identity for persons which relies only on memory claims’, Williams avers, ‘is just that “… being disposed to make sincere memory claims which exactly fit the life of …” is not a one–one, but a many–one, relation, and hence cannot possibly be adequate in logic to constitute a criterion of identity.’Footnote ¹³ Summing up his agreement with Williams, Phillips writes: ‘Thus, it seems, if a claim is to be borne out that so-and-so is so-and-so living again, it must be established that there is a one–one relation between the material bodies involved in the two spans of existence.’Footnote ¹⁴

In a later book, Phillips makes a passing reference to reincarnation, and there states that he is ‘not denying such beliefs’.Footnote ¹⁵ There, and in other places, Phillips is keen to maintain a distinction between, on the one hand, religious beliefs as they operate in the lives of actual religious believers, and on the other hand, the metaphysical constructions of philosophers which are liable to bear a merely terminological similarity to the religious beliefs in question. Phillips emphasizes this in, for example, his discussions of the term ‘soul’. ‘In philosophy of religion’, he writes, ‘we are often offered analyses which pay no attention to and, hence, fail to capture, the “soul” in the words of religious beliefs.’Footnote ¹⁶ Phillips describes these failed analyses as involving ‘a dislocation of language’. By this he means that certain terms or phrases, which have a vivid life in the beliefs and practices of religious believers, are abstracted from their contexts by philosophers. By uprooting forms of language in this way, philosophers inevitably overlook the connections with lived human activity that enable us to see the meanings that the salient concepts have.

With regard to his treatment of the notion of reincarnation in Death and Immortality, a charitable reader might suppose that Phillips is taking Williams to have discredited only a philosopher's reconstruction of this notion, and not the belief in reincarnation as it operates in believers’ lives. For to judge that Williams’ reduplication argument has somehow shown that billions of believers in reincarnation have fallen prey to a logical absurdity – that what appears to be a profoundly meaningful aspect of their lives is, in reality, a deep misunderstanding on their part – would hardly be in keeping with the kind of Wittgenstein-influenced approach to religious forms of life that Phillips espouses. Styling his (and Wittgenstein's) conception of philosophy as a ‘contemplative’ one, Phillips, in one of his last books, asserts that philosophy's disinterested ‘contemplation of the world’ does not seek to arbitrate ‘between our beliefs and convictions in the name of rationality.’Footnote ¹⁷ Such arbitration is precluded, Phillips maintains, because the distinction between what is rational and what is irrational is not independent of the contexts in which beliefs and convictions arise.

Whether it is rational for someone to hold a particular belief will show itself in the life of the individual, and in many cases the question of whether a belief is rational may be entirely out of place. It may be out of place because, for example, the belief goes too deep in someone's life to be intelligibly evaluated in terms of rationality: it forms part of the very framework within which assessments of what is or is not rational are made.Footnote ¹⁸ A consequence of this, for Phillips, is that the role of philosophy in the context of investigating religious beliefs ‘is not to justify, but to understand’,Footnote ¹⁹ and ‘To understand what it is to embrace a religious truth is at an infinite distance from embracing it.’Footnote ²⁰ So, too, is it an infinite distance from rejecting it. For to engage in contemplative philosophical inquiry is to seek neither to validate nor to undermine the modes of human life that one is examining. A deeper understanding – clarity as an end in itself, as Wittgenstein might put it – is the exclusive purpose.Footnote ²¹

A less charitable reader of Death and Immortality, however, might note that, in his remarks on reincarnation there, Phillips does not make any distinction between actual religious beliefs and mere philosophical constructions. He appears to toss the notion of reincarnation into the bag of ‘suggestions about the possibility of survival after death’ that ‘seem open to fatal logical objections’, without giving any attention to the contexts in which beliefs in reincarnation have their home. This lack of contextualization – and hence, we might say, this dislocation of language – is, I think, a conspicuous weakness in the first chapter of Death and Immortality. And since Phillips never returned to the notion of reincarnation, other than to make passing reference to it in the note I have mentioned, we must look elsewhere for more considered approaches to this topic.

2. Ethical implications of ‘same person’

Apart from accepting Williams’ conclusion, as Phillips does, there are other ways of responding to the contention that, even when he is the exclusive candidate for being (identical to) Fawkes, Charles cannot be (identical to) Fawkes, because our regarding him as such is contingent upon there not being one or more equally plausible candidates. Section 3 will consider a possible response that involves paying attention to actual reincarnation beliefs, including the belief that someone can be reincarnated as more than one person at the same time. In this section, meanwhile, I will discuss a response exemplified by David Cockburn, who questions the claim that, as he puts it, ‘how an individual is to be thought of and treated cannot depend … on what are not facts about him’.Footnote ²²

What Cockburn is questioning is the move in Williams’ argument that takes us from an acknowledgement that it cannot be the case that both Charles and Robert are Guy Fawkes, to the conclusion that, in a case where only one candidate is present, we still could not legitimately deem that candidate to be Fawkes. This move, it will be recalled, is thought by Williams to be licensed by the fact that, even in the absence of a second candidate such as Robert, the relation in which the single candidate, Charles, stands to Fawkes remains, logically, a many–one relation. That is to say, even when it is not in fact the case that there are more than one individual displaying the apparent recollective capacities that Charles displays, it nevertheless could be the case; and that logical possibility vitiates the propriety of regarding Charles as Fawkes reborn. It vitiates that propriety because it makes the plausibility of Charles's being regarded as Fawkes dependent upon something extraneous to Charles's relation to Fawkes; namely, it makes it dependent on there not being a second candidate.

While admitting that the principle in question – of thinking of and treating an individual only on the basis of facts about the individual himself – ‘sounds very plausible in the abstract’, Cockburn submits that ‘in practice we constantly violate it’.Footnote ²³ To illustrate this latter point, Cockburn cites the examples of (a) someone's being treated as the winner of a race depending ‘not only on how fast he ran but on whether anyone else ran faster’, (b) the severity of the punishment imposed on a drunk driver depending not only on whether it was he who knocked down a child, but on whether the child dies as a consequence of her injuries, and (c) someone's attitude towards his second wife changing as a result of discovering that, despite what he had previously believed, his first wife is not in fact dead.Footnote ²⁴ Examples of this kind may, of course, be multiplied. Someone's voting for one politician, Smith, may be largely dependent on the other candidates in the election having made appalling campaigning gaffes. In a case such as this, should we say that Smith was voted for because she ran a better campaign than her rivals and that her having run such a campaign is, after all, a fact about her? Well, we could say that. But the credibility of that statement will nevertheless depend on facts about her rivals’ campaigns.

A supporter of Williams’ argument could respond to the examples just given by noting that none of them bears upon the identity of the person concerned. Whether Jones is the winner of the race, and is to be treated as such, obviously depends on facts about other contestants as well as on facts about Jones. But whether Jones is, in fact, Jones – whether this particular athlete is the same one who, say, won the triathlon in the 2008 Olympic Games – cannot be dependent on facts about anyone other than Jones. It is the significance of its being, specifically, a relation of numerical identity between a person at one time and a person at an earlier or later time that Parfit is getting at when he formulates the relevant principle as the requirement that a future person's being me depends ‘only on the intrinsic features of the relation between us’ and not ‘on what happens to other people’.Footnote ²⁵

As I understand it, however, part of Cockburn's strategy in his article as a whole is to coax us away from the assumption that the notion of ‘identity’ at issue in this debate can be straightforwardly understood. Williams is concerned exclusively with ‘bodily continuity’ and ‘memory’, and is trying to nail down which, if any, of these two conditions is necessary, or sufficient, for its being logically correct to say of someone at one time that he is the same person as someone at an earlier time. Cockburn, meanwhile, is keen not to take it for granted that we know, in abstraction from any particular context, what it means to regard someone as the same person as someone at an earlier time; not to take it for granted that this must amount to only one kind of thing – only one way of regarding someone – in all possible situations. One of the central points that Cockburn stresses is that, in at least some cases and almost certainly in the case of relating to people whom one loves, the factors that come into play in relating to someone as that particular person will include emotional and ethical ones. What Cockburn explores in interesting ways are the possible implications of this latter consideration for what might be said of cases in which one or both parents of a child have spoken of that child as the ‘same person as’ (or: ‘the reincarnation of’) a previously deceased person.

The sorts of cases that Cockburn has in mind are those that have been copiously documented by the psychiatrist Ian Stevenson and his research associates. A typical case would be one in which, at a very young age, perhaps as young as two- or three-years-old, a child begins to speak and act in ways that indicate that the child regards herself or himself as having lived a previous life. In some cases, the ‘previous life’ is that of a sibling who died before the present child was born.Footnote ²⁶ Many people who reflect upon cases of this sort are liable to assume that if reincarnation has genuinely occurred, what must have happened is that some immaterial component of a human being has ‘transmigrated’ from one physical body to another. Such people may also assume that, before any useful discussion can take place concerning the ethical implications of regarding the present child as a reincarnation of the other, we first have to determine to the best of our ability whether it is really the case that the two children are continuous in the way that the notion of reincarnation implies. Cockburn does not deny the legitimacy of asking whether the living child is really (the same as) the one who died; rather, he aims to free us from the grip of ‘a certain picture’ that dominates our thought of what arriving at an answer to this question must involve, ‘a picture which is modelled on our understanding of scientific enquiry’.Footnote ²⁷

Imagining a scenario in which a married couple disagree over whether their new child is a reincarnation of the old, Cockburn contemplates how ethical considerations may feature in their disagreement, not as consequences of the child's being regarded as (or not being regarded as) her sibling's reincarnation, but as constitutive of what it means to regard her as such. Cockburn's intriguing suggestion here is that the ethical disagreement over how the child is to be treated – over whether, for example, it would be failing to acknowledge the present child's uniqueness and individuality to speak of her as the previous child – may be the very deepest level of disagreement; there may be no level that is ‘more basic’ than this.Footnote ²⁸ Those who are unfamiliar with this way of thinking are apt to misunderstand Cockburn's suggestion here. They are apt to assume that he is somehow denying the importance of the ‘factual’ question – the question whether the child really is the reincarnation of her deceased sibling – and replacing it with questions concerning what is best for the child and how she ought to be treated. To hear Cockburn's suggestion in this way is to miss its radical point. For the point is to raise doubts about what could be meant by ‘the question whether the child really is the reincarnation of her deceased sibling’ if this latter question is assumed to be amenable to being answered independently of questions concerning how she ought to be treated. ‘For it seems that, in a straightforward sense,’ says Cockburn, ‘all the facts might be in and yet there still be disagreement about whether this [child] is [the one who died]’.Footnote ²⁹

If we apply this way of thinking to Williams’ thought experiment, we may begin to notice how little can be decided on the basis of such a thought experiment when it is sketched as thinly as Williams sketches it. Among the questions that need to be asked would be ones concerning how those who come into contact with Charles behave towards him and whether those ways of behaving are appropriate. Moreover, we would need to take seriously the thought that, when addressing the latter question, there may be no ‘deeper’, ‘more basic’, level of facts about Charles’ true identity that can be appealed to than that at which the question of appropriateness itself resides. When faced with a question such as ‘Should Charles be held morally, and perhaps legally, responsible for deeds committed by Guy Fawkes?’, those who remain convinced that there is a deeper level of facts to be uncovered will want to say that this question cannot be satisfactorily answered until we have some reliable way of determining whether Charles really is (the same person as) Fawkes. But what the line of thought that we may pick up from Cockburn urges us to do is to at least entertain the idea that it may be precisely how we respond to questions concerning moral and legal responsibility that are (partially) constitutive of what we mean by declaring that someone is the same person as so-and-so.

To put it this way is not to suggest that we first have to decide how to treat Charles, and only then will we be in a position to address the question whether he really is Fawkes. Rather, it is to suggest that it is through, or in the process of, coming to some kind of agreement about how we should treat Charles that we will come to see his identity in a particular light, a light which may present him to us as Guy Fawkes, or as connected with Fawkes in some way, or, alternatively, as having no strong connection with Fawkes at all. How we do, or would, come to see him cannot be answered on the basis of a thinly sketched thought experiment. A question of this sort could begin to be addressed, I would suggest, only in relation to the kind of detailed scenario that we find in great works of literature, or in relation to real-life cases, with all their contingencies and messiness. Such contingencies and messiness do not well serve the objectives of philosophers who want, once and for all, to settle the issue of what conditions have to be in place for x to be identical to y. But there may be good reason to suspect that settling this issue in abstraction from the contingencies and messiness of life is neither a realistic nor an especially illuminating objective to pursue.

3. Multiple simultaneous reincarnation

One of the crucial assumptions underlying the conclusion that Bernard Williams draws on the basis of his thought experiment concerning Charles and Robert is that it makes no sense to regard two simultaneously existing people as numerically identical to one another. Given this assumption, we can say that, since it makes no sense to regard Charles and Robert as the same person, neither would it make sense to regard both of them as Guy Fawkes. This would make no sense because identity is transitive: if A is identical to B, and B is identical to C, then, necessarily, A is identical to C. Taken as an abstract logical principle, it is difficult to see how the transitivity of identity could reasonably be challenged. There remains, however, a question of whether, or to what extent, a logical principle such as this is applicable to actual reincarnation beliefs.

Part of Cockburn's response to Williams’ argument is, on my interpretation, to prompt us to see that, in the context of our emotionally and ethically inflected relationships with other human beings, what it means to regard someone as ‘the same person as’ (or: ‘a reincarnation of’) a particular individual may not be well captured by abstract logical principles. This kind of response can be taken further by attending to some specific forms that beliefs in reincarnation have taken within various human communities. Especially pertinent to the sort of scenario imagined by Williams is what has been designated by some researchers, including anthropologists, as the belief in ‘multiple simultaneous reincarnation’. In an encyclopaedia entry under ‘Rebirth, simultaneous’, Norman McClelland writes:

Similarly, writing of the reincarnation belief held by the Haida people of British Columbia, Margaret Blackman remarks that it ‘does not require that one be born after the death of the reincarnate kin. Furthermore, an individual can be reincarnated into more than one person at a time.’Footnote ³¹

Antonia Mills has noted the fact that belief in multiple simultaneous (or temporally overlapping) reincarnation within some Amerindian communities seems to be connected with the prestige that attaches to having a highly esteemed chief be reincarnated within one's family. Among the Gitksan people of Northwest coast America, for example, the ‘cases of multiple simultaneous rebirths … occurred when the previous personality was wished or willed in multiple families’.Footnote ³² Here, Mills could be read as proposing a reductive psychological account of the belief in question, in the sense that she is suggesting a motivation for holding it which appeals to wishes or desires of the believers as opposed to the cognitive- or truth-value of the belief itself. In other writings, however, Mills makes it clear that she is open to the possibility that beliefs in reincarnation could be true, or, at least, ‘that some paranormal process cannot … be ruled out’.Footnote ³³

My purpose in the present paper has not been, directly, to consider whether reincarnation beliefs are true, but rather to consider whether they make sense, and, if so, what sort of sense they make. Bernard Williams argues that a reincarnation case of the sort that he invites us to entertain in his thought experiment is in fact ‘logically impossible’ and hence that it would be ‘absurd’ to believe that such a phenomenon could occur. Especially pertinent to the belief in multiple simultaneous reincarnation is Williams’ insistence that,

Is it ‘logically impossible that two different persons should (correctly) remember being the man who did A or saw E’? In order to begin addressing this question, we might imagine two Gitksan families, each of which has a child who claims to remember being the same high chief. Are these families obliged by dint of the laws of logic to restrict themselves to speaking of the children only as claiming to remember being the high chief and not as really having been him? Suppose that the two families know one another, and hence know that there are at least two children who are claiming to have been the high chief. Are we forced to hold that, if they continue to speak of the children as really having been the high chief, they are being irrational and are failing to see the logical error that they are, purportedly, making?

It seems that we should say that these imagined Gitksan families are irrational only if we assume the meanings of expressions such as ‘same person’ and ‘different persons’ – and, crucially, what it means to speak of one person as the reincarnation of another – as being logically fixed independently of the roles that these and similar expressions have within the communities that use them. If it is a universal logic, abstracted from human life and activities, that determines whether a given way of speaking makes sense, then it may be the case that people whose ways of speaking appear to have a sense within particular contexts may, in reality, be merely talking nonsense. If, however, one takes it to be the very ways in which forms of language are used by human beings in the weave of their lives that imbue those forms of language with sense, then the contention that certain forms of language which play an active role in people's lives are ‘in reality’ meaningless will itself fail to have a clear sense.Footnote ³⁵

‘The logic the Gitksan use in no way makes it impossible for person A to “be” person B.’Footnote ³⁶ This contention of Mills’ may prompt us to contemplate the possibility that the concept of a person that is operative in the sort of argument constructed by Bernard Williams might not reflect the ways in which certain people, and certain communities, speak and think about themselves and others. If we were confronted with the two Gitksan families that I imagined above, one option would be to declare to them, on the grounds of an argument such as Williams’, that the way in which they are speaking of their children is absurd. It is absurd, we might say, because if the high chief were reborn as child A and child B, then both A and B would be identical with the high chief, and hence A and B would have to be identical with one another – and that is absurd. But it is far from obvious why the Gitksan families could not simply, and legitimately, deny that their concept of a person has that logical entailment. Could they not assert that their concept of a person (or: the concept of theirs whose verbal expression comes closest to being translatable as ‘person’) allows it to be the case that both A and B are reincarnations of the same high chief without its being the case that A and B are the same person as one another? Or, alternatively, could they not assert that A and B are the same person, notwithstanding the fact that A and B exist contemporaneously? Why could there not be a conception of personhood according to which a given person has, simultaneously, two or more bodies? There have, after all, been people who have believed it to be possible for certain saintly individuals to be located in two places at once.Footnote ³⁷ We may not share beliefs of this sort, but why should we reject them out of hand as absurd?

Context is of vital importance here. In a British court of law, for example, if there is strong evidence for believing that a man accused of burglary was, at the time of the burglary, several miles away from where it took place, this will count against the credibility of the accusation and is liable to prove decisive in favour of the man's defence. The idea that the man could have been in two places at once will not enter into the deliberations at all. In the autobiography of the Indian spiritual guru Paramahansa Yogananda, however, a scenario is described in which he and another person come to believe that a holyman named Swami Pranabananda was indeed present in two distinct geographical locations at the same time.Footnote ³⁸ Admittedly, they are amazed by this fact and refer to it as a ‘miracle’, thereby placing it outside the sphere of normal expectations.Footnote ³⁹ Nevertheless, the highly unusual nature of this fact does not, for them, preclude its possibility, for they do come to believe it. Yogananda supports the claim that this sort of thing can happen by directing his readers’ attention to a relatively well-known ancient text on meditation called the Yoga Sutras – in which, allegedly, ‘The power of appearing in more than one body’ is described as ‘a siddhi (yogic power)’ – and also by noting that the ‘phenomenon of bilocation has been exhibited in the lives of many saints down the ages.’Footnote ⁴⁰ Yogananda's interpretation of the relevant statements in the Yoga Sutras might be open to question, and some may want to invite him to provide examples of the ‘many saints down the ages’ to whom he alludes. Yet it remains the case that his tale of Swami Pranabananda gains its sense from, or in relation to, precisely these sorts of beliefs – beliefs that involve speaking of ‘bilocation’ or ‘appearing in more than one body’ as a genuine, religiously-significant phenomenon that has been documented in ancient scriptures and hagiographies.

The sort of ‘bilocation’ that Yogananda describes differs in various ways from the phenomenon of multiple simultaneous reincarnation believed in by members of the Gitksan and other native American and Inuit peoples. One significant difference is that the latter phenomenon is not – or, at least, has not traditionally been thought of as being – especially rare, and would not typically be regarded as anything ‘miraculous’, even though its possibility may be attributed only to certain highly esteemed individuals. Similarly, other reincarnation beliefs – such as the belief among the Kwakiutl people of British Columbia that hunters of land animals will be reborn as wolves and that sea hunters will be reborn as killer whales – would not be well described as a belief in a type of miracle, since, for these people, there is nothing especially exceptional about being reborn as a wolf or a killer whale; this is simply what happens to hunters (or: to the souls of hunters) of the relevant kinds.Footnote ⁴¹ Again, however, it is in connection with a network of other beliefs, values and practices that the beliefs that certain people can be reborn as more than one individual and that others can be reborn as wolves or killer whales have their place and meaning. As the anthropologist Michael Harkin observes, ‘it is important that we not simply extend implicit Western models of the person to other cultures’;Footnote ⁴² instead, if we are to enrich and deepen our understanding of the cultures concerned, and of the variety of forms that human concepts can take, we should endeavour to observe how people within these cultures speak about and conceptualize themselves and one another.

4. Concluding remarks

The article by David Cockburn that I discussed in section 2 ends with the following intriguing remark:

The sort of suspicion that Cockburn refers to here should, I think, be carried over into philosophical discussions of ‘personal identity’. As I noted at the beginning of this article, the philosophical literature on personal identity is replete with wonderful thought experiments. These frequently comprise speculations about whole-brain or brain-hemisphere transplants, amoebic fission of human beings, teleportation, and suchlike.Footnote ⁴⁴ These imagined scenarios are typically described in just enough detail for their respective philosophical inventors to derive the conclusion that they sought to derive. What tends to be seriously lacking is any sustained attentiveness to the multiplicity of forms that conceptions of persons can take in different cultural traditions. Any talk of ‘the concept of a person’ is, in my view, something of which we ought to be suspicious, whether it comes from a philosopher or from any other theorist. Social anthropologists are certainly not immune from such over-generalizations themselves, yet their detailed studies of the ways of life and forms of language engaged in by various cultural groups can serve the extremely important purpose of releasing us from the clutch of unwarranted assumptions – assumptions concerning what inferences can, or should, be drawn on the basis of philosophical thought experiments, for example.

Like Cockburn, I have in this article perhaps done little more than highlight the well-founded nature of the suspicion at issue. Yet, again like Cockburn, I take this to be a highly valuable task. It is when we overlook the broad repertoire of forms that human life and interactions with one another can take that we, like D. Z. Phillips in his early work, become prone to jump to hasty conclusions concerning the intelligibility of certain kinds of belief. In this article I have not urged readers to accept a belief in reincarnation as true, but simply to pause before dismissing it as evidently fallacious or illogical. Of course, if you are convinced that there is a single, correct understanding of what a person is, which is amenable to being defined in terms of the individually necessary and jointly sufficient conditions for the application of ‘the concept of a person’, then you will remain sceptical about the sorts of considerations with which I have been concerned. If, however, you remain open to the possibility that conceptions of what a person is arise, not within a vacuum, but within particular networks of belief and practice – within particular forms of life, which intersect and overlap with many other modes of human and non-human activity while nevertheless retaining their own cultural nuances – then you may concur with me that believing in reincarnation is very far indeed from believing something that is a logically impossible absurdity.