R. W. K. Paterson: Philosophy and the Belief in Life after Death. Macmillan Press, London, 1995. 223pp. £45.00. ISBN: 031212838X.

Book Review – Alan Gauld
(First published as “Philosophy and Survival: An Essay Review of R. W. K. Paterson’s Philosophy and the Belief in Life after Death.” Journal of the Society for Psychical Research, 1998;62:453-462, reproduced with permission)

The author of this book was formerly senior lecturer in philosophy at the University of Hull. Its aim is to give some account of the empirical evidence for life after death, and of the philosophical problems to which that evidence gives rise, in a manner accessible to "the general educated reader". It packs a great deal into a relatively short compass, but is none the less written with unfailing clarity.

The first chapter presents Paterson's own view of the nature of philosophical endeavour. The task of the philosopher (p. 1) is "to subject the concepts which figure in our beliefs to logical analysis, to an intensive scanning for hidden inconsistencies, obscurities, or ambiguities", and to logically evaluate the grounds on which those beliefs are held. Some of the issues tackled by philosophers "centre entirely on the possible relations between concepts and involve only arguments of a purely formal character", but others are "intellectually composite", and "also involve reference to various types of claimed empirical fact." These "composite" issues include the free will-determinism controversy, the nature of the relation between mind and brain, and the problem of whether or not human personality survives bodily death. In all these areas it is essential at the outset to scrutinize the logical coherence of the concepts involved and of the propositions advanced. If they pass this scrutiny, the answers become a matter for empirical enquiry; but even then further conceptual and logical issues are likely to arise over the proper interpretation of the findings.

After listing and dismissing certain 'basic limiting principles' which, following C. D. Broad, he suggests are widely subscribed to by contemporary scientists, and which would in combination rule out a priori the admissibility of any 'psychic' phenomenon whatsoever, Paterson moves on in chapter two to discuss the problems of personal identity.

He distinguishes between an ontological problem of personal identity, which is the problem of what "makes a person one and the same unique individual, however much he may change, and however closely he may resemble other individuals" (pp. 21-2), and an epistemological problem, which is that of "how we can satisfactorily pick out some present individual as being one and the same person as some past individual of whom we have knowledge." Clearly answers to these two problems have an immediate bearing on the questions of whether it makes sense to talk of individuals surviving bodily death, and of how we might recognize them again if they did.

Paterson divides attempted answers to the ontological problem into three main categories, physical, psychological and transcendental. Physical theories base personal identity on some physical set of facts about a person, for instance distinctive bodily features, facts about brains and brain functioning (which, however, tend on analysis to be facts about the supposed mental consequences of brain functioning), and facts about the presumably unique and uninterrupted spatio-temporal track of each person's body. Psychological theories invoke continuities of memory and of character; and transcendental theories look to an unique 'soul' or an unanalysable ego or T. Paterson investigates all these proposed solutions, and also glances briefly at Parfit's distinction between identity and psychological continuity (which may be partial or non-existent even within the history of the 'same' individual). He finds no approach fully satisfactory, and some wholly unsatisfactory (a conclusion with which I do not disagree). And it is certainly the case that there seems to be no impending consensus of opinion among philosophers.

One might suppose that failure to find an acceptable solution to the ontological problem would make it impossible to pursue the epistemological problem in any meaningful way. But Paterson argues otherwise. He points out that there are 'unanalysable' facts which we none the less do not hesitate to accept as facts. For instance (p. 44) although "we all understand what time is, we cannot give a clear explication of what it is; we cannot say what we mean when we speak of a 'past' event... and yet we and our hearers know perfectly what we mean." Likewise (p. 45) "even if we are unable to give a full and correct analysis of the claim that some disembodied person is numerically absolutely identical with the ante-mortem Winston Churchill, we understand what is being claimed and are entitled to weigh up such evidence as is available on behalf of this claim", even though this evidence may not be for 'strict' but only for 'relative' or 'partial' identity. By and large, and with one or two exceptions, the evidence which would be relevant to establishing someone's post-mortem identity with a certain pre-mortem individual will, Paterson holds, be the kinds of fact commonly accepted as establishing his pre-mortem identity from one day or week or year to the next.

By a natural transition, chapter three takes up the mind-brain problem. As a quick rule of thumb one might say of this problem that most of what philosophers say when criticizing the theories of other philosophers carries conviction; most of what they say in support of their own theories does not. Certainly it would be hard to dissent from Paterson's critical remarks about certain materialist theories of the mind. He examines epiphenomenalism at some length, and finally dismisses it as little more than a string of confused metaphors. He then tackles mind-brain identity theories of the 'type-type' variety. These theories postulate an identity between mental states or events of a certain specific category ('type'), e.g. all thoughts of Winston Churchill's cigar, and brain states or processes of a certain specific category. The proposed identity is not a logical identity, because there is no way in which the concepts we use when describing mental events can be reduced to or translated into the concept-set we use when describing physical events. Rather it is a contingent identity, one that we can discover empirically, as, for instance we might discover the identity of a lightning flash with a stream of electrons passing between cloud and cloud or cloud and ground. Though type-type theories derive a certain plausibility from their apparent harmony with modern neuroscience, Paterson finds them unacceptable on a variety of grounds, including the not infrequent recoveries and apparent relocation of function following brain damage, and difficulties over the very notion of 'contingent identity' in cases where (as with the mental and the physical) the types of phenomena concerned belong to 'different categories of reality*. He adds a couple of general objections to both identity theory and epiphenomenalism. He thinks that such theories are committed to brain-trace theories of memory, which are untenable on both logical and empirical grounds, and also that they cannot accommodate the facts of ESP and PK, which he regards as satisfactorily established. (Of course 'mentalistic' approaches to memory and to psi phenomena can easily run against difficulties analogous to those which confront materialistic ones—consider, for instance, image theories of memory and 'transmissive' theories of telepathy where the 'transmission' is held to be from mind to mind rather than from brain to brain.) Paterson seems to hold that both memory and psi phenomena are ultimate powers of the mind, not further explicable. "Knowledge and agency are intrinsic to mind" (p. 85) and memory (p. 90) is "an irreducible fact about our minds." What needs to be explained is therefore not how we remember or what the nature of ESP may be, but why we forget and why ESP is so infrequent and unreliable.

Paterson himself is a strong supporter of interactionist dualism, a theory which he regards as particularly hospitable to a belief in life after death. He does not, however, advance many considerations that directly support this position. Like most dualists he principally relies on the obvious inadequacies of other viewpoints. He adopts what he describes as a 'non-Cartesian' form of dualism, much influenced by Bergson, whose views on this topic have often had a particular appeal to psychical researchers (Bergson was, of course, president of the SPR in 1913). According to this theory the mind is "a dynamic continuant, a ceaseless flow of mental energy". The role of the brain is to detect and magnify "a tiny current of consciousness" which is already flowing (the exact force of this 'magnification' metaphor is obscure), to focus it, i.e. to restrict its field of activity by excluding many data, and to connect it with the motor mechanisms of speech and action by a process which falls "under the general heading of psychokinesis". Brain damage or atrophy will attenuate or impair the flow of consciousness. But again Paterson does not really argue for this position; he simply sets it out, which makes comment difficult.

Having set the scene by his discussions of personal identity and of the mind-brain problem, Paterson proceeds to consider the reasons that have been advanced for believing in survival (survival, he is careful to point out, does not necessarily amount to immortality). In chapter four he examines various philosophical, theological and moral arguments for believing in the immortality of the soul. He rejects most of them fairly quickly. The only one he takes more seriously is the so-called argument from the indivisibility of the soul—roughly the argument that the soul, being (as some interactionist dualists would maintain) one simple immaterial substance, is therefore "indestructible by any process involving decomposition, dissolution, or division" (p. 119)—but he does not go so far as to uphold it.

He next takes up, in chapter five, the evidence from psychical research, of which he gives a very fair, though necessarily brief, account. He surveys the literature on deathbed visions and on NDEs at some length, and passes on to a careful account of cases of veridical apparitions of the dead and the problems involved in interpreting them. He writes with a wide knowledge of the literature about cases of 'evidential' mediumistic communications, and ends by noting some of the evidence for reincarnation with special reference to the work of Ian Stevenson. His overall view of the empirical evidence for survival is that although survival of death has not been 'proved', and although each category of evidence for survival is a great deal less than conclusive, the cumulative evidence of all the various categories taken together shows survival of death to be overall distinctly probable.

In a final chapter Paterson raises the question of what an after-life might be like. He sets aside the numerous elevating and colourful descriptions of the next world and its inhabitants communicated through mediums, remarking (p. 201) that it "would be plainly impossible to construct any kind of map of the spirit world" on the basis of them, and that they leave a good many obvious questions unaddressed. He proposes instead (following H. H. Price) that any after-life that we may have is likely to be a kind of dream life, but a dream life perhaps influenced by other surviving persons with whom we may be in telepathic rapport. This world would be "neither purely objective nor purely subjective, but intersubjective" (p. 205). The possibilities and prospects opened out by such a view of the after-life may appeal to some, but to others they will serve as a reminder that there are many conceivable forms of after-life so awful that extinction would be a merciful alternative.

Paterson's tone and arguments are moderate throughout and he is well aware of the difficulties which confront the survival hypothesis; none the less it seems to me that in some ways he gives that hypothesis too easy a ride. He does so to an extent when discussing the empirical evidence for survival. Some of the cases he cites do not seem to me to measure up to the requisite standards of evidence; but that apart one might have expected a philosophical work to set out and assess the reasons for not believing in survival as fully and systematically as the reasons for accepting it. E. R. Dodds' classic paper (Dodds, 1934) might have served as a starting point. Even the rather wearisome 'super-ESP hypothesis' does not receive quite the consideration it deserves. Paterson recognizes that the survival hypothesis itself often necessitates postulating ESP (e.g. between medium and surviving communicator, or between and among the surviving spirit of the late Jones and the various percipients who collectively perceive his veridical post-mortem phantasm) of a degree that, by the standards of anything for which we have experimental evidence, would have to be accorded the prefix 'super'. But he seems to think that the ESP that would be required to explain away these cases without involving the agency of deceased persons would be more extraordinary still. If one is going thus to make a crucial distinction between a 'super-ESP' which is acceptable and what might be called 'superduper-ESP' which is not, a somewhat more extensive 'composite' philosophical-cum-empirical investigation of the criteria on which such a distinction might be based is surely called for.

Again, in his chapter on the brain-mind problem, Paterson bludgeons epiphenomenalism and type-type identity theories fairly effectively into the ground. But these days epiphenomenalism is an accusation that philosophers make rather than a position that they hold, while traditional type-type identity theories are almost as widely, if not quite as vehemently, rejected as dualism (though they may to an extent be enjoying a disguised revival as part of the currently burgeoning if frequently pseudo-scientific literature on the problem of consciousness). It might have been helpful to Paterson's likely readership if he had said a little more about various other materialist or quasi-materialist approaches to the mind-brain problem that have had considerable vogue in the last couple of decades, for instance 'functionalism' (particularly in its ‘computational’ form) and 'eliminative materialism', so energetically propounded of late by Paul and Patricia Churchland. (Some computational functionalists have even attempted to give accounts, in their own terms, of the possibility if not the likelihood of human survival of bodily death.) Both of these approaches are, in the opinion of their proponents, ways of sustaining materialist positions whilst escaping some of the difficulties confronting the traditional forms of identity theory. Room for some discussion of them might have been made by cutting down the perhaps needlessly extensive treatment of epiphenomenalism.

But the issue over which Paterson allows the survival hypothesis to escape most lightly (and it is a fundamental issue) is that of what he terms the ‘epistemological problem of personal identity’, that is, the problem of how we pick out some present individual as being the same person as some past individual. He holds that in general the sorts of evidence that will establish someone's pre-mortem identity are also the sorts that will serve to establish his post-mortem identity. But things are not quite so simple. I think it would be fair to say that in everyday life the 'epistemological' problems of pre-mortem identity very rarely arise. We do not generally have to ask ourselves on what grounds we might justifiably identify the Jones of yesterday with the Jones of today, and whether we have in fact got such grounds. But, as Paterson points out, problems of pre-mortem personal identity, involving assessment of various types of supposed evidence that two ostensibly different persons are or were in reality one and the same person, do sometimes seriously arise. Was Jeanne des Armoises really Joan of Arc? Was Perkin Warbeck one of the Princes in the Tower? Was Naundorff the same individual as Louis XVII? Was Arthur Orton Sir Roger Tichborne? Was Anna Anderson the Grand Duchess Anastasia? In these and kindred if less-known legal and historical cases (for a very useful selection of the legal ones see Rolph, 1957) it is necessary to enquire what might constitute adequate evidence for the underlying identity of the two individuals concerned, and whether such evidence exists (one might, indeed, in some circumstances ask similar questions concerning different phases of one's own career).

The evidence that has in practice been seriously entertained in such cases has usually been of kinds indicated by the various proposed solutions to the ontological problem of personal identity—bodily characteristics, veridical memories, personality characteristics, continuity of spatio-temporal track, the testimony of witnesses who knew the original person, even possessions once owned by that person and now owned by the current claimant. These categories of evidence are somewhat complexly interrelated. They may mutually support, or be necessary for, each other, and sometimes there may be conflict between them. But it is quite apparent that the central and predominant one, even though it is often unavailable, is that of spatio-temporal track. Where we do have this kind of evidence it prevails over all the others. Physical evidence has suggested that Anna Anderson and the Grand Duchess Anastasia had different spatio-temporal tracks. Should that be the case the matter is closed. If the remains of Sir Roger Tichborne were discovered in the wreck of the Bella that would finally scotch Orton's claims (if further scotching were needed) by showing that Sir Roger's track had terminated before Orton came forward as a candidate. And so on. And it is obvious why spatio-temporal track has this predominance. Two different individuals can look very much alike; one of them can cram himself with information about the other so as to pretend to possess his memories, can simulate his mannerisms, can even acquire some of his possessions. But two distinct individuals cannot, in the world as it is commonly accepted to be, share the same spatio-temporal track. In cases of doubtful identity in which evidence about spatio-temporal track is unavailable, there is scope for endless unresolved debate, which indeed we have had over some of the more notorious historical examples. And in these cases the force of the evidence cited is usually that the earlier and the later individuals must have had one continuous spatio-temporal track even though we can't document all phases of it.

Suppose now a case, still ostensibly a pre-mortem case, in which spatio-temporal continuity has been interrupted. An example might be the matter transporters of science fiction, in which an object, such as a human being, is disintegrated into constituent atoms, shot across space, and reassembled exactly as before (the idea seems to have originated from Conan Doyle's story The Disintegration Machine). Someone might raise the question whether the Jones who reappears at the reintegration station is the same person as, has a stream of consciousness to be identified with that of, the Jones who disappeared at the disintegration station. Ought we rather to say that the original Jones was destroyed body and mind upon disassembly, and that the reassembled Jones is simply an exact mental and physical replica? We could argue about this question indefinitely, citing, for example, the presumed orderly transmission and orderly reassembly of the bodily atoms as evidence supporting the continuity hypothesis, but, logically, it does not appear that we could ever arrive at a definite answer, one which would distinguish between Jones having maintained his identity despite disintegration and its merely being as
if he had done so. There might be a certain shortage of volunteers for transportation.

Rather similar considerations apply in the case of ostensible survival of bodily death. Again continuity of spatio-temporal track is lacking, and we have chiefly to rely on psychological evidence of identity—memories, personality traits, mental skills (even 'materializations', if there are such entities, are of very little value as evidence for survival unless backed by such evidence). I do not think it is meaningless (as some might hold) to raise the question of whether this kind of evidence can be given a survivalist interpretation. But it is hard to see what, in the absence of continuity of bodily track, could possibly count as distinguishing between Jones having survived the death of his body (though we don't understand how) and its being now and again transiently as if he had survived it (though again we can't make sense of it). In these circumstances peoples' beliefs as to the proper interpretation of the evidence are bound to be largely determined by their predilections and their views on related questions.

The root of the difficulties seems to be that whereas the apparent postmortem reappearance of Jones's characteristic memories, personality traits, skills, interests, purposes, even of his external appearance, in some sense constitutes evidence for his survival as a person (it is not easy to think of any other conclusion towards which it points), continuity of spatio-temporal track has become, at least in the setting of our own society, so predominant that it is now a criterion of continuing personal identity. The term 'criterion' has been much misused in recent philosophy. In dictionary usage a criterion is a standard by reference to which decisions are taken as to whether or not something is or shall be deemed to be the case. Criteria are set up by convention, either by a gradually evolving social convention or by a deliberately established artificial convention. They are adopted because they work, and because it is difficult for people to find grounds for disputing them. When being used qua criteria, they are no longer 'evidence' of that of which they are the criteria. For a broader or narrower, but still restricted, purpose the relationship between a criterion and that of which it is the criterion is taken as definitional, and within that context it becomes self-contradictory to affirm the criterion and deny that of which it is the criterion. None the less if C is a criterion of X it does not follow that the term 'C' is part of the meaning of the term 'X'. Thus for somewhat restricted legal purposes a certain level of alcohol in the urine might be accepted as a criterion of drunkenness. But being 'drunk' does not in more general usage mean or simply mean 'having more than a certain level of alcohol in the urine'. It means having imbibed too much alcohol and in consequence behaving in certain unacceptable ways. The criterion of drunkenness could be changed without seriously disrupting the concept of drunkenness.

To return now to the problem of personal identity and survival: When in the case of the apparent post-mortem manifestation of a certain person we are invited to interpret the phenomena as evidence of his survival, we are being invited to bypass or dispense with an established criterion of personal identity (that of continuous spatio-temporal bodily track) and instead accept evidence of identity (usually 'psychological' evidence) that conflicts with that criterion. The trouble is that if we downgrade the spatio-temporal track criterion, we are tampering with a deeply built-in and highly stabilising aspect of our web of social concepts and practices, an aspect which no doubt reflects contemporary science to a certain extent, but mostly reflects the world we have successfully grown up in and social practices that work within that world. Downgrading it could have somewhat disorienting knock-on effects, and is therefore a move likely to encounter much explicit and implicit resistance, whilst so far as the evidence for post-mortem identity is concerned, it would leave us in a state of ambiguity and conflict. For if we no longer accord pre-eminence to spatio-temporal track, but relegate it to being just one sort of evidence among others, we have to face the fact that when we examine the evidence bearing on the question of continued post-mortem identity parts of our materials seriously conflict with other parts, so much so that it is exceedingly difficult to locate any solid grounds for a decision. The evidence that seems, in some instances at least, to support a continuation of personal identity after death is very nearly overwhelmed by opposing considerations from everyday observation, from neuroscience, even from aspects of psychical research, which point to an ineluctable tie between the integrity and functioning of one particular body and nervous system and the integrity and functioning of one particular person, and suggest plausible counter-explanations of any apparent evidence to the contrary. Believers in survival (though Paterson is an exception) tend either to be ignorant of these considerations or else to brush them too easily aside. On the other hand there is no denying that the 'evidence for survival' includes some exceedingly odd things which are rather resistant to the more obvious attempts to undermine them. Dedicated disbelievers rarely acquaint themselves properly with the ins and outs of these refractory and puzzling cases, and tend to confuse the antecedent implausibility of the survival hypothesis with weakness of the evidence ostensibly supporting it. Arguments between supporters of a literal interpretation of the evidence for post-mortem survival and supporters of an 'as if' interpretation seem set to continue in their present unresolved and unresolvable deadlock more or less indefinitely.

Certainly the deadlock is not likely to resolved by the mere accumulation of more and more evidence of the kinds reviewed by Paterson (though I don't say that this would not be a valuable undertaking). Consider a parallel case: All the crew members bar one of the Star Ship Enterprise might be successfully beamed to the moon and back. This would be an accumulation of evidence that matter transportation is possible. But it would not be an accumulation of evidence favouring a literalist rather than an 'as if' interpretation of the apparent continuation of personal identity across a gap in bodily continuity, or vice versa. A thousand instances would not necessarily tell us more than a hundred or than ten. The real problems are over what is going to count as evidence of, or an acceptable criterion of, continuity (as distinct from replication) in these abnormal circumstances. Similarly a thousand more cases of veridical mediumistic communications would not necessarily help us decide between the literal and the 'as if' interpretations of the ostensible evidence for post-mortem survival of personal identity. A rational and generally acceptable decision in favour of the survivalist interpretation would only become possible if we were presented with new or hitherto imperfectly appreciated kinds of evidence for continued pre-mortem identity, linking with each other and with more traditional sorts of evidence, and this evidence were to force us to accept, or facilitate our accepting, the same sort of evidence, or related evidence, as evidence for continued post-mortem identity. And the trouble would be that it is not straightforwardly up to scientists and philosophers (still less to psychical researchers) to decree or engineer such a shift on the basis of current knowledge as they see it. Any shift would be at least in part accelerated or retarded by concordant changes in relevant social practice. Still a shift is, I suppose, just about conceivable.

Let us begin (as these days philosophers or indeed physicists often do) by imagining a world very different from our own, an Alice in Wonderland kind of world in which, for instance, the inhabitants might change shapes like Tarn Lin, or pass like insects through a series of metamorphoses, or disappear and reappear elsewhere like the Cheshire Cat, or exchange bodies as in The Great Keinplatz Experiment or Vice Versa. In such a world ideas about personal identity (that is, about pre-mortem personal identity) would inevitably be very different from those prevalent in our own. Psychological evidence, particularly but not exclusively evidence from memory, would obviously play a central role in attributions of personal identity, especially since we would all experience the shape-shifting and sudden transitions ourselves and know about them from the inside. Bodily characteristics and spatio-temporal track would only be of restricted value as evidence of personal continuity, and psychological evidence of identity might be elevated to 'criterial' status. And it would correspondingly be much easier for the inhabitants to accept psychological 'evidence' or 'criteria' of continuing post-mortem identity. Indeed death itself might become a somewhat elusive concept.

Now for psychological evidence of personal identity, including of postmortem identity, to attain prevalence, or criterial status, in our world, our world would obviously have to come to resemble, or be found in certain ways to resemble, the imaginary world we have been discussing. And this would not be achieved, as I have already remarked, just by piling up further survival evidence of the customary kinds. What would be needed would be, inter alia, much more evidence than we have so far obtained to suggest that even prior to bodily death one person may be associated with more than one body, or one and the same body with more than one person, and indeed that there may be brief phases of disembodiment as part of these games of bodily musical chairs. For instance we might look for cases in which a person undergoing an OBE appears as a mediumistic communicator (there are indeed a very few cases of ostensible mediumistic communications from the living); for cases like the Jasbir-Sobha Ram and the Sumitra-Shiva cases described by Stevenson and his collaborators in which one person has ostensibly disappeared from a seriously ill or comatose body and been replaced by a totally different person with a totally different set of veridical memories (suppose the Anna Andersen-Anastasia case had been of this kind?). If there were enough such cases, and enough cases of other and related sorts which one might imagine, e.g. cases in which deceased persons communicate successfully, announce their impending reincarnation, and then appear as predicted as one of Stevenson's children who ostensibly remember past lives, or cases of communicators who manifest with a continuous memory through a series of different mediums, or cases in which various communicators through various different mediums give coherent and consistent accounts of their present status and surroundings, and of their postmortem meetings with each other; then indeed (and especially if there were in addition rather marked, and one must add highly unlikely, changes in the general tenor of contemporary neuroscience) there might come about some softening up of the criterial relationship between continuity of spatio-temporal track and continuity of personal identity, and with it a strengthening of the grounds for accepting 'psychological' evidence or criteria for post-mortem continuity. 'Super-psi' and other 'as if' explanations of the evidence for survival might then be progressively sidelined as the evidence seemed more and more compatible with the purposes and activities of surviving 'apperceptive centres', to use Braude's convenient term (Braude, 1991, p. 78). Of course nothing would actually rule out the 'as if' interpretations; they would just be dismissed as vacuous and irrelevant to the issues, rather as the hypothesis that those around us are insentient automata is dismissed as having no bearing on the question of how we should treat our fellow human beings.

I do not say that any of this is likely, only that it is not inconceivable.

So much by way of a brief sketch of developments which might strengthen a literal interpretation of the ostensible evidence for post-mortem identity. I have not touched on other possible developments which might forever seal off such an interpretation, and leave us only with a choice between one or another form of 'as if' interpretation. But these last possibilities are so obvious and, many would say, so nearly upon us, that spelling them out would be superfluous.

To revert at last to the point I have been trying to illustrate. There is not, and there is not going to be, any simple way in which we can make the transition from straightforwardly accepting psychological evidence for continued pre-mortem personal identity to straightforwardly accepting the same sort of evidence for continued post-mortem identity.

I have argued at some length that Paterson has oversimplified various matters in a direction that tends to favour the survival hypothesis. But I do not think that this greatly detracts from the value of his book, which will serve for psychical researchers who do not know much philosophy as a helpful introduction to the philosophical problems of survival, and for philosophers who know little about psychical research as a guide to the relevant empirical facts, and for "the general educated reader" who does not know a great deal about either as a thoughtful and well-informed starting point. Those who have some acquaintance with both literatures will find many points worth reflecting on.

Regrettably its relatively high price, and the lack of a paperback edition, will probably prevent it from being as generally useful as it might otherwise be.

Alan Gauld
Department of Psychology
University of Nottingham
Nottingham NG7 2RD
England

References

Braude, S. E. (1991). First Person Plural: Multiple Personality and the Philosophy of Mind. London: Routledge.

Dodds, E. R. (1934). Why I do not believe in survival. Proceedings of the Society for Psychical Research, 42:147-172.

Rolph, C. H. (1957). Personal Identity. London: Michael Joseph.