C. D. Broad -Empirical Arguments for Human Survival

- C. D. Broad - Empirical Arguments for Human Survival

I MUST begin by saying exactly what I do and what I do not propose to discuss in this Chapter. I do not propose to discuss in any detail the special alleged facts (such as the Cross-correspondences in automatic writings) on which empirical arguments for human survival have been based. This is an extremely technical question which must be left to experts and would be out of place in a philosophical book. I do presuppose that the careful work of the Society for Psychical Research has elicited a mass of facts which may fairly be called "supernormal", in the sense that they cannot, if genuine, be explained on the usual assumptions of science and common-sense about the nature and powers of the human mind. And I do assume that a great many of the facts that come up to the extremely high standard of evidence required by the Society are "genuine", in the sense that they have been correctly reported and that they are not simply due to fraud or self-deception. I assume this on the basis of a fairly careful study of the literature; of a knowledge of the kind of persons who have controlled the policy of the Society and taken part in its investigations; and of some investigations of my own. I have, in fact, exactly the same kind of grounds for assuming the existence of genuinely supernormal phenomena as I have for assuming the existence of certain rare physical phenomena which are difficult to reproduce to order, and of certain rare diseases which competent doctors have described. I accept them on a mixed basis of authority and personal experience; and my authority is of the same kind and carries the same weight as the authority on which I accept the rarer and obscurer kinds of physical and medical phenomena as genuine. I do not think it is necessary to argue this point, because I have always found that those who deny it have not carefully read the relevant literature, have conducted very few careful investigations for themselves, and are ignorant of the intellectual calibre and the scrupulous accuracy of such men as Sidgwick, Gurney, and Podmore (to mention only the names of those who are no longer with us). Whenever we are told that "Science proves so-and-so to be impossible" we must remember that this is merely a rhetorical form of "Professor X and most of his colleagues assert so-and-so to be impossible". Those of us who have the privilege of meeting Professor X and his colleagues daily, and know from experience what kind of assertions they are capable of making when they leave their own subject, will, I am afraid, remain completely unmoved. I take human survival then to be one hypothesis among others to account for certain reasonably well established supernormal phenomena. The argument will be of the usual inverse-inductive type. Now, in such arguments we always have to consider the following points. (i) The antecedent probabilities of the various alternative hypotheses. (ii) And the completeness with which the various alternative hypotheses explain the special facts under consideration. If the antecedent probability of h1 be very much less than that of h2, then, even though h1, explains the special facts better than h2, it may be more prudent to try to make some modification of h2 rather than to put much faith in h1. I shall, therefore, begin by considering the antecedent probability of the hypothesis of human survival. The Antecedent Probability of Human Survival When we are considering the antecedent probability of a hypothesis put forward to explain certain special facts there are two points to be considered. (i) There is what may be called its "intrinsic probability". This depends on the structure of the proposition itself, and very little can be said about it here. (ii) There is the probability which the proposition has with respect to all known facts other than the special set of facts which it is put forward to explain. If p and q be two logically independent propositions, the proposition pq is intrinsically less probable than the proposition p. This is an instance of the first point. If a bishop falls down in the street it is antecedently more probable that this is due to a piece of orange-peel than to direct diabolic agency. For, although both hypotheses explain the observed fact equally well, the former fits in much better with the other facts which we know about the world than the latter does. This is an example of the latter point. I find myself quite unable to say much of importance about the intrinsic probabilities of human survival and its rival hypotheses. But there are a few logical points which are perhaps worth making. (1) Among alternative hypotheses to human survival which have been suggested we may mention (a) a very extended telepathy among living men, and (b) the action of non-human spirits who personate certain dead men. The second of these would seem to have the least intrinsic probability of the three hypotheses. For we have to postulate minds, for whose existence we have no other evidence, and to ascribe telepathic powers to them. The first hypothesis postulates no minds for whose existence we have not already independent evidence; but it has to ascribe to them telepathic powers of such great extent that we have little or no independent evidence for their existence. The hypothesis of human survival perhaps makes the minimum assumption of the three; since it merely postulates the continuance of something which we know independently to have existed, and it ascribes to this only such telepathic powers as we have reason to believe exist in embodied minds. (2) There is one very great logical difficulty which is inherent in the subject. We have not the least reason to believe that the hypotheses that have been put forward are exhaustive or even approximately so. Hence we have no ground for ascribing any very high antecedent probability to any one of them. We believe ourselves to know enough of the general structure of the material world to enable us to rule out all but a few hypotheses about the causation of a physical phenomenon. In such an unfamiliar region as we enter in doing Psychical Research we have not this advantage. There is just one other remark that I will make before w leaving this part of the subject. It is well known that many Roman Catholics and High Anglicans, not content with ascribing the phenomena to non-human spirits, ascribe them to "devils". Now I suppose that a "devil" means a non-human spirit who is morally much worse than the worst man. There appears to me to be absolutely nothing in the phenomena to warrant this hypothesis. (a) There is a certain amount of indecency in some automatic scripts. So there is in the writings of Petronius and in the conversation of many undergraduates; whilst Mr Gibbon informs us that "a learned prelate, now deceased, was fond of quoting ... in conversation" a passage from Procopius about the Empress Theodora which the historian prudently "veils in the obscurity of a learned language". (b) Most spiritualistic communications which are not merely trivial consist of elevated, but to my mind "twaddling", ethico-religious "uplift". If they be the communications of devils it must be admitted that most devils who communicate are decorous to the verge of dullness; and that the aphorism "Heaven for the climate, but Hell for the company" stands in need of considerable modification. (c) It may be admitted that to personate a dead man and raise false hopes in his friends and relations would not be the mark of a very high morality. It would be a somewhat heartless practical joke. But it is not necessary to be a devil in order to play heartless practical jokes; such things have been done before now by quite kindly but somewhat thoughtless undergraduates. (d) It may be admitted that a certain number of weak-minded people go mentally and morally to the bad through excessive indulgence in spiritualistic séances. The same may be said of excessive indulgence in alcohol or religion. And a devil who chooses this particular method of leading men to damnation when there are so many more profitable alternatives open to him must be extremely incompetent at his own business. (e) There is a certain amount of "roughness" and horse-play at some séances for physical phenomena; there is a great deal more after a bump-supper or at many political meetings. In fact, if we can judge of Hell from those denizens of it whom we meet, on this theory, at spiritualistic séances, we must suppose that it is very much like what I believe is called a "Pleasant Sunday Afternoon" at a Nonconformist chapel, enlivened by occasional bump-suppers. Its nearest earthly analogy would probably be a Welsh University; and I should suppose that those who pass directly from the one institution to the other must often fail to notice the transition. To sum up, from a fairly extensive reading of spiritualistic literature, and from a certain amount of personal experience of séances, I should say that the average "spirit" is morally no worse than the average Fellow of Trinity, though there is a very marked difference in the intelligence of the two. The motives which make the "devil-theory" so popular in ecclesiastical circles are tolerably obvious. In the first place, there is the perfectly legitimate desire to frighten one's congregation away from dabbling in practices which are very unlikely to do good to any of them and very likely to do positive harm to many of them. The second motive is probably just as strong, but is generally unrecognised or unadmitted. This is the objection which the members, and especially the officials, of all close corporations have to non-members who claim to perform the same functions. The objection of the orthodox churchman, and particularly the orthodox clergyman, to the spiritualistic medium is the same kind of objection which doctors feel towards "bone-setters" and trade-unionists feel towards blacklegs. It is necessary to disguise this to oneself and to others; and for this purpose the "devil theory" is very handy, just as doctors find it highly convenient to remind us of the deaths of patients under quacks and to forget that patients sometimes die under doctors. I propose now to consider whether there are any facts other than the special phenomena dealt with by Psychical Research which make the hypothesis of human survival antecedently probable. Although, as I have said, I do not think that such special propositions as the survival of man fall within the range of proof or disproof by metaphysical arguments, I can see of course that the antecedent probability of human survival will be greatly affected by one's general metaphysical position. If materialism or epiphenomenalism were strict metaphysical truth, survival, though perhaps still abstractly possible, would be in the last degree unlikely. If mentalism, in one of its forms, were strictly true, survival would not indeed necessarily follow. Lotze and Mr Bradley were mentalists; but they held quite consistently that their systems did not necessitate human survival and that it is on the whole improbable. Still, mentalism is decidedly more favourable to human survival than is the view of the world which is taken by common-sense, or by non-philosophical scientists, or by dualistic philosophers. Idealism (which I distinguish from mentalism, though most idealists have in fact been mentalists) is still more favourable to survival. For I take it that the essence of idealism is to hold that what we regard as the "higher" characteristics, such as life and consciousness, are fundamental categories which apply to Reality as such and are not just special and probably transitory features of certain specially complicated and probably unstable parts of Reality. It is possible to be an idealist and yet to regard human survival as false or highly improbable. This position was taken by Professor Bosanquet. But it must be admitted that idealism would favour the antecedent probability of survival. It might seem then that, in order to determine the antecedent probability of survival, it would be necessary to make up one's mind between various rival systems of metaphysics. I am certainly not prepared to do this. But I think I have a fairly good excuse. On my view no general metaphysical system can be proved deductively by reasoning from a priori premises. Idealism and materialism are just attempts to synthesise all the known facts; and their respective probabilities can be decided only by their respective success in doing this. There is then, in my view, no possibility of first deciding between alternative metaphysical systems on general grounds and then taking the system which we have accepted as a fixed datum from which to estimate the antecedent probability of survival. The question whether we probably do or probably do not survive the death of our bodies is just the kind of question that has to be answered before we can decide (say) between idealism and materialism or epiphenomenalism. What we must do then is to discuss the antecedent probability of survival on data which are common to all men, including the upholders of rival systems of metaphysics. And this means that we must consider the arguments for or against human survival which may be drawn from the constitution of the world as it presents itself to enlightened common-sense; for this is the common basis from which all the rival systems start. If we do this we may consistently use our result as one means of deciding tentatively between the various rival systems. Now, on the face of it, the most striking feature of the world as we know it in daily life is, for our purpose, that it does not present the faintest trace of evidence for survival. Continued action is a criterion of the continued existence of any substance; and this is conspicuously lacking after death. The body ceases to give the characteristic responses, and very soon it decays and loses even its characteristic shape and appearance. Hence the only evidence that we ever had for the existence of a man's mind has ceased abruptly; and, apart from the alleged facts investigated by Psychical Research, it has ceased for ever so far as our experience goes. We do indeed often believe in the continued existence of substances in spite of long periods during which neither we nor anyone else are aware of them by any of their usual signs. E.g., we believe that silver continues to exist though it be dissolved in nitric acid and kept for years as silver-nitrate. But in such cases we have reason to believe that at any moment we could restore a substance having the properties of the silver which we dissolved, and connected with it by identity of mass and continuity of spatial positions. Every such factor making for belief in the continued existence of dead men is lacking in our ordinary experience; and thus such a belief seems to have nothing whatever in its favour, and to be from a logical point of view a bare unmotived possibility. Yet of course, as a matter of history, this has seldom seriously militated against the belief in survival. Such a belief has been all but universal. Now, on the one hand, the mere universality of a belief is no proof of its truth. On the other hand, the fact that a belief has been widely held by ignorant and primitive men is no proof of its falsehood. Confronted then by a strong belief which seems to have arisen and persisted in spite of complete lack of evidence in its favour, we must consider what factors may have caused the belief, and whether any of them arereasons as well as causes. A primitive man would certainly not accept the statement that there is no evidence in ordinary experience for survival. He would claim to know of dozens of cases of men seen and heard after death; and he might even think that he had met with such cases in his own experience. Now, without prejudice to the genuineness of abnormal phenomena in general or to the possibility that they occasionally happen among savages, we may be quite certain that in most cases the primitive man is mistaken in thinking that there is any need to assume the continued existence of the dead to explain the phenomena which he would regard as evidence for survival. We may divide such phenomena into two classes. The first consists of those which are capable of a perfectly normal explanation; the second of those which would now be dealt with by Psychical Research. There is no reason to suppose that the latter will be more numerous or striking among savages than among civilised men. The first group provides no evidence at all for survival, since the facts have simply been misinterpreted. The second, supposing it to exist, contains no evidence  antecedent to Psychical Research; since, by hypothesis, it consists of precisely those phenomena which would now be treated by that science. Hence the primitive man had simply more causes, but no better reasons, for a belief in survival than we have; but a belief irrationally caused in him may have been handed on to us. No doubt experiences of fainting and sleeping helped the belief in survival. In these conditions the mind gives no external manifestations of its existence, and the body in many ways resembles a corpse. Yet consciousness returns; and, if we remember our dreams, we remember that it was not really absent when our bodies were giving no external signs of its existence. What more natural then than to suppose that at these times the mind leaves the body for a while and follows its own adventures, and that at death it leaves the body for good? But the differences between sleep and death make it impossible to accept this undoubted cause of a belief in survival as a valid reason in its favour. If, after dissolving a bit of silver several times in nitric acid and getting it back again, we one day dissolved it in something else and found that no efforts of ours could restore anything with the properties of silver, the inference would be obvious. It was reasonable to think that the silver survived the nitric acid treatment, because it could he restored; it would not be reasonable to conclude from this that it also survived the treatment after which nothing like it can be again obtained. If we choose to assume that it still exists, our assumption is an unmotived one. So once more we have a cause of belief which is not a reason for belief. Probably neither of the above-mentioned causes would have sufficed to produce an almost universal belief in survival. Both are to be regarded as interpretations of real or supposed facts in terms of this belief rather than as the original causes of it. The truth is that we have the greatest difficulty in actually envisaging the cessation of our own conscious life. It is easy enough to think of anyone else as having really ceased to exist; but it is almost impossible to give more than a cold intellectual assent to the same proposition about oneself. In making a will, e.g., containing elaborate provisions for the disposal of one's property after death, it is almost impossible (unless my experience be quite exceptional) not to think of oneself as going to be conscious and able to oversee the working of one's own bequests. I at least can continually catch myself in this attitude, and I should imagine it to be quite common even among people who are intellectually persuaded of their future extinction. Ought we to attach any weight to this primitive belief which nearly every one has in his own survival? The mere fact that it is held without reasons is no conclusive objection to it; for, unless some propositions can be known to be true without reasons, no proposition can be known to be true for reasons. We must, therefore, consider the belief on its merits without prejudice. Now it seems perfectly clear that it is not a self-evident proposition like an axiom, which becomes more certain the more carefully we inspect it. Nor can it be regarded as a postulate; i.e., as a proposition which, though not self-evident and incapable of either proof or disproof by experience, has to be assumed in order to organise experience and to furnish a motive for research. Certain propositions which we use in induction seem to me to be postulates in this sense; the proposition that John Jones will survive the death of his body seems to me to be quite plainly nothing of the kind. In fact I think that the belief represents nothing more profound than an easily explicable limit of our powers of imagination. Naturally all my experience of myself has been of myself as conscious and active. There have indeed been gaps during dreamless sleep or fainting fits, but consciousness has revived and the gaps have been bridged by memory. Again, at every moment I have been obliged for practical purposes to think of myself as going to exist at later moments; it is therefore a breach with the mental habits of a lifetime to envisage a moment after which the series of my conscious states shall have finally ended. This practical difficulty, due to habit, seems the sole and sufficient explanation of our primitive belief in our own indefinite continuance; and it obviously provides no evidence for the truth of that belief. I think then that we must conclude that a mere contemplation of the world as it appears in ordinary experience furnishes no trace of support for the belief in survival. Ought we to hold that the absence of all evidence for constitutes evidence against? This is a somewhat delicate question. Sometimes the absence of evidence for a proposition makes strongly against it, and sometimes it does not. If I look carefully round a room and, seeing no one, say: "There is no one in the room", my evidence is purely negative; but it is almost conclusive against the proposition: "There is someone in the room". But the fact that I did not see a tuberculosis bacillus in the room would be quite irrelevant to the question whether there was one there. Finding no evidence for a proposition is evidence against it only if the proposition be such that, if it were true, there ought to be some observable evidence for it. Now the proposition: "Some men survive the death of their bodies" is not precisely in the position of either of the two quoted above. I know enough about human bodies and about tuberculosis bacilli to be sure that one of the former could hardly be present in a room without my finding it, but that one of the latter could not be seen by the naked eye even if it were present. I know very much less about the conditions under which one human spirit can make its presence known to another; but I do know something about it. I am a human spirit connected with a body, and all other spirits of whose existence I am certain are in the same position. Setting aside the phenomena treated by Psychical Research, I know that one such spirit can make its presence known to another only by moving its own body, thence agitating the air or the ether, and thence affecting another human body. My friend dies; I remain alive and connected with my body. Communication with me, therefore, presumably requires the same complex and roundabout series of material changes as before. Its very complexity and indirectness make it not unlikely that, even if my friend has survived, some necessary link in this mechanism will have broken down. Hence the absence of evidence for his survival cannot be regarded logically as very strong evidence against it. The present position, therefore, is that at the level of ordinary experience there is not the faintest trace of evidence for survival, though there is a pretty general belief in it. The causes of this belief have been enumerated and seen not to be reasons. But the absence of evidence for the belief cannot be taken as strong evidence against it, in view of what we know about the means by which embodied human spirits have to communicate with each other. Is there at this level any  positive evidence against survival? I think that there are two sets of facts which impress common-sense and are interpreted in this direction. One is the apparently haphazard way in which men are born and die. Human beings are constantly brought into the world thoughtlessly and by mistake; many children live for a few minutes or hours and then die; many are born idiotic. The general impression produced is that the claim to permanence for creatures whose earthly lives begin and end in these trivial ways is somewhat ridiculous. An unwanted child is produced, let us say, in a drunken orgy; and in six weeks dies of neglect or is killed by its mother. Does it seem likely that a being whose earthly career is started and stopped by such causes is a permanent and indestructible part of the universe, or indeed that it survives the death of its body at all? The second fact which is felt to bear in the same direction is the continuity between men and animals. The bodies of each begin and cease to be animated by minds through precisely similar physical and physiological causes. No doubt the mind of any living man differs, not merely quantitatively, but also qualitatively from that of any living animal; still the most primitive men can hardly have differed appreciably from the highest animals in their mental endowments. Did Pithecanthropus erectus and does every Australian aborigine survive the death of his body? If they do, have not the higher animals almost an equal claim? And, if you grant this for cats and monkeys, will you not be forced in the end to grant it for lice and earwigs? If, on the other hand, you deny that any animal survives, on the ground that their minds are not complex or important enough to be permanent factors in the universe, how can you be sure that any man yet born has possessed a mind complex and important enough for survival? The two facts quoted above do, I am sure, exert a considerable influence against the view that men survive the death of their bodies. I am conscious that they affect me personally more than any others. But the question remains: "Have they any logical right to exert this influence?" I am inclined to think on reflection that the first argument is wholly fallacious It really involves the illegitimate introduction of a judgment of value into a question of fact. And the judgment of value is itself a rather superficial one. It is thought that, because the occasioning causes of birth and death are often trivial, therefore what seems to begin with birth and to end with death cannot be important enough to survive. But (a) you cannot argue from the triviality of a cause to the impermanence of its effect. (b) The cause is trivial only in the irrelevant ethical sense that it does not involve a considered and deliberate choice by a virtuous human being. There is really no logical transition from: "This is caused by the careless or criminal action or a human being" to: "This is the kind of thing whose existence is transitory". (c) When we say that the cause is trivial we make the common mistake of taking for the cause some necessary cause factor which happens to be specially noticeable or of special practical interest. The complete cause of the birth of a child or the death of a man must be of almost unthinkable complexity, whether the child be begotten or the man be killed carelessly or with deliberate forethought. This is true even if we confine ourselves to the material conditions; and we are not really in a position to say that the complete conditions of so singular an event as the manifestation of a new mind through a new body are contained in the material world. The second argument is of course of a well-known general type. It tries to show by continuity of cases that, if a man asserts one proposition, he ought in consistency not to deny a certain other proposition which he would like to deny. Arguments of this kind can be met in one of two ways. (1) We may point out that an argument from continuity is reversible, and that the direction in which one turns it is arbitrary. We might just as well argue by continuity from the supposed immortality of men to the immortality of earwigs as from the supposed mortality of earwigs to the mortality of men. The actual direction in which the argument is used presupposes that we are  already pretty certain that earwigs are mortal, and much more doubtful whether men are immortal. This no doubt is true. But it immediately raises the question: "Why are we practically certain that earwigs are mortal?" This question cannot be answered by considerations of continuity, but only by reflecting on the special peculiarities of earwigs. (2) When we raise this question two answers are possible. (a) We may find on reflection that we have no good reason for thinking that earwigs are unlikely to be immortal. In that case the argument from continuity to the case of men will prove nothing. Or (b) we may find that those characteristics of earwigs which make it very unlikely that they are immortal are obviously not present in men. In that case the argument from continuity will also prove nothing about men. At most it will show that it is difficult for us to say with confidence about certain intermediate forms of living being whether they are likely to be mortal or not. Let us then consider the question why we think it very unlikely that earwigs should be immortal; and let us also consider whether the reasons, whatever they may be, apply to men also. In the first place it might be said that an earwig's mind has very little value, and therefore it is unworthy to be a permanent factor in the universe. And it might be argued that it is therefore unlikely to survive. But (a) this would be an ethical argument of a kind which we have already dismissed. And (b) even if it were valid, it is obvious that most human minds are enormously more valuable than the mind of any earwig; so that it would not be inconsistent to think it likely that human minds are immortal and unlikely that the minds of earwigs are so. All that we should be entitled to say is (a) that it is not certain even that any human mind is valuable enough to be immortal; and (b) that, if it were certain, there would be intermediate cases, e.g., cats, about which the probabilities are about equally balanced. But the differences between the minds of men and those of the lower animals are never  mere differences of value. Presumably an earwig's mind has very little unity, complexity, or comprehensiveness. Now it is arguable that such a very simple mind is not very likely to survive bodily death. But (a) I do not think that what we know of nature suggests any straightforward connexion between unity and complexity on the one hand and stability on the other. Both the very simple and the highly comprehensive seem to be fairly stable, though for different reasons. The very simple, like the electron, is stable because of its comparative indifference to changes in external conditions. The highly unified and comprehensive complex, like the solar system, tends to be stable because it contains so much within itself that there is little left over to disturb it. It is therefore quite in accordance with what we know of the order of nature to suppose that the simplicity of the earwig's mind gives it a particularly good chance of survival. (b) Suppose, on the other hand, that we do hold that the simplicity of the earwig's mind makes it very unlikely to survive. Then we must admit that the human mind is enormously less simple and more comprehensive and highly unified. Hence it would be perfectly consistent to hold that the human mind is likely to survive because of its unity and comprehensiveness and that the earwig's mind is unlikely to survive because of its simplicity and poverty of content. Thus on neither alternative does the argument from continuity make it unreasonable to hold that the human mind is likely to survive. As before, all that we can legitimately conclude from the argument from continuity is (a) that it is uncertain whether any human mind even is complex and comprehensive enough to survive; and (b) that, if it were certain, there would be cases of intermediate complexity, e.g., cats, about which the probabilities would be nearly equally balanced. Again, some people no doubt shrink from admitting the possibility of survival to the lower animals out of horror at the immense number of minds which there would be if none, even of the lowest kind, died with the death of their bodies. This shrinking from mere numerical vastness seems to me to be childish. We have no reason to suppose that the universe is conducted in accordance with the Law of Parsimony; and it may well be that the world exhibits a profusion in the item of minds which would horrify the inhabitants of Aberdeen. Thus I do not think that this consideration makes it specially improbable that earwigs should be immortal. Lastly, the following argument might be used to suggest that the minds of the lower animals are very unlikely to survive the death of their bodies. The characteristic activities and experiences of animals seem to be specially and exclusively directed to preserving their own lives and those of their offspring. If we judge living things teleologically (and, in practice, it is hard to avoid doing this) it does seem that an animal accomplishes "all that is in it" when it succeeds in keeping itself alive long enough to produce young and to start them in the world. It is hard to see what "purpose" would be served by the individual survival of an earwig which dies at a reasonable age after bringing up a family of little earwigs. I do not know what weight to attach to such an argument as this. The principle of judging living beings and their parts in terms of a supposed "purpose for which they were made" is undoubtedly valuable as an heuristic method; and it is difficult to suppose that it does not in some way accord with the facts. But fortunately it is not necessary for our purpose to decide on the legitimacy of such considerations. For the position is this. (a) If it be not valid, the argument to show that earwigs are very unlikely to survive falls to the ground; and with it goes the argument from continuity to the probable mortality of human beings. (b) If, on the other hand, it be valid, the argument from continuity equally breaks down in another way. For it does seem as if human minds had many powers and faculties which are not merely directed to preserving the life of the individual and the species; and that the continued existence of certain human minds after the death of their bodies would "answer the purpose for which they seem to be made" in a way in which the continued existence of an individual earwig would not. Hence it would be perfectly consistent to hold, on the basis of this argument, that earwigs are most unlikely to be immortal and that men are quite likely to be immortal. As usual, the argument from continuity would raise a doubt only about certain intermediate cases, such as cats and dogs, where the probabilities might be about equally balanced. To sum up. The argument from continuity makes against the probability of human survival only on two conditions. (1) There must be some reason (and not a mere prejudice) for thinking that the survival of the lower animals is very improbable. And (2) this reason must not be the presence of some characteristic in the lower animals which differentiates them sharply from human beings. For, if our only reason for thinking it very unlikely that earwigs will survive be some characteristic in which earwigs differ profoundly from men, it will be perfectly consistent to think it likely that men will survive and that earwigs will not. The existence of a continuous series of intermediate forms between earwigs and men will prove nothing except that there are certain intermediate cases in which the probabilities for and against survival are about equally balanced. And there would not be the least trace of inconsistency in the position of a man who should be practically certain that earwigs are mortal and human beings immortal but should be quite unable to make up his mind about cats or kangaroos. Now, so far as I can see, these two conditions are never both fulfilled. The alleged reasons for thinking it very unlikely that earwigs are immortal either are no reasons at all or they obviously depend on characteristics in which human beings and earwigs differ profoundly. Hence I doubt whether the argument against the probability of human survival, drawn from the continuous series of living forms between men and the lowest animals, has any logical validity. The world then, as it presents itself to common-sense and everyday experience, offers no positive reasons for and no positive reasons against human survival. The only reason against it is the utter absence of all reasons for it; and we have seen that this is not a strong argument in the present case. Let us now enquire whether the more detailed investigations of science provide us with any grounds for deciding one way or the other. Science on the whole does not reverse, but merely amplifies and elaborates, the views of common-sense on the connexion of body and mind. We already knew that body and mind were intimately connected, and that injury to the former may gravely modify or to all appearance destroy the latter. The additional information gained from science may be summed up as follows. (i) More detailed knowledge has been got of the correlation between injuries to particular regions of the brain and defects in certain departments of mental life. Connected with this is the knowledge that many mental processes, which seem to common-sense to be almost independent of the body, have bodily correlates. (ii) We have gained the surprising information that, in spite of the apparent interaction of body and mind, the body and its material surroundings form a closed energetic system from the point of view of the Conservation of Energy. (iii) We know more about the detailed structure and general plan of the brain and nervous system. What bearing has all this on the probability of survival? We find bodies without minds; we never find minds without bodies. When we do find minds we always find a close correlation between their processes and those of their bodies. This, it is argued, strongly suggests that minds depend for their  existence on bodies; in which case, though survival may still be abstractly possible, it is to the last degree unlikely. At death there takes place completely and permanently a process of bodily destruction which, when it occurs partially and temporarily, carries with it the destruction of part of our mental life. The inference seems only too obvious. I think it is fair to say that our ordinary scientific knowledge of the relation of body to mind most strongly suggests epiphenomenalism, though it does not necessitate it; and that epiphenomenalism is most unfavourable to the hypothesis of human survival. It is, however, possible to put forward other theories about the mind and its relation to the body, which are consistent with ordinary experience and with scientific knowledge and are less unfavourable to survival than epiphenomenalism. I will call the first of these the "Instrumental Theory." The Instrumental Theory We must begin by drawing a distinction between the existence of a mind and its manifestation to other minds. On the Instrumental Theory the mind is a substance which is existentially independent of the body. It may have existed before the body began, and it may exist after the body is destroyed. For a time it is intimately connected with a certain body; and at such times it can get information about other things only by means of its body and can act on other things only by first moving its body. If the body be injured the mind may be cut off from certain sources of information about other things, and it may be prevented from expressing itself in certain ways; but otherwise it may be uninjured. It is certain that such a theory as this is consistent with a good many of the facts which are commonly held to prove the existential dependence of mind on body. Nevertheless, I think that, in this crude form, it cannot be maintained. Let us take the case of a man who is injured in a certain part of his brain, and for the time loses his power to remember certain events. It can hardly be maintained that, in any literal sense, he still remembers the events; and that all that has been damaged is his power of manifesting this knowledge to others by speech or writing. The latter case does sometimes arise, and it seems introspectively quite different from the former to the patient himself. Again, if the patient recovers these lost memories after a while, it seems to him that a change has taken place in the contents of his mind, and not merely a change in his ability to express to others what was going on in his mind before. We must suppose then that in such cases something more than the power to manifest one's knowledge to others has been injured. The only other alternative is to suppose that all such patients are lying and asserting that they cannot remember certain things which they actually are remembering if we reject this very violent alternative we must hold that in some cases an injury to the brain does actually deprive the mind of the power to remember certain events which it formerly could remember. Could a supporter of the Instrumental Theory square the facts with his view? He might say that the general power of remembering is unchanged; and assert that all that has happened is that the injury to the body has prevented certain past events from being objects of memory, as blindfolding a man would prevent certain present objects from being perceived. But in that case the mind is reduced to something which has merely certain very general capacities, and any particular exercise of these powers seems to depend on the body. Let us now take another example. We will suppose that a man is injured in the head; that before the injury he was of a cheerful and benevolent disposition; and that after the injury he is morose and liable to attacks of homicidal mania. Are we to say that the injury has made no difference to his mind; that this remains cheerful and benevolent; but that the change in his brain compels him to express his cheerfulness by scowling and his benevolence by attacking other people with carving-knives? This is scarcely plausible. And, if we accept it, we shall not be able to stop at this point. We shall have to conclude that it is impossible to tell what the character of anyone's mind really is. Lifelong philanthropists may be inwardly boiling with malice which some peculiar kink in their brains and nervous systems compels them to express by pensioning their poor relations and giving pennies to crossing-sweepers. Once more, the mind will be reduced to something with no definite traits of its own, such as benevolence or peevishness, but merely with certain very general powers to express itself in various ways according to the body with which it is provided. It seems to me that what is left of the mind when we try to square the Instrumental Theory with the known facts is so abstract and indefinite that it does not deserve to be called a "mind". The Compound Theory This suggests a modification of the Instrumental Theory, which I will call the "Compound Theory". Might not what we know as a "mind" be a compound of two factors, neither of which separately has the characteristic properties of a mind, just as salt is a compound of two substances, neither of which by itself has the characteristic properties of salt? Let us call one of these constituents the "psychic factor" and the other the "bodily factor". The psychic factor would be like some chemical element which has never been isolated; and the characteristics of a mind would depend jointly on those of the psychic factor and on those of the material organism with which it is united. This would allow of all the correlation between mind and body which could ever be discovered, and at the same time it is not open to the objections which I have pointed out in the ordinary form of the Instrumental Theory. Moreover, it is in accord with many facts which we know about other departments of nature. We know that chemical compounds have properties which cannot be deduced from those which their elements display in isolation or in other compounds. And yet the properties of these compounds are wholly dependent on those of their elements, in the sense that, given such elements in such relations, a compound necessarily arises with such and such properties. These properties do not belong to either of the elements, but only to the compound as whole. Now this does seem to accord fairly well with what we know about minds when we reflect upon them. On the one hand, it seems a mistake to ascribe perception, reasoning, anger, love, etc., to a mere body. On the other hand, as we have seen, it is almost equally difficult to ascribe them to what is left when the bodily factor is ignored. Thus the mind, as commonly conceived, does look as if it were a compound of two factors neither of which separately is a mind. And it does look as if specifically mental characteristics belonged only to this compound substance. It would be unwise to press the analogy to chemical compounds too far. So far as we know, when two chemical elements are united to form a chemical compound no permanent change is produced in the properties of either. It would be rash to assume that this is also true when a psychic factor is united with a bodily organism so as to give a mind. Both factors may be permanently affected by this union; so that, if they become separated again and continue to exist, their properties are characteristically different from what they were when the two first became connected with each other. Of course many different views would be antecedently possible about the supposed psychic factor. At one extreme would be the view that there is only one psychic factor for all minds. Different minds would then be compounds of this one psychic factor with different brains and nervous systems. Such a view would bear some analogy to Green's theory of the one Eternal Consciousness and the many animal organisms. But the psychic factor on our view would have no claim to be called a "Consciousness"; it would not perform those feats of relating and unifying sense-data which Green ascribed to it; and there is no reason to suppose that it would deserve honorific titles like "eternal", or be an appropriate object for those religious emotions which Green felt towards it. At the opposite extreme would be the view that there is a different psychic factor for each different mind. Then the question could be raised whether some or all of them can exist out of combination with organisms; whether some one psychic factor can combine successively with a series of different organisms to give a series of different minds; and so on. (It may be remarked that the view that the psychic factor cannot exist out of combination with organisms, and yet that the same psychic factor can be combined with a series of successive organisms, has a pretty close analogy to certain chemical facts. There are groups, such as NH4, CH3, etc., which are incapable of more than the most transitory independent existence. Yet one such group may pass successively from one combination to another, and may impart certain characteristic properties to each of these compounds.) Finally, there is an intermediate possibility for which there might be a good deal to be said. It might be suggested that the marked individuality of human minds indicates that there is a different psychic factor as well as a different bodily organism to each coexisting human mind. On the other hand, it might be held that there is only one psychic factor for the whole species of earwigs; and that the very trivial differences between the mind of one earwig and another are due simply to differences in their bodily organisms. It is obvious that only empirical evidence of a very special kind could help us to decide between these alternatives, even if we accepted the Compound Theory in its main outlines. Granted that the Compound Theory is consistent with all the facts which are commonly held to prove the existential dependence of mind on body, and granted that it is in better accord with the facts than the instrumental Theory, is there any positive evidence for it? We have a set of facts which point to the dependence of mind on body. One explanation is that mind depends on nothing but body, i.e., that mental events either  are also bodily events, or that at any rate they are all caused wholly by bodily events and do not in turn affect either each other or the body. The present explanation is that the mind is a compound of the body and something else, and that mental events and mental characteristics belong to this compound substance and not to its separate constituents. Both explanations fit all the normal facts equally well. But the Compound Theory is more complex than the Epiphenomenal Theory, and it would be foolish to accept it unless there were some facts which it explains and which the Epiphenomenalist Theory does not. Now I do not think that there is anything in the normal phenomena which requires us to suppose that a mind depends for its existence and functioning on anything but the body and its processes. We must therefore turn to the abnormal phenomena. Abnormal and Supernormal Phenomena I think that it is very important to begin by drawing a distinction which is too commonly neglected, viz., the distinction between Survival and mere Persistence. It seems to me that a great many of the phenomena which are held to point to the survival of particular human minds point only to the persistence of some factor which was a constituent of a human mind. We are not justified in saying that the mind of John Jones has survived the death of his body unless we have reason to believe that there is still a continuous stream of conscious mental states which may be said to be "further experiences of John Jones". We must suppose that this contains conations as well as cognitions, that it puts ends before itself and tries to realise them, and that it feels elation or disappointment according to its success or failure in doing so. No doubt such a stream of consciousness would be impossible unless past experiences modified later experiences; and no doubt we should not say that John Jones had survived unless he were able to remember some events in his life in the body. But these mnemic phenomena, though necessary to survival, are certainly not by themselves sufficient to constitute survival. If they occur alone, without the continuous stream of conscious cognitions, conations, and feelings, all that we have a right to say is that "some constituent of the mind of John Jones has persisted" and not that "John Jones has survived". Now it seems to me that the vast majority of mediumistic phenomena which are taken to suggest survival really suggest only persistence. The additional notion of survival is read into them because in our ordinary experience we do not find memories without a pretty continuous stream of consciousness filling the gaps between the memory and the event remembered. The cases that I have in mind are these. A medium goes into a trance. He is then supposed either to be in contact with the spirit of some dead man, or in rarer cases to be directly possessed by such a spirit. In either case he sometimes mentions incidents in the past life of the supposed communicator which are unknown to the sitter and can afterwards be verified. And in the latter case he sometimes exhibits in a very remarkable way some of the mannerisms and even the verbal intonations of the supposed communicator. The evidence for such phenomena is, in my opinion, good enough to make them worth serious consideration by philosophers. Now the ordinary spiritualist interprets such phenomena in terms of the Instrumental Theory; he supposes that a human mind is existentially independent of its body and just uses it as an instrument; that it leaves its body at death, goes on living its own life, and from time to time uses a medium's body for purposes of communication. But it seems to me that, apart from the intrinsic difficulties of the Instrumental Theory, the Compound Theory fits these supernormal phenomena on the whole much better. One thing which is highly characteristic of the communications of alleged dead men is their singular reticence about their present life, occupations, and surroundings. Such observations as are made by entranced mediums on these subjects seem to me to be extraordinarily silly, and to have every appearance of being merely the crude beliefs about the spiritual world which are current in mediumistic circles. Yet this nonsense is at times mixed up with traits which are highly characteristic of the supposed communicator, and with bits of detailed information about his past life which can afterwards be verified. Now, on the Compound Theory, we can suppose that the psychic factor may persist for a time at least after the destruction of the organism with which it was united to form the compound called "John Jones's mind". This psychic factor is not itself a mind, but it may carry modifications due to experiences which happened to John Jones while he was alive. And it may become temporarily united with the organism of an entranced medium. If so, a little temporary "mind" (a "mindkin", if I may use that expression) will be formed. Since this mindkin will contain the same psychic factor as the mind of John Jones it will not be surprising if it displays some traits characteristic of John Jones, and some memories of events in his earthly life. Since the bodily factor of this mindkin is the medium's organism, which is adapted to the medium's psychic factor and not to John Jones's, it will not be surprising if it shows many traits which are characteristic of the medium. And the reason why we can get no information about the present life and experiences of John Jones is that no such mind is existing at all. When the medium is entranced the psychic factor which was a constituent of John Jones's mind forms with the medium's body a mindkin which lasts just as long as the medium remains in trance. At intermediate times, on this view, all that exists is this psychic factor; and this by itself is no more a mind than John Jones's corpse is a mind. To explain the positive part of the phenomena it is plausible to suppose that  something has persisted, and that this something was an integral part of John Jones's mind. But it is an enormous jump from this to the conclusion that John Jones's mind has survived the death of his body. And the negative part of the phenomena strongly suggests that what has persisted is not a mind, but is at most something which in combination with a suitable organism is capable of producing a mind. Some of the facts of multiple personality would also be neatly explained by the Compound Theory. Of course, mediumistic phenomena are, in the first instance, cases of multiple personality. The peculiarity of them is that one of the personalities professes either to be a certain deceased human being, or more usually only to be in communication with one; and that, in some cases, there appear certain characteristic traits of this dead man, or knowledge is shown of some minute details in his past life. But ordinary multiple personality, such as that of the Beauchamp case, might be explained by supposing that the same organism can have two different psychic factors connected with it. We should then expect to find two minds having certain characteristic differences, and yet having a good deal more in common than two minds which differ in their organisms as well as in their psychic factors. Two personalities might be compared to two chemical compounds with one element in common, such as silver chloride and silver bromide; whilst two ordinary minds might be compared (say) to silver chloride and lead nitrate. I do not think, however, that ordinary multiple personality positively requires the Compound Theory for its explanation. We can never be sure that the organism is in precisely the same state when one personality is in control as it is when the other is in control. Hence it is possible that the facts could be explained on a purely epiphenomenalist theory. It is the apparent persistence of certain traces and dispositions after the destruction of the organism which seems to demand for its explanation something more than epiphenomenalism, and seems to suggest  at leastsomething like the Compound Theory. We must now consider: (a) whether there are any facts which require something  more than the Compound Theory to explain them; and (b) whether the facts that I have already mentioned could be explained with something less than the Compound Theory. It seems to me that we should have grounds for postulating the  survival of a mind, and not the mere persistence of a psychic factor, if and only if the communications showed traces of an intention which persisted between the experiments and deliberately modified and controlled each in the light of those which had preceded it. Now it is alleged that there are signs of this deliberate intention in the Cross-Correspondences which the Society for Psychical Research has been investigating for many years. If all or most of these came up to the ideal type of a Cross-Correspondence, I think we should have to admit that it looks as if a single intelligent being were deliberately trying in an extremely ingenious way to produce evidence of its continuous existence. The ideal Cross-Correspondence would be of the following form. Suppose three automatic writers in different places produce automatic scripts over a series of years. Suppose that they do not communicate with each other, but send their scripts from time to time to an impartial authority for comparison. Suppose that A, B, and C in their scripts get statements which, taken separately, are fragmentary and unintelligible to them; and suppose further that after such an unintelligible and fragmentary statement in A's script there comes an injunction to refer to what B and C are now writing or will shortly write or have written at some definite time in the past. Suppose that similar injunctions are found in B's and C's scripts after fragmentary and unintelligible passages in them. Suppose finally that when the impartial authority compares the scripts and follows the directions contained in them he finds that these separately unintelligible sentences combine to convey something which is highly characteristic of a certain deceased person who is alleged to be communicating. Then we should have a perfect instance of a Cross-Correspondence; and it would be difficult to resist the conviction that the phenomena are controlled intentionally by a single mind, which cannot be identified with the conscious part of the mind of any of the automatic writers. Unfortunately it is not clear to me that most of the alleged Cross-Correspondences accurately exemplify this ideal type. I also cannot help feeling suspicious of the enormous amount of learning and ingenuity which the impartial authority has to exercise in order to find the key to the riddle which the scripts set. Would not the same amount of patience, learning, and ingenuity discover almost as good Cross-Correspondences between almost any set of manuscripts? I do not say that this is so; but I should need a good deal of negative evidence, i.e., of failure to discover Cross-Correspondences between other manuscripts which were treated in the same way as these automatic scripts, before I was prepared to stake much on this argument for human survival. So far as I am aware, negative control experiments of this kind have not been tried. It is evident that they would be terribly laborious, and it is hardly to be expected that the same patience and ingenuity would be lavished on them as have been devoted to the interpretation of the automatic scripts in which positive results are hoped for. There is another remark to be made on the Cross-Correspondences. Suppose that they rendered it practically certain that  some mind other than the conscious minds of the automatists is controlling the experiments, can we feel any confidence that it is the mind of a certain deceased person who professes to be communicating? Is it not at least equally probable that it might be the unconscious part of the mind of one of the automatists or of one of the officers of the Society for Psychical Research? It would certainly be true to say that some of the automatists (in particular Mrs Verrall) were well aware of the problem of getting evidence for survival which could not be explained away by the hypothesis of telepathy between the living; that it must have occupied their thoughts a great deal; and that they must have had a permanent desire to devise some means of solving it. It is also true that the alleged communicators in the Cross-Correspondences had been well known in life to Mrs Verrall and to many prominent and active members of the Society who were not themselves automatists. Now I think that we may take the following propositions as reasonably well established. (a) That when a person is greatly interested in a problem this problem is often worked upon and solved by processes which are unconscious relatively to the part of the mind which is normally in control of his body. I need only mention in support of this the quite common experience of solving a problem while asleep, or the post-hypnotic calculations which I spoke of in an earlier chapter. (b) That it is extremely probable that telepathy can and does take place between the unconscious parts of living minds. In sittings with Mrs Leonard and other mediums I have met with clear cases of telepathy between myself and the medium when entranced. But I have noticed that these almost invariably involved past events of which I was not consciously thinking at the time. Thus the telepathic influence must have been due to mere "traces", or at most to processes of thought going on in my mind without my being aware of them, i.e., processes which were unconscious relatively to the part of my mind which normally controls my body. (c) That the unconscious part of the mind is often extremely willing to "oblige" the conscious part by providing "evidence" for what the conscious part wishes to believe. Now, if these three propositions be admitted, it is not unplausible to suggest that the unconscious part of the mind of one of the automatists worked out the problem of providing "satisfactory evidence" for survival and telepathically conveyed the fragmentary messages, which were to constitute the "evidence", to the other automatists. Personally I strongly suspect the unconscious part of Mrs Verrall's mind to have accomplished this feat. I am of course quite well aware that such a theory goes far beyond anything for which we have direct evidence; for it seems to imply that the unconscious part of Mrs Verrall's mind was capable of a kind of  selective telepathy, conveying so much and no more to one automatist and so much and no more to another automatist. But I must point out that, if we do not ascribe this power to any embodied mind, we have to ascribe it to the disembodied mind of the supposed communicator. So this much must be assumed in any case if we accept the interpretation which the investigators have put on the Cross-Correspondences. And, except on the principle of Omne ignotum pro magnifico, I do not see why we should think it more likely that the disembodied mind of a dead man should be able to exercise selective telepathy than that the unconscious part of the embodied mind of a living member of the Society for Psychical Research should be able to do so. In fact the hypothesis that the spirit of the late Dr Verrall is communicating involves the assumption both of an otherwise unknown power of selective telepathy and of an otherwise unknown substance, viz., a disembodied spirit, to exercise this power. The hypothesis which I tentatively put forward makes only the first of these two assumptions. It therefore has a greater intrinsic probability; and it seems equally capable of explaining the facts. I pass now to the second question. Could the facts which we have been considering be explained by something less than the hypothesis of a persistent psychic factor? It will be remembered that the facts to be explained are the revelation of certain details in the past life of a certain dead man, which are unknown at the time to the sitter and can afterwards be verified; or the occurrence of certain characteristic tricks of voice and manner in the entranced medium. Now it must be admitted that it is very rare for a detail about a dead man's past life to be verifiable unless it is known or has been known to someone now living. It must therefore be admitted to be theoretically possible that these phenomena are due to telepathy from the unconscious parts of the minds of living men who are remote from the place at which the sitting is being held. But, although this is conceivable, I cannot regard it as very plausible. It is very difficult to see what can determine the medium to select just those pieces of information from distant minds which are relevant to the supposed communicator. It is true of course that the sitter has generally known the communicator; and we should have to suppose that the presence of a man who has known X causes the medium to select from other minds bits of information about X and to reject bits of information about other men. On any view some selective action on the part of the sitter must be postulated, since in the main those who are supposed to be communicating when a certain man has a sitting with a medium are people whom the sitter has known. In my own sittings with Mrs Leonard, e.g., the alleged communicator has from the first been one particular man who was described with considerable accuracy and named with approximate (though not complete) accuracy at the first sitting. On the Compound Theory we should have to suppose that the presence of a certain sitter "attracts" the psychic factors of certain dead men who were known to him. On the purely telepathic theory we should have to suppose that the presence of the sitter causes the medium to "select" from various minds scattered about the world certain bits of information which are relevant to someone whom the sitter has known. Although this hypothesis is possible, there are, I think, two arguments which make slightly against it and slightly in favour of the Compound Theory. (1) On the purely telepathic theory it is difficult to see why mediumistic communications should not be as much or more concerned with one's living friends as with those who have died. This is not found to be so. On the Compound Theory this fact is explicable; for, on this hypothesis, the psychic factor of a living mind is already attached to a certain living organism, and this would presumably make it difficult or impossible for it to enter at the same time into the same relation with the organism of the entranced medium. I think that some weight must be attached to this argument, though it is not conclusive. The main interest and expectation of both sitter and medium is to get messages which purport to come from the dead and not from those who are still alive; and this might account for the fact that the medium "selects" bits of information about dead men, even on the purely telepathic theory. (2) The second argument is due to Dr Richard Hodgson. He used it against the hypothesis of telepathy from the sitter and in favour of the hypothesis that the messages are due to the disembodied spirits of dead men. I think that the argument can be adapted so that it can be used against the hypothesis of a more extended telepathy and in favour of the Compound Theory. The argument may be put as follows. Suppose that a number of sitters S1 ... Sn sit with a certain medium, and that a number of communicators C1 ... Cn profess to give messages through this medium. On the Compound Theory the adequacy or inadequacy of the communications which purport to come from a certain communicator C, through a given medium would presumably depend mainly on two things; (a) on the complexity of the psychic factor, and (b) on its adaptation to the organism of the medium. There is no obvious reason why the number and accuracy of the messages which purport to come from a given communicator through the same medium should vary much from one sitter to another; for the main function of the sitter, on this hypothesis,is simply to "attract" a certain psychic factor so that it enters into a temporary combination with the medium's organism. If this happens at all, the subsequent proceedings would seem to depend on the psychic factor and the medium rather than on the sitter. We should thus expect to find certain "communicators" who are good with most sitters, and others who are bad with most sitters; we should not expect to find certain sitters who are good with most "communicators" and others who are bad with most "communicators". On the telepathic hypothesis we should expect the opposite result. For, on this view, the sitter plays a much more active part. His thoughts and interests must determine the particular selection of information which the medium makes from a perfect rag-bag of living minds. And his power to do this would presumably depend on the peculiar endowments of his own mind and on its adaptation to the mind of the medium with whom he is sitting. On this hypothesis we should therefore expect that there would be some sitters who get good results from most alleged communicators through a given medium; and that there would be other sitters who get bad results from most alleged communicators through the same medium. Now Dr Hodgson had an enormous amount of experience of the results of sittings with Mrs Piper extending over many years. And he carefully studied them and classified them from the above points of view. His conclusion was that certain alleged  communicators gave copious and accurate information to most sitters; and that other alleged communicators gave fragmentary and incorrect information to most sitters. He did not find that certain sitters got copious and accurate information from most communicators; and that certain other sitters got feeble and fragmentary messages from most communicators. Thus, on the whole, the actual results are such as might be expected on the Compound Theory and are not such as might be expected on the theory of generalised telepathy from living minds. On the whole then I am inclined to think that there is slightly more to be said for the Compound Theory than for the other alternatives. Conclusion The view that the mind is existentially dependent on the organism and on nothing else is compatible with all the normal facts, and is positively suggested by them, though they do not necessitate it. And it is the simplest possible view to take. The theory that the mind merely uses the body as an instrument is difficult to reconcile with the normal facts; and it is doubtful whether there are any well-established abnormal phenomena that require it. The theory that the mind is a compound substance, whose constituents are the organism and what I have called a "psychic factor", is compatible with all the normal facts; though it is not suggested by them, and is more complex than the theory that the mind is existentially dependent on the organism and on it alone. This Compound Theory seems to be the minimum assumption that will explain certain fairly well attested abnormal phenomena. Of course, many people will unhesitatingly reject the alleged facts on which I have based the argument of the latter part of this chapter. I am pretty sure that they will be wrong in doing so; but I will confine myself to this remark for their benefit. Anyone who adopts the view that the mind is existentially dependent on the organism alone is taking up a position which is not necessitated by the facts which everyone admits, and which can hardly be reconciled with the very possibility of many alleged facts for which there is at least respectable prima facie evidence. Now this (I should have thought) is not a comfortable position to occupy. It compels one either to ignore all the phenomena in question, or to be continually occupied in explaining them away. The former course is not scientifically respectable; for it is certain that many people, quite as sensible as oneself and far more expert, have personally investigated these matters and have persuaded themselves of the genuineness of these phenomena and of the impossibility of explaining them completely by fraud or mistake. And the latter course may at any moment be barred by some fact which we simply cannot explain away. Now the Compound Theory has at least this merit. It is compatible with all the facts which everyone admits; it has nothing against it except a superstitious objection to dualism; and it leaves open the possibility that these debatable phenomena are genuine. At the same time it does not compel anyone to accept them. It is quite open for anyone to hold that the mind is a compound of the organism and of a psychic factor which is not itself a mind; and yet to doubt or deny that there is any conclusive evidence that a psychic factor ever persists after the destruction of the organism with which it was combined, or that if it does persist it ever combines even for a moment with the organism of some living human being to form a temporary mind. This seems to me to be the great merit of the Compound Theory. It leaves open possibilities, and allows us to investigate alleged facts without an invincible a priorprejudice against their possibility. And yet it allows us to be as critical as we like about each of these alleged facts, and about the evidence which is offered for each of them. I may remark in conclusion that the Compound Theory has certain advantages for those who favour the theory of metempsychosis, as Dr M'Taggart does. Instead of a single  mind whichanimates a successive series of organisms we should have a single psychic factor which combines with such a series of organisms to form a successive series of minds. There might be intervals during which a psychic factor has become dissociated from an organism which has died and has not yet entered into combination with an organism that is about to be born. During such intervals this psychic factor might produce those abnormal phenomena which the ordinary Spiritualist takes as evidence for the survival of a certain human mind, I do not know of any facts which strongly suggest metempsychosis; but it is a possible theory, and it has the advantage of dealing with the "origin" of the mind at conception as well as with the "end" of the mind at death. And it seems to me to be much more plausible when stated in terms of a persistent psychic factor, which is not a mind, than it is when stated in terms of a persistent mind which animates successively a series of organisms.