Proceedings of the Society for Psychical Research, 43, October 1935, pp. 397-438. Presidential Address. Read May 1, 1935. Reprinted in C. D. Broad Religion, Philosophy and Psychical Research (London: Routledge and Kegan Paul, 1953).

Normal Cognition, Clairvoyance and Telepathy

C. D. Broad

When the Society for Psychical Research did me the honour of making me their President they chose, presumably with their eyes open, a professional philosopher with very little first-hand experience of the subject. I think I shall be most likely to be of use to the Society in my presidential address if I stick to my last and speak as a philosopher.

All of us are aware that our subject differs from most others in the following important respect. It is much harder for us than for workers in other experimental fields to get any empirical facts or first-order generalizations established and universally admitted. No one doubts, e.g. that light is sometimes reflected and sometimes refracted; so the physicist can go on at once to seek for the laws of reflexion and refraction and the conditions under which such events take place. But contrast our position in respect of supernormal cognition. For my own part I have no doubt that telepathy among normal human beings happens from time to time. And it is quite clear to me that, in order to account for the information which is sometimes conveyed by good trance-mediums and automatic writers, a very extensive and peculiar telepathy among the living is the very least that must be postulated. Probably most, if not all, of those here would agree with me. But we know quite well that most scientists and the bulk of the general public would not admit this for an instant. And we know that this is not because they have looked into the evidence and found it faulty or have suggested plausible alternative explanations. They would no more think of looking into the evidence for telepathy than a pious Christian thinks of looking into the evidence for Mahometanism, or a pious Mahometan of looking into the evidence for Christianity. When we leave telepathy and pass to other forms of supernormal cognition there is no agreement even among ourselves. Many of us would say that non-inferential foreknowledge of an event is plainly impossible, and that no evidence could convince us of it. And many of us would feel that the modus operandi of pure clairvoyance or of non-inferential cognition of past events by a person who never witnessed them is so difficult to conceive that we could hardly be persuaded of the occurrence of such cognition.

Of course each of us is influenced to some extent by psychological causes, which are logically irrelevant, when he accepts or rejects an alleged fact or a suggested theory on the strength of evidence submitted to him. But an important logical principle is involved too. The degree of belief which it is reasonable to attach to an alleged fact or a proposed theory depends jointly on two factors, viz. (a) its antecedent probability or improbability, and (b) the trustworthiness of the evidence and the extent to which it seems to exclude all alternatives except the one suggested. On precisely similar evidence it would be reasonable to believe much more strongly that an accused man had cheated at cards if one knew him to be a bookmaker than it would be if one knew him to be an Anglican bishop, because the antecedent probability of the alleged event is much greater in the former case than in the latter. Now antecedent probability depends very largely on analogy or coherence of the suggested proposition with what is already known or reasonably believed about the subject matter with which it is concerned. Antecedent improbability depends very largely on lack of analogy or positive discordance with what is already known or reasonably believed.

The application of this to our subject is obvious. People have at the back of their minds a certain system of knowledge and belief about the nature and conditions of normal cognition. They suspect that the various kinds of supernormal cognition which have been alleged to happen would be utterly different in nature and would presuppose an entirely different kind of causation. They therefore regard the occurrence of supernormal cognition as antecedently very unlikely, and they demand for it evidence of such amount and such quality as they would not think of requiring for alleged facts of a normal kind. This attitude is, up to a point, perfectly reasonable, and it is impossible to say just where it ceases to be so. It seems to me that the whole situation would be very much clarified if the two following requests could be fulfilled. In the first place, we should like to have a clear and explicit statement of what may reasonably be regarded as well-established facts about the nature and conditions of normal cognition. Secondly, we should like to get from psychical researchers a moderately clear statement of what they understand by 'clairvoyance', 'telepathy', 'pre-cognition', etc., and some suggestions about the possible modus operandi of these forms of cognition if they do occur. If such statements were forthcoming, we might be able to see where precisely there is analogy or lack of analogy, coherence or discordance, between alleged supernormal cognition and admitted normal cognition. This would be a great advance on the present vague impression of oddity and upsettingness.

Now a professional philosopher, interested in Psychical Research ought to be of some use in this connexion. He ought at least to be able to make a moderately coherent answer to the first request, and he might be able to make a few suggestions towards answering the second. I propose to devote the rest of my address to these topics.

The forms of supernormal cognition which have been alleged to occur may be roughly classified as follows. We may divide them first into supernormal cognitions of contemporary events or of the contemporary states of things or persons, and supernormal cognitions of past or future events or the past or future states of things or persons. Under the first heading would come Clairvoyance and Telepathy. Under the second heading would come such knowledge of the past as was claimed by Miss Jourdain and Miss Moberley in their book An Adventure, and such foreknowledge as is claimed by Mr. Dunne in his book An Experiment with Time. We will call these 'Supernormal Postcognition' and 'Supernormal Precognition' respectively. Since Clairvoyance, if it happened, would involve no complications about other minds than that of the cogniser or other times than that at which he has his cognition, I shall begin with it. I shall then consider Telepathy. I shall not attempt to deal with Supernormal Postcognition or Precognition in this paper.


Suppose that a person correctly guesses the number and suit of a card in a new pack which he has never touched, and which has been mechanically shuffled so that no one else has the information in his mind at the time. If this were to happen often under test conditions, there would be a prima facie case for postulating pure clairvoyance. It would then be reasonable to raise the following question: 'Supposing that pure clairvoyance does occur, how far, if at all, is it analogous to ordinary sense-perception?' This is the question which I am now going to discuss.

Normal Sense-Perception

Plainly we cannot hope to answer this question until we have stated clearly what happens in normal sense-perception. I shall therefore begin by giving what seems to me to be, on the whole, the most reasonable account of this in view of all the known facts. We shall have to consider it in its psychological, its physiological, and its physical aspects. The subject is very complex and highly controversial, and I shall have to be rather dogmatic in order to be reasonably brief.

I think that the first point to be made is that there are several forms of sense-perception which are, prima facie, fundamentally different in nature. Philosophers have too often confined themselves to a certain one of them, viz. visual perception, in discussing the subject. It is essential that we should not make this mistake if we are seeking for analogies between clairvoyance and normal sense-perception. I begin, therefore, by dividing sense-perception into 'extra-somatic' and 'intra-somatic'. In the former the percipient seems to himself to be perceiving foreign bodies and events; in the latter he seems to himself to be perceiving the inside of his own body and processes going on in it. Now there are at least three important forms of extra-somatic sense-perception, viz. hearing, sight, and touch, which seem, prima facie, to be unlike each other in certain fundamental respects.

Sight and hearing agree with each other and differ from touch in that they seem to reveal to us things and events which are located at various distances out from our bodies. But hearing differs from sight in the following important way. When I say that I hear a bell I should admit that this is an elliptical expression. Strictly speaking, I hear a noise of a rhythmic booming kind which seems to be emanating from a distant place and coming to me in a certain direction. I take it that this place contains a bell, and that a certain rhythmic process in it is causing it to make the noise. On this point there would be no difference in principle between the account which an unscientific percipient would give of the experience as it seems to him and the account which a scientist would give of it from the standpoint of physics. But, when I say that I see a bell, I do not readily admit that I am using an elliptical expression, as I should admit that 'I hear a bell' is short for 'I hear a bell tolling'. I seem to myself to be directly and intuitively apprehending a remote coloured area which I take to be part of the surface of an independent foreign body. I may learn from the scientists that the situation, in its physical aspect, is very much like that which exists when I hear the bell. I may learn that certain rhythmic processes are going on in the place where the bell is, that these cause a disturbance to be emitted in all directions from this centre, and that this disturbance eventually travels to my body and produces a visual sensation. But, even if I accept this as proved, it remains a fact that the situation does not present itself to me in that way when I am having the experience. I continue to seem to myself to be directly apprehending the surface of a remote extended object and to be actively exploring it with my eyes. In this respect visual perception resembles tactual perception, except that the objects are perceived as remote from the percipient's body in the one case and in contact with it in the other.

We may sum up these likenesses and unlikenesses as follows. We may say that hearing is projective in its epistemological aspect, and is emanative in its physical aspect. We may say that sight is ostensibly prehensive and not projective in its epistemological aspect, but is emanative in its physical aspect. And we may say that touch is ostensibly prehensive in its epistemological aspect, and is non-emanative in its physical aspect.

Now the question at once arises whether sight and touch are really, as well as ostensibly, prehensive. We will now consider the two kinds of perception in turn. The mere fact that sight is physically emanative does not, as some people have thought, suffice to prove that it cannot be epistemologically prehensive. It is logically possible that the function of the light-waves which emanate from a distant object, strike the percipient's eye, and thus eventually affect his visual brain-centres, should be purely that of evoking and directing a cognitive act and not in the least that of producing or modifying a cognisable object. In fact the disturbance in the percipient's brain, produced by the light-waves, might simply cause his mind to apprehend directly the coloured surface of the remote object from which the waves emanated. If so, visual perception would really be prehensive. But, although this is logically possible, I think it may quite safely be dismissed as inconsistent with the facts taken as a whole. The argument for this conclusion is cumulative. Each kind of fact which seems to conflict with the view that visual perception is prehensive can, perhaps, be squared with it if we choose to make a complicated and ingenious enough supplementary ad hoc hypothesis. But these various supplementary hypotheses are logically independent of each other; and, when one takes them all together, the prehensive view becomes as complex and artificial and incredible as the Ptolemaic system of astronomy had become just before it expired.

I shall content myself with mentioning one particularly obvious difficulty. Light travels with a finite velocity. It is therefore possible that, when the light which started from a distant star reaches my eye, the star should have moved away from its original position, changed its original colour, or blown up completely. If sight were really prehensive the result of the light now striking my eye and affecting my brain would be that I now directly apprehend the surface of the star as it was when the light left it perhaps a thousand years ago. My act of direct acquaintance would thus have to bridge a temporal gap of a thousand years between the date of its own occurrence and the date of existence of its own immediate object. Yet the object which I see is most certainly perceived by me as simultaneous with my act of seeing it.

I conclude that visual perception, though ostensibly prehensive of external objects, is not really so. All the facts conspire to support the following conclusion. When I have a visual perception I seem to myself to be directly apprehending an area of a certain size and shape, coloured in a certain way, and forming part of the surface of a certain material thing at a certain position outside my body. But the shape and size and position which I perceive it as having, and the colour which I perceive as pervading it, are completely and finally determined, on the physical side, by certain processes which are going on at the time in a certain part of my brain. Provided that these processes are going on in this part of my brain, and that my mind is functioning normally, I shall have exactly this kind of visual experience no matter how the brain-process may have been set up, and no matter whether there is or is not an external body such as I seem to myself to be directly apprehending. If the brain-process has been set up by light which has travelled from an external source through a homogeneous medium to my eye, the visual perception will be as nearly veridical as it is possible for a visual perception to be. If it has been set up by light which has travelled from an external source but has undergone reflexions or refractions before reaching my eye, the visual perception may be highly misleading in many respects, but it will not be utterly delusive. If it has been set up by events in my own body, as in dreams or delirium, or by such abnormal causes as the suggestions of a hypnotist, the visual perception will be utterly delusive. Thus, even in the most favourable case, where there is or has been an external source and where the visual perception gives the percipient correct information about its shape, position, and physical state, the connexion between the act of perceiving and the external source is extremely remote. Even in this case the source and the processes going on in it are at most a remote causal ancestor of the visual perception and are never the immediate object of it. Thus there is always a certain element of delusiveness in even the most normal and veridical visual perception. For the percipient always seems to himself to be directly apprehending the surface of a remote object as it now is, whilst at best he is only cognizing very indirectly certain facts about an emitting source as it formerly was. Owing to the very great velocity of light the time-error is practically unimportant except when the source is at an astronomical distance from the observer. But ostensible prehensiveness, like original sin, is a taint which equally and systematically infects all visual perceptions, good, bad, or indifferent.

One important consequence of this is the following. Consider the statement: 'You and I are seeing the same part of the surface of the table.' There is no reason to doubt that such statements often record facts, and that they do this quite efficiently for most of the practical purposes of daily life. Nevertheless there is a suggestio falsi about them. They suggest that there is a certain part of the surface of a certain external body which you and I are both directly apprehending. But the fact which they record, when they do record a fact, is much more complex and of a very different kind. It would be more accurately expressed by the statement: 'This visual experience of mine and that visual experience of yours, though they are not prehensions of a common object, have a common causal ancestor in an emitting source outside our bodies.'

We can now turn our attention to tactual perception. As I have said, this is ostensibly prehensive in its epistemological aspect, and is non-emanative in its physical aspect. In tactual perception we must distinguish three factors. (i) Awareness of various sensible qualities, such as hotness and coldness, roughness and smoothness, etc. This may be compared with awareness of auditory qualities in hearing and colours in seeing. (ii) Awareness of shape and extent. This may be compared with the corresponding factor in visual perception. There is, I think, nothing like it in hearing. (iii) The experience of actively pulling and pushing foreign bodies which are in contact with one's own and making them move in spite of their varying degrees of resistance to one's efforts; the experience of trying to move them and failing because the resistance which they offer is too great; and the experience of being forced to move, in spite of resisting to one's utmost, by the thrust and pressure of other bodies on one's own. I will call this dynamic experience. I know of nothing analogous to it in any other form of perception.

It is this dynamical factor in tactual perception, and the systematic way in which variations in it are correlated with variations in the non-dynamical factors, which makes it difficult even for the most sceptical to doubt that tactual perception is really prehensive of external objects. We may admit at once that there is not here, as in the case of visual perception, a large coherent mass of facts which it is difficult or impossible to reconcile with the prehensive view. It might even be argued with some plausibility that, unless we really are directly acquainted with foreign bodies in the experience of active manipulation, we should never have seemed to ourselves to be directly acquainted with them in visual perception. But we must not let ourselves be rushed into accepting the prehensive view of tactual perception until we have noted one important fact which may bear in the opposite direction.

Tactual perception shares with sight and hearing a characteristic which we have not yet mentioned. It is transmissive in its physiological aspect, i.e. it depends on the existence and functioning of nerves which connect the periphery of the body to the brain and convey disturbances at a finite rate inwards and outwards. Now it is certain that the occurrence of a characteristic kind of disturbance in my brain is a necessary condition without which I shall not have a perception of myself as touching and interacting with a foreign body. The question is whether the occurrence of such a process in my brain is also the sufficient physical condition of my having such an experience. If it is sufficient I should have exactly the same tactual experience, provided that this process in my brain were to occur and that my mind were working properly, even if there were no foreign body in contact with my skin. If this were so, my tactual perceptions could not be prehensive. It is difficult to settle this question conclusively, because it is doubtful whether precisely that kind of brain-state which occurs when I am actually manipulating and struggling with a foreign body ever does arise from purely internal causes. But the fact that I can dream that I am struggling with a foreign body, though I am in fact doing nothing of the kind, certainly suggests that even the experience of active tactual manipulation may not be really prehensive.

My own tentative view is that tactual perception is probably not prehensive of external objects, but that, in spite of this, it justifies us in being practically certain that there are foreign bodies and that they do interact with our own bodies. It seems to me just conceivable, though extremely unlikely, that I might have had the kinds of experience which I describe as 'seeing' or 'hearing' foreign bodies even if there had been no foreign bodies or if they had never emitted light-waves or sound-waves to my body. But I find it almost impossible to believe that I could ever have had the kind of experience which I describe as 'pushing' or 'pulling' or 'struggling with' foreign bodies unless there had been foreign bodies and they had quite often interacted dynamically with my own body through contact. Granted that this has quite often happened, it is not hard to explain how occasionally, in dreams or delirium, I may have a close imitation of this experience although no foreign body is then interacting dynamically with mine.

There is one important point on which I want to insist before leaving the topic of extra-somatic perception. I have argued that, when we have the experience of hearing, seeing, or touching something, we are not in fact apprehending directly the foreign body, if such there be, which we say we are hearing, seeing, or touching. Now at this stage there is a risk of making a serious mistake. It might be thought that, because hearing, seeing, and touching are indirect and mediate, in the sense of being non-prehensive, they must be indirect and mediate in the sense that they involve inference. This would be a profound mistake. Even in the case of hearing I do not argue, from the fact that I am hearing a booming recurrent noise and from certain general principles of physical causation, that there is probably a bell tolling in a certain place outside my body. The fact is that my auditory experiences have been closely correlated with certain of my visual and tactual experiences in the past, and this correlation has established a persistent system of traces and dispositions in my mind. When I now hear a booming recurrent noise a certain part of this dispositional system is excited, and the auditory sensation is at once invested with an aura of acquired meaning in terms of a remote visible and tangible source. It is still more obvious that there is no element of inference in the experience which I call 'seeing this' or 'touching that'. I doubt whether we can account psychologically for the ostensible prehensiveness of visual and tactual perception by any process of acquirement of meaning through association in our early years. I think we must assume that visual and tactual experiences are taken by us, from the very first, as revelations of an external material world. No doubt all the later detailed development of this primitive vague conviction depends on the actual course of our experience and on the particular associations which are established in our early years.

So much for the purely psychological point. There is a logical point closely connected with it. Beliefs which were not reached by inference may be capable of being supported or refuted by inference. Now, in my opinion, something like the common-sense belief in a world of extended movable interacting bodies can be shown to be highly probable, on the basis of our auditory, visual, and tactual perceptions and their correlations, if and only if the following premiss is granted. Our primitive uncritical conviction that our visual and tactual perceptions are manifestations of an external material world, and that distinctions and variations in them are signs of distinctions and variations in it, must be allowed to have an appreciable antecedent probability. There is no way of proving this indispensable premiss. Some people may find it self-evident and count it as an axiom. I am content to take it as a postulate. We will call it the Postulate of Perceptual Transcendence.

Finally we must consider intra-somatic perception, i.e. the perception which each of us has of his own body, and of no other body, by means of organic sensations. Each of us is almost always aware of a general somatic background or field, which is vaguely extended and fairly homogeneous in quality throughout its extent. It is fairly constant in general character, though its specific tone varies from time to time. Such variations are recorded by expressions like: 'I am feeling tired', 'I am feeling well', 'I am feeling sick', and so on. No doubt the general character changes very slowly as we get older, and it may undergo profound and fairly sudden modifications in illness or at certain periods of normal life such as puberty. Against this fairly homogeneous and constant background there happen from time to time outstanding localized feelings which are independent of one's volition, e.g. a sudden twinge of toothache, a prolonged and voluminous stomach-ache, and so on.

We might compare the general somatic field to the visual field of which one would be aware if one lay on one's back and looked up at the sky when there is not much movement among the clouds. And we might compare the occasional localized outstanding toothaches, stomach-ache, etc., to the visual sensa which we should sense if there were occasional flashes of lightning, dark masses of cloud, and so on, in the sky.

Lastly, we must notice that, whenever we deliberately act upon or react against a foreign body, there are characteristic localized changes in the somatic field, connected with the pressures, tensions, and movements of our muscles and joints.

The following points are of special importance for us to notice. (i) Intra-somatic perception, like all other normal perception, is transmissive in its physiological aspect. If I am to have the kind of experience which I record by saying 'I am feeling a pain in my toe', it is not sufficient that there should be a process of a certain kind going on in my toe. It is necessary that a certain process should be going on in my brain. Moreover, we are told on good authority that persons who have had a limb amputated may yet have experiences of the kind which they would record by saying 'I have a pain where my amputated limb used to be'. It therefore looks as if the occurrence of a certain process in the brain were the final and sufficient physical condition of the occurrence of this kind of experience. If so, intra-somatic perception cannot be really prehensive of one's own body, however much it may seem to be so to the percipient. (ii) There is, however, no reason to doubt that the brain-process, which is the final and sufficient physical condition of an intra-somatic perception, generally arises from and corresponds in structure with a certain process in a certain other part of the percipient's body, such as his stomach or a tooth or a toe. Thus, although intra-somatic perception is probably not prehensive, there is no reason to doubt that it is generally veridical in outline if not in detail. (iii) One's awareness of one's somatic field as extended, and one's awareness of this or that outstanding bodily feeling as happening in this or that part of it, are, I think, psychologically quite primitive experiences. But the identification of this extended somatic field with the region occupied by one's body as a visible and tangible object, and the correlation of each part of the former a certain part of the latter, are, I am sure, products of early experience and association.

Before I leave the topic of normal perception I want to point out a certain analogy between sight and intra-somatic perception which seems to me interesting and important. So long as it is light and one's eyes are open, one really is directly apprehending something, though it is not what one uncritically takes it to be. This something is an extended, spatially continuous, variously coloured and shaded field, which is presented as a finite but unbounded whole. Outstanding coloured patches are presented as differentiations of this whole, not as independent elements, like bricks, out of which it is built. The mistake, which each of us makes is to identify this directly apprehended field and its differentiations with something public, neutral, and independent of him, viz. the ground, the sky, the surfaces of houses and trees, and so on. There really is a connexion between the two, but it is much more remote than we uncritically take it to be. I am going to sum up these facts about visual perception by calling it synoptic and macrocosmic. Now intra-somatic perception may be described as synoptic and microcosmic. It is synopic because the somatic field is presented as a whole, and the outstanding bodily feelings are presented as localized differentiations of this whole. It is microcosmic because, in apprehending it, one, does not seem to oneself to be apprehending a public neutral world of independent objects. On the contrary, one seems to oneself to be apprehending in a uniquely intimate way a certain particular object which is uniquely associated with oneself.

Touch, in contrast with sight and intra-somatic perception, gives us information piecemeal about foreign bodies and the surfaces of our own bodies. And, as we have seen, it makes us aware of bodies as dynamically interacting substances. Thus sight, touch, and intrasomatic perception severally supply their own characteristic contributions to our knowledge of ourselves and of foreign bodies. And it is only through their coexistence and their intimate co-operation that we acquire the general world-schema which is the common background of daily life and of natural science.

Clairvoyance and Sense-perception

Let us now turn from normal perception and consider an alleged cast of clairvoyance. It is essential to take something quite concrete and not to talk vaguely. I will suppose that a special pack of cards has been made on the following plan. Every card has for its face a white background on which are either squares or circles, but not both. Every card has black pips or red pips, but no card has a mixture of both. There are thus four suits, which we can call Red Squares, Black Squares, Red Circles, and Black Circles. Lastly, in each suit there are ten cards in sequence from ace to ten. The backs of all the cards are uniformly brown. Let us suppose that the percipient correctly guesses that the sixth card from the top of a new and mechanically shuffled pack of this kind is the eight of red squares. And let us suppose that such guesses of his have so often been right that we cannot ascribe his success to chance. Could we suppose that anything analogous to normal sense-perception is taking place?

To assert that a certain card is the eight of red squares is to assert three independent propositions, viz. that there are eight outstanding patches on the surface, that these are square in outline, and that they are red in colour. Now all these propositions could be known by sight to a person who could look directly at the front of the card in white light. This implies that there are eight square patches on the card, which differ physically from the background in such a way that they selectively reflect the red-stimulating lightwaves whilst the background reflects equally light of all wavelengths in the ordinary spectrum. Let us try to suppose that the clairvoyant gets his information by some mode of perception analogous to sight or hearing.

We shall have to suppose that the percipient's body is being stimulated by some kind of emanation from the front of the sixth card in the pack, although the back of the card is towards him. We shall have to suppose that the five cards which are on top of the selected one are transparent to this emanation, though they are not transparent to light. We shall presumably have to suppose that the five cards which are on top of this one and the thirty-four which are beneath it are all equally emitting radiation of this kind. Thus the emanation from the selected card will reach the percipient's body mixed up with the emanations from all the other cards in the pack. Next we shall have to assume that, although the emanation is not light, yet there is a characteristic difference between the emanation from the pips and the emanation from the background, correlated with the difference between red-stimulating and white-stimulating light-waves. Without this there is no hope of explaining how the clairvoyant can tell that there are pips and a background and judge the number of pips. Still less could we explain how he can tell the colour of the pips on the selected card. When we look more carefully into the last mentioned assumption we find that it is equivalent to the following supposition. We are, in effect, supposing that the physical difference between the pips and the background, which makes the former selectively reflect red-stimulating light-waves and the latter indifferently reflect a whole mixture of light-waves, is correlated with another physical difference which is concerned with another and unknown kind of emanation. This is certainly not very plausible.

We have not yet attempted to deal with the clairvoyant's knowledge that the pips on the sixth card from the top are square in outline. No assumption that we have so far made will account for this. If the face of the card were being looked at directly in white light, the light reflected from its surface would travel in straight lines to the percipient's eye. There it would pass through the pupil and be focused by the lens on the retina. There it would excite different parts of a certain area in various ways. The area as a whole, and the distribution of the excitement over it, would be geometrically a projection of the surface from which the light came. From this excited area, through the optic nerve, a corresponding pattern of excitement would be transmitted to the brain. At this stage the percipient would directly apprehend an outstanding oblong patch in his visual field, with a white background and eight red squares scattered about it. This he would automatically and uncritically, but erroneously, take to be the surface of the card. In order to have any analogy with all this we should have to assume that the emanation travels in straight lines through the medium between the card and the percipient's body, and that there is in his body some organ for collecting it and focusing it on a sensitive surface. I need hardly say that we know of no part of our bodies which could plausibly be regarded as such an organ. Moreover, the fact that we have had to assume that ordinary matter is transparent to this emanation makes it difficult to see now a material organ could collect and focus it. It is like being asked to construct a camera, or a telescope, or a microscope when the only material provided is clear transparent glass.

I have now dealt with the physical and physiological assumptions which would be involved in supposing that clairvoyant cognition is analogous to sight or hearing. It remains to consider the psychological aspects of this supposition. In the first place, we should have to assume that the ultimate result of this emanation being received by the appropriate organs, and of the disturbance being transmitted to the appropriate part of the brain, is that the clairvoyant directly apprehends a total sense-field of a characteristic kind. This experience must be analogous to the normal man's apprehension of his visual or his auditory field. So far as I know, there is no introspective evidence for the occurrence of any such experience in persons who claim to be clairvoyant. We should therefore have to assume that this peculiar kind of sensory experience belongs to a part of their mind which they cannot introspect in normal waking life.

Next we must assume that this peculiar sense-field is differentiated, and differentiated in a very special way. There must be in it an outstanding sensum which in fact corresponds to the sixth card from the top of the pack, and there must be in this sensum eight outstanding differentiations which in fact correspond to the eight pips on the face of this card. Moreover, there must be a certain determinate sensible quality in these eight outstanding differentiations which in fact corresponds to the visible squareness of the pips as they would appear to sight. There must also be a certain other determinate sensible quality in these eight outstanding differentiations which in fact corresponds to the visible redness of the pips as they would appear to sight. Although emanations are coming in on top of each other from all the cards in the pack, and presumably from the table and the walls too, we must assume that the sensum specially connected with the emanation from any particular card is distinct enough to be discriminated from the rest of the sense-field by the percipient if he pays enough attention. We must also assume that such a sensum has enough discriminable detail to display those features in the card which would appear to sight as a certain number of pips of a certain shape and a certain colour.

It must be admitted that this involves a very heavy draft on the bank of possibility. I think that the nearest known analogy is provided by hearing. The waves from a number of simultaneously sounding sources, such as the instruments in an orchestra, do come in on top of each other. Yet it is possible with practice and attention to discriminate the noise which in fact comes from one instrument from the noise which in fact comes from another. It is also possible to distinguish overtones, if one has an acute ear, in the noise which comes from a certain instrument. This analogy, though it is not to be despised, does not carry us very far. The noise which in fact comes from a certain instrument has no auditory quality which is invariably correlated with the shape or the colour which that instrument manifests to sight. The analogy would be a little closer if, when we looked at the various instruments, they appeared to be visibly vibrating at various rates and with various amplitudes. Then there really would be a systematic correlation between the auditory qualities of the noise which comes from a certain instrument and certain visible characteristics in the appearance which that instrument would present to sight.

We are not yet at the end of the psychological assumptions which we should have to make. It is not enough that there should be in the clairvoyant's peculiar sense-field a certain discriminable sensum which in fact corresponds to the sixth card from the top of the pack. If he is to answer our question: 'What is the sixth card from the top?' he must know or have reason to believe, with regard to a certain discriminable sensum in his field, that it corresponds to the sixth card from the top. Again, it is not enough that this sensum should have eight differentiations which in fact correspond to the eight differentiated areas on the card which appear to sight as eight red squares. If he is to answer our question, he must know or have reason to believe that the eight differentiations in this sensum correspond to eight differentiated areas on the card which would appear to sight as eight red squares. He must therefore know or have reason to believe, with regard to a certain sensible quality of these differentiations in this sensum, that it corresponds to visible squareness. And he must know or have reason to believe, with regard to a certain other sensible quality of these differentiations that it corresponds to visible redness. Unless the clairvoyant knew these facts he would be in much the same position as a man born blind who had acquired plenty of tactual experience and was then suddenly enabled to see. In the visual field of such a man there would be outstanding coloured patches which are in fact visual appearances of various things from which he has already received tactual sensations. And the visible shape of these visual sensa would in fact correspond to the tangible shape of the corresponding tactual sensa. But the newly cured blind man would not know these facts or have any reason to suspect them. So, if we were to ask him a question about an object which he has touched in the past and is no longer touching but is seeing for the first time, his visual experience would not help him in the least to answer it. It is not until his experiences of sight and touch have become correlated and associated, so that a certain kind of visual appearance has come to represent for him a certain kind of tactual appearance, that his newly acquired power of visual perception will enable him to answer our questions about external objects.

How could the clairvoyant acquire such knowledge or belief as we have had to assign to him? The extremely intimate association between sight and touch, which is established in infancy in all normal people, seems to provide the only helpful analogy. Here we must substitute for it an intimate association between sight and the peculiar kind of sense-experience which we have assumed the clairvoyant to possess. We shall have to suppose that all or most things which are visible also emit the peculiar emanation which gives rise to this peculiar kind of sense-experience when it reaches the clairvoyant's body. Arid we must suppose that every variation in the light reflected from bodies is correlated with a corresponding variation in this emanation. On this assumption, the clairvoyant will from infancy have been apprehending two co-existing and intimately correlated sense-fields, viz. the normal visual field and the peculiar sense-field connected with the emanation. This may be compared with the case of the plain man who apprehends from infancy a visual and a tactual field which are intimately correlated with each other. The difference is that the normal man is constantly aware of apprehending both the visual and the tactual field, whilst the clairvoyant in ordinary waking life is not aware of apprehending the peculiar sense-field connected with the emanation. In consequence of this constant and detailed correlation between the contents of the visual sense-field and those of the peculiar sense-field, in the clairvoyant's case, an intimate association will be established in his mind between the two, just as an intimate association is established in the case of the normal man between his visual and his tactual sense-fields.

When a normal man in the dark has a tactual sensation of a certain familiar kind, which has become associated through frequent past experience in the light with a certain kind of visual appearance, he is able to describe in visual terms the object which he is at present only touching and not seeing. Similarly, when the clairvoyant has a familiar sensation of his own peculiar kind, which has become associated through frequent past experience with a certain kind of visual appearance, he will be able to describe in visual terms the object which is evoking this sensation by its emanation but is at present hidden from his view.

It seems at first sight most unplausible to postulate in the clairvoyant's mind a whole special group of sensations of which he is totally unaware, and then to postulate that they are intimately correlated with his ordinary visual sensations and eventually become associated with the latter. Yet it must, I think, be confessed that a very similar postulate is unblushingly made by the most orthodox psychologists in trying to explain normal visual perception of distance and solidity. We are told a great deal by these scientists in this connexion about sensations of accommodation and sensations of convergence. W are told that these become so intimately associated with purely visual sensations that the minutest variation in the one represents to the percipient a corresponding variation in the other. But the fact remains that most of us at most times are quite unaware of these constantly occurring and continually varying sensations of accommodation and convergence. If we focus our eyes for a long time on a very small and very near object, we may begin to notice sensations of accommodation. If we indulge in elaborate and deliberate squinting, we may notice sensations of convergence. But it is only in these exceptional circumstances that such sensations are noticed or noticeable by the person who, presumably, is in fact never free from them. So orthodox psychologists are not in a position to cast stones at the postulates which would have to be made about the clairvoyant's special sense-field.

I have now enumerated and explained the various assumptions, -- physical, physiological, and psychological, which would have to be made if clairvoyance is to be regarded as a peculiar kind of sense-perception, emissive in its physical aspect, like sight or hearing. It must be confessed that they make a formidable list. But it is better to set them out fully and to face them squarely than to talk vaguely of analogies to wireless and television and 'the marvels of modern science'. Many people will be inclined, when faced with this list of necessary assumptions, to conclude that the attempt to make clairvoyance analogous to sight or hearing must be dropped.

Now, unless clairvoyance be analogous to a physically emissive form of sense-perception, like sight or hearing, it can hardly be analogous to any form of normal sense-perception. If we tried to compare it with touch, we should have to suppose that the clairvoyant's body is provided with invisible and intangible organs, supplied with sensitive spots on their surface and with conducting nerves. We should have to suppose that he can thrust these out and poke them between two cards which are and remain throughout the experiment, visibly in continuous contact with each other. And we should have to suppose that the square areas on the card which differ from the background by selectively reflecting redstimulating light-waves also differ from the background by giving a special kind of stimulus to the sensitive spots on this quasi-tactile organ. It seems hardly worth while to linger over these fantastic suppositions, or to consider what others might be needed in addition to them.

Perhaps some psychical researchers will welcome these conclusions. They will remind us that they have always insisted that clairvoyance cannot be analogous to any form of sense-perception, and they will feel that I have only been underlining the obvious. I cannot share their satisfaction. Have those who believe that clairvoyance occurs, and deny that it is analogous to any form of senseperception, any positive notion of its psychological nature or its modus operandi? If they have, it is most desirable that they should expound it. If they have not, they are just postulating what Locke would have called 'a something, I know not what'. Since their postulate will then have no discernible analogy or connexion with anything that is already known and admitted to be a fact, it will be impossible to assign a degree of antecedent probability or improbability to it. In that case we shall be unable to come to any rationally justified degree of belief or disbelief when they produce their empirical evidence, however impressive it may be.

Clairvoyance as Non-Sensuous Prehension of Physical Objects

The only intelligible positive interpretation which I can put on this view of clairvoyance is the following. Those who deny that clairvoyance is analogous to any form of sense-perception might suppose that the clairvoyant really does directly apprehend remote physical objects, as the ordinary man seems to himself to do in sight and touch. This supposition is, I think, prima facie intelligible. As I have said in discussing normal sense-perception, each of us really does directly apprehend something when he is seeing, hearing, etc. In seeing, e.g. one is directly apprehending an extended continuous variegated coloured field; though one uncritically mistakes it for something else, of a quite different nature, which one does not directly apprehend. So we can understand, in general outline at any rate, what we are being asked to suppose in the case of the clairvoyant.

But, as soon as we begin to consider the suggestion in detail, it becomes less and less intelligible. The card called the 'eight of red squares' is a physical object which, when suitably illuminated, reflects light-waves. If these reach the eye of a normal human observer, they stimulate it in a characteristic way, and at a certain stage in the process a characteristic kind of disturbance is set up in his optic centres. If and only if all this should happen, the card will be represented in the observer's visual field by an outstanding white oblong sensum with eight outstanding square spots on it. There is not the faintest reason to believe that the card itself, which is the locus of a remote causal ancestor in this long and variegated chain of events, has literally and intrinsically any colour whatever. That which corresponds in a physical object to the colour which it is perceived as having is presumably some special configuration or some rhythmic motion of its minute constituents, which causes it to reflect certain kinds of light-waves and to absorb others. If, then, the clairvoyant directly apprehends the card, as it intrinsically and independently is, he will not apprehend it as a thing with a white continuous surface on which there are eight square red spots; for it is almost certainly nothing of the kind. He might, perhaps, apprehend it as a swarm of very small colourless electric charges in very rapid rhythmic motion; for, according to the best information available at present to those of us who are not clairvoyants, this or something like this is what the card most probably is.

Now, if clairvoyants do directly apprehend physical objects as having those characteristics which scientists laboriously infer that they must have, they show no sign of being aware of their own knowledge. If they were, they could presumably put it, at least roughly and in outline, into words. They would then be invaluable helpers in all physical laboratories; for their information, artlessly expressed but 'straight from the horse's mouth', would suffice to head scientists off from plausible but false theories and to suggest fruitful lines of experiment and speculation. We shall have to assume, then, that the clairvoyant's direct apprehension of physical objects, as they intrinsically are, occurs in a part of his mind which is cut off from his ordinary waking experience.

The clairvoyant describes the unseen card in terms of colours, visible shapes, etc., and not in terms of electric charges, waves, and rhythmic motions. We shall therefore have to explain how he translates his direct apprehension of the unseen card, as it intrinsically is, into the colours, visible shapes, etc., which it would appear to have if it were being seen by a normal human being in daylight. It will be remembered that there is a rather similar problem for those who regard clairvoyance as a peculiar form of sense-perception. The suggestion which I made in that connexion might, perhaps, be modified to deal with the present problem. We shall have to suppose that the clairvoyant has, from infancy, been continuously though unconsciously apprehending directly all those objects which he has also been cognizing indirectly through sight and touch. Then we can suppose that an association would be set up between, e.g. the conscious experience of seeing an object as red and the unconscious experience of directly apprehending it as having that intrinsic characteristic which makes it selectively reflect red-stimulating light-waves. Suppose that, on some future occasion, such an object, though no longer visible, is still being directly but unconsciously apprehended by the clairvoyant. He will still apprehend it as having that intrinsic characteristic, whatever it may be, which has now become associated in his mind with the visual appearance of redness. Consequently the idea of it as a red-looking object will arise automatically in his mind, and he will announce that the unseen object is red.

I have now stated and tried to work out in some detail two alternative views of what clairvoyance would be if it took place. Neither of them is in the least attractive or plausible, but I know of no other alternative that is even intelligible. I hope that some of those who think that there is adequate evidence of clairvoyance will be inspired to suggest some other view of it which will be equally intelligible and much more plausible. Though I can offer no hint of a solution, I may possibly have given them some help by setting out elements of the problem in a clear and orderly way.


Telepathic Interaction

It is commonly assumed that one embodied mind can affect another only in an extremely roundabout way. It must first affect its own body; then this change in its own body must set up a series of physical changes which eventually affect another ensouled body; and, finally, this change in the other ensouled body must produce a change in the mind which animates it. Thus the process involves a psycho-physiological transaction at one end, a physiologico-psychical transaction at the other end, and a purely physical causal series between the two. A further restriction is commonly imposed on this general scheme. It is usually assumed that the process set up within the one ensouled body must issue in some overt macroscopic change of it, such as emitting a sound, making a gesture, or assuming a new facial expression; and it is assumed that this must affect the other ensouled body by sight, hearing, touch, or some such form of normal sensory stimulus. The wider assumption may be summed up in the following general principle: 'The only thing, other than itself, with which an embodied mind can directly interact is the brain and nervous system of the body which it animates.' If this be granted, the rest follows.

We can now imagine various stages in which the common-sense assumption might be given up.

(i) We might keep the general principle, but drop the further restriction which is commonly put on it. We might suppose that, in certain cases, the disturbance set up in A's brain by an event in his mind initiates a physical process of an emanative kind which travels out in all directions; that this may set up a disturbance in B's brain, if it reaches the latter; and that this disturbance in B's brain may affect his mind. On this view there need be no overt macroscopic change in A's body, such as emitting a noise, making a gesture, etc. And B's brain need not be stimulated through any of the ordinary sense-organs by what is happening in A's body. Yet the general principle about interaction will remain intact.

(ii) The next stage would be to drop one half of the general principle and to keep the other half. This would give two possible alternatives. (a) We might continue to assume that A's mind can directly affect only A's brain, and that B's mind can directly affect only B's brain. But we might now suppose that A's mind can, in some cases, be directly affected by disturbances in B's brain; and that B's mind can, in some cases, be directly affected by disturbances in A's brain. (b) We might continue to assume that A's mind can be directly affected only by A's brain, and that B's mind can be directly affected only by B's brain. But we might now Suppose that A's mind can, in some cases, directly produce disturbances in B's brain; and that B's mind can, in some cases, directly produce disturbances in A's brain.

(iii) Lastly, we might drop the general principle altogether. We might suppose that, in certain cases, one embodied mind can affect or be affected by another embodied mind directly, without any physiological or physical mediation. I propose to call the first alternative the 'Brain-wave Theory', and the third alternative the 'Theory of Direct Intermental Transaction'. Theories of the second kind might be called 'Theories of Extended Psychophysiological Interaction'. I cannot pretend that this is a 'snappy' title, but I think it is accurately descriptive.

If either of these three suppositions were ever realized in practice we should say that there had been a case of 'Telepathic Interaction'. If it were an instance of the Brain-wave Theory it would involve no supernormal interaction between mind and matter or between mind and mind. It would involve nothing but an unusual transaction between two brains and an intervening physical medium. If it were an instance of either form of the Theory of Extended Psycho-physiological Interaction it would involve supernormal interaction between mind and matter, but no direct interaction between mind and mind. The supernormality of the transaction would consist in the fact that an event in one man's mind directly affects or is directly affected by an event in another man's brain. If it were an instance of the Theory of Direct Intermental Transaction it would involve supernormal interaction between two embodied minds, but it would not necessarily involve any supernormal interaction between mind and matter.

If the Brain-wave Theory would fit the empirical facts, it would be preferable to the other two in respect of antecedent probability. But the general opinion of those who have studied the facts seems to be definitely adverse to this theory.

In favour of the Theory of Extended Psycho-physiological Interaction it may be said that we do know that each embodied mind directly affects and is directly affected by at least one brain and nervous system, though this kind of transaction has to be accepted as a completely mysterious brute fact. This one brain and nervous system is, of course, that of the one material system to which this mind stands in the peculiar relation of 'animating'. Now the theory under discussion is that this direct interaction between minds and brains, which is admitted to occur, is not necessarily or invariably restricted within these limits. Either the range within which direct interaction between a mind and a body is possible extends beyond the limits marked out by the relation of animation, or the relation of animation extends more widely than commonsense recognizes. The latter suggestion amounts to supposing that an embodied human mind may animate a material system which includes, in addition to one human body, parts of another human body which is animated by another human mind. This relation might be mutual as between two human individuals A and B. A's mind might animate a material system which includes, beside what we call 'A's body', a part of what we call 'B's body'; and B's mind might animate a material system which includes, beside what we call 'B's body', a part of what we call 'A's body'. In some cases of multiple personality it looks as if there were two minds simultaneously animating either the whole of a common brain and nervous system, or, at any rate, animating two parts of it which overlap each other. This at least supplies empirical support for the general conclusion that the relation of animation between minds and bodies is not always one-to-one. If two minds can animate one body, it may not be unreasonable to contemplate the possibility that one mind may animate one body and a bit of another body.

These speculations are, I know, very wild; but I make no apology for them on that account. The admitted relation of animation between the mind and the body of a normal human individual, and the admitted interactions between the two, are so mysterious that we are left with a wide field for legitimate conjectures. The situation is very different from that which faced us when we were considering normal sense-perception and alleged clairvoyance. We have a great deal of positive knowledge about normal sense-perception, in its physical, its physiological, and its epistemological aspects; so the field for legitimate conjecture is there much narrower.

Passing finally to the Theory of Direct Intermental Transaction, we must, I think, assign to it the lowest antecedent probability of the three typical theories. So far as I am aware, it is supported by no known analogy with admitted facts. We should, therefore, hesitate to resort to it unless the evidence rules out all theories of the other two types.

Telepathic Cognition

We have so far considered the possible causal relations between two embodied minds; we must now turn our attention to what primarily concerns us in this paper, viz. the possible cognitive relations between them. It is important to be quite clear that these are different problems, for the word 'Telepathy' seems often to be carelessly used to cover both supernormal causal influence of one embodied mind on another and supernormal cognition of one embodied mind by another. We have given the name 'Telepathic Interaction' to the former, and we will call the latter 'Telepathic Cognition'. Probably telepathic cognition would be impossible without telepathic interaction, but there is not the least reason why there should not be telepathic interaction without telepathic cognition. Cognizing or being cognized, on the one hand, and affecting causally or being affected causally, on the other, are utterly different relations. If either of them can be analysed, which is doubtful, it is certain that neither of them forms any part of the analysis of the other. So there can be no logical impossibility in two terms being related by one of them and not by the other. And, if it be granted that two minds could influence each other telepathically at all, it is quite easy to imagine that two minds which remained completely ignorant of each other might yet be in fact influencing each other frequently and profoundly by telepathic interaction.

Having made this distinction clear, we can now turn our attention to the cognition by one mind of another mind and its experiences. I shall begin by stating and explaining two principles which are commonly, if tacitly, assumed to apply to embodied human minds and their normal cognitions. The first is that one and the same experience cannot be owned by more than one mind. I do not think that anyone would question this. It is true that we sometimes use expressions which, if literally interpreted, would imply that one and the same experience is owned by several minds. We might, e.g. say of two people who both believe that Francis wrote the Letters of Junius that they both have the same belief about the authorship of the Junius letters. But we all recognize at once that such statements are not to be taken literally. One belief that Francis wrote these letters occurs in A's mind and not in Vs; another belief that Francis wrote these letters occurs in B's mind and not in A's. When we talk of the same belief occurring in two minds we mean that two beliefs, which stand in a common relation to one and the same fact, viz. the actual but unknown authorship of the Junius letters, are occurring and that one belongs to one mind and the other belongs to the other mind. A similar interpretation would have to be put on any statement that seemed to conflict with our principle. We will call this the 'Principle of Unique Ownership of Experiences'.

We come now to the second principle. It may be stated as follows. Any particular existent which can be directly apprehended by an embodied mind can be directly apprehended only by one such mind. Let us consider what kinds of particular existents a given embodied mind M can directly apprehend. They are (i) M itself, perhaps; (ii) some, if not all, of M's experiences; (iii) certain mental images; (iv) somatic sensa connected with the processes in M's body; and (v) certain visual, tactual, auditory, and other kinds of extra-somatic sensa. Of course the plain man would have included in this list something which we have not included, viz. the surfaces of certain foreign bodies and of his own body, and certain kinds of events happening from time to time in such bodies. And he would not have mentioned certain items which we have included, viz. various kinds of sensa. The cause of both these differences is the same, viz. the fact that the plain man mistakes what he directly apprehends in sense-perception for parts of physical objects and events in such objects. We have seen that he does not directly apprehend such particular existents, and so we have had to exclude them from our list. But we have also seen that he really is apprehending particular existents of some kind in sense-perception, and so we have had to introduce them into our list under the technical name of 'sensa'.

Now let us go through the list, and we shall see that, if it is exhaustive, it proves our principle. (i) Everyone would agree that normally no embodied mind but M could directly apprehend M. (ii) Everyone would agree that normally no embodied mind but M could directly apprehend any of M's experiences. (iii) Everyone would agree that normally no embodied mind but M could directly apprehend any mental image that M can directly apprehend. (iv) Everyone would agree that normally no embodied mind but M could directly apprehend the aches and pains and pressure-data and so on which arise in connexion with processes in M's body. (v) As regards extra-somatic sensa a difference of opinion might arise, but it would certainly be due to verbal confusion. A person might say: 'A noise is an extra-somatic sensum. Now we all know that M and N may both hear the same noise. So N can directly apprehend an extra-somatic sensum which is also being directly apprehended by M.' There is nothing in this argument. When M and N are correctly said to be 'hearing the same noise' each is directly apprehending a different auditory sensum. But these two auditory sensa are related in a certain characteristic way to each other, and they are manifestations of a common physical event at a remote common source. When the fact that normal sense-perception is not really prehensive of external objects is clearly understood and firmly grasped, and when the various verbal confusions which have arisen from its being ostensibly prehensive have been removed, we see that there is not the least reason to believe that, in normal life, N can ever directly apprehend any sensum which M can directly apprehend, or vice versa.

Now I think that, with the explanations which I have just given, it will be admitted that the above list includes all the various kinds of particular existents which any embodied mind, under normal conditions, could directly apprehend. And we have now seen, with regard to each of these classes of particulars, that any member of it which can be directly apprehended by any one embodied mind M cannot, under normal conditions, be directly apprehended by any other embodied mind. And so we reach our second general principle: 'Any particular existent which can be directly apprehended by an embodied mind can be directly apprehended only by one such mind.' I will call this the 'Principle of the Privacy of Prehensible Particulars'.

Before going further I will make some remarks on these two principles. (i) The Unique Ownership of Experiences is in a much stronger position than the Privacy of Prehensible Particulars. Many people would say that it is self-evidently impossible that one and the same experience should literally be an experience of two minds, no matter whether the minds were embodied or disembodied, in a normal or an abnormal condition, or what not. Without committing myself to this view, I must admit that it is highly plausible. Now the Privacy of Prehensible Particulars, as a general principle, is not in the least self-evident. We reached it simply by a process of enumeration and inspection, and there is no apparent absurdity in supposing that there might be exceptions to it. As we have seen, common sense does unhesitatingly take for granted that, in normal visual perception, one and the same particular can be, and often is, directly apprehended by several embodied minds. We rejected this, not in the least because it seemed intrinsically absurd or impossible, but because it was impossible to reconcile it with the relevant empirical facts taken as a whole. The outcome of this comparison between the two principles is that an alleged exception to the Privacy of Prehensible Particulars has an appreciable antecedent probability, whilst an alleged exception to the Unique Ownership of Experiences has far less, if any at all.

(ii) Some people have held that images and sensa are themselves experiences. Many others, who have not gone so far as this, have taken a view which may be roughly expressed as follows. They have held that a mental image can exist only as a logically inseparable factor in someone's experience of imagining it, and that a sensum can exist only as a logically inseparable factor in someone's experience of sensing it. If either of these views were accepted, we could replace the Privacy of Prehensible Particulars by the following principle: 'No embodied mind can directly apprehend anything but itself, its own experiences, and objects which are logically inseparable factors in its own experiences.' This principle does not seem to me to have any better claim to be self-evident than the Privacy of Prehensible Particulars. And I am not convinced that either of these two views about sensa and images is true. So I prefer to keep the second principle in the form in which I originally stated it.

(iii) Some people have held that, whenever a mind has an experience, it directly apprehends that experience. Others have held that, whenever a mind has had an experience, it could have directly apprehended that experience if it had attended, though it may not in fact have done so. If we accept either of these views, and combine it with the Privacy of Prehensible Particulars, the Unique Ownership of Experiences follows as a logical consequence. For suppose, if possible, that two minds, M and N, both owned a certain experience E. According to the view under discussion M could or would directly apprehend E, since E is an experience of M's. Similarly, on the view under discussion, N could or would directly apprehend E, since E is also an experience of N's. Therefore E could be directly apprehended by two different minds, which is contrary to the Privacy of Prehensible Particulars. So the supposition that E could be owned by two minds must be rejected, if the Privacy of Prehensible Particulars is to be retained and the view under discussion is to be accepted.

This result seems to me to be of logical interest rather than of practical importance. In the first place, the view that, whenever a mind has an experience, it directly apprehends that experience, seems to me obviously false. And the view that, whenever a mind has had an experience, it could have directly apprehended that experience if it had attended, seems to me quite uncertain. But, even if one or other of these doctrines were indubitable, it would still be a logical perversion to base the Unique Ownership of Experiences on it and the Privacy of Prehensible Particulars. For, as we have seen, the Unique Ownership of Experiences has some claim to be self-evident, whilst the Privacy of Prehensible Particulars has no such claim. We should therefore be basing the stronger of two propositions on the weaker. I conclude then that the two principles are best regarded as independent propositions.

(a) Telepathic Prehension

We have now stated, explained, and commented on the two principles which are assumed by common sense to govern the region with which we are at present concerned. We can look upon telepathic cognition as involving a real or apparent breach of one or other of these principles. Any breach of the Privacy of Prehensible Particulars would, ipso facto, be an instance of telepathic cognition. To be more precise, it would be an instance of what I will call 'Telepathic Prehension'. Under this heading would come the following five possible cases. (i) One mind directly apprehending another mind as a unit. (ii) One mind directly apprehending an experience which is occurring in another mind. (iii) One mind directly apprehending a mental image which is being imaged by another mind. (iv) One mind directly apprehending a somatic sensum which is being sensed by another mind and is the manifestation of a process going on in the body which that other mind animates. (v) One mind directly apprehending a visual, tactual, or auditory sensum which is being sensed by another mind in seeing, touching, or hearing an external object. Telepathic prehension of the first kind seems to be claimed for Mrs. Willett (see Lord Balfour's paper, Proc. S.P.R., Part 140, pp. 90-4). There are plenty of cases which look, prima facie, as if they were instances of the four remaining kinds. Are they really so?

In considering this question the first point to notice is the following. A breach of the Unique Ownership of Experiences would not be ipso facto an instance of telepathic prehension, for in itself it would not be an instance of cognition at all. It would best be described as an instance of 'Intermental Confluence'. But, if intermental confluence were to take place, telepathic prehension would almost certainly follow as an immediate consequence of it. Suppose, e.g. that, through mental confluence, N's experience of sensing a certain sensum or of imaging a certain mental image were also an experience of M's. Then M would be sensing or imaging the very same sensum or image which N is sensing or imaging. Now sensing and imaging are instances of directly apprehending. So M would be directly apprehending a sensum which N is sensing or an image which N is imaging. And, of course, the converse would also be true. So, if there were intermental confluence of this kind between M and N, there would necessarily be telepathic prehension of sensa or images by both M and N. This particular example can at once be generalized. If any experience which is a direct apprehension of a particular were, through mental confluence, owned by both M and N, M would be directly apprehending something which N is directly apprehending, and conversely.

We have seen, however, that intermental confluence would be ruled out by many people as self-evidently impossible. So we may now put the following question. Supposing that we rule out intermental confluence, is there any need to assume that telepathic prehension occurs? It seems to me quite unnecessary to assume this in order to account for successful experimental results in which one person conveys supernormally to another figures which he sees or draws, images which he calls up and fixes, or bodily feelings which he is experiencing. All that we need to suppose here is a particular form of telepathic interaction. It is enough to suppose that the occurrence of a certain sensation or imagination or bodily feeling in M's mind causally determines in N's mind the occurrence of a sensation with a similar sensum, or of an imagination with a similar image, or of a bodily feeling with a similar quality and feelingtone. In experiments it may generally be assumed that N knows that it is M, and no one else, who is trying to convey an impression to him. And it may generally be assumed that he knows roughly at what time M is going to try the experiment. Suppose that, at about the agreed time, N suddenly has a sensation or bodily feeling or becomes aware of an image. Suppose that there is no noticeable feature in N's surroundings at the time, or in his immediately previous train of thought, which would supply an obvious normal explanation for the occurrence of just that experience at just that moment. Then he will naturally suspect that the experience is caused by M, whom he knows to be experimenting at the time. So there is no need whatever to assume that N has any telepathic prehension of M or of M's experiences, however successful such experiments may be.

So far as I can see, it is quite possible that each of us may be often, or even continuously, influenced telepathically by other minds, and yet this fact might always have escaped notice. Suppose that an event in M's mind does in fact determine telepathically an event in N's mind. N will have no reason to regard this as an instance of telepathic interaction unless all the following conditions are fulfilled. (i) The effect on N must take the form of an experience which he can and does notice. Now the effect might equally well be a change in his mental dispositions, or be an experience which he does not or cannot notice.

(ii) This experience must be so discontinuous with his other contemporary and immediately past experiences and with his usual trains of association that he is surprised by it and is led to suspect that it is not caused normally. Now this condition would seldom be fulfilled. Very often I suddenly image an image, visual or auditory, which seems quite disconnected with my other contemporary and immediately past experiences and with my usual trains of association. But even I, who am professionally interested in such things, tend to dismiss it as just one more unexplained oddity in the workings of my mind. Most people are occupied for most of their lives in practical dealings with other people and things; so an experience of theirs would have to be very odd indeed before they would seriously raise the question whether it was or was not caused normally. Moreover, if an experience in N's mind be telepathically caused by an event in M's mind, the event in M's mind would never be the complete immediate cause of it. It would at most be one of the immediate necessary conditions. Another, and equally necessary, factor in the total immediate cause of this experience of N's would be the permanent dispositions, the acquired associations, and the contemporary or immediately past experiences of N himself. There is therefore no reason to believe that most telepathically caused experiences would be so outstanding and discontinuous as to attract the special attention of the experient.

(iii) Even if N notices this experience with surprise, and is led to wonder whether it may not be telepathically caused, he can get no further unless he can discover that, at about the same time, a certain other person was having an experience which was specially closely related to his own. Now this condition could not be fulfilled unless all the following conditions were also fulfilled. (a) M, the person who is in fact the telepathic agent in this transaction, would need to be known to N, the telepathic patient, or they would need to have common friends. Now it is obvious that M and N might be complete strangers. (b) The event in M's mind which telepathically determined this experience in N's mind would have to be an experience which M noticed and could describe to N or to their mutual friends. Now the event might not have been an experience at all; it might have been a change in the dispositional structure of M's mind. Or the event might have been an experience which M did not or could not notice. (c) There would have to be some specially intimate observable relation between M's experience and N's experience, which would make it reasonable to single out the former as a factor in the total cause of the latter. The only two relations that I can think of in this connexion are likeness and the relation of fulfilment to intention. The first would hold if the two experiences were alike in quality or if they were prehensions of similar objects. The second would hold if M's experience were that of intending to produce in N an experience of a certain kind, and if N's contemporary experience were in fact of the kind intended. Plainly there is not the least reason to suppose that either of these very special relations would hold as a rule between the telepathic cause-factor and the experience which it co-operates in producing. An effect may be extremely unlike every one of the factors in its immediate total cause. And most telepathic interaction may be entirely unintentional.

The upshot of the above discussion is this. If telepathic interaction takes place at all, it may well be a very common occurrence. But it will be noticeable only when a large number of independent and rather special conditions are simultaneously fulfilled. And, when these conditions are fulfilled, so that it does become noticeable, the experience which is telepathically produced in N will be very liable to be mistaken for a telepathic prehension by N of that experience of M's which is its telepathic causal determinant. It is easy to find analogies in the physical sciences to the situation which I have just shown to be possible about telepathy. Consider, e.g. ordinary magnetic forces, and the history of our knowledge of them. Such forces occur whenever electric charges are moving or electric forces are varying, and they pervade all space at all times and are profoundly important factors in the physical world. Yet they would hardly have been discovered had it not been for the happy accidents that the earth contains a good deal of the one element, viz. iron, which is very strongly susceptible to magnetic forces; that it contains natural magnets, viz. lodestones; and that it is itself a natural magnet. For centuries magnetism seemed to be a freak of nature which occurred exclusively in connexion with certain very special kinds of matter. Yet in fact it was all the time operating everywhere. And the very special characteristics which it displays in connexion with iron and with permanent magnets, masked its real nature almost as much as they revealed it.

I have now said all that seems necessary in support of my contention that experiments in telepathy, however successful they may be, would prove only telepathic interaction, of one or other of the three kinds which we distinguish as theoretically possible. They would not force us to abandon the Privacy of Prehensible Particulars and to postulate telepathic prehension. It remains to consider two other kinds of ostensibly telepathic phenomena, for which there is ample evidence, some of which is of excellent quality. The first is spontaneous telepathy, such as is reported in Phantasms of the Living. The second is the supernormal knowledge which mediums often display with regard to facts known to the sitter or to some other living person.

A good many cases of spontaneous telepathy can be regarded as similar in principle to the cases of experimental telepathy which we have already considered. Suppose that M, sitting in his dining room in a mood of intense depression, eventually takes poison, suffers great bodily pain, and dies. Suppose that there arise in N's mind, through telepathic interaction, visual sensations or visual imaginations very much like those which M is experiencing through normal visual perception of his surroundings. If N is familiar with M's dining-room, his telepathically induced visual experiences will naturally make him think of that room and of M. Suppose next that there arises in N's mind, through telepathic interaction, a feeling of intense depression very much like that which M is experiencing because of illness, financial trouble, or some other normal cause. It will be natural for N to connect together these two simultaneous abnormal experiences, and to suspect that there is something seriously wrong with M. Suppose finally that there arises in N's mind, through telepathic interaction, a sensation of intense bodily pain very much like that which M is experiencing in consequence of the action of the poison on his body. It will be natural for N to assume that M must be very ill and perhaps dying. If N should be asleep or in a dreamy state when the telepathic interaction takes place, it is extremely likely that the data supplied, and the normal associations which they excite, will be supplemented by a great deal of imagery. The whole thing may then be worked up into a vivid dream or waking hallucination, with the gaps filled in and the inconsistencies smoothed out correctly or incorrectly. No kind of telepathic prehension needs to be postulated here. Nothing need be assumed except the special kind of telepathic interaction, which we postulated to explain the experimental results, together with the normal workings of preformed associations in N's mind.

(b) Telepathic Discursive Cognition

It is doubtful whether all well-attested cases of spontaneous telepathy can be dealt with on these lines. And it is fairly certain that this cannot be the right explanation of the supernormal knowledge which mediums often display with regard to facts known only to the sitter or to some other living person. We may best approach the subject in the following way. There are at least two fundamentally different, though intimately connected, kinds of normal cognition, viz. prehensive and discursive. So far we have considered only the possibility of telepathic prehension, and we have found no direct evidence for it. Now it looks as if the mediumistic cases, and some of the spontaneous telepathy cases, might involve telepathic discursive cognition. I will now explain these statements and consider whether there is any reason to postulate such cognition.

The distinction between prehensive and discursive cognition is roughly identical with the familiar distinction between 'directly apprehending' and 'thinking about'. It is illustrated, e.g. by the difference between actually hearing a set of noises which form a tune and knowing or believing that this tune consists of a series of noises of certain pitches and durations following each other in a certain order. We may, of course, have discursive cognition about a particular which we are also directly apprehending; and the ground of our discursive cognition about it may be what is manifested to us in our prehension of it. But we can have discursive cognition about objects which we are not at the time prehending, about objects which we never have prehended, and about objects which we never could prehend. We can also have an experience which would properly be described as 'thinking of an x', e.g. a dragon, or 'thinking of the y', e.g. the King of the Fairies, although there may be nothing answering to the description 'an x' or the description 'the y'. But it would be impossible to have an experience which would properly be described as 'directly apprehending an x' or 'directly apprehending the y' unless there were something answering to the description 'an x' or to the description 'the y', respectively.

Discursive cognition consists in either knowing a fact or taking up one of a number of alternative cognitive attitudes towards a proposition which may be either true or false. Among these cognitive attitudes are included believing, disbelieving, opining, uncritically accepting, supposing, and probably many others. All such cognitive attitudes towards a proposition equally presuppose a more fundamental cognitive experience which may be called 'entertaining' the proposition. One and the same person may entertain the same proposition on many different occasions, and he may take towards it the same or different cognitive attitudes on different occasions. At one time he may doubt it, at another he may believe it, and so on. Again, several people may entertain one and the same proposition on the same occasion, and they may take various cognitive attitudes towards it. Smith and Jones may both believe it, whilst Brown doubts it and Robinson disbelieves it. (In saying these things I do not mean to imply that there is a peculiar class of entities called 'propositions'. I think it most likely that all the statements which I have just been making could be restated without introducing the word 'proposition' or any synonym for it. But the translations would be extremely complicated and verbose. The use of the word 'proposition' enables me to express in a reasonably simple verbal form what everyone admits to be facts about discursive cognition. No further excuse is needed for continuing to use it.)

There is one other general fact of very great importance which we must mention before we can profitably consider telepathic discursive cognition. At any moment far the greater part of any man's 'knowledge' or 'beliefs' or 'opinions' certainly does not take the form of experiences of knowing such and such facts or believing or opining such and such propositions. The truth about him is that he would have these experiences if he chose to direct his attention in a certain way, or if he were to be suitably stimulated. We may express this by saying that, at every moment of our lives, much the greater part of our knowledge, beliefs, and opinions consist of relatively permanent dispositions to know certain facts or to believe or opine certain propositions. It is always assumed that, to every such relatively permanent cognitive disposition, there must correspond some relatively permanent actual existent. This is generally supposed to be some actual modification of the structure of our minds or our brains, or to be some actual persistent unobservable process in our minds or our brains.

It is well to recognize that we know nothing at all about the intrinsic nature of the actual existents which are supposed to correspond to our cognitive dispositions. We do not know whether they are persistent structural features or persistent unobservable processes. And we do not know whether they are modifications of our minds or our brains or of both or of neither. All that we know of them is that they are produced and modified by our actual experiences, and that they are important factors in producing and modifying our experiences. There is very good reason to believe that the actual existents which correspond to the various dispositions of various kinds of matter are special peculiarities in the spatial arrangement and the motions of the ultra-microscopic particles of which bodies are composed. But, unless we assume that the actual existents which correspond to mental dispositions are themselves purely material, we cannot suppose that they are spatial arrangements or modes of motion of ultra-microscopic particles. Now it is extremely difficult to form any positive conception of purely mental structures or of non-introspectible mental processes which could plausibly be supposed to correspond to our mental dispositions. So we are between the horns of the following dilemma. If we put the correlates of all mental dispositions into the brain, we get a theory which is familiar and intelligible in outline but incredible when we come to consider it in detail. If, on the other hand, we postulate mental structures and non-introspectible mental processes as the actual correlates of our mental dispositions, we have no clear idea of what we are postulating and we run the risk of paying ourselves with words.

We are now in readiness to consider telepathic discursive cognition. Suppose that M knows the fact F or entertains the proposition P. The only normal way in which M's knowledge of F or his entertaining of P can cause another mind N to think of this fact or to entertain this proposition is the following. M must express the fact or the proposition by uttering or writing a sentence which expresses it in accordance with some conventional system of symbolization. N must hear or see or in some other way perceive with his senses either this spoken or written sentence itself or some reproduction of it, e.g. on a gramophone-record or in a book. Of course profound physical transformations may take place during the process which intervenes between M's utterance and the occurrence of the reproduction of it which N perceives; but a fundamental identity of structure must be preserved throughout, though it may be realized in very different media at different stages. This is well illustrated by telephonic or wireless transmission of speech. Next, the sentence which N eventually perceives must mean for him, in accordance with some system of conventional symbolization with which he is familiar, the same fact or proposition which M expressed by his original sentence. If N perceives M's sentence itself, it is essential that he should be familiar with the system of symbolic conventions which M uses. If N perceives only a reproduction of M's original sentence, this condition need not be fulfilled, but another will have to be substituted for it. M might express himself in French; and N, who knows no French, might still be caused to entertain the proposition which M was entertaining provided that N perceives a sentence which is an English translation of M's sentence. But, in that case, it is essential that there should have been a third person T, familiar with both M's and N's systems of conventional symbolization, who made a translation from one set of symbols to the other.

The following remarks are worth making at this stage. (i) M's knowledge of F or his entertainment of P may be an essential factor in causing N to think of F or to entertain P; and yet N may have no knowledge or thought of M or of M's cognitions. If N perceives and understands a sentence, and if he cares to reflect on the matter, he will indeed recognize that some mind or other must have entertained the proposition which this sentence means and must have expressed it in a sentence. And he will recognize that this event in another mind must be a causal ancestor of his own entertainment of this proposition. But N need not know or believe anything more definite about this other mind. (ii) Suppose that N perceives and understands a certain sentence, and also knows that it was uttered by M or is a reproduction of one of M's utterances. N will then know, or have very strong reason to believe, that the proposition which he has been led to entertain has also been entertained by M. But he may know nothing about M's cognitive attitude towards this proposition. If N has any beliefs on this subject, they may well be mistaken; as is abundantly proved by the occurrence of successful lies and political propaganda, which are taken by the duped hearer to express the knowledge or the beliefs of the lying speaker.

It is now easy to define the phrase 'Telepathically Induced Discursive Cognition'. Suppose that a certain mind N thinks of a fact F or entertains a proposition P at a certain moment. Suppose that N would not have done this unless a certain other contemporary mind M were knowing this fact or entertaining this proposition. Lastly, suppose that M's knowledge of F or his entertaining of P does not bring about N's thought of F or his entertainment of P by the normal process which we have just described. Either M never expresses the fact or the proposition in a sentence, or N never perceives the sentence or any reproduction of it, or N cannot understand the sentence or the reproduction of it which he perceives. If these conditions, positive and negative, were fulfilled, we should say that N was having telepathically induced discursive cognition of this fact or this proposition. And we should say that he was deriving this cognition telepathically from M's mind. Now it looks as if telepathically induced discursive cognition, in the sense just defined, were involved in some cases of spontaneous telepathy between normal people and in many cases of trance-mediumship. Can we say anything further about it?

(i) I suspect that some people have at the back of their minds a certain tacit assumption about the modus operandi of telepathically induced discursive cognition. It may be stated as follows. Suppose that N is cognizing a fact or a proposition, and that this cognition of N's is derived telepathically from M's mind. Then, it is assumed, N must be telepathically prehending M's cognition of this fact or proposition. And in so doing, it is further assumed, N will ipso facto be himself cognizing the fact or proposition which M is cognizing. To sum up the theory in a sentence: 'N's telepathically induced cognition of what M discursively cognizes depends upon N's telepathic prehension of M's experience of cognizing.'

I should very much hesitate to accept this theory. In the first place, we have so far found no reason to admit the occurrence of prehensive cognition by one mind of experiences belonging to another mind. Secondly, I would question the assumption that, if N directly apprehended M's experience of knowing the fact F or cognizing the proposition P, he would ipso facto be himself cognizing F. or P. It is, no doubt, true that a person could not directly apprehend his own experience of knowing a fact F or cognizing a proposition P unless he were knowing F or cognizing P. For, unless he were knowing F or cognizing P, there would be nothing answering to the description 'his experience of knowing F or cognizing P'. And, unless there were an experience answering to this description, he could not directly apprehend such an experience. But this argument will not lead to the desired conclusion if we apply it to N's prehension of M's cognitive experiences. The only conclusion to which it leads in this case is quite trivial. The conclusion is merely that, if N directly apprehends M's experience of knowing F or cognizing P, then M must be knowing F or cognizing P. The desired conclusion is that N must be thinking of F or entertaining P. And this certainly does not follow.

Now, if the fallacy which I have just indicated is avoided, there seems to be no reason to accept the assumption under discussion. Why should not N directly apprehend an event, which is in fact M's experience of knowing F or cognizing P, without realizing that the event which he is apprehending answers to this description? And, if this is possible, why should N ipso facto think of F or entertain P?

It might be plausible to maintain that N could not directly apprehend an experience of M's without ipso facto being aware of its psychological quality, e.g. without apprehending it as an experience of knowing or as one of believing or as one of doubting, as the case might be. But it is not plausible to maintain that N could not directly apprehend an experience of M's without ipso facto being aware of its epistemological object, i.e. of the fact of which it is a knowing or of the proposition of which it is a believing or a doubting. Yet, when telepathy takes place from M to N, the result is usually that N cognizes a fact or proposition which M is cognizing, but remains unaware of the psychological quality of M's cognitive experience. So there seems to be very little to be said in favour of the theory which we have been discussing.

Before we leave this theory there is one more remark to be made about it. If it were acceptable on other grounds, it could be applied to explain the apparently telepathic prehension by N of images which M is imaging or of sensa which M is sensing. The explanation would, of course, take the following form. N, it would be said, telepathically prehends M's experience of imaging the image I or sensing the sensum S. In doing this, it would be assumed, N ipso facto prehends the image I or the sensum S which is the object of M's experience. The general principle assumed is that, in prehending any experience which is itself a prehension of an object, one would be ipso facto prehending its object. I see no reason to accept this principle; and I have already tried to show that the results of experimental telepathy can be interpreted in quite a different way, which involves telepathic interaction but does not involve telepathic cognition.

(ii) I think that certain cases of telepathically induced discursive cognition could be explained on the same lines as the simple cases of experimental telepathy. Suppose that M knows the fact F or cognizes the proposition P. Although he does not utter or write a sentence which would express F or P in his own language, he may image a series of auditory or visual images corresponding to such a sentence. Certainly when I am thinking I often find myself doing this. Suppose now that a series of visual or auditory images, similar to these, were produced by telepathic interaction and imaged by another mind N. If N knew the language in which these image-sentences are composed, he would automatically entertain the proposition or think of the fact which they express in that language He would thus have been telepathically induced to entertain the proposition which M is cognizing or to think of the fact which M is knowing.

It must be noticed that this theory presupposes that N knows the language in which M would express himself if he were to speak or to write. It therefore could not explain how an Englishman could telepathically induce in a Frenchman, who knew no English, a cognition of a fact which the Englishman knows or a proposition which he cognizes. I do not know whether there is good evidence of telepathically induced discursive cognition in such cases. It would be a very important subject for experimental investigation.

(iii) Even if the explanation just proposed should be true of some cases of telepathically induced discursive cognition, I do not think that it could possibly cover all or most of them. In most cases it seems certain that the person from whom the cognition was telepathically derived was not thinking at the time of the fact or proposition concerned. And, if he was not thinking of it, he was a fortiori not imaging a set of spoken or written words which would express it in his own language. When N derives telepathically from M a cognition of a fact which M knows or a proposition which M believes, it is not usually the case that M is actually having an experience of knowing the fact or believing the proposition. Usually M's knowledge or belief is at the time purely dispositional, as most of our knowledge and our beliefs are at every moment. It is possible, of course, to evade this contention by saying that M must have been 'unconsciously' having an actual experience of knowing the fact or of believing the proposition at the time when the cognition is telepathically induced in N. This, however, would be a wholly gratuitous assumption, for which there is no independent evidence, and I shall ignore it.

The position, then, seems to be this. Suppose that N telepathically derives from M a cognition of a fact F, which M knows, or of a proposition P, which M believes. Then the operative factor on M's side will not as a rule be any actual cognitive experience which M is having at the time. The operative factor on M's side will usually be what we may call his 'potentiality of knowing F' or his 'potentiality of believing P'. By M's 'potentiality of knowing F' I mean that persistent modification of structure or process, whatever it may be, which ensures that, whenever M is suitably stimulated by a reminder, he will have an actual experience of knowing F. By M's 'potentiality of believing P' I mean that persistent modification of structure or process, whatever it may be, which ensures that, whenever M is suitably stimulated by a reminder, he will have an actual experience of believing P. I have already said that we know nothing whatever about the intrinsic nature or location of these assumed persistent modifications. We know them only as relatively permanent after-effects of actual experiences, and as relatively permanent cause-factors in producing and modifying subsequent experiences. Let us call them 'Experientially Initiated Potentialities of Experience'.

Now the normal rule is this. Any such potentiality which is a cause-factor in producing or modifying M's later experiences has been acquired from M's earlier experiences. I wish to point out that this is merely an empirical rule based on normal experience. Since we know nothing about the intrinsic nature or location of experientially initiated potentialities of experience, we cannot possibly see any kind of necessity in this or any other rule about them. It is logically possible that a potentiality which is an after-effect of M's past experiences should be a cause-factor in producing or modifying, not only M's future experiences, but also those of N. Many cases of telepathically induced discursive cognition seem to suggest that this logical possibility is in fact sometimes realized.

Let us begin by considering normal thinking. Here, as we have said, the only experientially initiated potentialities which affect a person's later experiences are those which were initiated by his own earlier experiences. In low-grade thinking, such as day-dreaming, it would seem that some one potentiality is activated by some very contingent experience of the thinker, and that this then activates another, and this in turn another, and so on, in an almost automatic way dependent on association by contiguity, similarity, etc. The result is a series of thoughts or images which have very little logical interconnexion; though the thinker himself, if he reflected on them, or a psychologist, if he performed a psycho-analysis, might be able to conjecture why the experiences had followed each other in this particular order. If, on the other hand, the person is actively pursuing a directed train of thought on some definite problem, those potentialities which would give rise to experiences relevant to the problem will tend to be stimulated and those which would give rise to experiences irrelevant to the problem will tend to be kept quiet. Even here the potentiality which would give rise to an experience highly relevant at a certain stage in the process often fails to be activated at the appropriate moment. And potentialities which give rise to irrelevant or misleading experiences often do get activated. Even when a process of thinking, directed to solving a certain problem, is eventually successful, the thoughts which are the stages in this process seldom arise in their proper logical order. The right logical order usually comes as a result of retrospective reflexion on the process by the thinker, followed by an act of rearrangement.

The point which I want to emphasize now is the following. When normal directed thinking is contrasted with normal low-grade thinking, it may fairly be called a 'voluntary' process. And it may fairly be said that the thinker 'deliberately selects', out of the mass of potentialities of experience which his past experiences have initiated, those which would give rise to relevant experiences if they were stimulated. But it is most important not to be deceived by such phrases. We must not imagine that the thinker perceives the various potentialities of experience, as a man might perceive a lot of ties and socks and shoes and pullovers in his bedroom, and then deliberately activates a certain selection from them, as a man might deliberately put on a certain tie, a certain pair of socks, a certain pair of shoes, and a certain pullover, in order to produce a certain colour scheme. The following analogy may make the fallacy quite plain. When the process of constructing a machine with one's hands is contrasted with blinking or jerking one's knee, it may fairly be called a 'voluntary' process. And it may fairly be said that the agent 'deliberately selects', out of a mass of potentialities of movement derived from his past bodily actions, those which would give rise to the relevant overt movements if stimulated. But he certainly does not perceive his own motor-nerves and muscles, select certain of the former, and decide to send such and such nervous impulses down the former in order to activate the latter in such a way as to make his fingers move as he wants them to do. He is perceiving and thinking of nothing but his hands and the materials with which he is working. He is desiring nothing but to make certain complicated movements with his hands against the resistance of the materials. This automatically, and in ways utterly unknown to him, sets up unfelt processes in unperceived nerves. And, in the main, these are in fact the appropriate processes in the appropriate nerves; since, in the main, the expected and desired overt movements result. To imagine that a thinker literally selects and deliberately activates those potentialities of experience which are relevant to the problem that he is trying to solve is like imagining that a manual worker literally contemplates his own brain and nervous system as if it were a complicated switchboard and deliberately presses such and such buttons. The thinker or the manual worker wills that a certain process of thought or bodily action shall take place; and automatically, in ways unknown to him, his volition initiates and sustains, among unobservable entities, unobservable processes which do in fact tend to bring about the desired process of thought or bodily action.

I have insisted upon this point about normal thinking because it has an important bearing upon telepathically induced thinking. It seems to me that there are two ways in which we are liable to make needless difficulties for ourselves in connexion with this subject. (i) We tacitly assume that potentialities of experience initiated by M's experiences must be located in M's brain or in M's mind; and similarly, mutatis mutandis, for N and for each other individual. (ii) We tacitly assume that, when a certain set of coexistent potentialities of experience are activated in such an order as to give rise to a certain coherent train of thought in M's mind, M must have contemplated a whole mass of coexistent potentialities and must have deliberately selected and activated this particular sub-group. Then we are faced with telepathy induced discursive cognition. We thereupon raise such questions as these. How can N contemplate potentialities of experience which are located in M's brain or in M's mind? How can N select from these just that sub-group which is relevant to his own problem at the moment? How can N activate this sub-group located in M's mind or brain? And, if N does this, why are the corresponding experiences produced in N's mind and not in M's?

Now these difficulties are at least lightened by the two following considerations. (i) Even if the potentialities of experience which are initiated by M's experiences are located in M's mind or M's brain, there is not the least reason to suppose that N would have to contemplate them and deliberately activate a certain selection of them. For we have seen that this is certainly not the way in which the set of potentialities which are relevant to a normal train of thought are activated by the mind in which that train of thought occurs.

(ii) We have very little ground for assuming that the potentialities of experience which are initiated by M's experiences are located in M's mind or in M's brain. If I say that an actual experience is located in M's mind, I know what I mean. I mean that it is one of M's experiences, and I know perfectly well what it is for a certain experience to belong to, or occur in, a certain mind. But experientially initiated potentialities of experience, whatever they may be, are certainly not themselves experiences. When I say that a certain acquired potentiality of experience is located in M's mind this can only be an abbreviated way of saying that it was produced by a past experience of M's and that it is a cause-factor in producing or modifying later experiences of M's. If the statement means anything more than this, I have no idea what it means. If, on the other hand, I say that it is located in M's brain, I must mean that it is a more or less persistent modification in the spatial arrangement or the movements of the ultra-microscopic particles in some part of M's brain. Now there are well-known empirical facts about the loss of a person's normal memories through injuries to his brain and his subsequent recovery of these memories which make it very difficult to accept this view of experientially initiated potentialities of experience. So the statement that potentialities of experience initiated by M's experiences are located in M's mind seems to be either metaphorical or meaningless; and the statement that they are located in his brain, if taken as the whole truth, seems to be difficult to reconcile with admitted facts about the effects of brain-injuries on normal experience.

We must therefore consider seriously the possibility that each person's experiences initiate more or less permanent modifications of structure or process in something which is neither his mind nor his brain. There is no reason to suppose that this Substratum would be anything to which possessive adjectives, such as 'mine' and 'yours' and 'his' could properly be applied, as they can be to minds and to animated bodies. The situation would be this. The modifications which are produced in this common Substratum by M's experiences normally affect only the subsequent experiences of M; those which are produced in it by N's experiences normally affect only the subsequent experiences of N. But in certain cases this normal causal 'self-confinement', as we might call it, breaks down. Modifications which have been produced in the Substratum by certain of M's past experiences are activated by N's present experiences or interests, and they become cause-factors in producing or modifying N's later experiences.

As we know nothing about the intrinsic nature of experientially initiated potentialities of experience, we cannot say anything definite about the intrinsic nature of the common Substratum of which we have assumed them to be modifications. As there is no reason whatever to think that such potentialities of experience are, or could be, themselves experiences, there is no reason whatever to suppose that the Substratum is a mind. On the other hand, it could hardly be any particular finite body. It does not seem impossible that it should be some kind of extended pervasive medium, capable of receiving and retaining modifications of local structure or internal motion. But I do not think that we have at present any adequate data for further speculations about its nature.