C. D. Broad

Professor of Philosophy at Trinity in 1923. In 1926 he became Lecturer in Moral Science, and served as Knightbridge Professor of Moral Philosophy from 1935 to 1953. He had Fellowships and honorary degrees in several countries. President of the Society for Psychical Research 1935-6 and 1958-60.

Saltmarsh's Investigation of Mrs Warren Elliott's Mediumship

 - C. D. Broad -

          IN THE present chapter I shall give a fairly full account of what I regard as a model investigation of a case of trance-mediumship. The medium was Mrs Warren Elliott, a professional. The investigation was conducted in a series of sittings with her by the late Mr H. F. Saltmarsh of the SPR towards the end of the 1920s. His report is to be found in SPR Proceedings, Vol. XXXIX.

General Account of the Experiments

I will begin by giving a general account of the nature of the experiments and of the characteristic features of Mrs Warren Elliott's mediumship.

(1) The Use of 'Relics'. Like a good many other mediums, Mrs Elliott finds it helpful or even necessary to have presented to her at the séance some article, e.g. a ring, a lock of hair, a letter, etc., which has been connected with the dead person with whom she is to try to make contact. We may call such an object a 'relic'. There is no plausible theory, so far as I know, as to how relics function. But it is a fact that the handling of an inanimate object does sometimes enable a medium in some way to make statements relevant to events in the life of some person or persons who are or have been connected, either directly, or indirectly through the intervention of some third party, with that object.

Now the sittings in this investigation fall into two groups, according to the way in which the relic was presented to the medium. (i) In one group the person who owned the relic would take it along with him or her to a sitting with the medium, would hand it to her at the beginning, and would receive it back from her at the end. These may be described as 'Owner-present sittings'. I will denote them, for short, by the phrase 'O-P sittings'. (ii) The other group of sittings were conducted in the following way. A number of persons were asked to send one relic apiece to the Secretary of the SPR at the Society's offices. On receipt of any such relic the Secretary would put it into an envelope, write a number on the outside, put the envelope with its contents into a certain locked cupboard, and enter the number (together with the sender's name and address) in a certain book. On the day on which a sitting of this kind was to be held, the person from the SPR who was to sit with Mrs Elliott would choose one of these envelopes at random, take it along to the séance, and hand it to the medium, just as the owner would have done at an O-P sitting. At the end the sitter would receive it back from the medium and return it at once to the Secretary of the SPR. Such séances as these may be described as 'Owner-absent sittings', and denoted for short by the phrase 'O-A sittings'. In two and only two cases the same relic was used at two O-A sittings.

Sometimes Mrs Elliott would open the packet, and sometimes hold it unopened in her hand. This seemed to make no significant difference to the degree of relevance or accuracy of her utterances. In all the sittings the owner's name was unknown to the medium. In the O-A sittings it was unknown to the sitter also. In the O-P sittings the owner was the sitter, and was accompanied by a shorthand-writer supplied by the SPR, who took down everything said by either the owner or the medium. In the O-A sittings the only sitter was the note-taker. Altogether four persons were employed as note-takers, but most of the note-taking was in fact done by two only of these. The names of the note-takers, with one exception, were unknown to the medium, though she knew that they came from the SPR. In the O-A sittings the owner of the relic was quite unaware that a sitting was being held at which the relic submitted by him or her to the SPR would be handed to the medium.

(2) Recording and Marking. Mrs Elliott had a control who called herself 'Topsy'. In all the O-P sittings the medium went into a trance, and the ostensible communications were through the Topsy-persona. That holds for most of the O-A sittings also. But in some of these Mrs Elliott, apparently in her normal waking state, would simply dictate her impressions to the note-taker. There was no significant difference in degree of relevance or of accuracy of the information conveyed at O-A sittings by these two methods.

After each sitting the shorthand record was typed out, and a copy was sent to the owner of the relic which had been used. (This would, of course, be identical with the sitter in the case of an O-P sitting.) The owner was instructed to annotate it, and then return it with the annotations to the SPR. Mr Saltmarsh would then go through each record and would note the various items mentioned in it. To each of these he would assign a reference number. Thereafter the annotated record was put into an envelope, on the front of which was written a complete list of the items referred to, with their reference numbers.

It was found that certain items tended to repeat themselves in the various records with abnormal frequency. Saltmarsh describes these as 'clichés'. He noted about one hundred of these, and entered them into a card-index. (An example of a cliché was the abnormally frequent mention of an injury to the leg or foot.)

Saltmarsh divided the ostensible communications into 'ante mortem' (A-M) and 'post mortem' (P-M), according as they referred to events alleged to have happened before or after the death of the ostensible communicator. He sub-divided each of these classes into 'physical' (Φ) and 'non-physical' (non-Φ). A reported event or circumstance was counted as physical, if and only if it were such that an ordinary observer, coming suddenly on the scene at the time and place referred to, would have witnessed it if it had existed or happened there and then. We thus have four ultimate sub-divisions in respect of subject-matter, viz. (A-M)-Φ, (A-M)-(non-Φ), (P-M)-Φ, and (P-M)-(non-Φ).

We come now to the scoring of the records. This was done in every case by Saltmarsh himself, on the basis of the comments made on the record by the owner of the relic employed at the sitting. For purposes of scoring Saltmarsh divided the statements into three groups, viz. 'vague' (V), 'definite' (D), and 'characteristic' (C). A V-statement is one of a thoroughly commonplace character, likely to fit a large proportion of persons. A D-statement is one which, while it might fit a fair proportion of persons, is not so commonplace as to be as likely as not to be true of any person chosen at random. A C-statement is one which would be most unlikely to be true of a person chosen at random. Saltmarsh assigned to any true statement 1 mark, if he judged it to be vague: 5 marks, if he judged it to be definite; and 20 marks, if he judged it to be characteristic.

The maximum possible score for a record was calculated by supposing that all the statements in it had been true, marking them in accordance with the above scale, and then adding the marks. The actual score would, of course, always be considerably less; for some of the statements would be false, and these would get no marks. Some records contained very few definite or characteristic statements. In such cases Saltmarsh made a deduction of 10 per cent from the gross actual score. The resulting net score was the one finally assigned in such cases. The mark eventually assigned to any record is the ratio of the net actual score to the maximum possible score, measured as a percentage.

Saltmarsh admits that simple addition is almost certainly an unsatisfactory way of combining the marks assigned to the various items in a record. He points out, too, that we must not regard the ratio of the actual net score to the maximum possible score for a record as representing the absolute value of the communications at the sitting in question. At best such percentages furnish a means of comparing the values of the communications received at different sittings and under various conditions. That was in fact the only use which Saltmarsh made of them.

The question naturally arises: How far can we rely upon the owners' annotations? Saltmarsh instituted the following check on them. He sent copies of the records of certain selected O-A sittings to persons quite unconnected with the owners of the relics used on those occasions. Each such person was asked to imagine that the record which he was to annotate was that of a sitting at which a relic contributed by himself had been used, and to annotate it on that assumption. Saltmarsh then marked these records, on the plan explained above, in accordance with the annotations of the pseudo-owners. He then compared the percentage scores thus obtained with those which he had assigned to the same records on the basis of the annotations made by the actual owners of the relics employed.

Two checks of this kind were carried out. In the first of them, all the records related to one and the same ostensible communicator, who claimed to be a certain airman killed in action. There were 53 records referring to this individual. Saltmarsh chose for the pseudo-owners six persons, all of whom he knew to have lost a son or a brother, of about the same age as the ostensible communicator, in rather similar circumstances to those in which he had died. He divided up the records among these six. The results were as follows. The aggregate of the scores for these records, derived from the annotations made by the actual owners was 4,107, whilst the aggregate derived from the annotations of the pseudo-owners was 452. The aggregate maximum possible score was 5,642. So the percentage score based on the annotations by the actual owners was 72.8 per cent, whilst that based on those of the pseudo-owners was only 8 per cent. If, however, we consider individual markings, we find that, with two of the pseudo-owners, the percentage scores based on their annotations do not differ so significantly from that based on the annotations of the actual owners. In the case of both these pseudo-owners, as it happened, the percentage score assigned on the basis of their annotations was the same, viz. 17 per cent. The scores assigned on the basis of the annotations of the actual owners were in one case 64 per cent and in the other 80 per cent.

The other check was conducted as follows. Saltmarsh had copies made of the records of 19 O-A sittings which had scored high percentage marks on the annotations of the actual owners. He made use of 15 pseudo-owners. He selected at random sets of 5 records out of these 19, and sent such sets to the pseudo-owners to be annotated. Of the pseudo-owners two annotated two sets of 5 apiece, whilst the remaining 13 each annotated one set of 5. Thus Saltmarsh received back in all 20 + 65 = 85 annotated copies of records of these 19 O-A sittings. Owing to his random selection of sets of 5, a set sent to one pseudo-owner on one occasion might overlap a set sent to another on another occasion. Thus, e.g., the record of one of the 19 O-A sittings was annotated by no less than 9 pseudo-owners, whilst the records of some others were annotated only by one pseudo-owner.

The aggregate maximum possible score for these 19 records was 5,554, and the aggregate actual score assigned by Saltmarsh on the basis of the annotations by the actual owners was 3,226. The percentage was therefore 58.1 per cent. Saltmarsh does not make it quite clear to me how he proceeded to get a comparable aggregate figure on the basis of the annotations made by the pseudo-owners. It seems obvious that, in the case of any record that had been annotated by more than one pseudo-owner, it would be necessary to take the average of the marks assigned to it on the basis of the annotations of each pseudo-owner to whom a copy of that record was submitted. In the case of any record that had been annotated by only one pseudo-owner, Saltmarsh would simply take the marks which he would assign to it on the basis of that person's annotations. The comparable aggregate score would then be the sum, for all 19 records, of the average scores assigned in cases of the first kind and the actual scores assigned in cases of the second kind. Whether Saltmarsh in fact proceeded in this way or not, the aggregate score that he gives in respect of the 85 annotations by his 15 pseudo-owners is 487. This is only 8.75 per cent of the maximum possible score of 5,554, as compared with the percentage of 58.1 per cent on the basis of annotations by the actual owners.

Here again, however, the aggregate result masks certain individual cases where the percentage score on the basis of annotation by a pseudo-owner came much nearer to that based on those of the actual owner. Thus, in the case of three pseudo-owners, each of whom annotated five records, the percentage scores based on their annotations were 23 per cent, 28 per cent, and 33 per cent, whilst those based on the annotations of the real owners were respectively 52 per cent, 59 per cent, and 46 per cent. Such differences seem hardly significant, though they are all in the same direction.

Everyone must decide for himself, on the basis of such figures as I have given, whether it is or is not reasonable to hold that the amount of agreement between the medium's statements and the relevant facts, as stated by the owners of the relics submitted to her, was significantly greater than might be expected through chance-coincidence. Saltmarsh admitted the crudity of the tests and the uncertainty of any inferences drawn from them. Later he and Dr S. G. Soal devised a much better general method of estimation, which, after being submitted to Sir Ronald Fisher and amended in certain respects by him, was published, under the title 'A Method of Estimating the Supernormal Content of Mediumistic Communications', in Vol. XXXIX of the SPR Proceedings. The problem has been attacked since then by several experts in the U.S.A. The reader may be referred to the following papers: Pratt, 'Toward a Method of evaluating Mediumistic Material' (Bulletin of the Boston SPR, 1936); Pratt and Birge, 'Appraising Verbal Test-material in Parapsychology' (Journal of Parapsychology, 1948); and Schmeidler, 'Analysis and Evaluation of Proxy Sessions with Mrs Caroline Chapman' (Journal of Parapsychology, 1958). For a startling example of how a complete stranger, when asked to annotate a mediumistic communication which seemed extremely appropriate to the peculiar circumstances of the actual sitter, may find it no less appropriate to himself and his own circumstances, the reader may be referred to a contribution by Mr Denys Parsons to SPR Proceedings, Vol. XLVIII, entitled 'On the Need for Caution in assessing Mediumistic Material'.

(3) The main Statistical Results. We can now pass to the main statistical results of the investigation. These may be roughly summarized as follows. (i) The percentage of true statements in O-P sittings is about half as much again as in O-A sittings (76.3 : 52 per cent). (ii) The number of post mortem and of ante mortem statements in the whole lot of sittings is roughly the same (P-M 1619, A-M 1592); but the former contain a somewhat higher percentage of true statements than the latter (P-M 74 per cent, A-M 66.3 per cent). (iii) The physical statements are considerably more numerous than the non-physical (Φ 2470, non-Φ 741). But they contain a lower percentage of true statements than the latter (Φ 66 per cent, non-Φ 84.5 per cent). Among the ante mortem statements those about the ostensible communicator are much more numerous than those about the sitter; but among post mortem statements the opposite is the case.

Saltmarsh's Conclusions and Reflexions

The results of the sittings can be considered from two points of view, viz. from the light which they throw on the psychological processes involved, and with regard to the question: Do they require us to postulate anything paranormal, and, if so, what? I am concerned here primarily with the former aspects of the case.

(1) 'Fishing', Groping, etc. It might well be thought that a medium, making a number of statements under the conditions described, would indulge in a certain amount of deliberate invention, and might further try to discover, by means of 'fishing', what statements would be likely to be appropriate.

Saltmarsh found no evidence for deliberate invention. At most there are the clichés, already mentioned; the occurrence later in a sitting of statements which would be easy inferences from what had gone before; and the occurrence of many commonplace statements which would fit most people.

As to 'fishing', this was of course impossible in the O-A sittings, since the sitter had no relevant normally acquired information for the medium to fish from. As regards the O-P sittings, Saltmarsh's comment is: 'Topsy does not fish, but she does grope'. What he means is this. Whatever may be the source of the information, it is certain that it comes to Topsy by means of visual symbols, which she has to interpret as best she can. Often she does not understand the symbolism, and then she is liable to offer alternative interpretations of it. Even when she has in fact hit on what seems to be the correct interpretation, she does not always realize that she has done so.

The Topsy-persona has certain mental or verbal habits, which show themselves in her utterances. (i) She gives certain names and initials much more frequently than they occur in contemporary England. Saltmarsh instituted a comparison between the frequency with which various male Christian names occurred in the communications and the frequency with which those names had occurred over the previous thirty-five years among the boys at a large public school. There were very great discrepancies, sometimes in excess and sometimes in defect. (ii) She has a habit of giving names that begin with the sound ROJ or ROD. (iii) She gives many highly fantastic names, and these tend to begin with the letter O. (It may be relevant that Mrs Elliott's maiden name was Ortner.)

In many sittings there is a great deal of completely unverifiable material given. Saltmarsh tried, as a check, a certain number of sittings at which the article handed to the medium as a relic was in fact something completely new. In all these sittings there was plenty of ostensible communication, similar in style and character to that given when there was a genuine relic. These utterances were completely irrelevant to the person contributing the pseudo-relic, and seem to have been wholly 'in the air'. It is, however, of some interest to remark that in most of these cases the Topsy-persona alleged, with regard to the ostensible communicator, that he or she was not connected with the object that had been handed to her. The successive items in the utterances on such occasions had very little interconnexion with each other. Saltmarsh was inclined to compare them to a series of hypnagogic images, rising from some stratum in the medium's subconsciousness. He suggests that these images are probably generated from traces left on the medium's mind by incidents at previous sittings or in her ordinary daily life. Such internally generated images form the matrix in all the sittings. In a successful sitting, with a genuine relic, the veridical matter (however it may originate) is imbedded in this matrix of reminiscent fantasy.

(2) Influences which might be operative. We can now consider various influences which might conceivably affect the results, and ask ourselves whether they in fact do so. Saltmarsh discussed in turn the owners of the relics, the note-takers, the nature of the relic submitted, and the condition of the medium at the time of the sitting. I will follow that order.

(i) Influence of Owners. We can divide the owners into three classes, viz. (a) bereaved persons strongly affected emotionally, (b) bereaved persons not strongly affected, and (c) non-bereaved persons. Each of these classes can then be sub-divided into three sub-classes, viz. (α) those convinced of survival, (β) those with an open mind on the subject, and (y) those convinced of non-survival. The results of the sittings can be classified as good, moderate, or poor. The question is whether there is any significant association between belonging to such and such a class of sitters and getting results of such and such a degree of goodness.

Now we must begin by allowing for the fact that a bereaved person whose emotions are strongly affected, or a person who is convinced of survival, will be likely, in annotating the record of a sitting which concerns himself, to stretch points which seem to favour the view that a deceased friend or relative has survived and is communicating. This applies both to O-A and O-P sittings. But we must remember, further, that at O-P sittings the owner is also the sitter. Now sitters of the two kinds just mentioned would be likely to give more encouragement to the medium, even if they do not inadvertently give away information; and that may suffice by itself to secure better results. Moreover, such sitters may also give hints by gestures or changes of facial expression. These would not get recorded by the note-taker, though verbal indiscretions would appear on the record and could be allowed for. After taking all this into account, Saltmarsh found no significant association, positive or negative, between an owner being of any one of his nine categories and the results with his relic falling into any one of his three grades.

The following negative fact is of some interest. No less than seven of the owners claimed to have some kind of 'psychic' gifts. There was no significant association, positive or negative, between this alleged characteristic of an owner and the goodness, indifference, or badness of the results claimed with his or her relic.

(ii) Influence of Note-takers. It will be remembered that most of the note-taking was done by a certain two of the four note-takers. It happened that one of these had contributed a relic, and that some of the best results obtained were at sittings at which that relic was used. Now at some of the sittings at which this person was note-taker, and relics belonging to other persons were used, it looks as if her special ostensible communicator 'intruded'. By this I mean that communications purporting to refer to that individual tended to mingle with those that purported to refer to the deceased person associated with the relic that was being used at the time. Apart from this, it was found that the Topsy-persona quite often referred to the contemporary circumstances of the person who was acting as note-taker, and that her references were often so apposite as to suggest sporadic telepathy from the latter.

(iii) Influence of the Relic. The question here is whether the medium could have been influenced in her utterances in a perfectly normal way by what she could see or feel of the relic. The answer appears to be decidedly in the negative. In some sittings the package was not opened by the medium at all. In others the sitting began with the package unopened, and later on the Topsy-persona, finding that she was not getting on well, would open it. There was no significant difference in the results in the two cases.

(iv) Influence of the Medium's Condition. Some of the most successful sittings took place when the medium was tired or disturbed by external noises. As already stated, she was in trance at all the O-P sittings, but not at all the O-A sittings. If one compares the results of the O-A sittings in which she was in trance with those of the O-A sittings in which she was in her normal state, there is no significant difference either in the proportion of true statements or in the distribution of the statements into vague, definite, or characteristic. If Mrs Elliott liked the sitter, the chances of a good sitting were increased; but such liking was no guarantee of a good sitting.

(3) Certain General Features of the Sittings. The following general remarks may be added about certain features in the sittings:

(i) It was of no use to put direct questions to the Topsy-persona, for it was seldom that any relevant and definite answer was given. She jumps about from one topic to another, and it was seldom possible for the sitter at the time, or for the experimenter reflecting at leisure afterwards, to discover the links between successive topics. That is true also, though to a lesser extent, of those O-A sittings at which Mrs Elliott was in her normal state.

It appears, therefore, that the links between successive topics are not those of obvious association. There may, of course, be associative links private to the Topsy-persona or to Mrs Elliott, as the case may be.

(ii) In this connexion it may be of interest to consider the clichés, i.e. words and ideas which recur with abnormal frequency, which are not mere commonplaces, and which yet have no obvious relevance to the owner of the relic or to the note-taker. Saltmarsh compares them to the recurrent dreams which many people have. He notes also that, in persons who have hypnagogic imagery, such imagery tends to be recurrent and typical of the individual concerned. He suggests that the clichés may originate partly in symbols whose meaning is not clear to the medium or to her control, partly in reminiscences of other sittings, and partly from normal associations due to certain conjunctions of experiences peculiar to the medium in her daily life.

(iii) A third feature to be considered is the so-called 'intrusions', which happen at certain O-A sittings. The essential point is that on such occasions communications appear to be coming from, or to be about, a certain recognizable communicator, but one who is altogether unconcerned either with the relic which is being used, or with its absent owner, or with the note-taker who is present and may be regarded as the sitter.

There were two fairly persistent intruders. One purported to be Mrs Dora Irving, the deceased wife of the Rev. W. S. Irving, a member of the SPR who had had frequent sittings with Mrs Osbourne Leonard, at many of which that medium had been ostensibly possessed by the Dora-persona. The other purported to be the deceased airman 'A', already referred to in the present chapter. Excellent ostensible communications, purporting to come from him, had been received at thirteen O-P sittings and seven O-A sittings in the present series, where a relic associated with him was used. He appeared as an 'intruder' in six other sittings, in which relics associated with other persons were used, and in which neither the owner nor the note-taker had any connexion with him. It may be remarked that both these intruders had shown themselves to be plausible and persistent ostensible communicators, in conditions where they were not intruding, e.g. in sittings with Mrs Leonard. Probably an ostensible communicator of whom that was not true would have failed to establish his or her identity sufficiently to be recognized if he or she were to intrude.

It is of interest to note that the Topsy-persona or normal Mrs Elliott, as the case might be, often recognized when the ostensible communicator was not connected with the relic submitted at the time. She had a characteristic symbolic experience on such occasions, viz. a visual image or a quasi-sensory visual hallucination as of the purporting communicator 'waving aside' the relic.

(4) The Nature of the Symbolism. This brings us to the general topic of the symbolism employed in Mrs Elliott's mediumship. Besides the example just given, we may adduce the following illustrations. Family relationship, or absence of it, between ostensible communicator and sitter is repeatedly symbolized for the Topsy-persona by ostensibly seeing the former as close beside (or, respectively, as remote from) the latter. Again, when she became aware somehow that two persons connected with a certain relic had been, as she put it, 'nearly married and then not married', she did so through the ostensible communicator showing her a wedding-dress and then letting it fall.

Sometimes the Topsy-persona is uncertain of the meaning of a symbol, which is presented to her in visual imagery or in quasi-sensory visual hallucination, and says so, giving two or more alternative possible interpretations. The following is an interesting case. It was characteristic of a certain ostensible communicator that the person whom he purported to be had had in this life a passion for rice pudding. He had been heard to say that he would willingly have one both at lunch and at dinner; and on one occasion he had said: 'You can give me more rice, I am never tired of it.' Now the Topsy-persona eventually got on to this singular taste of the ostensible communicator, but she did so in the following roundabout way. In the sitting at which the relic connected with this individual was used she approached this topic by an irrelevant and incorrect reference to 'India, or some hot place over the water, where there's lots of dark people'. Then came: 'Remember palms!' Then: 'He laughs, and says "Where bananas grow'. Then: 'He laughs, and shows Topsy a lot of rice. There must be a joke about lots of rice - "more rice, never tired"'. No doubt rice was associated in Mrs Elliott's mind with India. This called up by association dark people and palms and bananas and finally rice. And then at last came the correct interpretation of rice as a favourite article of diet with the ostensible communicator.

The symbolism generally depends on very natural, or very widely accepted conventional, associations of ideas, rather than on associations peculiar to the medium. Thus spatial nearness typifies close blood-relationship; black is a symbol for worry or sorrow or death; the waving away of a relic by the hallucinatory figure of an ostensible communicator is a sign that he is not connected with that relic; and so on.

The dramatic form of a sitting with Mrs Elliott controlled by the Topsy-persona differs in the following way from that of a sitting with Mrs Leonard controlled by the Feda-persona. Feda presents the communications as if she obtained them by listening to communicators who are seldom visible to her, and then reported them to the sitter. Topsy presents the communications as if she obtained them by watching communicators, who are inaudible to her but make gestures and exhibit visual symbols, which she describes and tries to interpret. The difference, of course, is not absolute. Feda claims sometimes to see the communicators, and Topsy claims sometimes to hear them.

The Topsy-persona always talks as if she literally saw the communicators, though not of course with her physical eyes, and as if they deliberately showed her this or that visible (though non-physical) object or scene, as a symbol of the information which they wish to convey. If this were accepted literally, it would imply that the communicators have 'astral' bodies; that the Topsy-persona herself has one, provided with the 'astral' equivalent of physical eyes; and that there are 'astrally' visible objects and scenes, to which the communicators can direct her attention. Saltmarsh is not prepared to accept this, but thinks it much more likely that the Topsy-persona is at such times subject to some kind of quasi-sensory hallucinatory perception of the visual kind, analogous to a vivid dream had by a normal person when asleep.

He thinks that the immediate source of these experiences is in every case some stratum of the medium's subconscious self, and that all the raw material of the hallucinations comes from there. But he admits that, in certain cases, we may have to suppose, in order to account for the relevance and the veridicality of the information conveyed by the symbolism, that the stimulus, which evokes and guides the phantasmogenic process, comes from some source other than any level of the medium's personality.

In support of his view that the immediate source of the symbolic experiences lies within the medium, Saltmarsh alleges that the Topsy-persona uses one and the same system of symbolism in ostensible communications from many different ostensible communicators. He alleges that that is true of the Feda-persona also, though the two use different symbols for the same idea. He argues that, if the symbols pre-existed in an 'astral' world and were literally shown to the control by the communicators, or if they were directly generated in her by the communicators as hallucinatory quasi-percepts, we should expect to find characteristic differences between the symbolism used when different ostensible communicators communicate through the same medium. Conversely, we should expect to find similarity between the symbolism used when one and the same ostensible communicator communicates through different mediums. (It should be remembered that several of the ostensible communicators in the sittings with Mrs Elliott had also ostensibly communicated at sittings with Mrs Leonard.)

It seems to me that the first prong of this double argument is somewhat blunted by Saltmarsh's statement that the symbolism used in the Elliott communications is based on natural, or very widely accepted conventional, connexions of ideas, rather than on associations peculiar to the medium. If that be so, the use of a common symbolism by a number of different ostensible communicators through Mrs Elliott might be due simply to the fact that the basis of the symbolism is common to almost everyone, including the medium, the sitters, and all the ostensible communicators.

(5) Saltmarsh's Theory of the Processes involved. However that may be, let us suppose, for the sake of argument, that the amount and quality of the veridical material in the ostensible communications makes it reasonable to postulate some remote source, other than any stratum of the medium's mind, as providing an evoking and guiding stimulus to her subconscious phantasmogenic powers. How are we, then, to think of the processes involved?

If we make this postulate, it is plain that we must assume that there is a certain stratum of the medium's mind which is the immediate recipient of this foreign influence. Saltmarsh calls this the 'Receptor Stratum'. We can then raise the question: What are the immediate products, in the Receptor Stratum, of the influence on it of sources of information foreign to the medium? Suppose we accept Saltmarsh's view that it is something in the medium herself which is directly responsible for the particular images or hallucinatory quasi-perceptions that eventually well up in the Topsy-persona or in normal Mrs Elliott and express in symbolic form the information received. Then we must suppose that such information is received and registered by the Receptor Stratum in some other form.

At this point we must distinguish between (i) information as to the emotions or bodily pains or pleasures of the ostensible communicator, and (ii) information as to physical states of affairs or as to experiences other than emotions or bodily feelings. The first case seems fairly simple, for the symbolism required is so obvious and natural. All that is needed is that the medium shall actually feel a kind of phantom bodily sensation, e.g. of pain in a certain part of the head, of suffocation, of palpitation, and so on, and that she should take this as a symbol of a corresponding affective experience in the ostensible communicator, of whom she is having an hallucinatory quasi-perception of the visual kind. There is considerable evidence that that was in fact the way in which such information was received.

But the mode of reception of information of the second kind raises much greater difficulties. Saltmarsh argued that such information must be supposed to be received and registered by the Receptor Stratum in some way which is altogether independent of formulation in audible or visible sentences, and of symbolization by imitative or metaphorical visual imagery or quasi-sensory hallucination. He summarizes this view in some extremely obscure sentences, which I will quote: '... The impression received by the Receptor Stratum is received as a meaning, i.e. the process is a purely psychical one, not expressed in language or other sensory impressions ... The Receptor Stratum lies below or beyond the point at which differentiation into various senses occurs; i.e. a meaning would not be visual, auditory, olfactory, sapid, or tactual, but just plain meaning' (the italics in this quotation are mine, and not Saltmarsh's).

I cannot pretend to feel at all certain that I understand what positive doctrine Saltmarsh intended to convey by these sentences. But I suggest that what may have been at the back of his mind is something that could be put as follows. It is a fact that a whole indefinitely large class of sentences, spoken, written, or merely imaged by various persons and on various occasions, may all have the same meaning. And it is a further fact that certain gestures, pictures, dances, etc., may express symbolically the same meaning as those sentences express verbally. We may therefore talk of 'the [common] meaning' of all those sentences and symbols. So far we are on safe ground.

But it is very natural to take the following step, viz. to take for granted that there must be a certain one peculiar entity, other than all the sentences and symbols which 'have the same meaning'; something which exists in its own right, and to which they all stand in a certain common relation, in respect of which it is what they all mean. We may use the word 'proposition' to refer to such supposed independently existent entities, provided we remember that this no more commits us to the belief that there are propositions, in that sense, than the use of the word 'dragon', to refer to supposed fire-breathing serpents, commits us to the belief that there are, have been, or will be such creatures. The step from the fact to the theory can be seen to be most precarious, when one contemplates it critically instead of taking it unwittingly. Yet I cannot help suspecting that, when Saltmarsh talks of the Receptor Stratum receiving information from an external source in the form of 'a meaning', and when he talks of 'just plain meaning', he is committed (though perhaps unwittingly) to this very theory. He is committed, I suspect, to the view that what is received is literally awareness of propositions, in the sense explained above, unmediated by any verbal or symbolic vehicle, whether sensory, quasi-sensory, or imaginal.

Now, quite apart from a strong feeling of doubt as to whether there are such entities as 'propositions', in the sense explained, I find it very difficult to swallow Saltmarsh's theory. Perhaps the following concrete example will bring out the difficulty that I feel. Suppose I wanted to convey to a certain individual on a certain occasion the information that some prominent public personage or other had lately died in England. If there were any language which I could speak or write and he could understand, I might utter in his presence or to him over a telephone, or write and have put before his eyes, a sentence in that language signifying that an event of that kind had happened. Failing that, if he and I were together in any town in England at the time, I might point out to him flags flying at half-mast on public buildings. If he were familiar with the relevant very widespread convention, he would equally well receive the information in that way.

Now, suppose that I had the power of evoking telepathically in his mind, without speaking to him directly or over a telephone, without putting any writing before his eyes, and without pointing to any symbolic object in his presence, a vivid image or hallucinatory quasi-perception as of that spoken or written sentence or as of a flag at half-mast on a building. Then there is no difficulty in principle in seeing that he could receive from me that particular bit of information on that occasion. The only difficulty would be in conceiving or imagining the process by which I could evoke telepathically in him the appropriate images or quasi-sensations. Given that I could do so, there is no difficulty in understanding how he would thereupon come to think of, and perhaps to believe, that some prominent public personage or other had lately died in England.

But what I find very hard to understand is how a person could, on a given occasion, come to think of, and perhaps to believe, something specific, although at the time there was not present to any stratum of his mind, even in the form of an image or an hallucinatory quasi-perception, any verbal formulation of it in a language that he understands, or any concrete symbol for it in a convention that is familiar to him. Yet that seems to be what Saltmarsh wishes us to try to contemplate, in the reception of information about physical facts or about non-affective experiences by Mrs Elliott's Receptor Stratum. If that was not what he had in mind, I cannot guess what was.

Perhaps all that we really need to suppose is something on the following lines. May it not be that the direct effect of the foreign source upon the medium is simply to set the relevant parts of her brain and nervous system in such incipient states as would, if they were to develop normally, lead to her uttering sotto voce certain sentences in her native tongue, or to her having auditory images of those sentences as spoken or visual images of them as written? And may it not be that those incipient brain-states do not in fact develop in her in that normal way? May it not be that they act, instead, merely as stimuli to some phantasmogenic stratum of her mind, which thereupon generates visual images or hallucinatory quasi-perceptions as of persons and things and scenes, which symbolize, in a way characteristic of the medium and her habitual associations, those ideas and beliefs which would have been verbalized if those incipient brain-states had developed in the normal manner?

In this connexion the following remark of Saltmarsh's should be noted. Suppose we regard the Topsy-persona, as he is inclined to do, as a kind of secondary personality of Mrs Elliott. Then we cannot identify the stratum or department of the medium's mind which generates the symbols with the stratum or department which functions as Topsy. For the latter often confesses herself uncertain as to the right interpretation of those symbolic images or quasi-sensory hallucinations which present themselves to her. (Compare the following fact. When one has a dream, as of oneself doing and suffering such and such things in such and such scenery, one's dream-personality at the time is just as ignorant as is one's waking personality later, of the sources and the possible significance of the dream-scenery and the dream-drama. So far from the dream-actor creating the dream-scenery and the dream-drama, he and they are alike products of something in the individual, of which the dream-actor is quite unaware.)

I pass now to another part of Saltmarsh's speculations. He makes considerable play with a rather elaborate analogy to the physical phenomenon known as 'osmosis', and he supplements this with certain other physical analogies, not necessarily consistent with this or with each other. The essential points of the osmotic analogy are these. The medium's mind is likened to a tube, closed at one end with a semi-permeable membrane, and containing a solution of some substance, such as sugar, in some solvent, such as water. The solution in the tube is compared to the normal contents of the medium's mind. The lowest layer of it, immediately in contact with the semi-permeable membrane, is compared to the Receptor Stratum. The tube is supposed to be dipped, with the closed end downwards, into a vessel containing a weaker solution of the same substance in the same solvent. This vessel and its contents are likened, respectively, to a foreign mind (possibly, though not necessarily, the surviving mind of some deceased human being) and to its contents. The dipping of the tube into the vessel is likened to the establishment of some kind of rapport between these two minds. In the physical case solvent would diffuse from the liquid in the vessel, through the membrane, into the tube, and would mix with the contents of the tube. This is likened to the conveyance of information from the external source to the Receptor Stratum of the medium's mind.

Now, as a general rule, I am strongly against professional philosophers badgering honest working psychologists or psychical researchers with niggling criticisms on tentative theories expounded in analogies. If, on the whole, they find a certain analogy useful to co-ordinate the facts observed up to date and to suggest questions that can be investigated by experiment or further observation, I would turn a blind eye to the fact that absurdities arise if the analogy be pressed too far. But I am afraid that, in the present case, the absurdity is patent at the very first move. It is of the essence of the osmotic process that what turns up in the tube is ipso facto removed from the surrounding vessel. Every particle of solvent that is added osmotically to the contents of the tube is osmotically subtracted from the contents of the vessel. But that is radically unlike the conveyance of information from one mind to another. Mind B does not ipso facto lose any item of information which it conveys to mind A. When an alleged analogy breaks down so fundamentally in principle, one doubts whether it can be worth while to pursue it in detail.

I am inclined to think that the only valuable feature in Saltmarsh's osmotic analogy is this. It suggests that the conveyance of information from the external source to the Receptor Stratum of the medium may be a process which is quite passive, in the sense that it is not deliberately initiated or directed by either party. It is important to bear this possibility in mind; for the word 'telepathy' is liable to call up a picture of one mind deliberately imposing certain ideas on another or deliberately exposing certain of its contents to the inspection of the other. And the word 'mind-reading' is liable to suggest one mind deliberately 'scanning' the contents of another mind. All these processes may, on occasion, take place. But it is well to have an analogy which allows for mere unintended and undirected 'leakage'; though, for the reasons given above, that analogy must not be pressed, and its positive associations must be firmly rejected.

If we must use a physical analogy, I think that resonance would be much better than osmosis. We might compare the medium's mind and that of the supposed external source of information to two stringed musical instruments; and we might liken the contents of a mind at any moment to the vibrations of those strings in such an instrument which happen to have been recently struck. Then, finally, we might compare the conveyance of information from the foreign mind to the medium's to the setting up by resonance, in certain strings of the former instrument, of vibrations similar to those at present going on in certain strings of the latter, through the two instruments being partially attuned to each other.

Saltmarsh makes several interesting suggestions which are quite independent of the osmotic analogy. One is that there would probably be a double process of elimination before the stage at which symbolic images or quasi-sensory hallucinations would arise in the consciousness of the control. The first elimination would take place at the level of reception; for it may well be that there is much in the content of the foreign mind to which the Receptor Stratum of the medium's mind cannot respond at all, or can respond only weakly or distortedly. The second would take place at the boundary between the Receptor Stratum and the phantasmogenic department of the medium's mind. Some of the content imparted to the Receptor Stratum might fit easily into the medium's innate mental structure or acquired habits and associations of ideas. Others might fail to do so, or might arouse positive resistance in her. The former would be likely to be represented with ease, and without much distortion, in symbolic images or quasi-sensory hallucinations. The latter might fail to get represented at all, or, if it were, might be represented by symbols so distorted that neither the control herself nor the sitters could guess what they meant.

Besides these gaps and distortions in the final product, due to this double process of elimination, there will inevitably be something which might be described as 'dilution' or 'contamination'. For any images or quasi-sensory hallucinations, which may have their ultimate origin in content imparted to the Receptor Stratum from a foreign mind, will certainly be accompanied, and often almost swamped, by dream-like fantasies, originating entirely within the medium from traces of her past experiences, as our ordinary nightly dreams and our waking reveries arise in us.

In view of the disconnected and fragmentary nature even of the information which turns out to be characteristic and true of a given ostensible communicator and to be outside the medium's range of normal knowledge and conjecture, and in view of the immense amount of padding in which it is imbedded, something like this account of Saltmarsh's seems to be a plausible picture of the psychological processes involved in Mrs Elliott's mediumship. Since a large proportion even of successful sittings with gifted mediums have the same characteristics, something like this view would seem to have a wide range of application to trance-mediumship. But, as Saltmarsh admits and emphasizes, there are occasional bursts of what look like deliberate coherent messages concerning specific and characteristic events or circumstances in the earthly life of the ostensible communicator or of some friend or relative of his or hers.

To fit such facts into the theory, we might have to suppose that, although the conveyance of information from the foreign mind to that of the medium is for the most part of the nature of 'leakage', or 'seepage', where such selection as takes place is automatic, yet that is not always the case. Occasionally, and for a short time, we might suppose, the foreign mind is able deliberately to determine that the Receptor Stratum of the medium shall respond to a certain selected coherent cluster of its own ideas. Even that will not suffice, unless the ideas thus evoked in the Receptor Stratum are such that the phantasmogenic mechanism of the medium's mind can readily produce such images or quasi-sensory hallucinations as will symbolize them in a way intelligible to the control or to the sitter. It may be that this is a condition over which the communicating mind has no control, and which it must just leave to luck. But it is also conceivable that it might occasionally and for a short time be able deliberately to influence this also.


"Lectures on Psychical Research" by C. D. Broad (London: Routledge & Kegan Paul, 1962.)


More articles by C. D. Broad

The Relevance of Psychical Research to Philosophy

Henry Sidgwick and Psychical Research

Empirical Arguments for Survival
Human Personality and the Question of its Survival of Bodily Death
Normal Cognition, Clairvoyance and Telepathy
Summing Up the Case for Psychical Research

Home | About Us | Latest News | Biographies | Articles | Photographs | Theory | Online Library | Links | Recommended Books | Contact Us | Glossary | Search


Some parts of this page 2012