W. H. Salter

William Henry Salter

1880-1969. Went to Trinity College, Cambridge, with a Classical Scholarship in 1899, took a first class degree in 1901, turned to read Law, and was called to the Barin 1905. Joined the Society for Psychical Research in 1916, to become a member of its Council three years later. From 1920 to 1931, a very difficult financial period, he served as Honorary Treasurer; and from 1924 to 1948 he was Honorary Secretary. He was President from 1947 to 1948. He made many contributions to the SPR Journal and Proceedings, and published two admirable books, Ghosts and Apparitions (1938) and Zoar (1961).

From Zoar, or The Evidence for Psychical Research Concerning Survival
(1961, Sidgwick and Jackson, London).

Chapter 3: Apparitions

- W. H. Salter -

Contents | Previous Chapter | Next Chapter

          ONE OF the earliest and most persistent views as to what follows on the death of the body of flesh and blood is that the dead continue to exist in another body closely resembling in appearance the one that has died, that can make itself seen and heard by the living, that can occupy a position in space and move, but that is not subject to all the limitations of ordinary matter. Though it can touch the living, it cannot usually be grasped by them and it can pass through solid walls and closed doors. Various names, with slightly different shades of meaning, such as "astral", "etheric", "metethedal", have been applied to this conception. I shall use the inclusive, non-committal term "quasi-material".

Until recent times this opinion rested almost entirely on apparitions seen or heard at about the time of the death, or after it, of some person whom they in some way resembled, to put it shortly on ghosts, which are among the oldest of human experiences. The earliest accounts which have come down to us are not reported at first hand by the percipient, but are poetic fictions doubtless based on the popular belief of the time. The earlier the narrative the closer it comes to real experience, as established by modern enquiry. Take for comparison two ghost stories told by two great poets, the second writing more than two thousand years after the first, the appearance of Patroclus in the 23rd book of the Iliad, and that of the dead king in Hamlet.

The ghost of Patroclus is almost completely realistic, and could be paralleled from many cases reported to the SPR during the last generation. It resembles the dead man in height, feature, voice and, significantly, in the clothes worn. After it has spoken, at a length of 23 lines, the only departure from realism, it slips from Achilles' grasp like smoke, leaving no trace behind: all perfectly normal.

Shakespeare's account is a great deal more impressive, and for that reason less realistic. The dead king is seen several times, often but not always in the same place. Where two or more persons are present he is seen sometimes by all of them, sometimes by one only. He gives a long and detailed account of the manner of his death, stating facts unknown to his hearer, but later found to be true. For every separate point in the story, the recurrence in a specified place, the "collective" perception, the transmission of true information unknown to the hearer, a parallel instance, well or moderately well established, could be found. But the story taken as a whole is just the sort of thing that every psychical researcher would give untold time, trouble and money to investigate, if only he could get the chance, which he never does.

If however he should he so lucky, there are several questions he would have to put to Bernardo, Marcellus, Hamlet and the rest. "Collective" cases, i.e., where an apparition is seen by more than one person at the same time, being uncommon and the psychology of them obscure, he would wish to ask Bernardo and Marcellus, which of them first saw the figure, which first recognised it as the dead King? Did he say or do anything that might have prompted the other to see or recognise it? Were the glimpses of the moon sufficient to give them a clear view? How much Rhenish had they drained before going on duty? All these questions could be put, and answered, without leaving the domain of the "normal" as already defined, however extraordinary the affair was.

As to the King's death, and the ghost's account of it, how much of that long speech did Hamlet get down on his tablets? Is he sure that he had not already heard rumours or entertained suspicions as to how his father died? What assurance has he that the death really happened as stated by the ghost? On the last point Hamlet would doubtless refer to Claudius's reaction to "the Mousetrap" and if he could meet the other two points equally well, the case would call for a paranormal explanation. There psychical research would have to stop. Ordinary enquiry as to the facts could go no further.

"Be thou a spirit of health or goblin damned,
Bring with thee airs of heaven or blasts from hell?"

That is a question which could only be answered on the supernatural plane, and therefore not one that psychical research would attempt to put. The case is in fact a good illustration of how a single occurrence may raise normal, paranormal and supernatural questions, and of the boundary between what is and what is not the proper province of psychical research.

It is clear, from the way the story is told, that it would have been accepted by Shakespeare's contemporaries, though with some reservations on the part of the more sophisticated. But he wrote on the eve of the Age of Reason when it became the fashion to decry as the product of vulgar credulity any narrative that did not harmonise with current scientific theory, or by hook or crook to explain it away somehow.

As a good specimen of this process one may take the experience recorded sixty-two years after it happened by Lord Brougham in his Memoirs. According to these he was, as a young man of twenty-one, travelling in Sweden in December 1799. On reaching his inn after a long day's journey he took a hot bath and, while lying in it, he saw a friend G., a former fellow student at Edinburgh University, sitting on a chair, looking calmly at him.

"How I got out of the bath I know not, but on recovering my senses I found myself sprawling on the floor. The apparition, or whatever it was, that had taken the likeness of G., had disappeared."

When G. and he had both been students they had drawn up an agreement, written with their blood, that whichever died the first should appear to the other. G. had gone to India, and after a few years Brougham had almost forgotten him.

Brougham could not bring himself to talk of the vision even to the friend travelling with him. In 1862 he wrote:

"I have just been copying out from my journal the account of this strange dream: Certissima mortis imago! And now to finish the story begun sixty years since. Soon after my return to Edinburgh, there arrived a letter from India, announcing G.'s death, and stating that he had died on the 19th of December!

"Singular coincidence! Yet when one reflects on the vast numbers of dreams which night after night pass through our brains, the number of coincidences between the vision and the event are perhaps fewer and less remarkable than a fair calculation of chances would warrant us to expect."

(See Phantasms of the Living, vol. I pp. 395, 6)

Now whatever precisely Brougham's psychological state was when he saw G., he was certainly not having an ordinary dream. It is not a common incident of dream-life that makes a man get out of a bath and sprawl unconscious on the floor. The frequency of dreams, by which Brougham. seeks to reduce his experience to the commonplace and normal, is therefore quite beside the point. It could only be possibly relevant if no distinction were drawn between vague and incoherent dreams on the one hand, and precise, realistic dreams on the other. The talk about "a fair calculation" is mere bluff, since the materials for such a calculation would be impossible to obtain, and if in fact such a calculation had been made and had shown the coincidences to be fewer than chance-expectation, there would still be a problem calling for explanation. That a former Lord Chancellor, the leader of the "March of Mind", should fumble his argument in this way, shows the strength of the then prevailing bias against anything that might now be termed paranormal.

The story, if Brougham's narrative inspires more confidence than his reasoning, is a good example of a veridical crisis-apparition, that is to say that by some means other than any of the ordinary means of communication it transmits visually facts not normally known to the percipient relating to a crisis involving another person who is visually represented. This type of case is so central to the whole problem of apparitions as to impel me to attempt this formal definition of it. Instances of the type occur frequently in the literature, where the diligent reader may look for them. Having set out Brougham's case with some fullness, I shall discuss the type rather than separate examples of it, except where these present special distinctive features. The type includes cases where the experience is auditory or tactile instead of, or as well as, visual.

Before however leaving Brougham's case it is to be noted that, however good an example it may be of a type of experience, it is an extremely bad example of how experiences should be recorded. He has left no record of the experience other than one written sixty-two years later. According to that record, he noted the vision in his diary, which was so far good, but made no attempt to get the date confirmed by his travelling companion. There is no independent corroboration of the letter from India announcing G.'s death, nor are any particulars given as to just what it said or the date Brougham. learnt of it.

Defects such as these and the manifold errors due to them discredited stories of ghosts and of other occurrences that could not be forced within the framework of current scientific doctrine, and brought on them the stigma of "anecdotalism", which is still, absurdly enough, imputed to accounts dealing with similar subjects, however carefully they may have been recorded, and however critically investigated. For the persistent disparagement of them did not prevent these things happening, and happening to persons who could not be lightly brushed aside as incompetent or credulous witnesses. When therefore the SPR was founded in 1882, it was natural that its very wide programme of research should include an enquiry into a subject in which much material lay ready to hand in the form of already reported experiences, and more could easily be obtained. But as an essential first step it was necessary to look closely into the evidential weaknesses that had brought the traditional ghost story into disrepute, and to formulate a procedure for eliminating them.

There are many collections of ghost stories and other "spontaneous cases", but there are three publications of the SPR which by reason of the number of cases set out, and the care and skill with which they were verified and analysed, are of exceptional importance. The first is Phantasms of the Living (1886) of which Edmund Gurney was the main author, with Myers and Podmore as coadjutors. Although every effort was made to verify the cases set out in it, some of them were too old to be unreservedly accepted. The main importance of the book arises from the analysis of the different sources of error that had discredited cases of this kind, and the putting forward of a hypothesis that would cover the cases, or most of the cases, that survived critical examination.

Next in time is the Report on the Census of Hallucinations (1894, Proc. X), in which a much larger number of cases, and recent cases, were examined in accordance with the principles established by Phantasms, so that it became possible to determine more exactly how far chance-coincidence was a sufficient explanation of correspondences, like that in Brougham's case, between experience and event. Then, after nearly twenty years in which knowledge of paranormal faculties had been greatly extended, Mrs. Henry Sidgwick collected and classified all the cases which had been privately printed in the Society's Journal but not published anywhere (1923, Proc. XXXIII). The lapse of time between this and the earlier collections makes it possible to judge how far the progress of enquiry supports, or fails to support, types of experience sparsely or dubiously represented in them.

The main sources of error, as shown in Phantasms, are

(1) faulty observation, due to the percipient's emotional state, his carelessness, or the conditions in which the experience occurs, such as poor light;

(2) absence of satisfactory confirmation of the experience, either by a written note made at the time, or by a statement made to an independent witness;

(3) faults of memory, especially where there is no sufficient record in writing;

(4) failure to verify with care the event with which the experience is supposed to correspond.

The main rules to be followed to cure these defects are simple to state but not always easy to apply or enforce. First, the essential points of the experience must be stated and independently confirmed before the corresponding event is known to the percipient; second, there must be satisfactory evidence that at the time he stated his experience he had no normal knowledge of the event nor could have rationally inferred it, a matter on which one may sometimes have to rely on the percipient's word; third, the event must be verified to show that the correspondence is real.

The next point to be considered is whether, where a case is evidentially sound, or is only formally and superficially defective, the correspondence of experience and event can reasonably be assigned to chance-coincidence. An analysis by the authors of Phantasms of the 5,000 (approximately) waking experiences collected in response to a questionnaire, and of a similar number of dreams collected in the same way, suggested that something more than chance-coincidence was at work. The Report of the Census, with records of 17,000 waking experiences to analyse, came to the same conclusion. The Committee that drafted the Report pointed out that the question would only be settled by selecting a coincidence between two definite events and seeing how often it would occur by chance, and how often it actually does occur. The two events they selected were:

(1) visions of a recognised person seen by a waking percipient within twelve hours of the death of that person, the death being neither known to the percipient nor expected by him; with such visions were grouped auditory and tactile experiences of the same sort:

(2) the death of the recognised person.

The report sets out fully the elaborate statistical process that the analysis involved. It can be summarised by saying that correspondences between these two events were found to be 440 times as numerous as might have been expected if nothing but chance had to be taken into account.

Both in Phantasms and in the Census Report there are many interesting tables of statistics, showing e.g. the relative frequency of visual and auditory experiences, and of the realistic, semi-realistic and symbolical types; of the relations between agent and percipient: of the proportions of male and female percipients, and so on. The most important of these is the table in the Census Report that shows that about one in ten of the 17,000 persons asked whether, to put the question shortly, they had had a paranormal experience when fully awake, answered yes; that one in twenty of them had seen a realistic apparition; and one in thirty a realistic apparition of a recognised person.

Where however the question is whether an occurrence is fortuitous or not, statistics are of help only when the enquiry has been planned so as to lead up to quantitative assessment, as has been done in many recent experiments in ESP. In other lines of enquiry, however, such as those that relate to spontaneous experiences and sittings with mediums, statistics are of little value. Such material falls within certain general classes, from which some general principles can be deduced, but the details vary from instance to instance so as to defy quantitative appraisal, and the variation may be as significant as the instance's conformity with a general type.

But after making allowance for this, the figures of the Census Report may be accepted as showing the veridical death-coincidences are not fortuitous. What then is the cause of them? The traditional answer was that the dead person was locally present in a body having some, but not all, of the qualities of ordinary matter, including that of being perceived by the senses of the living. But this view presents many difficulties. In well-established cases, for instance, the apparition leaves no trace behind at the end of the experience.

Apparitions are usually clothed and are sometimes accompanied by visionary objects resembling those they used in their former life. Have all these things also quasi-material counterparts that continue their previous association with the quasi-material body of the dead?

There is no clear-cut division between apparitions and a large number of other non-fortuitous experiences to which the hypothesis of a quasi-material body cannot readily be applied. One may mention visions that are only in part realistic, as where a man in a hotel bedroom saw a portrait of his father occupying the frame of a picture at a time when his father was dying many hundreds of miles away; or symbolic visions, as where a mourning band "seen" round a top hat, but not in fact there, portended a death; or dreams, such as those collected in Phantasms; or experiences that are never "externalised" to any of the senses, intuitions that may be just as definite and veridical as a realistic apparition.

While none of these difficulties may be conclusive against the quasi-material hypothesis, which is in fact held by some serious students, it has been accepted by most psychical researchers that these experiences take place "in the mind's eye", but are none the less objective. as corresponding to something real but external to the percipient's normal knowledge or expectation.

The authors of Phantasms advanced the hypothesis that apparitions(1) were telepathic impressions transmitted by the person represented in them, called the "agent", to the percipient's mind and externalised as a visual hallucination. This last word, it may be desirable to emphasise, means simply the "apparent perception of external object not actually present" (Concise Oxford Dictionary), although often carelessly used to imply weakness of mind. This view would bring them into line with those semi-realistic, symbolic and intuitional experiences and dreams with which they share the quality of being veridical, and would get over the other difficulties mentioned with regard to the hypothesis of the local presence of a quasi-material body.

(1) For the sake of simplicity I omit reference to the auditory and tactile cases, which are in a general way parallel.

The hallucinatory view of apparitions was not new; the telepathic view of them was, hazardously so in fact, as at the date of Phantasms the experimental evidence for "thought-transference", to use the term then current, left much to be desired both as to quantity and quality. The position is now quite different, and the literature on experimental telepathy is voluminous. Some experiments have been conducted with "free" material, the agent choosing any one out of an unlimited number of "targets" for the percipient to aim at. The best known experiments of this type were those in which Gilbert Murray was the percipient, and the targets were incidents of real life, imaginary episodes, pictures, scenes from books. For reports of them see Proc. XXIX, 46-109; XXXIV, 2 12; Journal XXXII, 29.

In other experiments the number of targets is limited, e.g., five geometrical symbols. It is therefore possible to estimate precisely the probability that the results of a series of attempts at them were or were not due to chance. This is not possible with "free" material, and the method is accordingly better adapted to proving that telepathy is a real faculty, especially to proving it to people who rightly or wrongly distrust their ability to judge the evidence by their own commonsense. The best known investigators who have employed this method are J. B. Rhine in America and S. G. Soal in the United Kingdom, each of whom has recorded his results in several books.

If instances of telepathy are, as I believe them to be, common enough, the concept is so at variance with the generally accepted principles of science, that no methods of enquiry into it ought to be neglected. The quantitative method is the more conclusive as to the reality of telepathy, the qualitative the more informative as to how it works.

The conflict with general scientific opinion is due to the absence of any physical mechanism to account for the transmission from agent to percipient. It has indeed been held by some students that transmission is effected through "waves" of some kind. It is not fatal to this view that the "waves" cannot at present be specified, as new forms of radiation are constantly being discovered. But there are other, more fundamental, objections. No one has ever pointed to any organ in the human body capable of transmitting or receiving any sort of wave over more than trifling distances. Most serious of all, every normal mode of transmitting messages depends on some prearranged code understood by bath parties. If there were such a telepathic code it would have to be one capable of putting over complex ideas and elaborate mental images, as in experiments such as those with Gilbert Murray, or in some of the cases of crisis-apparitions. What then is the code? Who formulated it? How did it become intelligible as between agents and percipients having no normal knowledge of it?

Then there is the question of whether it is affected by the distance between agent and percipient. It is effective over great distances, as some of the spontaneous cases show, apparitions of persons dying in Australia being seen in England. But the relative effectiveness over long and short distances can only be tested by experiment. Comparison of long-distance and short-distance experimental results suggests that there is some reduction in effectiveness in the former, but there are too many variables involved to make the comparison conclusive. It cannot however be confidently asserted that telepathy infringes the "law of inverse squares".

The early conception of telepathy was that it was a one-way process between a single agent and a single percipient. This is much too simple to fit the material with which the psychical researcher now has to deal, Some spontaneous cases, for instance, are reciprocal, the agent being also a percipient, and the percipient an agent. Cases of this type were, when Phantasms appeared, so rare that the authors doubted their genuineness. The type was however well-established when Mrs. Sidgwick wrote her report in Proc. XXXIII. She cites, for instance, a reciprocal dream experience of two friends who had vivid dreams on the same night in which they each thought they stood in a dark wood, where the other also was: one of them shook a tree, the leaves of which turned into flame. Commenting on these cases, Mrs. Sidgwick wrote (p. 419):

"I think the kind of union of minds, the thinking and feeling together, here shown may be regarded as the type or norm of telepathic communication to which all other cases conform in varying degrees."

She added that it was in "collective" cases, in which several percipients shared the same experience, that this could perhaps be most clearly seen, and she called it transfusion rather than transmission of thought. Tyrrell in his Apparitions (1943, revised edition 1953) argued that even in cases that were neither "collective" nor reciprocal the dramatic presentation of the telepathic impulse implied some collaboration between agent and percipient.

Collaboration would be subconscious on the percipient's part, and in most, perhaps in all cases on the agent's. In some instances of crisis-apparitions a conscious desire to communicate with the percipient is shown by the agent's crying out the percipient's name. In other cases there is no such direct evidence. A desire to communicate may be inferred from the fact that a message is actually transmitted, but the impulse seems to have been purely subconscious, and even where some conscious desire is shown it may well have been made effective by subconscious activity.

This book will be largely occupied with accounts of the many and varied functions of the subconscious, but for the present it is sufficient to make it clear in what sense that word is used. It is desirable to give a name to all that region of the mind which lies outside of immediate awareness and beyond the reach of easy recall to conscious memory. The word "subliminal" was used by Myers and many others with much the same meaning, but had the disadvantage of confusion with "sublime" for those ignorant of its etymology, and of importing, for those better informed on this point, the metaphor of "threshold", which to English people has little relevance. It is more important to emphasise the distinction between "subconscious" as used in this book, and "unconscious" as used by the Freudians. That there is a region of the mind having the qualities they describe as distinctive of the "unconscious" is, I think, established, but these are not characteristic of the "subconscious" as a whole, except as regards conative activity, which is an important element in both conceptions.

The argument of Phantasms at one other point rested on slight evidence. It had to he shown that telepathy, if accepted as a real faculty, could produce realistic apparitions visible by a percipient unaware of the attempt. A few successful experiments are quoted in Phantasms, in only one of which was the agent's intention to project a vision of himself confirmed by independent testimony before the experiment was made. One of the cases was later found to be a hoax. A few fairly good cases were reported to the SPR in the latter years of the 19th century, but Mrs. Sidgwick writing in 1923 could find no reports later than 1900. The experimental evidence is good enough to exempt it from summary rejection, but hardly good enough to support so important a part of the argument. It is to be hoped that the further experiments now contemplated will throw more light on this problem, but if it is a correct view that in the spontaneous cases it is the agent's subconscious that is effective, then failure in consciously directed experiments would not further weaken the argument, since there is good reason to believe that the subconscious has powers exceeding those of the conscious mind.

The argument of Phantasms, notwithstanding this weakness, has been generally though not unanimously accepted by psychical researchers, so far at least as regards single apparitions seen by a single percipient, which are much the most common type of experience. I share without reserve the majority view. But there are other, rarer, types, to which the argument cannot without difficulty be made to extend, and these will be discussed in the next chapter.

Contents | Previous Chapter | Next Chapter



Contents | Chapter 1 | Chapter 2 | Chapter 3 | Chapter 4 | Chapter 5 | Chapter 6 | Chapter 7 | Chapter 8 | Chapter 9 | Chapter 10 | Chapter 11 | Chapter 12 | Chapter 13 | Chapter 14 | Chapter 15 | Chapter 16

Home | Intro | News | Investigators | Articles | Experiments | Photographs | Theory | Library | Info | Books | Contact | Campaigns | Glossary | Search


Some parts of this page The International Survivalist Society 2004