Robert Thouless

Dr. Robert H. Thouless

Educated at Corpus Christi College, Cambridge, where he was awarded his PhD in 1922. He then went on to become a Lecturer in Psychology at Manchester, in Glasgow, and again in Cambridge. Here he became the Reader in Educational Psychology. He was President of the British Association's Psychology Section in 1937, and published a number of books connected with this subject. President of the SPR from 1942 till 1944.

Anecdote and Experiment

 - Robert H. Thouless -

          THE BASIC materials from which psychical research starts are stories which people tell of extraordinary and unexpected things having happened to them: of dreams that have afterwards been fulfilled, of convictions that they know who has rung them up on the telephone before they have lifted it up, of having seen people, either living or dead, at places where they could not have been, and so on. Such stories do not form a very good basis for a scientific study of the paranormal since their value as evidence depends on the reliability of the people telling them and on the circumstances in which they are told, whether, for example, soon after or long after the event.

The weakness of such stories as evidence may he illustrated by one reported in the early days of psychical research. The following story, for example, is to be found in the first volume of the Proceedings of the Society for Psychical Research (Barrett et al., 1882-3):

On September 9th, 1848, at the siege of Mooltan, Major-General R   , C. B., then adjutant of his regiment, was most severely and dangerously wounded, and supposing himself dying, asked one of the officers with him to take the ring off his finger and send it to his wife, who, at the time, was fully 150 miles distant, at Ferozepore.

On the night of September 9th, 1848, I was lying on my bed, between sleeping and waking, when I distinctly saw my husband being carried off the field, seriously wounded, and heard his voice saying, 'Take this ring off my finger, and send it to my wife.' All the next day I could not get the sight or the voice out of my mind. In due time I heard of General R    having been severely wounded in the assault on Mooltan. He survived, however, and is still living.

It was not for some time after the siege that I heard from Colonel L   , the officer who helped to carry General R    off the field that the request as to the ring was actually made to him, just as I had heard it at Ferozepore at that very time.- M.A.R.

Certainly the incident described may have happened exactly as recorded, and it may have been a genuine case of thought-transference or some other mode of extra-sensory perception. To one convinced of the reality of ESP on other grounds, this seems not unlikely, particularly since the incident was one involving emotional stress which would seem to be favourable to the occurrence of ESP. As evidence of the reality of ESP, however, the story is without value.

The narrator of the story was the wife to whom it happened, so the story is at first-hand. If we ask when the account was recorded, the answer does not seem to be so satisfactory. It appears that, from being an adjutant, R    had been promoted to Major-General, which implies the lapse of many years. To be of value as evidence, it would be necessary that the incident should have been recorded before the confirming evidence was known to the percipient, otherwise memory may distort the incident to make it more striking and more evidential than it really was. Moreover, it is to be supposed that the officer's wife was often anxious about him when he was in action, and we do not know how often the anxiety led to semi-hallucinatory experiences of seeing him wounded; semi-hallucinatory experiences are not uncommon in the condition between sleeping and waking. Knowledge of how often Mrs R    had such experiences is necessary for a judgment as to how likely is the chance coincidence of the experience with the occurrence. That both the wounding and the experience took place on the same day seems striking, but it is unfortunate that the time of the experience was not independently recorded. In recollection of events in the past there is a tendency to remember them in a form which makes them fit into whatever theory one has formed to account for them, whether this theory is of the guilt of an accused person or of the paranormality of an event. This fact makes coincidence in time a somewhat doubtful piece of evidence unless it is corroborated by independent recording of the time of the event and that of the ostensibly paranormal experience. Mrs R   's narrative uses the term 'at that very time' which suggests a close coincidence of time but it seems unlikely that the attack took place during the night when Mrs R    was trying to go to sleep; this phrase may be merely an effect of the tendency to systematise a memory of the past.

The most striking piece of evidence is certainly the giving of the ring which is hardly likely to be an element fabricated by the uncertainties of memory. The sceptic might, however, ask whether it was a matter understood between husband and wife that the ring would be sent if he were seriously wounded. The sceptic might also ask whether the wife's experience of the ring having been sent had been recorded before she heard by normal means that it had been sent.

To ask such questions as these is not necessarily to adopt an attitude of disbelief to the story; this seems not unlikely to be true. They do, however, imply doubt as to the evidential value of the story. The conscientious psychical researcher must be very unwilling to admit this or similar stories as evidence.

It is easy to say what one would require of a spontaneous observation of an apparent case of thought-transference before accepting it as watertight evidence. The person who experienced it should have recorded it with a note of the time of the occurrence and delivered it to some reliable person before he had the opportunity of discovering whether it corresponded to any actual event. The person to whom the event occurred should also have made a record of it together with its time and delivered the account to some other person before hearing of the experience of thought-transference. The evidence would then be derived by a comparison of the two records.

It is also easy to see why few, if any, spontaneous cases fulfil these conditions. Those who claim to have experiences of thought-transference in their daily lives do not generally know which experiences are of this nature until they later learn of the event of which they appear to have had paranormal knowledge. None of us can be expected to record all the hunches we have and to give these records to someone else to take care of on the chance that they may turn out to be examples of some kind of paranormal knowledge. So the observational evidence we have is generally vitiated by the fact that the records were not made until after the verification of the supposed paranormal experience, with all the possibilities of distortion of memory which result from thinking and talking about the event afterwards.

How much the passage of time may distort the memory of an event supposed to be paranormal is shown in extreme form in a case published in the early years of the Society for Psychical Research(1). This was an account of an apparition seen by Sir Edmund Hornby, a judge of the consular court in China (Proc. S.P.R., 1884). Judge Hornby reported that on an occasion eight years earlier, he had seen a journalist in his bedroom in the early hours of the morning, that he had told the journalist the judgments he had decided to give next day, and that he had heard next morning that the journalist had died in his own house at about the time that Hornby had the experience of seeing him. This seemed to be a case of an apparition seen at about the time of death, but subsequent enquiry showed that nothing of the sort could have happened since the date of the journalist's death was months before Hornby reported having seen him and that no judgments had been given by Hornby on the actual day of the death Judge Hornby agreed that, if these facts turned out to be true, his memory must have played him the most extraordinary trick. So apparently it had.

(1) This story is only to be found in the first printing of vol. ii of Proc. S.P.R. On reprinting it was replaced by another case. This is unfortunate. One learns also by one's errors, and such learning is not helped by obliterating errors.

No doubt this is an extreme example of falsification in memory, but the fact that so much falsification can take place reinforces the case against supposing that knowledge about the paranormal can be much advanced by the mere collection of anecdotes about alleged paranormal events. This will not even provide convincing evidence of the reality of the paranormal, still less will it by itself give theoretical understanding of the nature of the paranormal.

From realisation of the fact that one cannot advance towards scientific understanding of the paranormal by the mere collection of anecdotes, there are two possible directions of advance. One is to look at the problems raised by the anecdotes in order that one may devise experiments in order to solve them. The other is to improve the conditions in which the anecdotes are obtained in the hope that they may become reliable sources of information. One can find out, for example, when the alleged incidents were recorded, what independent corroboration there is for them, improve conditions of recording by inviting people to send in records of what, for example, they imagine to be precognitive dreams, so that records are out of the sender's hands before their fulfilment, etc.

The first of these two methods of advance is the one the present book is about; it is the way that would be most likely to commend itself to anyone with a scientific training. It is a method which has been widely used in the sciences, and which has been conspicuously successful in providing understanding of how things happen. Thus people had seen thunderstorms for many years without understanding their nature; then physicists did experiments in which they produced sparks between electrically charged objects and laid bare the theoretical explanation of thunderstorms. It is true that the spark produced in the laboratory is a trivial event compared with a thunderstorm, but it is an event of the same kind, and, because its conditions can be observed and measured, it can be the source of a scientific understanding which can never be gained merely by looking at thunderstorms. In the same way, the experimenter in psychical research may try to replace such sporadic events as an officer being wounded in an attack by some trivial event not known to the percipient and then see whether it can be correctly reported by the percipient.

Critics of the use of the experimental method in psychical research often say that the trivial activities we ask experimental subjects to perform, such as guessing cards or reproducing hidden drawings, may not reveal paranormal capacities since these may only work in conditions of emotional stress between people united by a bond of affection. This is a criticism which seems less important to the practical research worker than to the arm-chair critic. If it were wholly true that ESP only takes place in situations of emotional stress, this would certainly be a grave difficulty in the way of devising a fruitful experiment in psychical research. If, as seems to be the case, it is only partly true, it creates a difficulty in successful experimenting which challenges us to find better and more fruitful ways of experimenting. It is a surprising fact that experiments in which people are asked to do trivial and uninteresting things like guessing cards do succeed much better than the above criticism would lead one to expect. It may well be that more people would show ESP capacities if they were tested by means of tasks in which real emotional forces were involved; such tasks are not easily adapted to experimental conditions. Meanwhile experimental tasks of the trivial type have given us a lot of information about the paranormal which will be discussed in later chapters.

Both refinement of observations and the use of experiments were methods explored in the early days of psychical research. For many reasons, however, the main weight of the activity of the early psychical researchers was in the direction of accumulating more and better records of spontaneous cases. One of the earliest S.P.R. publications was Phantasms of the Living which was mainly the work of Gurney (Gurney et al., 1886). This contains accounts of some experiments on telepathy but it is mainly a collection of anecdotes which had been tested as far as possible by personal interviewing of those reporting the experiences. It is implied that further progress in psychical research should be by accumulation of more stories with better attestation of the adequacy of the conditions of their reporting. Gurney's idea seemed to be that experiments were of value as proving the reality of telepathy, but that this was only a preliminary step towards the task of elucidating its nature by the examination of reports supplied by those who claim to have had paranormal experiences of various kinds. The later book on survival by F. W. H. Myers followed the same pattern (Myers, 1903).

Both of these books were magnificent collections of stories of the paranormal. The idea behind Phantasms of the Living of classifying and ordering the stories that people tell of the marvellous was a bold and original one for the execution of which later psychical researchers are heavily in debt to the authors. Yet to one trained in the methods of experimental science, it may seem that this work set psychical research on an unprofitable course, particularly in its grave underestimate of the importance of experiment as a method of advancing theoretical understanding.

In the experimental sciences, experiment is not merely a method of confirming what takes place; it is above all the method of testing and guiding theoretical advance. If a theoretical possibility is suggested by observation or by a preliminary experiment, the experimentalist asks: 'How can I devise an experiment to find out whether this is true or false?' This was not how most of the early psychical researchers thought, they accumulated anecdotes and based theories on them but did not take the further step of devising experiments to see whether the theories were true or false. They do not seem generally to have realised that the essential requirement of a fruitful scientific theory is that it should lead to observable consequences that can in principle be tested by experiment.

I do not think it is unfair to say of the early psychical researchers that they were not experimentally minded. This is hardly surprising when it is noticed that, intellectually distinguished as many of them were, few had had any training in scientific research while many of them had had no scientific training at all. It is true that Crookes was an experimental scientist of considerable distinction, and he showed a wholly experimental attitude towards the paranormal phenomena of Home (Crookes, 1874). But, by the time the S.P.R. started its work, his research interests were no longer in this field and he seems to have played no part in the early S.P.R. experimenting. Barrett was a Professor of Physics, but in Dublin, so he could not play any considerable part in the early experimenting. In the early days there was no one with research knowledge of the biological sciences, which would have been more closely relevant to the experimental problems of psychical research than was that of the physical sciences. It was, of course, too early for an experimental psychologist to have been available; the first laboratory of experimental psychology was opened in Leipzig in 1879, only three years before the founding of the S.P.R. It is unfortunate, however, that no experimental physiologist was in a leading position in the Society; perhaps they were unduly influenced by the unfavourable opinion of psychical research that had been expressed by the most influential physiologist of that time, Helmboltz.

It would, of course, be absurd to blame the early psychical researchers for their lack of experience of scientific experimentation. Rather one must admire the boldness and determination with which people whose training lay in classics or philosophy tackled this unfamiliar field. It does, however, suggest an explanation for some of the defects of these early experiments, both for their restricted aims and for the inflexibility of their design. The general tendency was to make large numbers of observations under identical conditions. This might be the best way of accumulating evidence of the reality of some paranormal phenomenon provided that the conditions were adequate to ensure that the paranormal process in question was the only possible explanation of the results obtained.

To discover something about the nature of the paranormal process concerned requires a different kind of experimental design with systematic variation of the experimental conditions. This the early experimenters neglected to do, so the experimental results were generally somewhat uninformative. It is, for example, surprising to discover that in the early experiments on thought-transference, there was little systematic attempt to interfere with various possible sensory channels of communication in order to see whether one of these was being unwittingly used. Also there was no attempt to test experimentally the hypothesis, that was taken for granted in all the early 'thought-transference' experiments, that the condition for a successful response by the experimental subject was the fact that someone else was thinking of what he had to do. Thus the authors of Phantasms of the Living, discussing some of Richet's experiments on guessing cards, remarked that the guiding condition which makes the percipient guess right 'could be nothing else than the fact that, prior to the guess being made, a person in the neighbourhood of the guesser had concentrated his attention on the card drawn' (Gurney et al., 1886). This, however, is by no means a necessary conclusion from the success of the experiments. It is a hypothesis to explain that success which may be right or wrong, and its rightness or wrongness could only be tested by comparing results when someone is concentrating his attention on the card to be drawn with those obtained when no one knows what card is drawn. It is a mark of the inflexibility of the early experiments that no test of this hypothesis was then made, and the question was not answered until it was experimentally tested by J. B. Rhine half a century later. Then it was found that the success of the percipient in guessing the right card did not depend on anyone else concentrating on it or even knowing it; he could succeed just as well if no one knew which was the right card (Rhine, 1934).

Although the early investigators had carried out experiments, psychical research was not at that time primarily an experimental science. The tendency was to regard the accumulation and validation of large numbers of reports of ostensibly paranormal events as the right way of finding out about psi, while experiments served the subordinate purpose of providing convincing evidence of the occurrence of telepathy or other psi-phenomena. A more completely experiment-oriented approach was set into motion in 1927 when the first university laboratory for the experimental study of the subject was started at Duke University in North Carolina under J. B. Rhine. From that time, parapsychology, like other branches of scientific research, used experiment as a means not merely of verifying the fact of psi but of finding out about its nature and properties.

One of the results of the emergence of experiment as a means of parapsychological investigation was a tendency to neglect such collections of reports of spontaneous cases as are to be found in Phantasms of the Living. In 1948, however, J. B. Rhine pointed out that such collections still had a useful role even for the experimentalist (Rhine, 1948). If they were no longer to be appealed to as proofs of the reality of the paranormal or as means of settling theoretical questions about its nature, they remain important as means of suggesting hypotheses which may afterwards be tested by suitably designed experiments. It is indeed important that the experimental psychical researcher should be well aware of what people report of their ostensibly paranormal experiences in order that his experimental researches may be in close touch with the problems of psi as they appear in the outside world and should not be a mere laboratory game. There are numerous examples of questions posed by reports of spontaneous psi experiences which a mere study of these reports cannot solve. Some of them, it must be admitted, cannot yet be solved by experimental methods either; they remain problems for future experimental solution. One may ask, for example, whether, in telepathic communication between two individuals, the active role is played by the sender or the receiver of the communicated information, whether it is a case of one of the individuals sending a message, or of the other individual getting hold of a piece of information in someone else's mind. Stories of spontaneous telepathic experiences raise this question; a study of a large collection of spontaneous cases may suggest an answer to it. But whether the answer is right or wrong can only be settled when someone devises and carries out an experiment which will give one result if one of these alternatives is right and another result if the other alternative is right. So far as I know, such an experiment has not yet been done. It is one of many challenges to the ingenuity and skill of future experimenters.

If this is accepted as the proper role of reports of spontaneous cases of ostensibly paranormal events, we need no longer attach to the question of rigid verification the importance that was properly attached to it by the earlier psychical researchers. If we are to draw theoretical conclusions from our collection of spontaneous cases, we must be very sure that none of them are distorted by failure of memory or by lying. If we are only going to use them to indicate possible conclusions that are to be tested afterwards by experiment, there is no reason for being so particular. If the stories are sometimes erroneous, the worst that will result from this will be that experimenters may be misled into trying an experiment that does not yield the expected result. This is a fairly common occurrence in experimental work anyway and it is not a serious one.

There is a further point to be considered about the question of rigid verification of records of spontaneous cases. If records of spontaneous cases are scrutinised carefully and all those are rejected which do not satisfy rigid criteria of adequate recording, that is, rejecting all those that were not recorded on the spot, or of reported apparitions that were not seen by at least two persons, etc., we shall, probably, cut out eighty per cent or more of the material reported. We shall then no longer have a complete sample of what is reported, and the very restricted sample we have may be distorted by the process of selection we have adopted. Apparitions seen by only one person may have different properties from those seen by more than one person; there may be differences in kind between the sort of paranormal event that is immediately recorded and the sort that is not. For the experimental psychical researcher who wants records of spontaneous cases to suggest possibilities and not to prove conclusions, it may be more important to have a representative sample than to have one that is free from errors. It is probably impossible, in any case, to ensure complete freedom from errors, and the effort to attain this unattainable ideal may result in a large number of genuine but poorly attested cases being rejected, with great loss to the experimentalist who wants the records to suggest as many problems as possible for experimental testing.

The first person to carry out the project of using records of stories of the paranormal in this way was Dr Louisa Rhine (1961). She started making a large collection of case records in the Duke University Parapsychology Laboratory with the idea that this material was to be used to provide suggestions for research but not as proof of anything. Supplementary validation was not required so long as material seemed to have been communicated in good faith by apparently sane individuals. They were intended as collections of what people report, not of what is known to happen.

In a series of articles in the Journal of Parapsychology Mrs Rhine has used her collection of cases in order to see what indications it gives for experimental testing. The first of these was in 1951, when she considered the proportion of cases in which the person having the paranormal experience was convinced of its paranormality, having a dream for example of some future event and being convinced by it that the event was going to take place (Rhine, 1951). Such conviction is common in spontaneous cases; it has been rarely checked so far in the laboratory. Mrs Rhine also ends that there are certain conditions in spontaneous cases under which conviction of paranormality is greater. These may suggest types of experiment in which such conviction can more easily be found. Here, as elsewhere, the supporting experimental study has not yet been made, but Mrs Rhine's work provides a rich field from which future experimentalists can extract problems for laboratory investigation.


BARRETT, W. F., GURNEY, E., and MYERS, F. W. H. (1882-3), 'First Report On Thought-Reading', Proceedings of the Society for Psychical Research, i, 13-34 (30-31 for paragraphs quoted).

CROOKES, W. (1874), "Researches in Spiritualism", London.

GURNEY, E, MYERS, F. W. H., and PODMORE, F. (1886), "Phantasms of the Living", London (republished abridged, New York, 1962).

MYERS, F. W. H. (1903), Human Personality and its Survival of Bodily Death, London (republished abridged, New York, 1961).

RHINE, J. B.. (1934), "Extra-Sensory Perception", Boston.

RHINE, J. B. (1948), 'The Value of Reports of Spontaneous Psi Experiences', Journal of Parapsychology xii, 23 1 -5.

RHINE, LOUISA E. (1951), 'Conviction and Associated Conditions in Spontaneous Cases', J. Parapsych., xy, 164-91. 

RHINE, LOUISA E. (1961), Hidden Channels of the Mind, New York. Proc. S.P.R. (1884) 'Fourth Report of the Literary Committee', ii, 180.


Related Material

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


Some parts The International Survivalist Society 2003