Uri Geller - a bibliography - homepage


John Taylor

Viking Press, 1975 - ISBN 0670684708

Experiments on Geller

Literally hundreds of “tests” have been conducted on Uri Geller – on theatre stages, before television cameras, in aeroplanes, taxis, hotel rooms, restaurants and innumerable other places. The vast proportion have been carried out by newspaper reporters, television personalities, and friends and acquaintances of Geller. Very few have been held under what might be called scientific conditions, or with scientists directly involved. Even when scientists have expressed interest in carrying out experiments with Geller, as they have done recently in Britain, it has proved extremely difficult to reach him, surrounded as he is by a protective entourage of intermediaries whose interest may lie in values other than scientific ones. And naturally enough, the commercial interest arouses one’s suspicion. In these circumstances it is a hard job for the scientist to obtain any data which will stand up to the scrutiny of his fellows. Since very few scientists have had access to Geller, the possibility of repeating experiments under carefully controlled conditions appears remote.

In fact, there have been two series of tests on Geller, which go part of the way towards satisfying the strict criteria needed to dispel our suspicions of fraud. The experimental conditions have been specified in enough detail to allow other researchers to duplicate them. But during these early experiments it did not prove possible to monitor and describe Geller’s psychological and physiological state. This is generally a difficult problem – how to delineate this state for any person is still not clear.

Puthoff and Russell Targ, two physicists trained in laser research. Most of the sessions were filmed or videotaped, so that any kind of mass hypnosis which might have occurred could be detected; no materials used in the tests were supplied by Geller. Some of the experiments were scrutinized by a magician on television monitors for possible sleight of hand procedures (none were in fact detected); the experiments were performed on what is termed a “double-blind” basis wherever possible, which means that neither Geller nor the experimenter could know the correct answer beforehand.

The first set of experiments took place over a period of six weeks in November and December of 1972, involving a variety of tests, some with astoundingly successful results. One of these was dice-guessing. A single dice was placed inside a closed metal box, which was then vigorously shaken by one of the experimenters and placed on a table. Nobody could see how the dice had fallen. Geller then tried to guess which face of the dice was upwards. He made eight guesses – and every time he was right. But he would not hazard two further tries as he did not feel confident that his perception was clear enough.

If we exclude these two refusals as true attempts, then the odds in favour of his guessing by chance the upper face of the dice eight times in succession are about one in a million. In assessing the probability, it seems justifiable not to take the other two throws into account since they were not even attempted. Only if Geller had tried and failed in these two attempts would they have to be counted.

The next test was to guess in which of ten identical aluminium film cans was hidden a certain object. After an assistant had put the object in one of them, the cans were placed in random positions on a table. The assistant then left the experimental area. Geller then entered the room with experimenters, who equally had no idea which can contained the object. Geller either passed his hands over the cans or simply looked at them. He then indicated which cans he thought were empty, and these were removed; when only two or three cans remained Geller stated which can he thought contained the object.

In this experiment Geller was successful each of the twelve times he made a guess. As in the previous dice-guessing test, he declined to guess on two occasions, having difficulty once with a metal ball-bearing wrapped in paper and once with a sugar cube. Both of these “failures” can be neglected for the same reason as before. He was successful in finding water (twice), steel ball-bearings (five times) and small magnets (five times). The odds of his twelve correct guesses being due to chance were again one in a million, a small number indeed.

Two experiments were next performed  to test Geller’s powers of psychokinesis. In the first of these, a precision balance had a one-gram weight placed on its pan before being covered with a glass jar. The balance generated an electrical output voltage proportional to the force applied to it, and the voltage was recorded on a chart in the form of a continuous strip. On several occasions Geller made the balance move, and a displacement was recorded on the strip chart. The movements were ten to a hundred times larger than could be produced by striking the table on which the balance stood, hitting the glass case covering the balance pan, or by jumping on the floor (none of which Geller, under close observation throughout, could have done). On one occasion the deflection on the chart was equivalent to a decrease in weight of the balance pan of about one and a half grammes; on another, to an increase in its weight of nearly a gramme; each happening for about one-fifth of a second.

The second main psychokinetic experiment was that of seeing if Geller had a magnetic field. He passed his hands near the probe of a Bell gaussmeter, an instrument which measures such a field. Geller was able to cause a full-scale deflection of the instrument a number of times (recorded on film) showing that he had an effective magnetic field at least half as strong as the earth’s. This effect may not, in fact, have been due to any magnetic field he himself possessed, but to some direct interaction he achieved with the measuring apparatus, such as affecting its electronic circuitry. The experimenters are said to have also watched the movement of iron filings lying on a sheet of paper when Geller’s hands were nearby; this does seem to mean that there was perhaps some magnetic field caused by Geller’s presence.

He briefly tried his hand at metal-bending, but because of his need to touch the metal the effect was not taken seriously. Nonetheless, the actual force required to bend some metal rings to the extent that Geller had distorted them was found to be the considerable amount of 150 pounds.

The remaining tests were all devoted to telepathy. Seven simple pictures were drawn on filing cards before Geller arrived at Stanford, and each was sealed in an envelope by an outside assistant. At the start of the experiment an envelope was selected, opened by the experimenters and the picture in it identified. The scientists then went into the experimental room and asked Geller to guess what was the drawing inside the envelope. He guessed with almost complete accuracy in each of the seven cases. Even more dramatic was Geller’s performance in the second set of experiments at Stanford Research Institute, lasting eight days in August, 1973. The drawings and their senders were all some distance from Geller during these tests, so that the possibility of his receiving any clues was zero. Furthermore, Geller was either kept in an electrically shielded room or else the pictures he was trying to guess were drawn on the East Coast, three thousand miles away, thereby making the surreptitious use of something like radio signal apparatus almost impossible. The drawings in the first experiment were obtained by choosing at random from a dictionary any noun that could be drawn. This was done after Geller had gone into the shielded room, the sender and drawing staying always outside. The first word chosen was “fuse” and the subject drawn was a firecracker. Geller said he saw “a cylinder with a noise coming out of it” and his drawing looked like a drum with a number of other cylindrical objects. The second word chosen was “bunch”, and so a bunch of grapes was drawn. Geller said he saw “drops of water coming out of the picture”, then mentioned “purple circles”. He then drew a bunch of grapes, exactly the same in number, twenty-four, as in the original drawing and of nearly identical shape.

Geller and Dr Puthoff were then locked up in a shielded room together, half a mile away from the experimenter’s office. The next drawing was of a devil with a trident; Geller found this difficult, and produced three drawings, composed of Moses’ tablets with the Ten Commandments inside the earth, with a trident outside; of an apple with a worm coming out of it (with a snake also in the picture); and finally one which was a mixture of the previous two with God inside. The difficulty Geller had in drawing the devil may well have been cultural.

One of the experimenters was then incarcerated in the shielded room, while Geller stayed outside with the other experimenter. The picture Geller had to guess was of a solar system; he drew a picture which was a very close resemblance, with Saturn and it’s rings and a central glowing sun. The next task was to guess, while outside the shielded room, a picture placed inside it drawn earlier by a scientist who did not belong to the experimental group nor was present during the test. The power being tested here was that of clairvoyance, of being able to “see” what would be to normal senses a completely invisible object; the lack of involvement of the person who had drawn the picture ruled out the possibility of telepathy – i.e. of any thought transference. Geller was unable to succeed. Two further tests were tried with Geller’s brain waves being monitored whilst he was in the shielded room, and he failed again. His next two attempts were only moderately successful, but the third was totally so. The original drawing was of a seagull in flight; he said he saw a swan flying over a hill and drew several birds, one almost identical with the original. He was also very close to the picture of a kite drawn by a computer on a television screen; he was about 150 feet way from the original, which was in a different room. He was relatively successful in guessing the picture of an arrow through a heart stored in the computer memory, though he failed in another such case. This was not a clear-cut test of clairvoyance, however, since there were several people in the computer room who knew the nature of the stored target. There were two attempts at long-distance communication, one resulting in some similarity, but the other not so close to the mark. To summarize, out of fifteen drawings which Geller tried to guess, he was undeniably successful in seven of them, to some extent in four others, and made no attempt with the remaining four. This is not a bad score. We should expect that the mere chance of guessing what had been drawn by the experimenters in most of these cases would be almost zero. The bunch of twenty-four grapes is here a very good example. One way of assessing this is to consider the original drawing as consisting of a set of seven rows of grapes, containing, 1,3,3,4,6, 4 and 3 grapes respectively (starting at the bottom). Geller’sbunch of grapes has corresponding numbers 1,3,4,4,5,4,3, the third row having one too many, the fifth row one too few. If we assume that these seven figures were guessed at random, each being chosen from the digits 0 to 9, the odds for Geller’s choice being identical with the original drawing in five of the numbers by chance are one in one hundred thousand. This calculation overlooks the fact that the other two numbers, which were wrong, were also very close. Nor does it take into account that Geller guessed the number of rows correctly; this introduces another factor of at least ten in the odds against chance (since it is unlikely that there would be more than about ten rows in a bunch of grapes). But then it is necessary to include the chance of it being a bunch of grapes in the first place – “purple circles” as Geller described them. There could have been at least a hundred objects to chose from, if not more, and that they had to be in a bunch and not in a line or other shape is also improbable by a factor of at least ten. Taken all in all, the odds for Geller guessing the details of this picture by chance alone are at least as low as one in a thousand million, and this is definitely an underestimation: a remarkable result, and one only probable in terms of some form of communication.

The major portion of the research with Geller on remote perception of graphic material appeared in Nature, one of the world’s most respected scientific journals, in October 1974, along with similar results on other subjects. There was an accompanying editorial comment which criticized the report as “weak in design and presentation,” giving “uncomfortably vague” details of various safeguards and precautions introduced against the possibility of conscious or unconscious fraud on the part of one or other of the subjects, and “not concentrating in detail and with meticulous care on one particular approach to extrasensory phenomena.” At best, it was concluded, the tests were more “a series of pilot studies … than a report of a completed experiment”. In spite of these considerable criticisms the paper was published by Nature, partly to put the results “in more reasonable (p.54) perspective” than the considerable advance publicity attending the paper had given it, and also to allow other scientists to gauge the quality of the particular research. Criticism was much heavier from other areas, highlighting supposed inadequacies in the experimental procedures which could have allowed deception to occur. Some of these criticisms have since been found to be invalid. In particular, the supposition that some sort of electromagnetic device was secreted by Geller in his teeth or elsewhere in his body has been ruled out, because it would have required collusion on the part of some of the experimenters, as well as being impossible to achieve by present technology. Nor could the dice used have contained small radio transmitters signalling which side was up, unless they had been put there by the experimenters themselves.

Targ and Puthoff themselves summed up their results as follows: “As a result of Geller’s success in this experimental period, we consider that he has demonstrated his paranormal ability in a convincing and unambiguous manner.” From the calculation made above for the bunch of grapes, this conclusion would seem to be fully justified. There appears to be no chance of fraud in the experiment, unless there wasgross collusion between all concerned. If such a charge is to be made, then there must be good backing for it. From other evidence presented in the first chapter, together with the filmed evidence described earlier, there seem no grounds at all for suspecting fraudulent behaviour. Nor could the collusion have only been on the part of the three principal characters; other people were involved in the various experiments. This, allied with the disinterested character of the investigation, renders any such charges unfounded and we need not consider them further. The results obtained at the Stanford Research Institute are in some ways very difficult to accept.

The scientific tests which Geller was subjected to at Stanford did not probe deeply into metal-bending, owing to the possibility that he might have been causing it by physical pressure. To avoid any likelihood of this it is necessary to do one of two things. One the one hand, tests must be carried out in which such force as (p.58) is applied during the bending is measured while the experiment is in progress. If less force were used by Geller during a bending session than would be needed to cause the metal to bend by the amount it actually had, then some unknown, non-mechanical force would have been responsible and the Geller effect authenticated. This approach is at present under way, but requires delicate measuring devices embedded in pieces of metal to register the amount of force actually applied. A more direct test, which precludes the question of the mechanical force applied during bending, is to require that metal be bent without Geller actually touching it. If he can achieve that, then the Geller effect is indeed beyond current scientific understanding. Geller did not achieve any well-authenticated bending at a distance during his time at Stanford, but on 2 February 1974, during one of his visits to England, a successful test of this kind was carried out with Geller by myself.Pieces of metal (aluminium and copper), strips of various types of plastic, some single crystals of potassium bromide which were long enough to be stroked like pieces of cutlery, various wire mesh tubes and a sealed glass tube containing a strip of aluminium – all these were used to try out Geller’s powers. In addition, a small but sensitive Geiger counter – to detect radioactivity – and a primitive detector of ultra-violet radiation were included in the apparatus.

The various strips of metal and plastic and the sealed glass tube were laid a few inches apart on a metal sheet. A strip of aluminium, placed inside a wire mesh tube with its end firmly fixed, was also laid out. The objects had been prepared in the metallurgical department of King’s College, London, and there was no chance of Geller having been in contact with them before the experiment. Two of my colleagues were also in the room with Geller, acting as observers. The various strips of metal and plastic were carefully scrutinized at the beginning of the test to confirm that they were straight. First of all, Geller tried to bend a metal rod without (p.59) touching it, but he did not succeed. It was then observed that one of the aluminium strips lying on the tray was now bent, without, as far as could be seen, having been touched either by Geller or by anyone else in the room.

To see whether Geller could repeat the metal-bending feat of the Dimbleby programme, he was then handed a teaspoon which had been brought along with the other materials. I held the bowl end while Geller stroked it gently with one hand. After about twenty seconds the thinnest part of the stem suddenly became soft for a length of approximately half a centimetre and then the spoon broke in two. The ends very rapidly hardened up again – in less than a second. There was also, as far as could be determined by touch, a complete absence of heat at the fracture. This sequence of sudden softening and complete loss of cohesive strength, breakage, and then rapid hardening was almost identical to that observed when the fork broke during the Dimbleby programme. Here, under laboratory conditions, we had been able to repeat this remarkable experiment. Geller could simply not have surreptitiously applied enough pressure to have brought this about, not to mention the pre-breakage softening of the metal. Nor could the teaspoon have been tampered with – it had been in my own possession for the past year.

Then Geller gently stroked a single crystal of potassium bromide about two centimetres long, and it split into two pieces within ten seconds. It was difficult to assess the force that had actually been applied to the crystal, but subsequent tests have shown that such crystals cannot be broken by gentle stroking alone. To show, of course, that pressure was not the cause of the crystals braking and that this was a paranormal effect, it would be necessary to measure how much pressure Geller had in fact applied.  Geller also stroked a thin wooden strip with no result. When he held his hands above a blue plastic strip this became discoloured. Such discolouration is normal when bending these plastic strips, though Geller was not in fact able to bend it without touching it. (p.60) After this series of tests the objects on the tray were re-examined. It was found that the last five centimetres at one end of the aluminium strip in the closed wire mesh tube was now bent with a radius of curvature of about five centimetres. One should bear in mind that Geller was continually under the scrutiny of the two observers. He could not unseen have opened the sealed tube containing the aluminium strip and interfered with it. Indeed, he was occupied trying to bend other objects at that time. Furthermore, there was no evidence of any tampering with the end of the tube.

Afterword, autumn 1975

Since he started off so much of the controversy, it is appropriate to commence with Uri Geller himself. Concern has been expressed, especially by conjurers, that Geller achieved his results by fraud. Various professional magicians have claimed that they can reproduce the spoon-bending and watch-starting and-stopping phenomena. One, James Randi, even convinced the editorial staff of the British parapsychologicaljournal Psychic Researcher of his paranormal powers, as well as some of my colleagues at King’s College in London (I was not present at the time). He at no time claimed that he used paranormal methods to deceive, and has recently published a book describing how Geller managed his tricks.

Similar explanations of Geller’s powers have also been published by the British popular science weekly, New Scientist, giving details as to how the frauds must have been achieved. Since Randi’s book is not yet available I can only comment on the New (p.178) Scientist discussion. It concentrated on the telepathy tests carried out at Stanford Research Institute and on the metal-bending tests at Birkbeck College, London University. In the former, various miniature radio transmitters and receivers were claimed to have been used to send the necessary information to Geller from an informant. There is the technical difficulty of the size of such devices, but assuming that they could have been so small as to be inserted, as it was claimed, in Geller’s teeth (he has no cavities, nor ever visited a dentist, he recently told me) there is the major problem that the informant in certain cases would have had to be one of the scientific team involved in the test. Thus the charge of fraud could only stick if it included the scientists, and so has to be discounted unless all such evidence in this area is rejected. Even so, care must be taken here; no guarantee of impunity for cheating scientists can rest in the repeatability of their tests by other independent groups. The metal-bending tests at Birkbeck were criticized as being poorly controlled with too much confusion. But later tests appear to have avoided that difficulty, especially in the breaking of a crystal without direct contact by Geller, when professor John Hasted claimed Geller could not have touched it at all.

Repeatability of metal-bending has occurred with Geller, at least at a certain level. Two tests have been reported on Geller, which seem to have satisfied the people concerned, both with conjurer’s training. One of these, W.E. Cox, reported tests of both key-bending and watch-starting. Two keys were bent by Geller, one of which was held down on a flat surface by Cox. In the other case, “Geller had been handed this key, like the first, but returned it to me in a perfectly straight condition. My forefinger pressed against the toothed end. Geller stroked only an inch of the handle, and this time a bend slowly appeared near that end, a full inch away from my finger. It conspicuously continued until it reached 36 degrees. Again, there was no noticeable pressure upwards against my finger, and the time required was less than a minute.” Geller also started a watch which had been (p.179) specially prepared. Cox’s conclusion was that fraud could be ruled out, nor had the fellow magicians he had consulted on this disagreed with him.

The other positive test was conducted by the magician Arthur Zorka and one of his colleagues; in it Geller caused the nylon reinforced handle of a forged steel fork to explode when he touched it, and also accurately guessed several drawings made by the magicians or even only thought by them. They concluded "…there is no way, based on our present collective knowledge, that any method of trickery could have been used to produce these effects under the conditions to which Uri Geller was subjected."

Naturally enough, the subject himself has something to add on this in his book My Own Story. It makes interesting reading, especially since the straightforward attitude presented, consistent with that Geller has shown in numerous media interviews, adds weight to the suggestion that he uses no deception.

The case for spoon-bending has been lent weight by the discovery of others, especially children, who have come forward claiming powers similar to Geller’s. Here again deception occurs, though it can be spotted more readily than in the case of a practiced magician. In my own investigations I have seen children who used methods which were somewhat suspicious, and Geller himself remarked on this to me. I have not continued working with cases of that sort as soon as the possibility of fraud has occurred.Recently two accounts of such deception have been reported, one by the New Scientist and the other by scientists from Bath University.

The New Scientist account was on the David Berglas/Daily Express spoon-bending competition (with a £5000 prize given by Berglas). Several of the children were seen bending the spoons mechanically. Thus “When attention was diverted to a child at one end of the room, a boy at the other end glanced around the room; confident that no one was watching he used two hands and all his strength to bend a fork.” The Bath investigation (p.180) used one-way mirrors to observe a group of six children. Of these, five were caught clearly cheating; the sixth was unable to achieve anything.

The difficulty with the Bath tests was that no clear attempt was made to exclude fraud; thus the use of deception seemed to have been positively encouraged. For example, in the Bath tests “the observers in the room were instructed to deliberately relax their vigilance at intervals after the first twenty minutes.” This also seems to have been the case with the Berglas/Express tests, for, as it was said, “it was not possible to watch all the children constantly.” Even so, it was felt “Some bent cutlery in a way not readily explicable by the observers.” This feature of childhood deception is clearly one to reckon with most carefully. As I said earlier, I had already experienced some of this, yet the evidence presented earlier in the book was taken in a way to guard against any such deception by child or adult. Tests were done either with the pressure being measured or contact being excluded completely by metal being placed in sealed containers.

This article contains copyrighted material that has not been specifically authorized by the copyright owner. It is being made available for the general purpose of criticism, comment, news reporting, teaching and/or research. I believe that use of this material is covered under the terms of "fair use". If you wish to use this copyrighted material for purposes other than that provided by law, you must obtain permission from the copyright owner.

Uri Geller - a bibliography - homepage