Part 1: The First Two Years

By James Randi

When it was announced in 1979 that noted engineer James S. McDonnell, board chairman of McDonnell-Douglas Aircraft and devotee of the paranormal, had awarded a $500,000 grant to Washington University in St. Louis, Missouri, for the establishment of the McDonnell Laboratory for Psychical Research, it seemed the ideal opportunity to initiate an experiment I had contemplated for some time. It was designed to test two

major hypotheses.

Parapsychologists have been lamenting for decades that they are unable to conduct proper research due to the lack of adequate funding, but I felt strongly that the problem lay in their strong pro-psychic bias.

 

The first hypothesis, therefore, was that no amount of financial support would remove that impediment to improvement in the quality of their work. Moreover, I have always been in accord with many others in the field – such as Stanley Krippner, current president of the Parapsychological Association – who insisted that qualified, experienced conjurors were essential for design, implementation, and evaluation of experiments in parapsychology, especially where deception – involuntary or deliberate – by subjects or experimenters, might be possible. So the second hypothesis was that parapsychologists would resist accepting expert conjuring assistance in designing proper control procedures and, as a result, would fail to detect various kinds of simple magic tricks. U.K. Parapsychologist Trevor Finch had even directly suggested that skeptics try to introduce a conjuror into a lab disguised as a psychic.
Certainly my plan seemed to be in accordance with the expressed needs of the parapsychological community.

 

The director of he McDonnell Lab was physics professor Peter R. Phillips, who had a decade of interest in parapsychology behind him. In the press he declared that the lab intended to investigate psychokinetic metal bending (PKMB) by children. Accordingly, I asked two young conjurors who had been in touch with me by mail, and had expressed an interest in my work as a skeptic, to write the McDonnell lab claiming psychic powers. Our experiment was to be code-named “Project Alpha.”

 

We learned that the lab had considered some 300 applicants who contacted them in response to notices in the media. Both my colleagues Steve “Banachek” Shaw (herein referred to as his stage name Banachek), an English immigrant hospital employee in Washington, PA, and part time mentalist, and Michael Edwards, a student in Marion Iowa, and well-known there as a magician, were the only McDonnell lab subjects chosen from that rather large group of applicants. They were 18 and 17 years old, respectively when they began the project.

We had established well in advance of the beginning of Project Alpha that at a suitable date we would reveal the deception. Also, the subjects agreed that, if they were ever asked directly by an experimenter if they were using tricks, they would immediately answer, “Yes, and we were sent here by James Randi.” They would then answer any and all questions concerning their involvement.

 

Even before the boys were tested at the lab, I sent Phillips a list of eleven “Caveats” concerning tests done with human subjects. For example, I warned him not to allow the subjects to run the experiments by changing the protocol. Similarly, I suggested capricious demands by subjects might well be the means of introducing conditions that would permit subterfuge. He was warned that reports of conditions should be very precise, assuming nothing. Above all, I urged that a conjuror be present. To that end, I offered to attend the McDonnell lab tests at my own expense, without any requirements that I be credited with any participation, or even attendance, in subsequent reports.

 

From the very beginning, the researchers ignored the rules I had suggested. As in other investigations, the “gifted subjects” took over running the experiments. They threw minor tantrums (inspired by similar events reported to have taken place at the Stanford Research Institute when Uri Geller was examined there in the 1970’s) whenever conditions were not to their liking.

 

Though I had specifically warned Phillips against allowing more than one test object (spoon or key, for example) to be placed before a subject during tests, the lab table was habitually littered with objects. The specimens were not permanently marked, but instead bore paper tags attached with string loops. Banachek and Edwards found it easy to switch tags after the objects had been accurately measured, thus producing the illusion that an object handled in the most casual fashion had undergone a deformation.

 

During one type of telepathy test, a subject would be given a sealed envelope containing a picture drawn from a target pool. Left alone with the envelope, the subject would subsequently surrender the envelope to the experimenter, who would examine it for signs of tampering. The subject would then announce his selection for the target pool. This series of tests was quite successful – though not overly so, because the boys realized that 100 percent might be suggestive of trickery. They purposely minimized their success. The method was easy. Since the envelope was “sealed” only with a few staples, they removed them, peeked, and then replaced the staples through the original holes! In one case, Michael lost two staples, and to cover this he opted to open the envelope himself upon confronting the experimenter. The breach of protocol was accepted. The subject had been allowed to shape the experiment.

 

In other ESP tests, significant results were obtained only when on of the subjects was aware of the target drawing and was allowed to watch a TV monitor while the other tried to duplicate the drawing. The laboratory investigators decided, in their official report on the tests, that communication between the two by any means other than telepathy had been ruled out, since “though it might seem suspicious that the most significant scores were obtained under just that condition which might have permitted collusion . . . we feel that any hypothesis of normal communication is very unlikely: even the best of our hits are not consistent with verbal cueing, but rather exhibit consistent resemblances of form without any semantic relation.”

 

What the experimenters could have been told, if they had been willing to listen, was that the best of conjurors’ “mind-reading” tricks are accomplished by a “hot-and-cold” system of communication having nothing to do with actual verbalization. Results obtained therefore appear much more striking in nature, and seem to be what might have been achieved as a result of a “telepathic” transference. No amount of acting can simulate the actual difficulty experienced by the operators of such a system.

Though no communication took place during these tests, the lack of “consistent” hits referred to in the quote above would have provided just the required conditions for acceptance, had the experiment been successful. As it was, another common conjuror’s ploy was utilized; giving an edge to the results that caused some excitement among the scientists. That “edge” would be eliminated, however, by proper double blind evaluation of the data.

 

One rather naïve experiment, conducted with Banachek, involved a small slab of clear acrylic plastic in which a shallow groove had been cut. Into this groove was placed a thin (about 1/16”) metal rod a few inches long that fit loosely, flush with the surface. It was believed, and so stated that it was not possible to remove the rod from the groove by hand without either overturning the slab or using a tool of some sort. Banachek was asked to stroke the metal with his finger and cause it to bend. He quickly discovered that the rod tilted up and out of the groove when he pressed down on one end, the flesh of his finger having squeezed into the groove. He simply removed the rod unnoticed, bent it slightly, and re-inserted it into the groove, lying it on its side, since the groove was wide enough to accommodate the bend. Then he stroked and rotated the rod 90 degrees to make it appear to bend up and out of the groove. The feat was deemed impossible by trickery.

 

In another run of experiments, involving an electronic setup, the boys were asked to shorten the capacity of tubular electric fuses. A current would be passed through a fuse and gradually increased until the fuse blew out, and that value would be recorded. They obtained excellent results in this test, seeming to cause premature rupture of the fuses through mental influence. The reason for their success was simply that they were allowed to handle the fuses freely. They were able to re-insert the same blown fuse repeatedly! The circuit had been designed so that if a fuse were “open” – already blown – the instruments did not reveal the open circuit until the current was somewhat advanced. Thus it seemed that the fuse had blown quite prematurely. Also, the boys found that by pressing down upon one end of the fuse in its holder, or merely by touching it briefly while passing a hand over it, the instruments recorded unusual results that were interpreted by the experimenters as marvelous effects.

 

Banachek suggested that he might be able to affect a video camera and while seated before one, staring into the lens, he gestured mysteriously over the instrument. The picture twice suddenly “bloomed” brightly, the image swelling and subsiding. This was recorded and subsequently shown in the official McDonnell lab film. It seemed to the researchers that this event was not possible by any but paranormal means: yet Banachek had simply reached forward and turned the “target” control on the side of the camera – twice – unseen by the lab personnel.

 

The McDonnell Lab videotape showed the subjects causing a light-weight paper rotor perched atop a needle-point to turn – in either direction, at their will – while mounted inside what was called a “bell jar.” The terminology was misleading indeed. A bell jar is a piece of laboratory equipment designed to be hermetically sealed to a base-plate, and usually capable of being evacuated of air. The actual item used was a glass dome, the cover of a cheap clock, placed upon a wooden base with a slot to receive the edge of the dome. A layer of aluminum foil that settled in the recess made a further seal.

The boys demonstrated that a static-charged comb was not able to cause a deflection of the rotor because of a special anti-static coating sprayed upon the dome. But when they replaced the dome, it was an easy matter to drop a small ball of aluminum foil into the slot, causing the dome to tilt slightly and provide a gap. By simply blowing surreptitiously at the base of the dome, the boys could cause air to circulate within and thus turn the balanced rotor. Changing body position and blowing from a different angle changed the direction of air movement, and thus the direction of rotor movement. Since the small ball of foil matched the layer of foil, it was invisible and could easily be removed after the experiment – especially since the boys were allowed to handle everything freely.

 

One device developed at the laboratory for testing the Alpha subjects consisted of an overturned aquarium bolted and padlocked to a stout table. Objects would be put inside and left overnight. Since the locks on the doors were of excellent quality, and Phillips wore the padlock and door keys around his neck, security was thought to be absolute. It was not. Banachek and Edwards simply left a window unlocked, and returned to the premises at night. There were several ways to open the sealed aquarium, and they were free to do anything they pleased with the contents, which were discovered in the morning by lab personnel to have been bent, twisted, broken, and moved about by mysterious paranormal forces.

 

A part of the aquarium test used a shallow box in which dry coffee-grounds were spread in a thin layer. Small cubes and other objects were placed therein, and were found to have spelled out strange cabalistic symbols when examined in the morning. This evoked much wonder among the investigators.

 

Later in the progress of Project Alpha, the amateur magician who originated the sealed-aquarium system, and who even now proclaims himself of being flimflammed in his specialty of designing untamperable sealed containers, tried to improve upon the aquarium by providing inviolable bottles. As we shall see, he failed even more spectacularly.

There is no question that the lab personnel believed that the boys were psychic. They did. It was this belief that made the deception exceedingly easy, and it was clear that, had the two entered the arena as conjurors, they could never have gotten away with all they did. The lab personnel further crippled themselves by referring to the kids as “gifted subjects,” even inventing the term Psychokinete to apply to them. Simple tricks, performed under very informal conditions of control, were declared PK events, and careless descriptions of circumstances surrounding the performances were written up. These factors certainly added to the sympathetic atmosphere in which the subjects were operating.

 

Another factor that led researchers down the garden path was their total, unquestioning acceptance of, and the belief in, the work of their fellows in the field. Even the most doubtful results, seriously questioned and in some cases thoroughly denounced by colleagues, were embraced by the investigators when it matched their needs. It is apparent that many parapsychological investigators never do house-cleaning to get rid of the obvious trash, and the clutter that results makes it impossible to obtain a clear picture of just what the their problem is.

 

Any minor remark of claim made by the subjects that seemed to fit an outside researcher’s notion of reality or fulfilled some expectation was further evidence to the laboratory investigators that they were dealing with the real thing. For example, Banachek and Mike complained about electronic equipment putting out “bad vibes,” not only to satisfy this established bit of mythology, but also to minimize proper video observation. Also, they were careful to mention that in early childhood both had experienced electric shocks, after which they had become aware of their psychic powers. Though not useable as strict evidence, acceptance of these tidbits further deepened the quicksand into which the researchers continued to sink.

 

All through the three-year period that Banachek and Mike were at the McDonnell lab, I continued to write Professor Phillips offering to attend experimental sessions as a consultant. Phillips seemed quite confident that he could not be deceived, however, and did not accept my offer. Then, in July of 1981, I “leaked” broad hints of Project Alpha at a magician’s conference in Pittsburgh. Eleven days later, I heard that some rumors had reached the McDonnell lab. This had been done in an attempt to alert the parapsychologists. Instead, the rumors were reported to Banachek and Mike at the lab as great jokes. They had not asked if there was any truth to them.

Just previous to this event, Phillips had for the first time actually written to me for assistance. He asked if I would be prepared to supply him with a videotape of fake PKMB being performed, along with a revelation of how it had been done. He intended to show it at the forthcoming August meeting of the Parapsychological Association in Syracuse. I immediately agreed to do so, and within a few days and I had a number of performances from my videotape library in which I was shown bending and breaking keys and spoons as well as doing some convincing “ESP” tricks. I supplied two sound tracks, one the original and the other a running commentary describing in detail the method I used. I threw in , for educational purposes, an episode with Uri Geller in which he is seen to use exactly the same method of Spoon-bending performance, and is caught on tape doing so.

 

I felt that rumors of Alpha would reach Phillips at about the same time he had my videotape and that he would be able to examine both his evidence and mine in light of the possibility that the collusion rumor was true. In return for my participation, I asked Phillips if I might have a copy of the McDonnell lab videotape of the Alpha subjects that had been prepared for showing along with my tape at the upcoming PA meeting. He agreed to do so: and, just days before the convention, I received his tape. I drew up a detailed analysis of the tricks shown there, pointing out that positively unmistakable evidence of deception was contained on their tape.

 

At the convention, Phillips showed my tape and his own. An active rumor began circulating that Phillips and I were working together to discredit the PA, and it was widely believed. It was no surprise that the parapsychologists with little enthusiasm received his announced findings – though some of them, Walter Uphoff and William Cox in particular, were ecstatic. Cox, never one to entertain any doubts, had written Phillips a month earlier objecting strenuously to his intention of showing the videotape I had prepared. He apparently felt that it would not be good to introduce any doubts whatsoever into the proceedings.

 

A formal report on the two subjects, prepared by the McDonnell lab and distributed at the convention, was hastily recalled, and modifiers (“apparently,” “seemingly,” and “ostensibly”) were inserted at the appropriate points in the text. It was reprinted and once again distributed. In somewhat a state of shock, Phillips was cornered by me after the workshop, and I insisted upon showing him and Mark Shafer, his principal researcher, where the tape showed evidence of fraud. Visibly shaken, the two thanked me for my efforts, and I parted from them reasonably sure that they had been impressed enough to change their ways.

 

Upon my return from the convention, I contacted Banachek and Edwards, and informed them that Phillips was now very suspicious, and that Project Alpha was probably about to end.