Harry Collins, a University of Bath sociologist, is best known for his extreme relativistic philosophy of science (see my review of Frames of Meaning, which he coauthored, in the Fall 1983 issue of Free Inquiry), and for having caught a group of spoon-bending children at cheating. The New Scientist ( June 30, 1983) printed his "Magicians in the Laboratory: A New Role to Play," in which he discusses what he calls the "vexed relationship" between magicians and psi researchers. His article contains many misconceptions about magic; but, before detailing them, first a sketch of his views.
Randi's recent Project Alpha, Collins writes, has reminded us again of how easily psi researchers can be hoodwinked. Because the history of paranormal research has been riddled with fraud, Collins wisely recommends that, no matter how innocent a subject may appear, experiments must be designed on the assumption that the subject is "a notorious cheat." Unfortunately, he adds, completely fraud-proof tests are impossible because there is no way to anticipate new methods of cheating. Since magicians know standard ways, they can be enormously useful as advisors. But because they are not much better than nonmagicians in spotting new methods they are of little value as observers.
He feels that magicians should not be allowed to monitor experiments because they are usually unfriendly toward psi research and have a vested interest in seeing psychics discredited. Collins doesn't mention the belief of most parapsychologists that hostile observers inhibit psi phenomena, but even aside from this he thinks magicians would have a damaging effect on experiments if they were allowed to monitor them.
How, then, can conjurors help? One way, Collins says, is by breaking their code of secrecy and explaining to researchers how cheating can be done. If magicians are unwilling to do this, they should serve as "protocol breakers," by demonstrating the same paranormal phenomena under the same controls applied to the psychic. If they fail to break the protocol, this "would act as a certificate of competence in experimental design."
Misconception 1: Collins fails to distinguish stage performers from