The Duke Zener Card ESP Series

The whole design rests on one number. A Zener deck holds twenty-five cards, five each of circle, cross, wavy lines, square and star, so a subject calling all twenty-five in order will average five hits by chance. Everything the Duke laboratory published is a statement about how far a subject's score departed from five per run, expressed in standard deviations and accumulated across many runs. That is a different kind of claim from anything psychical research had made before.

The Society for Psychical Research had spent half a century arguing about the credibility of a witness to a single striking event, a dispute with no natural end, because each side could always question the other's judgement about a story. Rhine replaced the story with a tally. He also separated conditions that had previously been run together, testing clairvoyance by having nobody look at the target, telepathy by having an agent concentrate on it, and precognition by asking for the order the deck would take after a later shuffle, and treating these as three claims needing three designs. Whether anything real was measured is contested. That the question became one about procedure and arithmetic rather than about the honesty of a witness is not.

Three objections arrived almost immediately and none was frivolous. The first concerned the cards themselves: early Zener decks were printed thinly enough that a symbol could sometimes be made out from the back under the right light, and the edges of a card carry information about how it has been handled. The second concerned recording, since in much of the early work the person who knew the target also wrote down the call, which allows small unconscious corrections to accumulate in one direction. The third concerned selection, both of which subjects to keep testing and of when to stop a series.

To these was added the plain fact of failed replication, most publicly the Princeton series run by W. S. Cox, over twenty-five thousand trials with 132 subjects and no effect at all. Rhine's response was to change the procedure rather than to argue: opaque screens between the parties, mechanical shuffling, duplicate independent scoring, greater physical distance between subject and cards. He also founded the Journal of Parapsychology in 1937 so that criticism and negative results would have a home inside the field. Critics answered, and still answer, that the effect shrank each time a control was added, which is the pattern an artefact produces and not the pattern a phenomenon produces.