In 1933–1934, Duke University researchers tested whether Hubert E. Pearce Jr. could identify hidden cards while separated from the person handling them. Their reported scores made the Pearce–Pratt distance series an important case in the debate over extrasensory perception. The dispute was not simply about arithmetic. It concerned whether the procedures ruled out ordinary access to information, how the buildings had been arranged, and what could responsibly be inferred decades afterwards.
Map location: Approximate area: Durham city reference from the 2025 Census Gazetteer. The experiments took place at Duke University. This broad city-area pin does not identify an original testing room, a viewing position or the measured distance between the participants.
.jpg){kind=link}
What Rhine reported in 1934
J. B. Rhine’s Extra-Sensory Perception, chapter VII, describes Pearce as a Methodist ministerial student at Duke and J. G. Pratt as Rhine’s assistant. The chapter discusses varied tests rather than one unchanging procedure. Its closing addition reports Pratt handling face-down cards in Duke’s Physics Building while Pearce made corresponding calls in the library, over 100 yards away. A card rested on a book for each minute. Rhine lists twelve 25-card runs, totalling 300 trials, with an average of 9.9 correct calls per run against a chance expectation of five.
That account also says the cards were subsequently moved to the Medical Building, more than 250 yards from Pearce. The longer-distance work was still proceeding when the passage was written. Elsewhere in the chapter, Rhine reports poorer distance results in earlier work with C. E. Stuart. The 1934 account therefore should not be presented as the final report of every later run, or as a statement that all experiments involving Pearce and distance were successful.
The later combined results
Rhine and Pratt’s 1954 review, available here in an Institut Métapsychique International French translation, combines four subseries conducted between August 1933 and March 1934. It reports 558 correct identifications in 1,850 trials, against 370 expected by chance. The series involved 74 runs of 25 cards. The authors considered the results evidence of ESP, while acknowledging that a single subject over a limited period could not establish how other people would perform or predict when the same performance could be repeated.
The review describes signed duplicate records placed in sealed envelopes for Rhine, usually delivered before Pearce and Pratt compared their sheets, with procedures also covering exceptions. Rhine joined Pratt as an observer in the final subseries. The authors’ confidence in these precautions is part of the historical record. Their account also describes changes during the project: one subseries was extended after variable scores attracted interest. The final total consequently belongs to an evolving programme, rather than a single immutable trial plan fixed before the first result was known.
How the separation was supposed to work
Eliot Slater’s 1968 review of C. E. M. Hansel’s critique gives a useful account of the procedure under dispute. Pearce and Pratt synchronised watches before separating. Pratt worked through a shuffled and cut pack, keeping the cards face down during the run, then recorded their order afterwards. Pearce wrote his guesses separately. Copies went to Rhine. This arrangement was intended to prevent the two participants from adjusting their records together to manufacture matches. Slater nevertheless accepted Hansel’s objection that Pearce’s whereabouts were not continuously supervised.
Keeping cards face down during the guessing interval and making separate records address different risks. The first limits what the handler knows at that moment; the second helps protect the subsequent comparison. Neither precaution, considered alone, establishes where an unobserved participant remained throughout a session. Equally, identifying that gap does not establish that anyone exploited it. This distinction separates a criticism of an experiment’s evidential strength from a demonstrated accusation against a participant.
Pratt and Gardner disputed the sightlines
In a July 1966 published exchange, Pratt challenged Martin Gardner’s acceptance of Hansel’s proposed peeking route. Pratt said a plan marked as not to scale misplaced the windows and that the supposed view from a room across the corridor would not work. Gardner replied that the original report should have documented the layout adequately. He also argued that a corridor route remained a concern even if the particular across-corridor reconstruction failed.
The disagreement was therefore more specific than a contest between a large score and disbelief. It concerned what could have been seen, from where, and under which historical arrangement of doors and windows. Alterations made between the experiments and later inspections complicated that reconstruction. The exchange also included an argument over Pearce’s responses to later correspondence. Such retrospective willingness or unwillingness to discuss a test cannot substitute for observing the original procedure. The published letters show competing interpretations; they do not supply a contemporaneous recording of Pearce leaving his station.
A demonstration was disputed too
The September 1966 continuation sharpened the disagreement. Pratt said he had checked the corridor himself and found that the windows did not line up as Hansel’s plan suggested. He also disputed the relevance of Hansel’s later peeking demonstration, describing it as a different room, with a temporary partition and a crack above a door, rather than a repetition of the original Pearce–Pratt conditions. Gardner continued to question the precautions and the documentation of alterations. Neither side accepted the other’s account as settling the experiment.
A successful demonstration of an ordinary way to obtain answers can expose a weakness without reproducing every feature of an earlier test. Its force depends on how closely the demonstrated opportunity matches the disputed opportunity. Conversely, showing that one proposed route is impractical would leave other specific criticisms to be assessed. That is why the room details matter, while a statement that the experiment had simply been repeated and exposed would go beyond this exchange.
Hansel’s later criticism and the statistical question
In his own 1987 commentary, “Experimental evidence for paranormal phenomena”, printed pages 590–591, Hansel returned to Pearce’s lack of continuous supervision and the walking distance between the stations. He said his earlier critique had proposed several possible ways to obtain sight of the targets. He also questioned differences between reports and the reporting of less successful work. These remain Hansel’s criticisms: the passages do not constitute a confession by Pearce or an observation of him using a particular method.
The reported excess of correct answers is a fact about the published results. Interpreting it requires assumptions about how those results were produced. A small probability calculated for chance guessing is not, by itself, the probability that ESP exists, nor the probability that fraud occurred. The historical question is whether ordinary information, recording problems and changes of procedure were excluded sufficiently well to support the proposed extraordinary explanation. Reading the investigators and their critics together preserves both the unusual reported score and the unresolved argument over controls.
Sources and related reading
The sources used here are Rhine’s 1934 book, particularly chapter VII’s closing distance account, read in a transcription; the French translation of Rhine and Pratt’s 1954 review, rather than an inspected original English printing; the July and September 1966 exchanges; Slater’s original 1968 review; and Hansel’s 1987 commentary. No original trial sheets were inspected for this case. Compare the Rhines’ Lady Wonder investigations, or browse Spiritual / Psychic.