Parapsychology

What it is

The scientific study of apparent anomalies of information and influence — collectively psi:

• Telepathy — information passing between minds without a sensory channel
• Clairvoyance — information about distant objects or events
• Precognition — information about the future
• Psychokinesis — direct influence of mind on matter

It is distinguished from the traditions elsewhere on this site by intent: it proposes to settle these questions experimentally rather than to assert them.

Where it came from

The Society for Psychical Research was founded in London in 1882 by a group including the Cambridge philosopher Henry Sidgwick and the classicist Frederic Myers, explicitly to examine such claims "without prejudice or prepossession". Its early membership included several future Nobel laureates and three future presidents of the British Association.

J.B. Rhine at Duke University from the 1930s moved the field into the laboratory, introduced the term extrasensory perception, and ran enormous numbers of card-guessing trials with the five-symbol Zener deck. In 1969 the Parapsychological Association was admitted as an affiliate of the American Association for the Advancement of Science, where it remains.

The Stargate Project ran from 1972 to 1995, funded by US defence and intelligence agencies, on remote viewing. A 1995 review commissioned as it closed produced a genuinely instructive disagreement: the statistician Jessica Utts concluded that a statistical effect had been established; Ray Hyman agreed the numbers were anomalous but attributed them to methodological flaws rather than psi. The programme was closed as having produced nothing of operational use.

The PEAR laboratory at Princeton ran from 1979 to 2007, chiefly on whether intention could bias random number generators. It reported effects of extraordinary smallness — on the order of one part in ten thousand — accumulated over millions of trials, and was criticised for optional stopping, for a lack of independent replication, and for the difficulty of interpreting an effect that small.

How it is used

To argue that materialism is incomplete; and, from the other side, as a case study in how a field can persist without producing a result.

What we can and cannot say

We can say the central fact is the absence of replication. After a hundred and forty years, no psi effect has been demonstrated in a way that independent laboratories can reliably reproduce. That is the standard the field set for itself, and it has not been met.

Effects appear, generally small and generally just above chance. They shrink as controls tighten. They fail to appear in the hands of investigators who expect them not to. And no theoretical framework predicts when they should occur.

And we can say the single most consequential episode in this field's history is not a psi result at all.

In 2011 the social psychologist Daryl Bem published "Feeling the Future" in the Journal of Personality and Social Psychology, one of the discipline's leading journals. Across nine experiments he reported evidence of precognition — most memorably, that participants' performance on a memory test appeared to be improved by studying the material after the test.

The paper used entirely standard methods, passed normal peer review, and was published in a first-rank venue.

The response reshaped psychology. Stuart Ritchie, Richard Wiseman and Chris French ran replications, found nothing, and had their paper rejected — including by the journal that published Bem, which did not publish replications. Wagenmakers and colleagues published a Bayesian reanalysis arguing the data did not support the conclusions. And the wider discipline was forced into an uncomfortable question: if the standard methods of psychology could produce nine studies supporting precognition, what were they producing everywhere else?

The answer, worked out over the following decade, was the replication crisis. Bem's paper is routinely cited as one of its triggers, alongside contemporaneous fraud cases, and the reforms that followed — preregistration, larger samples, registered reports, the reformed treatment of p-values — are now standard.

A large multi-laboratory preregistered replication of Bem's paradigm, published in 2023, found nothing.

So the most useful thing parapsychology has produced is a demonstration that mainstream psychology's methods could generate a confident false positive. That is a real contribution, and it came from a field most scientists dismiss without examination.

We can say the field has also been methodologically ahead of its neighbours. Blinding, randomisation, preregistration and adversarial collaboration were taken seriously in psi research before they were routine in psychology — because psi researchers knew their results would be attacked, and built accordingly.

We cannot say psi has been disproved. Absence of a replicable effect after long and serious search is strong evidence, and it is not proof, and honest sceptics say so. What can be said is that the burden has not shifted, that a hundred and forty years is a long time, and that the most valuable thing to come out of the search was a lesson about everyone else's methods.

Further reading

• Utts and Hyman's 1995 Stargate reviews, read as a pair.
• Daryl Bem, "Feeling the Future" (JPSP, 2011), and Wagenmakers et al.'s response.
• Etienne LeBel and colleagues on what the Bem episode did to the discipline.