Sensory leakage
Early card tests allowed subtle cues — marks on the backs of cards, the tester's face — that later protocols removed.
Tested, replication failed
Clairvoyance
Clairvoyance is the claimed ability of seeing objects, places or events beyond the reach of the senses. Status: tested, replication failed.
What people describe
Clairvoyance is described as seeing things beyond the reach of the eyes — a hidden object, a distant place, an event happening elsewhere. People report flashes of imagery, symbols that need interpreting, or scenes that play out like a short film behind the eyes.
How it entered the record
In the 1930s J. B. Rhine at Duke University tested clairvoyance with Zener cards — five symbols, one to guess — and reported above-chance scores that later researchers could not reproduce under tighter controls. From the 1970s until 1995 the US government funded remote-viewing research later known as the Stargate Project; an independent evaluation in 1995 concluded it had never produced intelligence that was useful in practice, and the programme was closed.
The experiment
The ganzfeld experiments — Bem and Honorton meta-analysis, 1994
How it was run
The receiver sits in uniform red light with halved ping-pong balls over the eyes and white noise in headphones, then picks which of four images the sender was looking at.
Sample
Around 350 sessions in the 1994 meta-analysis; hundreds more since
Result
Hit rate reported at roughly 32–35% against a 25% baseline. Milton and Wiseman then pooled 30 further studies in 1999 and found no effect at all.
The disagreement is not about whether 32% beats 25% — it does. It is about whether the 32% survives when the people running the study expect nothing.
Claim
Seeing objects, places or events beyond the reach of the senses
Word from
French clair "clear" + voyance "vision"
Status
Tested, replication failed
Where the word comes from
Entered English in the 1840s from French mesmerism.
What psychology offers instead
Vague imagery
Descriptions loose enough to fit many targets are judged as hits once the real target is known.
Publication bias
Positive studies are more likely to be published, so meta-analyses can show effects that vanish in pre-registered replications.
How to test it fairly yourself
Have someone else choose a target at random from a set of four images, seal it where nobody in the room can see it, and record your description before the set is shown. Over many rounds chance alone gives 25% hits; anything convincing needs to be clearly above that.