The Ganzfeld Protocol

The ganzfeld, German for "whole field," was not invented for psi research at all. Perceptual psychologist Wolfgang Metzger developed it in the 1930s to study how the visual system behaves when given a completely uniform, featureless field to look at, typically achieved today with translucent halved ping-pong balls taped over closed eyes under red light, paired with white or pink noise piped through headphones.

Parapsychologist Charles Honorton adapted the setup in the early 1970s on the theory that reducing ordinary sensory input might make a subject more receptive to a faint telepathic signal, the same logic behind older "quiet room" and hypnagogic-state ESP experiments. In the standard ganzfeld trial, a "receiver" sits in the mildly deprived state for roughly half an hour while, in a separate isolated room, a "sender" concentrates on one randomly selected image or video clip out of a set the receiver has not seen; afterward the receiver is shown that same small set of possible targets and asked to identify which one the sender was viewing, giving a baseline chance hit rate of about twenty five percent with four choices. Honorton and his collaborators ran and compiled dozens of such studies through the late 1970s and early 1980s, reporting a pooled hit rate consistently above that chance baseline.

Skeptical psychologist Ray Hyman reviewed Honorton's pooled ganzfeld results in the early 1980s and identified a list of methodological concerns, including inadequate randomisation procedures, the possibility of sensory cues leaking between sender and receiver rooms, and selective reporting of successful studies. Rather than simply talking past each other, Hyman and Honorton produced a formal 1986 joint communique, agreeing that the accumulated data showed an effect too consistent to attribute to chance or to selective publication alone, while continuing to disagree over whether that effect reflected psi or some subtler, uncontrolled artifact, and jointly specifying the tighter randomisation and automation standards any future study would need to meet before either side would treat it as decisive.

Honorton's subsequent autoganzfeld series, run with computer-controlled randomisation and target selection built to satisfy the 1986 standards, again reported an above-chance hit rate. Since then, larger and more rigorously pre-registered replications, including a well-powered multi-lab series conducted in the 2000s, have produced mixed and generally weaker results, and no outcome has been accepted by both sides as meeting the bar the 1986 communique itself set. The ganzfeld therefore stands, uniquely among psi protocols, as a case where both proponents and critics agree on what a decisive result would look like and simply have not agreed that one has yet arrived.