Telepathy

← Encyclopedia

Telepathy is the claim that one mind can reach another directly: that a thought, a feeling, the moment of a death can pass between two people with no word spoken and no signal sent. The wish underneath it is old and entirely human — to be known without saying; to learn, across any distance, that someone loved is in danger. The word itself is young, and it was built to sound like science.

It was coined in 1882 by Frederic W. H. Myers, a classical scholar and a founder of the Society for Psychical Research, formed in London that February to investigate mesmeric and spiritualist claims in a scientific spirit. Myers introduced it together with a companion word that did not last, telaesthesia; the two were meant, in his phrasing, to cover all cases of impression received at a distance without the normal operation of the recognized sense organs. The society’s first committee on the subject bore the plainer name Thought-Transference; Myers’s Greek compound — tele, at a distance; patheia, feeling — soon displaced “thought-reading.” The relabeling was deliberate, and it worked: it detached the idea from the séance room and dressed it for the laboratory. It also did something subtler. As the literary historian Roger Luckhurst has shown, the new word gathered under one apparently neutral, quasi-physical heading a scatter of older categories that had never been equivalent to one another — and in gathering them, it made a single secular object where there had been many. The hope the coinage carried was one critics later named openly: if living minds could touch without the body’s channels, minds outliving the body became easier to think. Establishing telepathy, not open-ended experiment, was arguably the society’s real aim from the start.

What the word gathered

The phenomena are far older than the term, and they were never one thing. The deep antecedent in the Western philosophical tradition is sympatheia, the cosmic sympathy of the Stoics and the Neoplatonists: the doctrine that the universe is a single living being whose parts, however distant, answer to one another as the organs of one body do. Plotinus draws on it in the Enneads (IV.3–5, IV.9) to account for action at a distance of every kind — magic, prayer, divination, even seeing and hearing across a gap — and at IV.9.3, where he asks whether all souls are one, later scholars have read a possible glance at what would now be called telepathy. But sympatheia is a cosmology, not a faculty. It explains why minds might touch by making the whole world one organism; it does not isolate “thought-reading” as a power some people have and others lack. Reading it as “ancient telepathy” is exactly the flattening the 1882 coinage invites and the historical record resists.

The same caution applies across the materials Myers’s word absorbed. Divination, oracular knowledge, and the veridical dream — the classical and biblical apparatus of foreknowledge — the Victorians re-sorted into “precognition” and “telepathy,” categories their sources did not recognize. The folk substrate of the wraith, the fetch, and the doppelgänger — the figure of an absent person seen at the hour of a crisis — Phantasms of the Living would later try to systematize as a telepathically generated hallucination. Highland second sight (an dà shealladh), documented by Martin Martin in 1703 and by a line of Scottish writers after him, was a regional category of involuntary vision with nothing of mind-to-mind transfer in it. The mesmerists supplied the most immediate proto-telepathy: the “rapport” and “community of sensation” reported between magnetizer and entranced subject, the apparatus the SPR inherited directly. And the contemplative traditions had long described attainments of mind-knowing that the Victorians were tempted to read as a single faculty but that are, in their own settings, distinct soteriological doctrines: Patañjali’s para-citta-jñāna, “knowledge of other minds,” arising from a disciplined concentration on another’s mental state (Yoga Sūtra III.19) — where the classical commentators are careful to say the yogin knows the state of the other mind but not its object; the Buddhist ceto-pariya-ñāṇa, the encompassing of others’ minds, among the higher knowledges; Sufi firāsa, intuitive discernment; and the Christian hagiographers’ cardiognosis, the reading of hearts. Each was embedded in a path and a cosmology of its own. Telepathy did not name a kind that all of these were instances of; it manufactured one, retrospectively, out of their remains.

The Victorian program: a striking residue, a fatal critique

The SPR was constituted on 20 February 1882, growing out of an autumn-1881 conversation between the journalist Edmund Dawson Rogers and the physicist W. F. Barrett. Henry Sidgwick, the Cambridge philosopher, was its first president; the early circle drew in Edmund Gurney, Myers, Frank Podmore, Eleanor Sidgwick, and later the physicists Oliver Lodge and William Crookes. Its stated aim was to examine these problems without prejudice or prepossession, in the spirit of exact inquiry the established sciences observed; deliberately, it suspended rather than asserted the “spirit hypothesis,” and that suspension split it from organized Spiritualism.

The first evidence arrived fast and collapsed almost as fast — and the two collapses are routinely merged, though they are distinct cases. The Creery sisters, the five daughters of a Derbyshire clergyman, were tested from 1880; Barrett’s enthusiasm and the society’s first reports treated their card- and name-guessing as proof of thought-transference, until the results “declined” over eight years and two of the girls were caught using a code of coughs and small movements, and confessed. Separately, the young mesmerist G. A. Smith and the journalist Douglas Blackburn produced striking transmissions of drawings and numbers in 1882–83, with hand-contact permitted, under Gurney’s auspices. In articles beginning in 1908 and culminating in a 1911 confession, Blackburn declared the whole corpus a hoax — the basic evidence of the truth of thought-transference, he said, had been bogus from the start, the product of codes and the reading of involuntary muscular cues. Smith denied it to the end, and the details of the confession have been challenged; either way the affair became a standing lesson in how experimenters are fooled. The stage knew things the laboratory was slow to learn: the era’s celebrated “thought-readers,” examined, proved to be reading muscles, not minds.

A subtler legacy of the same milieu is rarely noticed outside the history of statistics. The philosopher Ian Hacking has argued that the SPR’s thought-transference experiments were the birthplace of randomization in experimental design — the now-mandatory practice of assigning targets by chance to defeat the guesser’s bias. The discipline that would later be used to demolish the field’s results was, in part, invented inside it.

The society’s most ambitious early work was collection rather than experiment. Phantasms of the Living (1886), written mainly by Gurney with Myers and Podmore, analyzed 701 first-hand accounts of apparitions — many of them “crisis apparitions,” the figure of someone later found to have been dying or in danger at that hour — and read them as hallucinations telepathically generated by the distant mind. The Census of Hallucinations followed in 1894, its report drawn up by Eleanor Sidgwick: about seventeen thousand people canvassed, of whom 1,684 affirmed a vivid waking hallucination, and a committee identifying roughly thirty well-attested death-coincidences — a rate the researchers calculated at some 440 times what chance alone would yield, and read as confirming the telepathic theory of crisis hallucinations. Myers’s Human Personality and Its Survival of Bodily Death (1903) gave the most serious non-reductive synthesis the field produced, the “subliminal self,” a threshold below ordinary consciousness through which such impressions might rise; it had a documented line to William James’s Varieties and the American SPR, founded in 1885. The objections came at once and have never really changed: C. S. Peirce held that no scientific conclusion can rest on anecdote, and others noted that the cases lacked contemporaneous record and that memory, exaggeration, and deceit had nowhere been excluded. The shape of the next century is already visible here: a serious effort, a striking residue, and a critique that the controls were not equal to the claim.

A distinct Continental wing pursued the same questions through probability. In France, the physiologist Charles Richet — a future Nobel laureate in medicine — published in 1884 an early statistical treatment of card- and picture-guessing, reporting, for instance, 510 successes against a chance expectation of 458 over 1,833 trials, and pressing the calculus of probabilities into psychical research before almost anyone else. Gurney called the paper a likely permanent landmark even as he warned that Richet had pushed the mathematics past what it could bear. Richet would later coin métapsychique, the French name for the whole enterprise, in 1905; René Warcollier carried the French drawing-transmission tradition forward in La Télépathie (1921), where he compared the supposed transfer to a “mental radio” — a phrase that would soon cross the Atlantic.

From stories to statistics: the laboratory turn

The next move was from spontaneous cases to controlled trials, and the center of the field shifted to J. B. Rhine’s laboratory at Duke University, where he and Louisa Rhine worked under the psychologist William McDougall. Using a deck of twenty-five cards bearing five symbols — the Zener cards, designed by his colleague Karl Zener — Rhine fixed the chance baseline at one in five and ran trials by the tens of thousands. His Extra-Sensory Perception (1934) reported more than ninety thousand trials and concluded that ESP was an actual and demonstrable occurrence; it popularized both “extrasensory perception” and “parapsychology” (the latter borrowed from Max Dessoir’s older German coinage, which Rhine duly credited) and made card-guessing a household phenomenon after a favorable notice in the New York Times. The Pearce–Pratt distance series, run with the divinity student Hubert Pearce, produced the laboratory’s most celebrated above-chance runs.

The controversy that followed was statistical, and its partial settlement is constantly misreported. In 1937 the Institute of Mathematical Statistics, through its then-president Burton Camp, held that Rhine’s statistical methods were sound — that if there was anything wrong with the experiments it must be sought in the conditions, not the mathematics. This was an endorsement of the arithmetic and emphatically not of the conclusion that ESP exists; the two are routinely conflated. The rest of the arc repeated the first one at scale: the early high scores are now discounted, since sensory leakage and loose handling could account for them; the high scorers vanished as Rhine tightened his procedures; and the field eventually abandoned card-guessing altogether.

The word, meanwhile, had outrun the evidence into the wider culture. Upton Sinclair’s Mental Radio (1930) reported his wife reproducing target drawings under conditions critics judged uncontrolled, and carried a guarded preface that Albert Einstein had drafted for an unrealized German edition — a careful note vouching for Sinclair’s good faith while pointedly raising a non-telepathic explanation, unconscious suggestion between intimates, and so no endorsement of telepathy at all, however often it is quoted as one. Freud addressed thought-transference in several papers, intrigued and uneasy by turns.

A generation after Rhine, the Maimonides Medical Center in Brooklyn ran a dream-telepathy program (1964–1978) under the psychiatrist Montague Ullman, joined by Stanley Krippner and, later, by a young researcher named Charles Honorton whose name would dominate the next phase of the field. Sleepers were monitored by EEG and woken from REM sleep to report their dreams while a sender in another room concentrated on a postcard-sized art print, and blind judges matched the transcripts to the targets. Several series were striking; none was independently replicated, a thinness even sympathetic reviewers concede. Sleep laboratories were also expensive, and the cheap successor became the battleground.

The ganzfeld and the meta-analytic war

The successor was the ganzfeld — German for “whole field,” the name of a condition of uniform, featureless sensory input. A receiver rests in mild sensory homogenization, halved ping-pong balls over the eyes under red light and white noise in the headphones, and free-associates aloud for half an hour while a sender elsewhere views a randomly chosen image; the session’s target is then picked from four candidates, so chance is one in four. (The paradigm is treated in its own right under experimental parapsychology.) By the mid-1980s some forty studies existed, and Honorton argued they were sufficient evidence for psi: his 1985 meta-analysis of twenty-eight studies reported a combined Stouffer’s Z of 6.6, with twelve of the studies independently significant and no correlation between a study’s methodological quality and its outcome. Ray Hyman disagreed, and the two published dueling analyses of the same database. Hyman found that every study exhibited at least one of twelve methodological flaws — optional stopping, sensory cuing from handled targets, the file-drawer effect, flexible analysis — and that the apparent effect dissolved under correction.

What happened next is rarer than either finding. In 1986 the two issued a joint communiqué. They agreed, in language carefully negotiated, that there was an overall significant effect in the database that could not reasonably be explained by selective reporting or multiple analysis; they continued to differ over whether it constituted evidence for psi; and they specified together the stricter, computer-automated standards by which future experiments should be judged. Two adversaries had looked at the same data, refused to agree, and refused equally to pretend the disagreement could not be stated exactly. The exchange remains a model of adversarial honesty, and part of why the ganzfeld is still the case everyone argues about.

The new experiments came. Honorton’s own automated studies — 106 hits in 329 sessions between 1983 and 1989, a rate of 32 percent against the 25 expected — were published with the social psychologist Daryl Bem in Psychological Bulletin in 1994 as evidence of a replicable anomaly; Honorton died at forty-six, nine days before the paper was accepted. Hyman replied that the significance came disproportionately from the new dynamic video targets, that the static pictures scored near chance, and that one laboratory under one leader is not independent replication. Doubt reached backward too: Susan Blackmore’s report of protocol irregularities at Carl Sargent’s Cambridge laboratory touched a database in which Sargent’s and Honorton’s labs had supplied roughly half the significant studies, and Sargent withdrew from the field. In 1999 Julie Milton and Richard Wiseman meta-analyzed thirty post-communiqué studies from seven independent laboratories and found nothing: the ganzfeld, they concluded, “does not at present offer a replicable method for producing ESP in the laboratory.” Proponents answered that the selection had excluded the successes and departed from protocol. In 2010 Lance Storm, Patrizio Tressoldi, and Lorenzo Di Risio analyzed a fresh tranche of ganzfeld studies and reported a hit rate of 32 percent again, with a fail-safe of roughly two thousand four hundred unpublished null studies needed to reduce the ganzfeld result to chance; Hyman answered that the consistency had been manufactured by merging databases and trimming outliers, and that a meta-analysis is not a confirmation.

The argument at the top of the literature

A parallel shock to the field’s credibility came from one of its own champions. Bem’s 2011 paper “Feeling the Future,” reporting nine experiments in which subjects appeared to be influenced by stimuli they had not yet seen, was published in a flagship social-psychology journal and became, against its author’s intent, a catalyst of psychology’s replication crisis. If standard methods could certify an effect as impossible as precognition, the reasoning ran, the methods themselves were broken. Eric-Jan Wagenmakers and colleagues showed that the evidence vanished under a Bayesian reanalysis; Stuart Ritchie, Wiseman, and Christopher French reported three failed replications. The episode is widely credited with making psychologists confront optional stopping, p-hacking, and the file-drawer problem across their whole discipline — the field’s reckoning detonated, fittingly, by its most contested claim.

The deepest disagreement now runs at the very top of the literature, and it is a disagreement about where to begin. Etzel Cardeña argued in American Psychologist in 2018 that the cumulative evidence for psi is comparable to that for accepted phenomena in psychology and cannot be readily explained away by study quality, fraud, or selective reporting. Arthur Reber and James Alcock answered in the same journal in 2020 that the claims simply cannot be true — that effects with no possible physical mechanism can have no ontological status, the data therefore flawed whether or not the flaw can be located. Each side surrenders a different symmetry: one starts from the evidence and follows it; the other starts from the priors and refuses to be moved. Behind the priors stands the question of mechanism. The one physically plausible carrier, electromagnetism, is far too weak at any biologically credible strength to cross the distances reported; and the parapsychologists’ own data make the problem worse, since the claimed effect does not fall off with distance the way any real signal must. Mainstream psychology’s settled verdict — delivered by a National Research Council review in 1988, which after some 130 years of research found no scientific justification for the phenomenon, and unmoved since — treats the residue as the signature of selective publication and flexible analysis, with ordinary psychology supplying the experiences: underestimated coincidence, confirmation bias, the vividness of the remembered hit and the silence of the forgotten misses. The field’s own defenders concede an awkward regularity that fits neither story cleanly — the ganzfeld works extraordinarily well for a handful of experimenters and hardly at all for everyone else.

The folklore outlives the studies

All the while the experience runs ahead of the evidence and outlasts every negative result. Science fiction made telepathy its most durable paranormal ability — the hunted telepath-elite of Slan, the telepath-saturated society of The Demolished Man — until cyberspace offered fiction a more plausible wiring for mind-to-mind contact and the trope quietly receded. The lived folklore is older than the word. Twins were credited with a strange mutual sympathy in John Wesley’s journal a century before Myers and in Dumas’s Corsican Brothers (1844); about a third of twins still report the connection, though controlled tests find nothing. And in late 2024 a podcast claiming telepathy in nonspeaking autistic children, The Telepathy Tapes, ranked for several weeks as the most-listened-to show in both the United States and the United Kingdom — resting on facilitated-communication methods that controlled message-passing tests discredited in the 1990s, in which accuracy collapsed to chance the moment information was hidden from the facilitator. The pattern repeats at every scale and across a hundred and forty years: the spectacular claim, then the cueing critique.

What Myers did in 1882 was less a discovery than a translation. He took an old constellation — the sympathy of souls, the crisis-apparition, the mesmeric rapport — and re-dressed it for a statistical age: communion became transmission, spirits became signal, a faculty was minted where there had been a cosmology. What no one designed is what the word became afterward: the longest controlled experiment ever run on an esoteric claim, fourteen decades of effects that appear and are then squeezed, retreating as the controls tighten or, as defenders insist, persisting a few stubborn points above chance. Both readings are still in print, in the same journals, answering each other.

Scholarship and the experimental record

The historiography of the coinage rests on Roger Luckhurst’s The Invention of Telepathy 1870–1901 (Oxford University Press, 2002), the load-bearing account of telepathy as a datable Victorian cultural object rather than a perennial category; Janet Oppenheim’s The Other World (Cambridge University Press, 1985) and Alan Gauld’s The Founders of Psychical Research (1968) supply the institutional history of the SPR. Ian Hacking’s “Telepathy: Origins of Randomization in Experimental Design” (Isis 79, no. 3, 1988, pp. 427–451; https://doi.org/10.1086/354775) traces the unexpected debt modern experimental method owes to the thought-transference trials.

The modern experimental literature divides along the meta-analytic fault line. The pivotal détente is the joint communiqué of Ray Hyman and Charles Honorton, “A Joint Communiqué: The Psi Ganzfeld Controversy” (Journal of Parapsychology 50, 1986, pp. 351–364; reviewed in the Society for Psychical Research’s account of the controversy at https://psi-encyclopedia.spr.ac.uk/articles/ganzfeld-esp/). The case for a replicable effect is made in Daryl J. Bem and Charles Honorton, “Does Psi Exist? Replicable Evidence for an Anomalous Process of Information Transfer” (Psychological Bulletin 115, no. 1, 1994, pp. 4–18; https://doi.org/10.1037/0033-2909.115.1.4), and renewed in Lance Storm, Patrizio E. Tressoldi, and Lorenzo Di Risio, “Meta-Analysis of Free-Response Studies, 1992–2008: Assessing the Noise Reduction Model in Parapsychology” (Psychological Bulletin 136, no. 4, 2010, pp. 471–485; https://doi.org/10.1037/a0019457). The null case is set out in Julie Milton and Richard Wiseman, “Does Psi Exist? Lack of Replication of an Anomalous Process of Information Transfer” (Psychological Bulletin 125, no. 4, 1999, pp. 387–391; https://doi.org/10.1037/0033-2909.125.4.387).

The contemporary verdict-level exchange runs between Etzel Cardeña, “The Experimental Evidence for Parapsychological Phenomena: A Review” (American Psychologist 73, no. 5, 2018, pp. 663–677; https://doi.org/10.1037/amp0000236), and the categorical reply of Arthur S. Reber and James E. Alcock, “Searching for the Impossible: Parapsychology’s Elusive Quest” (American Psychologist 75, no. 3, 2020, pp. 391–399; https://doi.org/10.1037/amp0000486). The replication-crisis episode is anchored by Daryl J. Bem, “Feeling the Future: Experimental Evidence for Anomalous Retroactive Influences on Cognition and Affect” (Journal of Personality and Social Psychology 100, no. 3, 2011, pp. 407–425; https://doi.org/10.1037/a0021524), with the methodological rebuttal of Eric-Jan Wagenmakers and colleagues, “Why Psychologists Must Change the Way They Analyze Their Data: The Case of Psi” (Journal of Personality and Social Psychology 100, no. 3, 2011, pp. 426–432; https://doi.org/10.1037/a0022790), and the failed replications of Stuart J. Ritchie, Richard Wiseman, and Christopher C. French, “Failing the Future: Three Unsuccessful Attempts to Replicate Bem’s ‘Retroactive Facilitation of Recall’ Effect” (PLoS ONE 7, no. 3, 2012, e33423; https://doi.org/10.1371/journal.pone.0033423). The mainstream institutional assessment is the National Research Council’s Enhancing Human Performance: Issues, Theories, and Techniques (National Academy Press, 1988; https://www.nationalacademies.org/read/1025/chapter/13), whose parapsychology subcommittee, chaired by Hyman, found no scientific justification for the phenomenon.

→ In the library: Plotinus — Enneads IV.9 'Are All Souls One?' (MacKenna) · The Yoga Sutras of Patanjali (Johnston, 1912)

→ Related: Mesmerism · J B Rhine · Modern Ganzfeld Experimental Parapsychology · Thought Transference Thought Reading Experimentation · F W H Myers · Victorian Psychical Research Spr · Charles Richet · Precognition · Scottish Second Sight An Da Shealladh