Overview

Most scientists treat the mind as something that happens entirely inside the skull. Signals come in through the senses, the brain processes them, and behavior goes out. On this view, one person cannot directly sense another person’s gaze or intentions without some physical signal traveling between them. The extended mind hypothesis challenges that picture. It proposes that minds are not sealed inside bodies but reach outward, making genuine perception at a distance possible.

Two very different ideas share this name, and it is worth keeping them apart. Philosophers Andy Clark and David Chalmers proposed an “extended mind thesis” in 1998 to explain how people offload thinking onto notebooks, smartphones, and other tools. That idea is mainstream cognitive science and has nothing paranormal about it. The parapsychological version, developed most fully by biologist Rupert Sheldrake, goes much further. Sheldrake argues that minds extend through what he calls morphic fields. These are invisible organizing fields that connect organisms to each other and to their environments. On his account, this field-based extension is what makes telepathy, the sense of being stared at, and pet-owner anticipation possible.¹²¹

The Sense of Being Stared At

The most widely tested prediction of the parapsychological extended mind hypothesis is simple: people can tell when someone is looking at them, even when they cannot see the looker. This is called scopaesthesia. The feeling is familiar to almost everyone, but familiarity is not evidence. Sheldrake and colleagues designed controlled experiments to find out whether the effect is real or just a compelling illusion.⁴²

In the basic setup, one person sits with their back to another. The looker follows a random sequence of staring and not-staring periods. The person being looked at presses a button to guess which condition is in effect. Because there are only two options, chance performance is 50%. Across a very large number of trials, Sheldrake’s research program has reported hit rates modestly but consistently above that level.⁴²

Sheldrake has also collected hundreds of accounts suggesting the effect is directional. When people or animals sense a stare, they often turn to look straight at the starer rather than scanning around at random. A survey of 960 such accounts found that roughly two-thirds showed this directional quality. Sheldrake interprets this as evidence that visual attention flows outward from the eyes toward its target. He calls this theory visual extramission.¹⁷

Telephone and Telecommunication Telepathy

A second major research program tests whether people can sense who is about to call them before the phone rings. This is called telephone telepathy. The setup is straightforward. A participant is told that one of several pre-selected callers will ring at a random time. The participant guesses who it is before answering. Because the caller is chosen randomly, chance performance is predictable. Sheldrake and colleagues have run this kind of test in supervised lab conditions, in filmed experiments, and through automated online systems.⁶³

A 2025 meta-analysis combined results from 26 separate telecommunication telepathy experiments. The overall hit rate was meaningfully above chance. The effect was stronger when the caller and receiver shared a close emotional bond. Participants who said they relied on intuition rather than deliberate guessing also scored higher. The effect held up across phone calls, emails, and text messages.⁶³

“Those who said they took an intuitive approach had an above-chance hit rate of 56.7%” in telecommunication telepathy tests, and emotional bonds between caller and receiver were associated with hit rates roughly 19% above chance.⁶³

Morphic Fields and Learning

The extended mind hypothesis predicts not just person-to-person effects but also collective ones. If morphic fields store the habits of a species, a skill should become easier to learn as more people practice it. This would hold even for people who have no contact with earlier practitioners. Researchers tried to test this prediction using the online word game Wordle.⁴

Wordle is played by millions of people worldwide each day, all solving the same puzzle. If morphic resonance is real, players who attempt the puzzle later in the day should find it slightly easier. Earlier players would have already built up a collective field of practice. Three studies tested this idea. The results were inconsistent. One study found a small pattern matching the prediction. The other two did not, and one actually found the opposite pattern. The researchers concluded that if morphic resonance exists, it is a subtle effect that may be easily disrupted by other factors.⁴

Skeptical Critiques and Discussion

Critique 1: Staring-detection hit rates are explained by response bias, not genuine perception

Schmidt (2008) found that when staring-detection data are adjusted for participants’ tendency to favor one response over another, the apparent gap between “being stared at” and “not being stared at” conditions disappears.⁶⁶

Sheldrake and Smart (2008) argue that their repeated-testing studies used randomized sequences and trial-by-trial feedback. Control trials with no staring produced chance-level results. They say this rules out simple response bias as a complete explanation.⁴² Sheldrake and Smart (2024) also contend that multiple studies using proper randomization and no feedback have produced significant positive results.²⁷

Evidence strength: contested

Critique 2: Anecdotal case collections document belief in scopaesthesia, not the phenomenon itself

Brugger (2024) argues that listing nearly a thousand professional accounts of staring detection shows how widespread the belief is, but provides no evidence that the phenomenon objectively exists. He cites work by Guterstam and colleagues showing scopaesthesia functions as a belief rooted in folk theories of vision, not as a genuine perceptual ability. His conclusion: “experimental evidence for scopaesthesia as an objective perceptual phenomenon is currently absent.”²⁴

Sheldrake and Smart (2024) responded that Brugger misrepresented the methodology of the 2003 staring study. They say he falsely claimed it used trial-by-trial feedback and pseudorandom sequences when neither was present. They argue that studies meeting rigorous design standards have still produced significant positive results. Dismissing all evidence on philosophical grounds, they say, is not a scientific response.²⁷

Evidence strength: contested

Critique 3: Positive results in Sheldrake’s experiments reflect inadequate blinding and selection effects

A persistent concern across Sheldrake’s research programs is that participants or experimenters may accidentally signal the correct answer through subtle cues. When an experimenter knows the condition being tested, they may unintentionally influence the participant’s response. This is why experiments use blinding: the person recording responses does not know what the correct answer is. Critics argue that many of Sheldrake’s studies have not achieved adequate blinding, and that positive results cluster in less tightly controlled conditions.⁶⁶²⁴

Sheldrake and Smart point to studies using automated randomization, physical separation between starer and subject, and independent error-checking. In the 2008 repeated-testing study, the error rate across 2,800 trials was 0.1%, and net errors amounted to 0.04%. They argue that procedural mistakes were not driving the results.⁴² The telecommunication telepathy meta-analysis also found that effect sizes in selected-participant studies were larger than in unselected samples. The authors argue this fits a genuine psi effect rather than a flaw in how the studies were run.⁶³

Evidence strength: moderate (concern is legitimate; rebuttal is partial)

Critique 4: The morphic field concept is not scientifically testable in any decisive way

A broader concern is that morphic fields are defined so flexibly that almost any result can be accommodated. When the Wordle studies found contradictory results across three studies, the authors suggested that morphic resonance may be “easily inhibited by conscious cognitive processes” and may require “more refined experimental designs.” A hypothesis that retreats to new qualifications each time a prediction fails is difficult to falsify.⁴

Sheldrake and colleagues acknowledge the inconsistency in the Wordle results directly and do not claim the morphic resonance hypothesis was confirmed. They treat the mixed findings as a reason to refine the experimental design rather than as proof of the hypothesis.⁴ The staring-detection and telephone telepathy programs do generate specific, quantitative predictions. These can be tested and, in principle, falsified by chance-level results across large samples.

Evidence strength: weak (the falsifiability concern is serious and not fully resolved)