Blog · Morphic Resonance
A Cambridge-trained biochemist with a Royal Society fellowship proposed that nature has memory — and Nature, the journal, wondered aloud about burning his book. Four decades on, Sheldrake is still running experiments and his critics are still unconvinced. Here are the books, the claims, the strongest critiques, and the only response Ashta considers adequate: test it properly.
Rupert Sheldrake is not a fringe figure who wandered into science — he is an establishment scientist who walked out of it, and that friction is the whole story. Born in 1942, he took a double first in natural sciences at Clare College, Cambridge, won the University Botany Prize, spent a year at Harvard on a Frank Knox Fellowship, and completed a Cambridge PhD in biochemistry in 1968. As a Fellow of Clare College funded by a Royal Society research fellowship, he co-developed (with Philip Rubery) the chemiosmotic model of polar auxin transport — a genuine, still-cited contribution to plant physiology. From 1974 to 1978 he was Principal Plant Physiologist at ICRISAT in Hyderabad, working on crop physiology for the semi-arid tropics. Then he left to write a book that would make him, depending on who you ask, either a visionary or the textbook case of a scientist gone astray.
A New Science of Life (1981) proposed morphic resonance: the idea that nature has a kind of inherent memory. On this hypothesis, the regularities of the world are less like eternal laws and more like habits; self-organising "morphic fields" shape the form and behaviour of molecules, organisms and societies; and each new member of a species draws on — and adds to — a collective memory of its kind, across space and time. The hypothesis came with testable predictions: newly synthesised compounds should crystallise more easily everywhere once they have crystallised somewhere, and rats should learn a task faster once rats elsewhere have learned it.
The reaction was unforgettable. John Maddox, editor of Nature, ran an editorial titled "A book for burning?", calling it "the best candidate for burning there has been for many years" — though the nuance matters: Maddox did not literally call for burning, he argued the book was an intellectual aberration dressed as science. Thirteen years later he went further on television, comparing the case to the Pope's condemnation of Galileo and calling Sheldrake's ideas "heresy" — a word Sheldrake's defenders have been quoting gratefully ever since, because dogma is supposed to be the other side's vice. The mainstream objections, though, are substantive, not just rhetorical: morphic fields have no proposed physical basis, the hypothesis is flexible enough to absorb failures, and the suggestive evidence has conventional explanations — the famous rat-learning trend is attributed to design flaws, and easier crystallisation to seeding and better technique.
For readers who want the full arc: A New Science of Life (1981) and The Presence of the Past (1988) lay out morphic resonance; The Rebirth of Nature (1991) argues for a living rather than mechanical view of nature; Seven Experiments That Could Change the World (1994) proposes cheap, do-it-yourself tests of taboo questions; Dogs That Know When Their Owners Are Coming Home (1999) reports the animal-anticipation research; The Sense of Being Stared At (2003) collects the staring and "extended mind" experiments; The Science Delusion (2012, published in the US as Science Set Free) reframes ten assumptions of materialist science as open questions; and Science and Spiritual Practices (2017) and Ways to Go Beyond (2019) examine what contemplative practices measurably do. His recent output has been journal papers and dialogues rather than new trade books.
Nearly everyone has turned around to meet someone's eyes. Sheldrake turned the intuition into a protocol: a starer either looks or does not look at a blindfolded subject's back in randomised sequence, and the subject guesses "looking" or "not looking." Across very large informal datasets he reports small but persistent above-chance hit rates — a few percentage points above the 50% baseline.
The critiques are specific and instructive. Marks and Colwell argued the published randomisation had learnable structure — too many alternations, too few long runs — so subjects given trial-by-trial feedback could pick up the pattern rather than any stare; without feedback, their replication found nothing. Others proposed simple response bias. The best formal summary is the Schmidt and colleagues meta-analysis in the British Journal of Psychology (2004): across fifteen remote-staring studies using physiological measures, a small significant effect — which shrank to nothing in the highest-quality subset. "Statistically significant but fragile under rigour" is the honest one-line verdict, and it is exactly the kind of verdict that justifies a cleaner, bigger, pre-registered test rather than a verdict-by-shouting. Sheldrake, to his credit as an empiricist, has published rebuttals and continued refining protocols — including automated tests — rather than retreating from testability.
The Jaytee case is the cleanest illustration in all of this literature of how the same data can honestly yield opposite conclusions. Jaytee, a terrier belonging to Pam Smart, seemed to go to the window around the time Pam set off home. Sheldrake videotaped hundreds of returns at randomly chosen times and reported the dog at the window vastly more during Pam's homeward journeys than at other times. Richard Wiseman ran his own filmed trials — invited by Sheldrake — applied a pre-set "first signal" criterion, and declared the effect refuted. Sheldrake replotted Wiseman's own raw data and showed they display the same rising pattern as his. Wiseman's reply: a dog simply visits the window more the longer its owner is away, no telepathy required. Both analyses are public. The dispute is not about what the dog did; it is about which success criterion was fair — which is why Ashta fixes its criteria in advance, in public, before any data exist.
Sheldrake's best-known live claim is telephone telepathy — participants guessing which of four nominated callers is ringing, with reported hit rates around 40% against a 25% baseline. We cover that literature, its critiques, and Ashta's own large-scale replication through the Spanda app in a dedicated evidence review. In 2013, his TEDx talk "The Science Delusion" was removed from the main TED channels after complaints from skeptic bloggers and review by TED's science board; TED relocated it to a discussion page with caveats, critics called it censorship, and the "banned TED talk" became a permanent part of his public story — over ten million views later, arguably the most effective distribution decision TED never intended. He remains active into the 2020s: recent papers on staring through video, a 2025 meta-analysis of telephone telepathy, ongoing public dialogues, and participatory online experiments hosted at sheldrake.org.
Here is the point Ashta cares about, and it cuts both ways. Sheldrake's consistent position is that these questions should be settled by controlled experiment, not by taboo — and on that, we agree without reservation. His critics' consistent position is that when the controls tighten, his effects weaken — and that is a factual pattern the record partly supports, which is why our stance is replication under pre-registered, blinded conditions with the null as default. If the extended mind is real, it will survive good controls. If it is not, the honest experiments he has always called for will be what shows it. Either way, he posed questions worth the price of testing — see our research lineage and what we measure.
