A New Science of Life

Rupert Sheldrake's A New Science of Life proposes that the regular forms and behaviours of organisms are not fully explained by genes and proteins, and that they are instead sustained by formative causation — the idea that a pattern, once repeated, becomes easier for later similar patterns to take up through morphic resonance. This article covers the argument as it appears in the book's treatment of morphogenesis, behaviour and learning: the molecular-similarity problem that motivates the hypothesis, the proposed cross-location rat-learning experiment, the extension of morphic fields to behaviour, and the human-learning tests collected in Appendix A.

The problem the hypothesis is meant to solve

Sheldrake's starting point is a mismatch between molecular and organismal difference. He cites protein comparisons showing that the average human polypeptide is more than 99 per cent identical to its chimpanzee counterpart, and DNA comparisons of non-repeated sequences showing an overall human–chimpanzee difference of only about 1.1 per cent. He also notes that species within the genus Drosophila, and different species of mice, differ from one another more than humans differ from chimpanzees. Svante Pääbo, director of the chimpanzee genome project, is quoted to the effect that the genome does not reveal why humans and chimpanzees are so different. Sheldrake draws the conclusion that "the contrasts between organismal and molecular evolution indicate that the two processes are to a large extent independent of each other," and asks what, if not genes and proteins, accounts for the difference.

This molecular argument is the strongest and most independently sourced part of the excerpt. It is a real observation about the relationship between genetic similarity and phenotypic divergence. It does not, on its own, establish morphic resonance; Sheldrake treats it as the motivation for looking elsewhere, not as proof of his mechanism.

Formative causation and morphic fields

The proposed mechanism is that forms and behaviours are repeated because they resonate with similar forms and behaviours in the past. Morphogenetic fields organize biological form; behavioural and motor fields extend the same logic to patterns of movement. Inheritance of these fields is said to depend on the same factors discussed for morphogenetic fields, and the strength of the effect is expected to scale with the number and duration of past repetitions.

A biological illustration Sheldrake offers is hybrid behaviour in lovebirds. Fischer's lovebird carries nesting strips in its bill; the peach-faced lovebird tucks them among its feathers. Hybrids tear the strips normally but then behave inconsistently — sometimes tucking, sometimes carrying in the bill, and even erecting the lower-back feathers while carrying in the bill as if attempting to tuck. Sheldrake reads this as two parental behavioural fields of comparable strength pulling in incompatible directions. The anecdote is illustrative rather than evidential, and the book does not present it as a test.

The proposed rat-learning experiment

The concrete methodological core of the excerpt is a cross-location learning test. Large numbers of rats would be trained on a single task at one location. At a second location, hundreds of miles away, the learning rate of rats on several different tasks would be monitored at regular intervals, for example monthly. The task chosen for mass training at the first location would be selected at random from that set, and the start of training would also be randomized — say, four months after the regular testing began at the second location. Experimenters at the second location would be told neither which task had been selected nor when training had begun.

If the learning rate for the selected task rose markedly at the second location only after training began at the first, Sheldrake argues, this would be strong evidence for formative causation. He notes that Crew and Agar's group, repeating McDougall's rat work, found their rats learning considerably faster at the outset than McDougall's rats had when he first began — a pattern he suggests might reflect exactly this kind of effect.

Sheldrake also concedes a structural feature of the design: a positive result would not be fully reproducible, because rats in any repetition would themselves be influenced by morphic resonance from the original experiment. To demonstrate the effect repeatedly, the task or the species would have to be changed each time. This is a tension between the hypothesis being testable and its being falsifiable in the ordinary sense.

Learning, motor skills and the confound problem

Sheldrake extends the argument to human motor learning. If a pattern of behaviour brings an individual into resonance with everyone who has performed it before, then learning should become progressively easier as more people acquire a skill. He suggests this might be detectable in recently originated motor patterns — riding a bicycle, driving a car, skiing, playing a video game — where cumulative resonance would be growing.

He immediately qualifies the claim: even if reliable quantitative data showed learning rates rising, the interpretation would be complicated by improved machine design, better teaching methods and higher motivation. Only specially designed experiments holding those factors constant could yield persuasive evidence. He also states plainly that the hypothesis cannot account for the origin of new patterns of behaviour in the first place; creativity, he says, lies outside the scope of natural science and can only be addressed on metaphysical grounds.

Appendix A: the human-learning experiments

The appendix collects tests in which participants are asked to learn or recognize material that has, or has not, been widely encountered before. The logic is that real, culturally established material should be easier to process than invented or scrambled equivalents if morphic resonance contributes.

• Hebrew words (Schwartz). Participants were shown real and scrambled Hebrew words and later asked to guess which was which. Conscious performance was no better than chance, but participants reported greater confidence about the real words, which Schwartz interpreted as an "unconscious pattern recognition effect."
• Persian words (Alan Pickering, Hatfield Polytechnic). Participants viewed a word for ten seconds and then drew it. Independent judges, blind to condition, rated reproductions of real words as significantly more accurate than scrambled ones.
• Persian and Arabic student replications (Nigel Davidson, Geraldine Chapman). Both reported similar positive results.
• Morse code (Arden Mahlberg). A new code was constructed by reassigning dots and dashes to letters. Participants who did not know Morse learned genuine Morse significantly more accurately than the invented code. The letters S and O were excluded because of familiarity with the S.O.S. pattern.
• Hiragana (Suitbert Ertel, Göttingen). Participants saw nine hiragana characters projected for eight seconds, then marked them among twenty on an answer sheet, over six trials. Recognition generally improved trial by trial. Ertel predicted that characters would be recognized more readily upright than inverted, since millions of Japanese readers encounter them in the normal orientation, and reported that this is what he found. In a further test, participants could not consciously distinguish genuine hiragana from artificial characters invented by a graphic designer, yet remembered the real characters better.
• The "crucial" inverted-fake test (Ertel). Ertel compared real characters upside down with fake characters upside down, arguing that rotation should not affect fake characters because resonance plays no part in them either way up. Sheldrake reports that the results were confusing and Ertel's interpretation hard to follow.
• Repetition priming (Zoltan Dienes, Oxford). Ninety participants judged whether letter strings were real words or non-words. A "shared" word set was shown to all participants; a "unique" set was shown only to every tenth participant, creating 80 "boosters" and ten "resonators." If resonance were at work, response speed to shared stimuli should improve relative to unique stimuli. All trials were run in a controlled environment with distinctive visual, olfactory and auditory cues to maximize resonance between participants. The outcome was positive and statistically significant: the more often a non-word had been seen, the faster later resonators responded to it. When Dienes attempted to repeat the experiment at the University of Sussex, there was no significant effect.
• Anagram experiments (Ertel). Two further "new-field" experiments were run through a magazine, using anagrams such as "Seterleirei" for "Reiseleiter," with readers asked to find the normal words. Sheldrake notes that simple crosswords did not provide a good test because many clues were not in fact new.

How strong is the evidence?

The excerpt is candid about its own limits. The molecular comparison is well sourced and stands on its own. The learning evidence is weaker: small samples, several student projects, one explicit failed replication (Dienes at Sussex), one result the experimenter himself found confusing (Ertel's inverted-fake test), and Sheldrake's own concession that improved equipment, teaching and motivation confound the bicycle/car/ski/video-game trends. The book does not claim these tests are conclusive, and the resonance interpretation of each belongs to Sheldrake and the named experimenters rather than to the underlying data.

Two internal tensions are worth keeping visible. First, the rat experiment is presented as strong evidence yet is described as inherently non-reproducible, which sits awkwardly with ordinary standards of falsifiability. Second, the Dienes Oxford positive and Sussex null are reported side by side without resolution. Externally, the molecular-similarity argument motivates a non-genetic inheritance mechanism, but similarity of genes does not by itself establish morphic resonance; the book keeps these as separate steps, and readers should too.

Related work in this wiki

This source extends the existing Sheldrake cluster, particularly morphic-resonance-experimental-tests, habits-of-nature and laws-of-nature-critique, and it overlaps with the earlier treatment of the same book in . It also sits alongside the reading guide to The Presence of the Past (), which develops the same formative-causation program. The pattern of small positive studies plus a named failed replication is consistent with how this wiki treats Sheldrake's staring, homing and prayer evidence elsewhere.

Open questions

Whether the cross-location rat-learning experiment was ever actually carried out is not established by this excerpt.

Keep reading

• Morphic Resonance: The Nature of Formative Causation — Develops a related question.
• The Presence of the Past: Morphic Resonance and the Habits of Nature — Develops a related question.
• The Sense of Being Stared At: And Other Unexplained Powers of the Human Mind — Develops a related question.