The Real Creation Story — E. O. Wilson, The Social Conquest of Earth (Long Now × Exploratorium, 20 Apr 2012)

Edward O. Wilson · 2012 · 1:32:10 · ▶ Watch

Summary

The central question. Gauguin’s three questions: where do we come from, what are we, where are we going. Wilson says religion and academic philosophy have not answered them, and that science now can.

The thesis. “there is a real creation story of humanity and one only and it is not a myth”. The story runs like this:

The human condition began with eusociality: groups where adults raise young together and some individuals give up their own reproduction for the group. It is very rare, about two dozen origins in all the history of life that we can read, once in the human line. Yet the few lineages that got it (ants, termites, humans) dominate the land.
Every known case followed the same path. The last pre-adaptation is a defended nest to which a parent forages and brings food. When the young stay instead of leaving, eusociality begins.
The human path: down from the trees, upright, more meat, fire and a campsite, a surge in brain size, and then a strange new force: group-against-group competition.
Two instincts came out of that: an urge to form groups and a genius at reading other people’s intentions.
Because selection acts at two levels at once (individuals within a group, and groups against groups), we are built with a permanent internal conflict between selfish and cooperative drives.
The older explanation for such altruism, kin selection, is wrong. “we now have got to clear the wreckage of kin selection off the road”.

The Q&A covers what eusociality means, ant super-colonies and why they do not take over the world, whether humans are still evolving, whether the nest step can be tested in the lab, how you would falsify group selection, Pinker’s Better Angels, the future of eusociality, the decline of creation stories, how to survive in Alabama as a non-believer, staying active at 82, and why he would study microbes if he started again.

Close. Brand ends on microbes and the unknown frontier: “there’s a great deal of science yet to come” is how he puts it, and Wilson’s last long answer is about soil bacteria, the human microbiome and antibiotic resistance.

Stance to note. Wilson is an unabashed synthesiser and a self-declared atheist who still wants to work with religious people on conservation. He is gentle with people and blunt with theories: kin selection is “wreckage”, creation stories are a “killer” in the sense that they bind people into tribes. He openly says the group-selection question “is not settled”, then presents his side as if it were.

Map — talking points

#	Point
1	Brand’s introduction: Wilson’s string of revolutions (ants, island biogeography, sociobiology, biodiversity)
2	Gauguin’s three questions; every religion has its own incompatible creation story
3	Philosophy gave up; science is left with the job. “There is a real creation story”
4	Defining eusociality; it is rare (about two dozen origins)
5	Eusocial insects dominate the land; humans too
6	Two mysteries: why so rare and so late, and what force drove it
7	Deep time: insects in the coal forests, dinosaurs that never grew big brains
8	The “evolutionary maze”: pre-adaptations and a final step
9	The final step is a defended nest with food brought home; shrimp and sweat bees as examples
10	Why most species stall at the threshold: group selection must beat individual selection
11	Beetles that herd larvae: the wrong way to eusociality. Chimps, gorillas, wild dogs at the threshold
12	Hominin steps: leave the trees, walk upright, ape cultures
13	Australopithecines; the brain-size surge from 3–2 million years ago; meat
14	Campsites, fire, cooked food, Homo erectus; the breakout from Africa 60,000 years ago
15	Inter-group aggression as the “signature behaviour” of early humans
16	Individual-level vs group-level selection, with a bird-colour example; termite warfare; Darwin on group selection
17	Multi-level selection: selfish within groups, cooperative between them
18	Two instincts: forming groups, and reading intentions
19	The Nowak–Tarnita–Wilson argument against kin selection; “not settled”
20	Why it matters: the door to human biology is open; history needs prehistory needs biology
21	A conflicted species forever; is this the only route to advanced intelligence anywhere?
22	Sensory constraints, complexity preferences, habitat preferences (savanna, water, a view)
23	Close: “clear the wreckage of kin selection”
24	Q&A: what “eusocial” means; why not more super-colonies; Argentine and fire ants
25	Biodiversity as insurance; is humanity still evolving; homogenisation and new diversity
26	Can the nesting step be tested? Sweat-bee experiments; “spring-loaded” pre-adaptation
27	Maths of multi-level selection; permanent instability; conflict resolution
28	Pinker, violence, gene-culture co-evolution
29	Winning by using resources best; ant distributed intelligence
30	How to falsify group selection; the haplodiploidy “test” that failed
31	Genetics and the future: finding the “spring-loading” genes; personality heritability
32	The future of eusociality; decline of creation stories; The Creation, evangelicals, Revelation
33	Surviving in Alabama; advice to young scientists; “never retire”; Vanuatu
34	Why microbes: soil, the human microbiome, antibiotic resistance, disease as invasion

Putting it together

The mental model in one paragraph. Wilson tells a story with a single spine: the nest. A parent that builds a defensible home and brings food to it creates a place where grown children may stay. If staying pays for the group against rival groups strongly enough to beat the pull of individual advantage, eusociality arises. It has happened about two dozen times, and each time the result dominated its world. In our line the nest was the campsite, the fuel was meat and fire, and the rival groups were other bands. Selection acted on individuals inside the band (making us selfish) and on bands against each other (making us cooperative, tribal and good at reading minds). That two-level push gives us our permanent inner conflict, and the best we can do is understand it. This story is used to replace kin selection, which Wilson says is mathematically wrong, and to retire religious creation stories as tribal flags. The rest of the talk applies the same lens to ants, to biodiversity and to microbes. The weak points to remember: the hominin “nest” is inferred, early human war is contested, kin selection and multilevel selection often give the same maths, and the specialists mostly rejected the attack on kin selection.

Three tensions to keep open

Kin selection or multilevel selection, or both? Equivalent bookkeeping or a real difference in causes? Birch and Okasha (2015) is the neutral guide.
Genetic or cultural group selection? Wilson wants genes for tribal cooperation. Henrich, Richerson and Boyd find the same pattern in cultural evolution, which needs no new genes and is more widely accepted.
A “real” story or one story among several? Science supplies the facts about origins, but whether it can supply the meaning that creation stories supply (belonging, purpose, obligation) is the question the talk raises and does not answer.

Reading path

Start: Wilson’s own The Social Conquest of Earth (2012), chapters on the “maze” and on group selection. Then Birch and Okasha, “Kin selection and its critics” (2015), open access in BioScience. Hamilton’s 1964 paper in plain summary (any textbook).
Then: Nowak, Tarnita and Wilson (2010) and the 137-author reply (2011), read side by side. Pinker’s Edge essay “The False Allure of Group Selection”. Sober and D. S. Wilson, Unto Others (introduction). Hrdy, Mothers and Others (ch. 1–3). Haidt, The Righteous Mind (the last third on group selection).
Deeper: Okasha, Evolution and the Levels of Selection. Maynard Smith and Szathmáry, The Major Transitions in Evolution. Queller and Strassmann (2009). Henrich, The Secret of Our Success. Bowles (2009) on warfare and cooperation. Hölldobler and Wilson, The Superorganism. Durkheim, Elementary Forms, book 1. Blaser, Missing Microbes.

Part 1 — Before you read

Key terms

Eusociality. “Truly social”. Wilson’s version in this talk: groups lasting over generations in which adults raise young together and “enhance the complexity and success of the social system”, with a division of reproduction and labour based on altruism. The classic textbook trio is: cooperative brood care, overlapping generations, and a reproductive division of labour (some individuals, “workers”, give up or reduce their own breeding). Humans are a loose fit to the strict version (we do not have sterile castes), which is a standing objection. Wilson’s usage is wider. Check which one a source uses.

Altruism. Watch out: two meanings.

Field	What “altruism” means
Evolutionary biology	An act that lowers the actor’s reproductive success and raises another’s. A technical, behaviour-only definition with no motive
Everyday and moral philosophy	An act done out of concern for another, usually with an intention
Wilson’s use of “altruistic” in the talk is the first kind, though he slides into the second when he talks about “virtues”.

Individual selection vs group selection. Individual: genes spread because they help the individual that carries them out-reproduce others in the same group. Group: genes spread because groups that carry them out-compete other groups. Wilson’s “bird colour” example is the first. His “termites at war” example is the second.
Multi-level selection (MLS). Both at once. Wilson’s slogan: individual selection makes “selfish behaviour”, group selection makes “cooperative behaviour”. The 2007 slogan of D. S. Wilson and E. O. Wilson, “selfishness beats altruism within groups; altruistic groups beat selfish groups”, states it†.

Kin selection / inclusive fitness. Watch out: a name for a theory and a name for a method.

Term	Meaning
Kin selection (Maynard Smith, 1964)	Selection favouring a gene for helping because relatives share copies of that gene
Inclusive fitness (Hamilton, 1964)	A way of counting an individual’s fitness: its own offspring plus the extra offspring its help causes in relatives, weighted by relatedness
Hamilton’s rule	A gene for helping spreads if r × B > C (relatedness × benefit to recipient > cost to actor)
Wilson’s attack in this talk is aimed at all three together. His critics say the three are different things and that the maths he attacked (an inclusive fitness calculation under strict assumptions) is only one use of the general idea.

Pre-adaptation (exaptation). A trait that evolved for one job and makes a later jump possible. Wilson calls the steps toward eusociality pre-adaptations. Careful: the word is retrospective. You only call a step a pre-adaptation after you know where the path ended.
Spring-loaded. Wilson’s term: a species already has a trait that turns into a big leap in new circumstances. His example is nesting bees that are territorial. Put two in one cramped nest and one becomes the dominant “queen” and the other a worker.
Evolutionary maze. His metaphor. A species wanders by luck through possible genome-and-environment states. Selection pushes, but there is no direction. Reaching eusociality is passing through a series of doors.
Superorganism. A colony that works as one organism, with its own distributed intelligence. Wheeler used the word for ants in 1911 (from memory). Wilson and Hölldobler’s 2009 book is The Superorganism.
Supercolony. A very large ant “colony” with many queens and no hostility among its nests, like the Argentine ant in California. Not the same as a superorganism.
Consilience. Wilson’s 1998 word for the joining of the sciences and humanities; not said in this talk but his closing is its argument: “history makes no sense without prehistory”.
Sociobiology. Wilson’s 1975 name for the biological study of social behaviour, including humans. Brand calls it “highly controversial at the time later proved to be quite right”. That is Brand’s view. Many scholars would say it was partly right and partly overreach.

People named (a reading list)

Person	Why they come up	Start with
Stewart Brand	Host. His own record: Whole Earth Catalog, Long Now	The Clock of the Long Now (1999)
Paul Gauguin	The painting’s three questions	D’où venons-nous ? Que sommes-nous ? Où allons-nous ? (1897–98, Museum of Fine Arts, Boston)
Charles Darwin	Group selection in Origin (workers) and Descent of Man	The Descent of Man (1871), ch. 5
William James	”history is a bloodbath"	"The Moral Equivalent of War” (1910)†
Martin Nowak, Corina Tarnita	The 2010 mathematicians who challenged inclusive fitness	”The evolution of eusociality”, Nature (2010)
Bert Hölldobler	Wilson’s co-author on ants and The Superorganism	The Ants (1990), The Superorganism (2009)
Steven Pinker	His Better Angels is raised in Q&A	The Better Angels of Our Nature (2011)
William D. Hamilton (not named)	Author of kin selection / inclusive fitness	”The genetical evolution of social behaviour” (1964)
George C. Williams (not named)	Killed naive group selection in 1966	Adaptation and Natural Selection (1966)
George Price (not named)	The maths that unifies both levels	”Extension of covariance selection mathematics” (1972)
David Sloan Wilson (not named)	Modern MLS advocate, no relation	Unto Others (1998, with Sober)

Wilson’s own frame

Wilson (1929–2021) was a Harvard entomologist, an expert on ants who became the most public biologist of his generation. The milestones, from memory†, place the 2012 book in his career:

The Theory of Island Biogeography (1967, with Robert MacArthur): the species–area relation, which became the basis of conservation planning.
Sociobiology: The New Synthesis (1975): the last chapter applied the ideas to humans, which set off the “sociobiology wars”.
On Human Nature (1978), Genes, Mind and Culture (1981, with Charles Lumsden), Biophilia (1984), The Diversity of Life (1992), Consilience (1998), The Creation (2006), The Superorganism (2009, with Hölldobler), the novel Anthill (2010).
In 2010 he co-authored the Nature paper against inclusive fitness, and in 2012 published The Social Conquest of Earth, the book behind this talk. He wrote Letters to a Young Scientist next (the talk says he is writing it,; published 2013, from memory).

So in this talk he reverses himself. His 1975 book built on kin selection. He says so openly: “I’m guilty of that too” (paraphrase; the caption is garbled).

The current debate (live search, Nov 2026)

Inclusive fitness is still standard. The 2010 attack drew a 137-author reply in Nature in 2011 (Abbot et al., “Inclusive fitness theory and eusociality”), and later reviews argue that the critics conflated the general theory with one modelling method. The Birch and Okasha paper “Kin selection and its critics” (BioScience, 2015) is the best neutral summary (PDF). Evolutionary biologist Jerry Coyne’s blog reaction to Wilson’s later claims is one example of the hostile reception (Why Evolution Is True, 2013). A Scientific American blog on the fight gives the 2012 mood (The Good Fight).
Multi-level selection is alive but split. The “cultural group selection” camp (Richerson, Boyd, Henrich, Bowles) argues that human groups compete and that culture, not genes, carries the traits (Richerson et al., Behavioral and Brain Sciences). Pinker’s Edge essay “The False Allure of Group Selection” is the best known sceptical case (Edge). Neither side has won. The search results here were mostly commentary and blogs, not a systematic review.
The war question underneath. Wilson needs lots of group conflict in the Pleistocene. Anthropologists still argue about how common deadly inter-group violence was; see Turchin’s summary (The war over war).
Religion’s decline has levelled off, not continued. Wilson predicts a decline of organised religion and creation stories. Pew’s 2023–24 Religious Landscape Study found 29% of US adults unaffiliated and 62% Christian, but also said the mix has been fairly stable since about 2020 (Pew, Feb 2025). So his prediction is half true: big drop since 2007 (78% Christian then), now a plateau.
The microbiome frontier he points to has moved fast (see idea 10). Not searched in detail; from memory.
This talk is in the Long Now canon. The seminar page and a “Wilson seminar primer” are at longnow.org and the primer.

The ideas

Wilson spoke to a general audience in 49 minutes and compressed about fifty years of argument. Eight families of ideas meet in this talk. Here they are laid out, including ones he does not name.

 ALTRUISM PROBLEM ──────────┐                    ┌──── LEVELS OF SELECTION
 Darwin · Fisher/Haldane    │                    │     Wynne-Edwards · Williams
 Hamilton · Trivers · Hare  │                    │     Price · Wade · Sober/DS Wilson
 Haplodiploidy, monogamy    ▼                    ▼     Okasha · Nowak/Traulsen
                    ┌──────────────────────────────────┐
 MAJOR TRANSITIONS ►│  WILSON 2012: eusociality arrives │◄ HOMININ EVOLUTION
 Wheeler · Buss ·   │  by a nest + group selection;     │  Isaac · Lovejoy · Wrangham
 Maynard Smith ·    │  humans are conflicted because    │  Hrdy · Dunbar · Tomasello
 Szathmary · Queller│  we are selected at two levels    │  Boehm · Out of Africa
                    └──────────────────────────────────┘
                      ▲              ▲                ▲
  TRIBAL PSYCHOLOGY & │  SOCIOBIOLOGY  │  CREATION STORIES │  BIODIVERSITY &
  WAR                 │  & ITS POLITICS│  & RELIGION       │  MICROBES
  Tajfel · Sherif ·   │  Gould/Lewontin│  Durkheim · Boyer │  MacArthur · Elton
  Haidt · Bowles ·    │  Sahlins ·     │  Atran · D.S.     │  Tilman · Woese ·
  Keeley · Pinker     │  Lumsden · Boyd│  Wilson · Norenz. │  Lederberg · Blaser

Where each school of thought is covered:

The altruism problem: from Darwin to Hamilton’s rule: idea 2
Levels of selection: the long quarrel about groups: idea 5
Major transitions and the superorganism: idea 7
Hominin evolution: the steps Wilson summarises: idea 3
Tribal psychology and the question of war: idea 4
Sociobiology and its politics: idea 6
Creation stories, religion and science: idea 10
Biodiversity, ant societies and the microbial frontier: idea 11

1. “A real creation story”: why science, not religion or philosophy

In short. Wilson opens with Gauguin’s three questions. He says religions cannot answer them, since each has its own creation story and they cannot all be correct. Philosophy long ago gave up on the meaning of life. That leaves science, which he says offers one real creation story of humanity that is not a myth, pieced together like a jigsaw from many fields.

🎙 In the talk. Wilson opens with Gauguin’s three questions: where do we come from, what are we, where are we going. He argues that religion cannot answer the first two because “each organized religion and there are hundreds of them has its own different creation story in competition with all the others”. Each was written when its authors knew almost nothing about the universe, and each claims to be the right one. On this he is blunt: “they cannot all be correct no two can be correct”. Academic philosophy, he says, “long ago gave up trying to answer the overall question what is the meaning of life”, and after the decline of logical positivism the philosophers “scattered” into areas science had not colonised. So the job falls to science, which “promises” this: “there is a real creation story of humanity and one only and it is not a myth”. He builds it from many fields (genetics, neuroscience, ecology, archaeology, psychology, history) and calls it “the piecing together like a jigsaw puzzle”. It is, he says, his own reading of the primary literature of the last 20 years, not an idea from his head.

📚 Deeper theory

Why “creation story” is a loaded phrase. Wilson takes the vocabulary of religion and applies it to science. That is a rhetorical choice with a history. Mircea Eliade and Joseph Campbell used “creation myth” for stories that tell a group who it is. By saying “real creation story … not a myth” he claims the same job (an account of origins that orients a people) but with the authority of testing. Compare Consilience (1998), where he argues that the evolutionary epic is the best myth we have†.
The logical positivism remark. Logical positivism (the Vienna Circle, 1920s–30s; Carnap, Schlick; Ayer’s Language, Truth and Logic, 1936) held that only testable statements and logical truths are meaningful. It collapsed by the 1960s under attacks from Quine (“Two Dogmas of Empiricism”, 1951), Popper and Kuhn. Wilson’s claim that philosophers then fled to the margins is a caricature of a much richer field (philosophy of biology, philosophy of mind and ethics are all active, and the philosophers Okasha and Sober are central to the very debate in this talk). Treat it as a rhetorical simplification.
Gauguin’s three questions are older than Gauguin. They echo the Gnostic-style formula “who we were, what we have become, where we were, where we are going” attributed to Theodotus†. Gauguin painted them from right to left across a Tahitian scene: birth, adult life, old age. Wilson uses the painting only as a frame. The fact that he gets Gauguin’s biography slightly wrong (see warning box) is a small sign of how fast the opening is meant to be.
Sciences as one story: consilience. Wilson’s long-running claim is that knowledge from different fields can be joined (consilience, a word from William Whewell, 1840). This is a methodological position: explanations should be continuous from the physical to the social. Critics call it reductionist imperialism: social science and the humanities are asked to accept biology as their foundation. Compare the “social construction” tradition (Berger and Luckmann; Sahlins).
The role of tested knowledge. He says science gives “knowledge of the real world that can be tested and shared with every person”. This is the Popperian/Mertonian ideal: public, repeatable, falsifiable. It becomes important when Brand asks how group selection could be disproved (idea 5).

Where others push back

Religion’s role is not mainly to answer scientific questions. Gould’s NOMA, anthropological work (Geertz, “Religion as a Cultural System”, 1966) and theologians argue that creation narratives convey meaning, obligation and belonging rather than competing causal accounts. A scientific account of origins does not do those jobs†.
A “one and only” story is itself contested. Evolutionary scenarios for human origins are revised often (the age of Homo sapiens, the role of interbreeding with Neanderthals and Denisovans, the dating of fire use). A story that keeps changing is a feature of science. It is a problem for any claim to a single settled “creation story”.
The “story” is partly Wilson’s selection. Another evolutionary biologist, using the same primary literature, could stress cooperative breeding (Hrdy), cultural evolution (Henrich), or kin selection, and tell a different tale.

What he leaves out. The question “where are we going” (Gauguin’s third). Wilson ends by discussing the near future of eusociality and religion in the Q&A, but the talk itself tells a story of the past. The book’s last sections on the future of humanity and the environment are not covered here.

Check yourself

Why does Wilson think that religion and philosophy cannot answer “where do we come from”?
What is the difference between a creation myth and a scientific origin account in Wilson’s usage? Where might the line blur?
Which of his three claims about philosophy would a philosopher of biology dispute?

2. Why eusociality is so rare: the nest, the maze and the spring

In short. Eusociality means groups over generations where adults rear young collectively, with division of labour based on altruism. It has arisen only about two dozen times, yet those lineages rule the land. Wilson explains this with an evolutionary maze, a random walk through states under selection. The last step he finds in every case is a complicated nest.

🎙 In the talk. Wilson defines eusociality as the existence of “groups over multiple generations in which adult members rear the young collectively”, with a division of reproduction and labour based on altruism. It has arisen “only very rarely”: he counts “some two dozen” origins across the history of life, in insects and crustaceans, twice in African mole-rats, and once in the human line. Those few lineages rule the land: ants and termites can make up much of the biomass in a habitat, and “they absolutely rule the invertebrate land and living environment”. This sets two questions that biologists, he says, had not asked together: “why is it so very rare and why did it take so long to originate an evolution”, and “what kind of selection pressure drove it”.

The deep-time story: insects covered the Palaeozoic coal forests, then the Permian crash wiped out many orders, and none of them were eusocial that we know of. Dinosaurs had free hands and the capacity for bigger brains, but never got them. Ants and termites arrive in the Mesozoic and take over by the early Cenozoic.

His explanation is the evolutionary maze. A species wanders by luck from state to state under the push of natural selection, with no direction, “in the surrounding forces”. Looking back, the steps on the successful paths are pre-adaptations. The last one, found in every case he can trace, is “the building of a complicated nest defensible” to which a female, a mated pair or a small group returns with food for the young. Usually the grown young leave. In rare cases “the young do not leave they stay” and then you have a primitive eusocial society. This might take “a single mutation silencing the tendency to disperse”. He gives sponge-dwelling snapping shrimp (five independent origins) and sweat bees as examples.

Why do most species at the threshold not cross it? Because the new group must beat solitary rivals, and “group selection has to be powerful in order to come overcome the effects of individual level selection”, which has shaped these lines for hundreds of millions of years. Herding beetles that move their larvae around without a nest are “the wrong way”; the nest is “a pretty strict requirement”. Chimpanzees, gorillas and African wild dogs are near the threshold but have not crossed.

In the Q&A, asked whether the nesting step could be tested in short-lived species, Wilson says it has been, starting with the Japanese: force two nest-founding sweat bees into a space with room for only one nest and “one of them starts digging”, becomes the dominant female and then the queen, and the other forages. He calls this “spring-loaded adaptation or spring-loaded pre adaptation”. Territoriality, built in to nesting insects, turns into dominance and a queen and worker split. In this way one can “create” simple eusocial colonies in the laboratory.

📚 Deeper theory

The standard classification. Wilson (1971, The Insect Societies) used three criteria: care of young by others, overlapping generations, reproductive division of labour. Later authors rank societies on a gradient (solitary, subsocial, communal, quasi-social, semi-social, eusocial, and then Michener’s and Sherman’s “highly eusocial”). The sweat bees (Halictidae) are the textbook case of a polymorphic family, with solitary, communal and eusocial species, sometimes within one species. Sherman, Jarvis and Alexander (1991, 1995; naked mole-rats) and Duffy (snapping shrimp Synalpheus, 1996 onward) supply the vertebrate and crustacean cases Wilson counts†. The count of origins depends on definition. Strict counts give fewer, looser ones give more.
The nest as “fortress defence”. Queller and Strassmann (1998) and Alexander, Noonan and Crespi (1991) developed the “fortress defender” idea: species that defend a valuable, defensible resource (a nest in wood or a gall, a sponge) benefit from staying with the parent. This is the best-known version of the “defended nest” step. Wilson’s contribution is to make it the universal final pre-adaptation and to add group selection as the driver.
The counter-example Wilson mentions. Beetles that herd larvae without a nest “never” became eusocial. A later case may weaken the rule: the Australian ambrosia beetle Austroplatypus incompertus is reported to be eusocial, and it does live in galleries in wood, so it actually supports the nest rule†.
Why a ‘mutation’ can be enough. A gene that reduces dispersal is a plausible single change. In bees, differences in a few regulatory genes shift behaviour between solitary and social forms in the same species (for example work on Lasioglossum albipes by Kocher and colleagues)†. Wilson’s talk predicts that “the genes that actually caused the step” will be found. As of this page, genomic work on ants and bees has found that eusocial lineages share changes in gene regulation more than single “eusocial genes”†.
Pre-adaptation, exaptation and the ‘spring’. The term pre-adaptation can mislead because it sounds as if evolution aims at the result. Gould and Vrba (1982) proposed exaptation for a trait co-opted for a new use. Wilson’s “spring-loaded” is close to this. It also looks like evolutionary capacitance and developmental plasticity: a latent response is revealed when the environment changes (West-Eberhard, Developmental Plasticity and Evolution, 2003, the “genes are followers” argument).
Haplodiploidy is missing from the nest story. Wilson reaches for kin selection later and for a strict rule of nest first. Hamilton’s followers would say that the nest sets the opportunity, and relatedness sets whether the young “choose” to stay and help. The two accounts are not exclusive.
Dinosaurs and brains. The remark that dinosaurs “might have” become something like the creature in the picture refers to Dale Russell’s thought experiment of the “dinosauroid” (1982)†. It is speculative. Dinosaur brain–body ratios were modest; birds are the surviving dinosaurs and do show large brains in some lines (crows, parrots).

📚 Roots: The altruism problem: from Darwin to Hamilton’s rule

The problem. If natural selection favours whatever leaves more offspring, why would a worker ant give up her own reproduction, or a soldier die for the nest? Darwin called sterile insect castes a difficulty that “at first appeared to me insuperable, and actually fatal to my whole theory” in the Origin (1859), chapter 8 (quote from memory)†. Wilson’s talk is a long answer to the same question.

The chain of thinkers

Charles Darwin, Origin (1859) and Descent of Man (1871). In the Origin he saved the theory by saying selection could act on the family (the colony), since sterile workers are relatives of fertile queens. In the Descent he argued that tribes with brave, loyal, self-sacrificing members would beat other tribes, even if within each tribe the selfish did better. This is the “group selection” Wilson says Darwin “embraced”. Darwin’s two arguments are different (colony or family versus tribe), which matters later.
R. A. Fisher and J. B. S. Haldane (1930s). Fisher (1930) suggested family-level selection for distasteful insects. Haldane’s joke is famous†: he would lay down his life for two brothers or eight cousins, because that would preserve as many copies of his genes.
W. D. Hamilton, “The genetical evolution of social behaviour” (1964). A gene for helping spreads if rB > C. r is the chance that the helped individual carries the same gene by common descent. This is the mathematical form of “kin selection” (John Maynard Smith named it in 1964). Hamilton also pointed to the odd genetics of ants, bees and wasps.
Haplodiploidy. In Hymenoptera, females come from fertilised eggs (two sets of chromosomes), males from unfertilised eggs (one set). Full sisters share 75% of genes by descent (all of their father’s genome is identical, plus half their mother’s), but only 25% with brothers, and 50% with their own offspring. A worker “should” prefer to raise sisters over daughters. Wilson called this “Newtonian” in its elegance.
Robert Trivers and Hope Hare (1976). Used haplodiploidy to predict that workers would bias investment towards sisters (a 3:1 female-to-male ratio). Some data fit, others do not. It became one of the best-known tests, and Wilson says the theory “failed” as the cases grew.
E. O. Wilson, Sociobiology (1975). Adopted Hamilton’s framework for the evolution of sociality, which is why he can say “I’m guilty of that too” in 2012.
The monogamy turn: Hughes et al., Science (2008)†. A reconstruction of ancestral states suggested that in the lineages that became eusocial (ants, bees, wasps) the ancestral females mated with a single male, which makes workers’ relatedness to siblings high whether or not haplodiploidy exists. This supports the kin-selection story by a different route. Wilson’s talk does not mention it.
Nowak, Tarnita and Wilson, “The evolution of eusociality”, Nature (2010). Argued that inclusive fitness theory is not needed and is mathematically limited, and that eusociality is better modelled as a population-genetic problem with a nest, a gene for staying, and selection at two levels.
Abbot et al., “Inclusive fitness theory and eusociality”, Nature (2011). The reply, signed by 137 researchers: the 2010 paper misunderstands the theory. Further replies followed from Gardner, Bourke, Rousset and Lion, and others; Birch and Okasha (2015) summarise the exchange.

The thread. Hamilton gave altruism a gene-level rule, and ants and bees looked like the star case. Over the next 40 years the cases got messier (termites are not haplodiploid; many eusocial species fall outside the neat pattern). Wilson reads the messiness as refuting the theory. His critics read it as showing that relatedness is one factor among several, including ecological benefit and cost, which Hamilton’s rule already puts in B and C.

Where others push back

The sample is small and hand-counted. “Two dozen” origins gives little statistical power to claim that all of them followed one path. Selection of cases and the definition of eusociality matter a lot.
Relatedness data. Studies of the ancestral mating system (Hughes et al. 2008) suggest high relatedness is common at origins; Wilson’s critics say that it is hard to argue that kin selection played no role.
Humans are not a clean case. Humans do not have sterile castes or a queen. The claim that humans went through the same maze is an analogy by Wilson, and cooperative breeding theorists (Hrdy) point to a different route (many carers, not a defended nest).
“Group selection must be powerful” is partly circular. Wilson sees the rarity as evidence for a strong group-level force. Others read the rarity as evidence that the conditions for kin-based altruism seldom arise.

What he leaves out. The fitness costs to workers in different species; secondary losses of eusociality (some bees reverted to solitary living, from memory), and reproductive skew models (Vehrencamp; Keller and Reeve) that predict when subordinates stay and help.

Check yourself

What is the last pre-adaptation, and what single behavioural change crosses the threshold?
Why does Wilson say group selection must be “powerful” for a eusocial group to survive?
How does the sweat-bee experiment test the maze idea, and what would count as it failing?

3. The hominin path: down from the trees, meat, fire and the camp

In short. Wilson traces the steps from the split with chimpanzees about six million years ago: out of the trees, then upright walking with free hands on the savanna. Brains stayed small until about three to two million years ago, then grew fast, while one line ate more meat. He says we can surmise but have no direct proof that they shared meat at campsites. By Homo erectus he reports campsites with controlled fire.

🎙 In the talk. For humans Wilson traces “the steps the pre humans took” from the split with chimpanzees about six million years ago. Step one: “get out of the trees”. Chimpanzees did not go all the way: they are mainly knuckle-walkers, though they have cultures (stone-throwing, digging tools) that differ between regions. Step two: upright walking and free hands in savanna and grassland, seen in the early form he shows from 4.4 million years ago, which has a chimp-sized brain. The australopithecines then radiate into several species, mostly vegetarian, with brains of only 400 to 500 cubic centimetres. Had aliens landed three million years ago they would have seen “no special mention note of it”. The change comes from about three to two million years ago, when “the brain size after remaining static for all that period of time suddenly began to increase”, “unlike any other evolution in rate of a complex organ”. One line began to eat more meat. Chimps get a small share of calories from meat, humanity much more; Wilson flags his own figures with “if I’ve got these figures right”. He then says: “we can surmise but have no direct proof” that they worked out of campsites and shared meat there. By Homo erectus, about two million years ago, he reports campsites with controlled fire. Getting fire is “not hard to do”, since “all you got to do is pick up burning branches” from lightning fires. Animals killed by fires gave cooked meat; teeth, skull and gut changed. The campsite is the nest: “some went out and they divided labor”. Then the breakout from Africa about 60,000 years ago, along the Nile corridor, across Asia to New Guinea and Australia. He says that our ancestors then looked much as today’s hunter-gatherers do, with the same cognitive ability: “not in cognitive ability” have we changed.

📚 Deeper theory

The ‘home base’ is the pivot and also a weak point. Glynn Isaac’s hypothesis (1978) read clusters of stone tools and bones at sites like Olduvai and Koobi Fora as camps. Lewis Binford (1981) and others argued many clusters came from carnivores, scavenging and natural accumulation, not human sharing sites. Fire and hearths clearly support central-place foraging much later (Gesher Benot Ya’aqov, Israel, about 790,000 years ago, and Wonderwerk Cave, South Africa, about one million years)†. If the earliest ‘nest’ is not secure, Wilson’s key pre-adaptation is inferred more than shown. He admits this: “we can surmise but have no direct proof”.
Meat, scavenging and the brain. Evidence for stone-tool butchery goes back about 2.6 to 3.4 million years (Gona, Dikika; from memory, the oldest claim is contested). The expensive tissue hypothesis and the cooking hypothesis (Wrangham) offer different energy routes to a larger brain. A third, the cooperative breeding account (Hrdy), says the real constraint was the cost of raising slow-growing, large-brained children, so a network of carers had to come first. Another is the cultural intelligence hypothesis (Herrmann et al. 2007, Henrich): big brains were an outcome of the benefit of learning from others.
Brain size is a messy signal. The sharp rise “from three to two million years ago” is clear in the fossil series, but modern analyses (for example work on Homo naledi, with a small brain and complex behaviour claims, 2013 onward) show that brain size and behaviour do not match simply. The naledi finds were announced after this talk†.
Chimpanzee culture and violence. Wilson uses chimps as the reference. Goodall’s reports of the Gombe ‘four-year war’ (1974–78) and Wrangham and Peterson’s Demonic Males (1996) are the usual background for an ancient root of intergroup killing. Bonobos, the sister species, are far less violent, which is an argument that the shared ancestor may not have been aggressive. Wilson mentions “the two species the chimpanzee and the vulnerable” (a garbled bonobo) only in passing.
Out of Africa as a bottleneck story. Genetic diversity drops with distance from Africa, which fits a serial founder effect (Ramachandran et al. 2005). Interbreeding with Neanderthals and Denisovans was only discovered in 2010 (Green et al., Reich et al.). So the line “all of our ancestors looked sixty thousand years ago something like this” is a simplification. (Neanderthal DNA is found in non-African people today, about 1–2%; from memory.)
Language and symbolic culture. The talk jumps from the breakout to inter-group violence. A fuller account would add language, art and ornaments, which are among the late-appearing signs of ‘modern’ behaviour (the ‘human revolution’, Mellars; or gradual accumulation, McBrearty and Brooks)†.

📚 Roots: Hominin evolution: the steps Wilson summarises

The problem. What changed in the six million years since the split from the chimpanzee line? Wilson’s list follows the standard narrative with a few emphases.

Bipedalism. Ardipithecus ramidus (“Ardi”, White et al., Science, 2009), Australopithecus afarensis (“Lucy”, 1974) and the Laetoli footprints (about 3.6 million years old) show upright walking long before big brains. Owen Lovejoy (1981; and his reading of Ardi in 2009) argued that bipedalism freed the hands to carry food to a mate and young: a pair-bonding story. Wilson does not use it; he leaves the cause of upright walking open.
The brain surge and meat. Brain size roughly doubled between early Homo and Homo erectus. Leslie Aiello and Peter Wheeler (1995), the expensive-tissue hypothesis: guts are costly, so a better diet that allows a smaller gut pays for a bigger brain. Wilson’s version: more meat.
Cooking and fire. Richard Wrangham, Catching Fire (2009) argued that cooking made food easier to digest and drove the evolution of Homo erectus. This is the link Wilson draws. The archaeological evidence for controlled fire that early is thin and contested (see the caption warning).
The campsite as “nest”. Glynn Isaac’s “home base” hypothesis (1978) said early hominins carried food to a base and shared it there. Later archaeologists have questioned it, but it is the model Wilson uses: a defended site to which foragers return. This is the key move of the whole talk. He turns the textbook camp into the eusocial “nest”.
Cooperative breeding. Sarah Blaffer Hrdy, Mothers and Others (2009) argues that human infants were raised by many carers (“alloparents”), which selected for social-emotional traits including mind-reading. This is a rival and partly complementary account of how care became collective. It does not need war.
Social brain and gossip. Robin Dunbar (1990s): primate brain size tracks group size, and language replaced grooming as the glue. Wilson’s “we worry all the time” and “gossip all the time” sit in this tradition.
Shared intentionality. Michael Tomasello, The Cultural Origins of Human Cognition (1999) and A Natural History of Human Thinking (2014). Humans uniquely share goals and attention. Wilson’s “reading intention” is the same skill under another name (also called “theory of mind”, Premack and Woodruff 1978).
Egalitarian bands. Christopher Boehm, Hierarchy in the Forest (1999). Hunter-gatherer groups suppress would-be bullies by coalition. A “reverse dominance hierarchy” is a way groups keep individual selfishness in check. It is a candidate mechanism by which “group selection engenders cooperative behaviour” could work in practice.
Out of Africa. The genetic and fossil picture of a dispersal of Homo sapiens from Africa, with the main wave dated to roughly 60,000 to 70,000 years ago (the number in the talk,). Older sapiens fossils (Jebel Irhoud, Morocco, about 300,000 years) and earlier exits are now known†.

Where others push back

The just-so risk. Gould and Lewontin (1979) warned against fitting a neat adaptive story to fossils. A chain of steps that all point to one outcome (eusociality) is attractive for exactly this reason.
Different order of steps. Pair-bonding (Lovejoy), cooperative breeding (Hrdy), or the shift to extractive foraging (Kaplan et al.) can each be put first.
Dating claims. Controlled fire at two million years, a fully bipedal Ardipithecus, and the sharp brain surge are all disputed in details (see the caption warning).
Meat figures. The headline contrast between chimps and humans depends on which hunter-gatherer groups one samples; many foragers get most calories from plants and honey.

What he leaves out. Sex differences in foraging, the role of grandmothers (Hawkes), the origins of language, and the long Pleistocene climate swings (Potts’ variability selection) that affected populations.

Check yourself

Why is the “campsite” step essential to the eusociality argument?
Name two rival explanations for the brain surge, and say what evidence each would need.
Which of Wilson’s hominin claims does he himself flag as a surmise?

4. Tribal instincts, war and “the signature behaviour”

In short. Wilson calls aggression between groups the signature behaviour of early modern humans. He names two instincts: the urge to form groups and see one’s own as better, and constant evaluation of others. The same social skills build bonds and also deceive and manipulate. They carry both the basic virtues and willingness to sacrifice.

🎙 In the talk. Wilson calls “aggression between groups” “the signature behavior of the groups of the newly evolved and spreading humans”. He gives the Yanomamö as a well-known but not unique case: “periods of extreme aggression” run through human history, and quotes William James: “history is a bloodbath”. He adds cave art from 32,000 years ago, with “abundant evidence of killings of people men primarily by spear” which he reads as raids.

He then says humans have “two powerful clearly instinctual traits” that are so common “they’re like air and water part of nature”:

The urge to form groups, easily and arbitrarily, then to see one’s own group as better. In experiments, teams of volunteers “in short order” think themselves “smarter and more trustworthy”.
Constant evaluation of one another, “reading intention”, at which “we are geniuses”. “human beings are obsessively interested in other human beings”. It builds networks that help survival and also “generates … so many of the human traits”, the subject of fiction and literature.

The same skills serve “to form bonds to form groups to cooperate” and also “to deceive manipulate control that’s what so much of human social activity consists of”. They carry the “basic virtues” and the “willingness to the altruistic and self-sacrificing”. In the Q&A he adds that our “obsession with group identity” hobbles us, that we “will always form groups” and will not get “one world with one big collectivity”.

📚 Deeper theory

The minimal-group experiments (Tajfel, 1971) show that group favouritism needs very little. They do not by themselves show that it is genetically selected. It could be a learned or cognitive by-product. Wilson’s step from “universal” to “selected by group competition” is an inference.
Parochial altruism. The pairing of love for the in-group and hostility to the out-group appears in models (Choi and Bowles 2007) and in experiments on oxytocin and in-group trust (De Dreu et al., 2010; from memory, results debated). This is the closest scientific support for Wilson’s “two instincts working together”.
What counts as ‘war’. Raiding with few deaths, feuds, and organised battles are different. Jebel Sahaba (Sudan, roughly 13,000 years old, many skeletons with projectile wounds) is a Late Pleistocene case; Nataruk (Kenya, about 10,000 years, announced 2016, after this talk) is disputed†. Europe’s Palaeolithic record shows few clear war scenes. So Wilson’s cave-art reading goes beyond the evidence, though inter-group killing as such was plausibly common somewhere.
The war-or-peace argument. Rousseauians (Fry, Ferguson) say organised war grew with settlement, storage and hierarchy. Hobbesians (Keeley, Pinker, Gat) say it is ancient. Both sides use evidence on forager death rates, which vary a lot. For Wilson’s argument to work, one needs war to have been frequent enough and group-structured enough to make groups the unit of selection.
Tribalism as modern politics. Haidt (2012) and Greene (2013) apply the same logic to modern politics. Amy Chua’s Political Tribes (2018) is a later popular version (from memory). The links between Wilson’s evolutionary story and current polarisation are tempting but not proven.
Reading minds as a double-edged skill. The ability to model others’ intentions supports cooperation, but also deception (“Machiavellian intelligence”, Byrne and Whiten, 1988). Wilson’s pairing “bonds … and deceive” draws on this.

📚 Roots: Tribal psychology and the question of war

The problem. Wilson says humans have an instinct to form groups and that early humans were warlike. Are both true, and do they work together?

Hobbes versus Rousseau. Hobbes (Leviathan, 1651): life before the state was a “war of all against all”. Rousseau (1755): people were peaceful until property. Every modern argument is a version of this. Wilson sides with Hobbes’ picture of conflict but puts the cause in biology, not in the absence of government.
Muzafer Sherif’s Robbers Cave (1954). Two groups of boys at a camp became hostile once they competed, and cooperative once they had a shared goal†.
Henri Tajfel (1971), the minimal-group paradigm. People assigned to groups by a coin toss, or by liking a painter, still favoured their own group. This is the kind of experiment Wilson describes: arbitrary teams, quickly seeing themselves as “smarter and more trustworthy”. Tajfel and Turner’s social identity theory (1979) built on it.
Jonathan Haidt, The Righteous Mind (2012), and Joshua Greene, Moral Tribes (2013) extend the point: morality “binds and blinds”. Haidt explicitly uses group selection and a “hive switch”.
Napoleon Chagnon (Yanomamö) and Lawrence Keeley, War Before Civilization (1996). Evidence that small-scale societies often have high rates of death in conflict. Wilson names the Yanomamö as one famous case. Critique and counter-evidence: Douglas Fry, Beyond War (2007), Brian Ferguson, and the controversy over Chagnon (Tierney, 2000) argue that war is much younger or that Chagnon’s figures are distorted by contact with outsiders†.
Choi and Bowles, “The coevolution of parochial altruism and war” (Science, 2007), and Bowles (2009). Models showing that, with the death rates seen in forager societies, genes for in-group altruism and out-group hostility could have spread. The strongest scientific support for Wilson’s scenario. Disputed because the death-rate inputs are disputed.
Steven Pinker, Better Angels (2011). Violence has fallen over history. Pinker also publicly rejected group selection (the Edge essay, 2012). Wilson says he has not read the book.
Boyd and Richerson, Culture and the Evolutionary Process (1985); Henrich, The Secret of Our Success (2016). Group competition may favour cultural traits (norms, religions) that pass through learning, not genes. This is the mainstream version of group selection in the human case and avoids many gene-level objections. Wilson touches gene-culture co-evolution only when Brand raises it.

Where others push back

Inference from universality. A trait that is universal might be cultural or by-product, not an adaptation for group conflict.
Group conflict need not be violent. Competition between groups could be about trade, status and migration. Selection may reward groups with better tools and norms, which is a cultural process (Henrich).
Human peacefulness. Evidence that humans readily cooperate across groups (trade, marriage exchange) and that small bands are fluid, with people moving between them, suggests that strict ‘group’ units were weak (Hill et al. 2011, from memory).
Individual benefits of conformity. Pinker and others argue that tribal feelings can be explained by individual-level advantages of coalitions, with no group-level selection.

What he leaves out. The evidence on reciprocity (Trivers 1971; Axelrod 1984), reputation and punishment (Fehr and Gächter 2002), which many scientists think explain much human cooperation without group selection.

Check yourself

Which two instincts does Wilson name, and what experiment does he cite for the first?
Why does the strength of Wilson’s argument depend on how common war was?
Name one way tribal psychology could exist without having been selected by group conflict.

5. Multi-level selection against kin selection: the technical heart

In short. Individual selection within groups breeds selfishness, and group selection breeds cooperation. Wilson says this multi-level selection explains social evolution. He attacks kin selection, saying Nowak and Tarnita showed its mathematical basis is unsound. He admits it is not settled and that science is not done by polls.

🎙 In the talk. Wilson sets up the two types of selection with a bird-colour example: individual selection shifts gene frequencies when individuals compete with others in the same group. Group selection shows in social insects, where “some termites are the most warlike of all known creatures” and colonies destroy neighbours. He cites Darwin’s use of group selection to explain the worker caste in the Origin and humans in the Descent. The sum is multi-level selection: “overall individual selection within groups engenders selfish behavior”, and group selection “engenders cooperative behavior”.

The attack on kin selection comes: “two mathematicians of the first rank”, Martin Nowak and Corina Tarnita, showed that “the basic assumption of the inclusive fitness theory are mathematically unsound” (spanning). He explains kin selection as altruism among close relatives who share genes. Then: “unfortunately that is mathematically incorrect”. The argument caused what he calls a “fluttering of the dovecote”, but after 18 months “none have addressed the mathematical argument” and there is no critique of the multi-level alternative. He adds the caveat: “in terms of opinion oh this is not settled” and “the science is not conducted by polls”. He closes the talk with “we now have got to clear the wreckage of kin selection off the road”.

In the Q&A, Brand asks how group selection could be falsified. Wilson answers with the haplodiploidy “test”. Kin selection was said to be proved because most eusocial cases are in the Hymenoptera, where “workers more closely related in shared genes to each other than they are to mother”. To him this “was Newtonian”. But when other eusocial species were found without that genetics, “that was a test it failed” and yet “they kept on believing it”. For group selection, he admits “we have yet to develop rigorous tests of group selection in the field”, but points to the finding that faster-growing colonies win (“rapid growth of colonies clearly gives that superiority”) and to the “so far invariant pathway” to eusociality as evidence. He expects a “whole new mathematical theory of multi-level selection”, since earlier kin selection ran “without any mathematician trying to plunge” into its assumptions.

📚 Deeper theory

What Nowak, Tarnita and Wilson (2010) actually did. They argued that Hamilton’s rule holds only under restrictive conditions (additive, pairwise, weak selection) and that a standard population-genetics model of a gene for staying in the nest (with relatedness arising from the colony structure) predicts eusociality without invoking “inclusive fitness”. They concluded that inclusive fitness theory “is not needed” and “does not provide additional insight”. (Paraphrase of the paper’s claim; from memory.)
The rebuttal. Abbot et al. (2011) said the paper (a) confused the general theory (Hamilton’s rule can be derived for general conditions via the Price equation) with a specific modelling approach, (b) ignored empirical tests of kin-selection predictions, and (c) misdescribed the evidence. Gardner, West and Wild, Bourke, Rousset and Lion gave separate replies. Allen, Nowak and Wilson (2013) responded. So the claim in the talk that nobody had “addressed the mathematical argument” is not right, though one can argue the replies did not accept its premise.
The equivalence point (the one to remember). Kin selection, inclusive fitness and MLS are different bookkeeping for the same population genetics, in many cases. Hamilton himself (1975) showed this. Marshall (Social Evolution and Inclusive Fitness Theory, 2015) is a textbook treatment. So a fair summary: Wilson is right that group-level thinking helps in cases where relatedness is hard to define, and that Hamilton’s rule is not a universal law; his critics are right that this does not make relatedness unimportant.
Empirical tests that favour kin-selection thinking. Sex-ratio studies (Trivers–Hare); policing of worker egg-laying in honeybees, where workers eat each other’s eggs when queens mate with many males (Ratnieks); the monogamy result (Hughes 2008). Wilson’s talk does not mention any of these. The later literature has also tested the 2010 claims directly: Liao, Rong and Queller (2015) found that relatedness correlates with the evolution of worker-sterility across many eusocial lineages†.
Group selection tests that Wilson could cite. Colony-level selection experiments (Wade 1977, Goodnight and Stevens 1997), the evolution of ‘organismality’ (Queller and Strassmann), and the Price-equation analysis of field data. They exist but are fewer, which is why Wilson says rigorous field tests are lacking.
Falsifiability of the two views. Popper’s criterion says a theory should forbid something. Kin selection forbids helping non-relatives under many conditions; MLS in its general form (Price) forbids little, because it can describe any case. That is one reason critics call it a re-description, not a competing theory. Wilson’s alternative, a specific model of the nest with two levels, is falsifiable, and some of its specific predictions have been tested.
How the fight became public. The 137-author letter, the rare use of “Nature” for a sharp public reply, and later popular reviews of the book (including Dawkins, 2012, in Prospect, and Coyne’s) turned it into a culture war inside biology†. See the links in Part 1 for reaction.

📚 Roots: Levels of selection: the long quarrel about groups

The problem. At what level does natural selection “see”? The gene, the individual, the group, the species? This is a conceptual and mathematical dispute, not just a factual one.

Darwin on tribes (1871), above.
V. C. Wynne-Edwards, Animal Dispersion in Relation to Social Behaviour (1962). Argued that animals restrain their own breeding “for the good of the group”. A naive form of group selection.
John Maynard Smith, “Group selection and kin selection” (Nature, 1964), and G. C. Williams, Adaptation and Natural Selection (1966). The demolition. Groups with altruists are open to cheats: one selfish mutant out-breeds its neighbours, so within a few generations the group is selfish. Group selection needs groups to go extinct faster than cheats take over. Williams said that individual-level explanations should be preferred (parsimony). Richard Dawkins, The Selfish Gene (1976) spread this gene’s-eye view to the public. This is the “recent years … disfavored” consensus Wilson mentions.
George Price (1970, 1972). The Price equation splits any change in a trait into a between-group part and a within-group part. It is a bookkeeping identity, not a model. This is the mathematical heart of multilevel selection, and it is what Wilson’s “selfish within, cooperative between” framing restates. Hamilton (1975) took it up and showed that kin selection can be seen as a case of group selection, with the groups being families.
Michael Wade (1977) and others†. Lab experiments on flour beetles in which groups were bred for low or high population size, showing a group-level response to selection under the right conditions.
Sober and D. S. Wilson, Unto Others (1998); D. S. Wilson and E. O. Wilson, “Rethinking the theoretical foundation of sociobiology” (Quarterly Review of Biology, 2007). The modern MLS manifesto. The 2007 paper is the immediate source of the talk’s framing.
Samir Okasha, Evolution and the Levels of Selection (2006). The philosophical standard. Okasha separates MLS1 (selection of particles within collectives, as the Price equation tracks) from MLS2 (selection of whole collectives, with their own fitness). Many fights are people talking about different ones.
Traulsen and Nowak (PNAS, 2006) modelled MLS in finite populations. Wilson says “I think that I will see a whole new mathematical theory of multi-level selection develop”.

The key conceptual point for this talk. Most specialists now accept that kin selection and MLS are mathematically equivalent for many cases (both partition the same covariance). The argument is then over which framing is more useful and less misleading. Wilson’s position is stronger than “MLS is a nicer description”. He says kin selection is wrong. That is why he met such resistance. (This is my reading of the debate, from memory; Birch and Okasha 2015 is the paper to check.)

Where others push back

“Mathematically incorrect” overstates. The rule rB > C is a valid result under stated assumptions; the 2010 paper shows its limits. Calling it ‘unsound’ is stronger than most mathematicians who checked it would say (from memory; the exchange is technical).
Timing. Eighteen months is short in science; the replies came within months, not years.
“Science is not conducted by polls” is true, but a 137-author rebuttal is not a poll; it is a set of arguments.
Wilson’s own positions shift. He leaned on kin selection in 1975 and says so. Being right now does not make the earlier work wrong in its time, and critics note that his reversal came with Nowak, an expert in mathematical models, not with a field-data programme.

What he leaves out. The equivalence of the frameworks, the monogamy result, the sex-ratio data, and any mention of the 137-author letter. Also the cultural group selection route, which would give him human evidence without needing a gene for group cooperation.

Check yourself

State Hamilton’s rule in words. What does r stand for?
In what sense can kin selection and multilevel selection be “the same maths”? What does Wilson say instead?
What makes a theory falsifiable, and which of the two views does that test hit harder?

6. A permanently conflicted species: instability, ethics and the future of violence

In short. Because the two levels of selection pull in opposite directions, Wilson says we are a conflicted species and will stay one. Pure individual selection would dissolve society, and pure group selection would make us angelic robots. There is no ideal society to move toward. We can only mute the conflicts, through understanding ourselves and building cooperation and tolerance.

🎙 In the talk. Because the two levels of selection pull in opposite directions, the result “is unstable we are a conflicted species”, and “we will be species forever because we cannot go to either extreme”. Pure individual selection “would dissolve the society”; pure group selection “would make us angelic robots”. This, he says, explains why people are conflicted “particularly just within their own families but in their own group and also internationally”. He wonders whether this is the only route to advanced intelligence and social organisation anywhere in the universe: “maybe that’s the only way that it’s ever occurred”.

In the Q&A he adds that “we’re permanently unstable we’re permanently conflicted”. There is no ideal society to move toward. “what we have is a muting self-understanding and a muting of the conflicts”; we should understand ourselves “not as errant primates that just haven’t found its way to an ideal social condition” and work through “conflict resolution and mechanisms for more cooperation and tolerance”.

Brand reads a question on Pinker’s Better Angels and the decline of violence. Wilson says plainly “but I haven’t read his book yet”. Brand says Pinker documented it thoroughly and asks whether genes are involved; Brand suggests (this is Brand’s reading, not Wilson’s) that groups that cooperate well beat those that do not. Wilson’s answer concerns fitness in general: “who uses the resources best and fastest is the species that can displace other species”, with ants as the example of success by organisation, not by violence alone.

On the future: “we’re going to put a lot more emphasis on conflict resolution” and “a lot less emphasis upon idealistic ideologies”. In an earlier answer he said the Argentine-ant style of unity is not for us: “we’re not going to have one world with one big collectivity”. And he is wary of “the price of are all becoming wimps” (Brand’s words in the caption) with a “lovable humanity”.

📚 Deeper theory

The old idea of a divided human nature. Wilson restates a very old picture in biological form: Plato’s tripartite soul, Augustine’s two cities, Kant’s “crooked timber of humanity” (from the Idea for a Universal History, 1784), and Freud’s Civilization and Its Discontents (1930). The twist is that he gives the conflict a mechanism (selection at two levels) and says it is permanent. Compare Haidt’s “we are 90% chimp and 10% bee” (The Righteous Mind, 2012; from memory).
A hybrid, not a pure type. Humans fall between a strict eusocial colony (no internal conflict, as in ants) and a pure individualist species. Queller and Strassmann’s “organismality” scale would rank humans low on cooperation and high on conflict. Wilson’s “instability” is the same point.
The ethical implication he draws is modest. Understanding the conflict will not abolish it. This is close to Isaiah Berlin’s value pluralism and to the tragic view of politics (Reinhold Niebuhr, Moral Man and Immoral Society, 1932: groups are more selfish than individuals). It is not a recipe. He hopes for self-understanding.
Is-ought. Hume’s guillotine (Treatise, 1739) says one cannot derive moral conclusions from facts alone. Wilson’s Consilience argued that ethics could be grounded in biology; philosophers such as Peter Singer (The Expanding Circle, 1981) accept that evolution explains our moral emotions but argue reason can extend them. In this talk Wilson stays on the descriptive side.
The decline of violence and the biology question. Pinker explains the decline by the state, commerce, literacy, and the “expanding circle”, not by genetic change. In the exchange, Brand wonders if genes might be “selected for” gentleness. Evidence for such a recent genetic change is thin. The decline is mostly regarded as cultural. Richard Wrangham has a competing view on self-domestication (The Goodness Paradox, 2019; from memory), in which groups killed their most aggressive males, lowering reactive aggression, which is a group-selection-like process.
Anything else evolving? In the Q&A Wilson says “progressive” evolution is “just not occurring”, but that new genotype combinations are growing (see idea 8). Population geneticists such as John Hawks et al. (2007) argued that adaptive evolution has accelerated in the last 10,000 years†. Wilson’s phrase ‘progressive’ evolution is narrow (bigger brains, better looks), so the two claims do not necessarily conflict.

📚 Roots: Sociobiology and its politics

The problem. Wilson’s career since 1975 has been shaped by one fight: can biology explain human social behaviour without justifying injustice?

The 1975 reaction. A letter from the Sociobiology Study Group (including Stephen Jay Gould and Richard Lewontin) accused the book of supporting biological determinism. In 1978 a protester poured water over Wilson at a scientific meeting (from memory; widely reported)†. Brand’s remark that sociobiology was “later proved to be quite right” takes Wilson’s side of that fight.
Gould and Lewontin, “The spandrels of San Marco” (1979). The critique of “adaptationist storytelling”: not every trait is an adaptation. A relevant warning for the hominin “just-so” story in the talk.
Marshall Sahlins, The Use and Abuse of Biology (1976). An anthropologist’s critique: human culture cannot be reduced to kin-selection arithmetic. Sahlins returned to the attack on Wilson in 2012–13 and resigned from the National Academy of Sciences†. Note: Sahlins and Wilson both reject kin selection, for opposite reasons. Sahlins thinks culture is not biology; Wilson thinks multi-level selection is better biology.
Gene-culture coevolution. Lumsden and Wilson, Genes, Mind and Culture (1981); Boyd and Richerson; Kevin Laland. Culture and genes shape each other, as lactose tolerance did in herding people, an example Wilson uses.
Evolutionary psychology (Cosmides and Tooby, The Adapted Mind, 1992). Modular mental adaptations. Tooby is also a leading opponent of Wilson’s group selection claim. The “instincts” Wilson lists fit the evolutionary psychology tradition, but his explanation for them (group selection) is the part its leaders reject.

Where others push back

Permanence. A permanent, balanced conflict is a strong claim. Boehm, Henrich and others describe institutions (norms, punishment, religion, states) that shift the balance a lot. If culture can do this, the “instability” is not a fixed fact but a variable.
Gradual moral change. The abolition of slavery, falling homicide rates, and expanding rights are evidence of large change in behaviour with no change in the genes.
The ‘wimps’ worry. Brand’s remark ties peace to weakness. Data on societies that value cooperation (Scandinavia, say) do not show that they are less successful in competition, though the comparison is loose.

What he leaves out. Institutions: law, markets, states, federations and the evidence for how they change cooperation (Ostrom’s commons work, 1990; North, Wallis and Weingast’s “limited access orders”, 2009; from memory). Also: what, practically, “conflict resolution” would look like at scale.

Check yourself

Why does multilevel selection predict a permanent conflict rather than a stable compromise?
How does Wilson’s view differ from the claim that humans simply need better education to reach peace?
What would count as evidence that a fall in violence has a genetic part?

7. Ants as the model: superorganisms, supercolonies and distributed intelligence

In short. Eusocial means true: a society with division of labour, altruism, overlapping generations and shared care of young. Humans differ from insects through cultural overlay. Some ants become supercolonies through a single mutation, but they appear mostly in disturbed areas. Ants also work as superorganisms with distributed intelligence.

🎙 In the talk. Brand asks about herds and flocks: what is special about eusocial? Wilson: “it just means true”, a society with division of labour, altruism among competing individuals, two or more generations and cooperative care of young. Humans differ from insects because of “this immense overlay of cultural modification of our behavior”.

Brand then asks why humans do not become one supercolony like the Argentine ant that “lately invaded the west coast”. Wilson describes how some species, including the Argentine ant and the fire ant, went from colonies “always at war” to supercolonies “by a single mutation”, which silences distinguishing one colony from another and tolerating many queens. Colonies smell alike: “every colony of ants has it has a peculiar its own peculiar blend of hydrocarbons”, “like you know they precise bouquet of a wine”. They lose that sense and the multi-queen limit. Why does the world not fill up with supercolonies? Because they “usually appear only in disturbed areas”; in natural woodland there may be 10 to 40 ant species, each with its own defence methods, and a supercolony cannot “penetrate easily”.

Later he explains ants as true superorganisms, “operating like one big organism”, with “distributed intelligence”. Some ants remember five food sites and the time of day to visit them; scouts know their part of the territory; others nurse or build; the result “enormously enhances the competitive power productivity and the competitive strength of a colony over an individual organism”.

📚 Deeper theory

A supercolony is the opposite of a superorganism. In a normal ant colony, relatedness is high and non-members are enemies. In a supercolony, workers cooperate with huge numbers of unrelated nestmates. The invasive supercolonies are a puzzle for kin selection, since relatedness within them is close to zero (Helanterä et al. 2009 review; from memory). Wilson uses them for a different purpose: as a model of what unification would look like, and why it fails in diverse systems.
Why they succeed in disturbed areas. In native habitats, aggression among nests and the competition with other ants limit them. After introduction they escape their enemies and parasites, drop internal aggression and win by sheer numbers (Tsutsui and Case; Holway et al. 2002; from memory). In California the Argentine ant displaced native ants over large areas.
The ‘single mutation’ detail. The fire ant social structure is controlled by a “social chromosome” carrying the Gp-9 gene (Krieger and Ross 2002); the supergene was characterised in detail in 2013 (Wang et al., Nature), after this talk†. The mechanism involves workers killing queens that do not carry the right allele, so it is not simply a silenced recognition system. Wilson’s description is simpler.
Distributed intelligence and stigmergy. Pierre-Paul Grassé (1959) coined stigmergy: workers coordinate by changing the environment (pheromone trails) rather than by messaging. Deborah Gordon’s work (for example on harvester ants) shows that task allocation arises from encounter rates (from memory). Swarm intelligence became an engineering field (Dorigo’s ant colony optimisation, 1992; Bonabeau, Dorigo and Theraulaz, 1999). The same idea gives an alternative reading of “collective intelligence” in human groups (Hutchins’ Cognition in the Wild, 1995; Malone and Woolley).
Learning in ants. Wilson says ants learn simple mazes more slowly than rats, “twice the time”. The point is that individually limited animals combine knowledge across individuals. The honeybee dance (von Frisch; Nobel 1973) is the classic communication system in this picture.
War among ants. Wilson stresses that termites and ants fight. Field studies of army ants, weaver ants and Pheidole species show that large colonies win contests (the ‘Lanchester’s laws’ work by Plowes and Adams, 2005, from memory), which supports the claim “rapid growth of colonies clearly gives that superiority”.

📚 Roots: Major transitions and the superorganism

The problem. Cells joined into organisms, organisms into colonies. How do cooperating units stop competing and become one individual?

William Morton Wheeler, “The ant-colony as an organism” (1911). The colony is the organism. Wilson trained in this tradition at Harvard.
Alfred Emerson (termites, 1930s to 1950s) (from memory). Developed the “superorganism” idea for termites.
Leo Buss, The Evolution of Individuality (1987). Multicellular bodies result from cell lineages being forced to cooperate; the germ and soma split is the original queen and worker division.
John Maynard Smith and Eörs Szathmáry, The Major Transitions in Evolution (1995). A list of transitions (genes into chromosomes, cells into multicellular organisms, individuals into eusocial colonies). In each, what used to reproduce alone now reproduces only as part of the larger whole. Eusociality is the most recent transition in animals.
David Queller and Joan Strassmann, “Beyond society: the evolution of organismality” (2009). The test of an “organism” is cooperation and low conflict. By this test many ant colonies are organisms and human societies are not.
Hölldobler and Wilson, The Superorganism (2009). The immediate source of the talk’s ant material.
Thomas Seeley, Honeybee Democracy (2010)†. Bee swarms choose a nest site by a distributed vote. A clean example of the “distributed intelligence” Wilson praises.

What this adds. In the transitions view the key move is suppressing conflict inside the unit. Wilson takes the opposite line for humans: we never completed a transition, so we are permanently pulled both ways. That matches Queller and Strassmann’s point that human societies are not organisms. But they would not put the cause where he puts it (see idea 3 and idea 4).

Where others push back

Analogy limits. Human groups are not colonies: our members reproduce, move between groups and have their own goals. The “superorganism” analogy was used to defend ugly politics in the past (organic-state ideas of early 20th-century nationalism), which is a reason for caution (from memory).
Supercolonies and kin selection. Cases like these are a challenge to a simple relatedness account, but they are rare and mostly invasive (a sample with a bias).
“Smart” ants. The credit for colony cognition varies; many results come from simple rules, not memory-rich individuals.

What he leaves out. The costs of ant society: parasites, social parasitism (slave-making ants), and worker policing of reproduction, which are the sort of internal-conflict data that bear on his claim about humans.

Check yourself

What is the difference between a superorganism and a supercolony?
Why do supercolonies thrive mainly in disturbed places?
What is stigmergy, and how does it relate to Wilson’s “distributed intelligence”?

8. Humanity’s future biology: homogenisation, diversity and the limits of “progress”

In short. Wilson says humans are not evolving progressively: no bigger brains and no better looks. The change now is homogenisation between places, with more variation inside communities, which he likes. He expects the genetic basis of traits such as empathy and aggressiveness to become clearer. He says we will learn what makes us human.

🎙 In the talk. Asked whether humanity is still evolving, Wilson says there was a lot at the time of the breakout, mostly “resistance to disease and the ability to acquire new food products”, for example “independent origin of adult lactose tolerance in Europe Eurasia and also in Africa” which allowed herding as a livelihood. But “progressive evolution it’s just not occurring” and “we’re not getting bigger brains we’re not getting any better looking”. What is happening is homogenisation: variation between places is falling as people travel and intermarry, “theoretically we’ll end up with people in Beijing and in Stockholm being the same statistically”. At the same time variation inside communities is increasing: “we are creating a large number of new human genotypes”. He likes this, since “our species … depends upon your diversity”, and expects “new forms of genius”, jokingly “maybe someone will run a two hour marathon and maybe it won’t be from Kenya either”.

He also looks ahead to human genetics: a time when “we’ll be able to get some idea of the genetic basis” of personality traits such as empathy, aggressiveness and assertiveness, with heritability “as much as one half” in twin studies. He says “we will learn what makes us human”, and that identifying the ‘spring-loading’ genes in social insects is the next step.

📚 Deeper theory

‘Race’ and human variation. Lewontin (1972) found that about 85% of human genetic variation lies within populations and about 15% between continental groups. Wilson’s remark that variation was “Geographic” and is now shifting inwards matches this general picture in direction. The claim that people in Beijing and Stockholm will be ‘the same statistically’ is a limiting case that needs centuries of mixing and ignores local adaptation (skin colour, altitude, lactase). Treat as rhetorical.
Is human evolution continuing? Yes by most population geneticists: lactase persistence (Tishkoff et al. 2007 found several independent African mutations, which supports Wilson’s example), high-altitude adaptation in Tibetans (EPAS1, 2010), malaria resistance, and many weaker signals. Hawks et al. (2007) claimed a recent speed-up. Others (Sabeti; Pritchard) find strong signals but argue about interpretation. Natural selection also continues through differential fertility in modern populations (Stearns et al. 2010, from memory). Wilson’s ‘progressive’ qualifier is the safe part of his claim.
New genotypes and “genius”. Mixing creates combinations but not necessarily better ones. Whether it produces more extreme performance depends on how traits are built from many genes (polygenic architecture). The marathon joke touches the real question of why East African runners are fast (a mix of altitude training, culture and some genetics; debated). As an aside: the record has fallen since (Kipchoge ran 1:59:40 in 2019 in a non-record event; official records moved below 2:01 in recent years)†.
Personality heritability. Twin studies show roughly 40 to 50% heritability for the Big Five (Bouchard and McGue 2003; Polderman et al. 2015 meta-analysis of twin studies, from memory). The genes found so far by genome-wide association studies explain little (“missing heritability”). Wilson expects that these genes will eventually be mapped has been only partly met: many tiny effects, no simple ‘empathy gene’.
Gene-culture co-evolution is the link. The lactose example shows that culture (dairying) selected genes. Wilson uses it to say that biology and history are one process (also idea 9’s theme of ‘history needs prehistory’).
Ethics of genetic knowledge. The history of eugenics (Galton; the US sterilisation laws; Nazi Germany) is why critics of sociobiology worried about claims on heritable traits. Wilson in 2012 does not engage with this, though Sociobiology and Consilience did, at length (from memory).

Where others push back

Geneticists would say variation is not shrinking that fast. Population structure remains strong in much of the world; marriage is still local for most people.
Selection in modern populations is not zero and can be fast on traits linked to fertility, disease and diet.
‘Genius’ is not just genes. Training, opportunity and culture dominate outcomes; claims about “new forms of genius” mix genetics with sociology.

What he leaves out. Genome editing and assisted reproduction, which were already moving fast by 2012 (CRISPR-Cas9 was first described as a programmable tool later that year)† and would make ‘evolution’ partly a human choice.

Check yourself

Why can the human gene pool become more uniform between places and more varied within them at the same time?
Which of Wilson’s examples show that humans evolved after the breakout?
Why has “find the gene for trait X” worked poorly for personality?

9. Beauty, habitat and arts: the sensory shortcuts

In short. Wilson says human senses make people less rational than they think. Some patterns arouse the brain most, which he links to designs in art and logos. In cross-cultural tests people choose to live on a rise, with a view over parkland and near water, and pay high prices for it. He calls this a residual habitat preference.

🎙 In the talk. Briefly at the end of the prepared talk, Wilson says that human sensory limits make people less rational than they think. Some patterns “arouse” the brain most, and he links that to the complexity of designs in so-called primitive art and modern logos. He then cites “cross-cultural tests” showing that when people can choose where to live they pick three features: height (“on the rise”), a view over parkland or savanna, and water nearby, and they will “pay extremely high prices” for it “regardless of whether there’s any utilitarian value to it”. He says it is instinctive and “a residual habitat preference” that animals show and need for survival. He also says we show irrational behaviours that are “the residue of that long prehistory”.

📚 Deeper theory

Savanna hypothesis. Orians and Heerwagen (1992) argued that humans prefer savanna-like scenes (scattered trees, open grass, water). Balling and Falk (1982) found that children of 8 to 11 preferred savanna to other biomes, even in places without savanna. Critics note that preference may come from familiarity and from learning about what is typical in a given culture (Ulrich’s stress-recovery studies, 1984, support nature in general rather than savanna specifically; from memory).
Prospect-refuge theory. Jay Appleton (The Experience of Landscape, 1975) proposed that we like places from which we can see without being seen. Wilson’s “on the rise” and “look out” fit.
Biophilia. Wilson (1984) and Kellert and Wilson (eds., 1993): an innate tendency to attend to life and lifelike processes. Evidence includes quick learning of fear for snakes and spiders (Öhman) and benefits of green space on health. The hypothesis is broad and tested only in parts.
Complexity and arousal. Berlyne’s (1971) work on aesthetics found that preferences peak at moderate complexity. Wilson links it to alpha-wave blocking in the brain. The studies he alludes to concern visual patterns with a certain amount of redundancy (he gives about 20% in Consilience)†. Art historians and anthropologists note that tastes in decoration vary widely and that “complexity” is hard to measure across cultures.
Evolutionary aesthetics. Denis Dutton, The Art Instinct (2009), argues that the arts have evolutionary roots; Ellen Dissanayake (What Is Art For?, 1988) argues for ‘making special’. Others (Pinker, How the Mind Works, 1997) called the arts ‘cheesecake for the mind’, a by-product. The disagreement mirrors the eusociality fight: adaptation versus by-product.
Why this is in a talk on eusociality. It supports Wilson’s plea that “prehistory” leaves marks on present choices. He is using it to sell a research programme, not to prove the main thesis.

Where others push back

Culture and class shape what people say they prefer and what they pay for. High property prices for views may reflect status.
Savanna evidence is mixed. Cross-cultural tests do not always find a savanna preference, and exposure explains a lot.
By-product accounts. A preference for water and views can be explained by safety and resource cues without a dedicated adaptation.

What he leaves out. Any link from this section to group selection. It is a side-show in the prepared talk, included because the book covers it.

Check yourself

What three features does Wilson say people choose in a habitat?
What is the difference between an adaptation and a by-product, with an example from art?
How could you test whether a preference is innate and not learned?

10. Religion, creation stories and the future of belief

In short. Wilson expects organized religions to decline, especially their creation stories. Stories get set in stone, people join groups and demand fidelity, and attacking the story feels like attacking the tribe. He says he is an atheist, but that believers and he are natural allies in saving biodiversity, which he calls creation. Some believers, he says, see earth as a way station and do not worry about species.

🎙 In the talk. Asked about the future of eusociality, Wilson says “we’re just going to find out how to get along together” and expects more “conflict resolution” and fewer “idealistic ideologies”. He expects “a decline in organized religions but especially with reference not transcendent or spiritual seeking but rather the creation stories”. Why? Because the stories get “set in stone and people then belong to groups and they feel they have to have complete fidelity to that creation story of their tribe”, and when you question evolution “they get hostile why because you’re attacking their tribe”. He expects the decline “in spite of what’s happening in the United States right now”.

Brand points to Wilson’s book The Creation, written as letters to a pastor, which is gentle to Christians, and contrasts it with this talk. Wilson explains: “I said right from the start I’m an atheist”, different worldview, but “we are natural allies” in saving biodiversity, which he calls “creation”. He had support from the Mormon Church leadership, visited Southern Baptist strongholds, but “it didn’t go very far”. For many Christians, he says, the earth is a “way station”, the far right “believes that you should not worry about the environment”. About Revelation he says it shows “all the signs of paranoid schizophrenia”, and that extreme believers think “we should be worrying about saving souls” and not species. The point is not to defend the lecture’s tone; he says “it’ll get me in trouble”.

On Alabama: “I’m very popular there and one of the reasons I’m popular is that I love them”. He grew up there, and after being baptised as a child (“I went under the water when I was a kid”) the community expects him to “come back into the fold” before he dies, and “I’m not going”.

📚 Deeper theory

Is it really the creation story that creates hostility? Wilson’s mechanism (a story that defines the tribe makes doubt feel like betrayal) is Durkheim plus social identity theory. Research on ‘sacred values’ (Atran and Ginges) and on ‘identity-protective cognition’ (Kahan) supports the idea that people defend group-linked beliefs even against evidence. Kahan’s work on cultural cognition (2010s) found that the more scientifically literate people on both sides were more polarised on contested topics like climate change; this suggests tribal identity, not ignorance, drives resistance to the science†.
Will creation stories decline? Pew’s 2023–24 survey finds a big fall since 2007 and then a plateau: 62% Christian, 29% unaffiliated, a mix “fairly stable” since about 2020 (Pew). Gallup’s long series on belief in creationism (about 40% of US adults saying God created humans in their present form within the last 10,000 years) has hovered near the same number for decades, then slipped in recent surveys†. So Wilson’s forecast has direction right but timing and extent open.
‘Nones’ are not secularists. Many unaffiliated people hold spiritual or religious beliefs. Wilson’s distinction between “organised religion” and “spiritual seeking” is closer to this than to the New Atheists’ line.
The evangelical ‘creation care’ alliance. Wilson’s strategy has partners: the Evangelical Environmental Network, Katharine Hayhoe’s climate outreach, and the 2006 ‘Evangelical Climate Initiative’. The Southern Baptist Convention formally debated environmental resolutions in 2007–2008†. The obstacle he names, end-times theology (dispensationalism), is real in American evangelicalism, though many evangelicals reject it.
Reading Revelation. Scholars (Pagels 2012; Collins; Barr) place it in the Roman imperial context as a coded protest against Rome. Wilson’s psychiatric diagnosis is a joke, a provocation, and also a case of what he warns about: judging a text from outside its own terms. Brand’s own question had already pushed him to say it is a theme “in the book”.
A friendly atheist. Wilson’s talk shows a method: state your difference first, find the shared goal, and keep the relationship. It echoes work on bridging (Putnam) and the practice of ‘accommodating’ dialogue between science and religion (compare Gould’s NOMA or Francis Collins’ BioLogos), though Wilson would reject accommodation on facts.
His own religious history. Wilson grew up Southern Baptist and left through reading Darwin (from his memoir Naturalist, 1994)†. His warm treatment of Alabama is consistent with that history.

📚 Roots: Creation stories, religion and science

The problem. If every religion has its own creation story and each claims to be uniquely right, what are such stories for, and what happens when a scientific story competes?

Émile Durkheim, The Elementary Forms of Religious Life (1912). Religion is society worshipping itself. Sacred symbols bind a group. Wilson’s claim that believers react badly “because you’re attacking their tribe” is Durkheim in evolutionary dress. He does not cite him.
Mircea Eliade and Joseph Campbell treated creation myths as the template for ritual and meaning; Wilson reads them as pre-scientific guesses.
Stephen Jay Gould, Rocks of Ages (1999), “Non-Overlapping Magisteria”. Science covers facts; religion covers meaning. Wilson rejects the separation: the creation story is a factual matter. He does leave room for “transcendent or spiritual seeking”.
Daniel Dennett, Breaking the Spell (2006), Pascal Boyer, Religion Explained (2001), Scott Atran, In Gods We Trust (2002). Cognitive science of religion: religion as a by-product of ordinary mind-reading and agent-detection. This is another reading of Wilson’s “reading intention” instinct.
David Sloan Wilson, Darwin’s Cathedral (2002). Religion as a group-level adaptation. The most direct bridge to this talk, but D. S. Wilson thinks religion is functional for groups, while E. O. Wilson regards particular creation stories as harmful.
Ara Norenzayan, Big Gods (2013). Moralising gods helped large groups cooperate. Another group-level function story.
Robert Bellah, Religion in Human Evolution (2011). Religion from tribal to axial forms.
Wilson on religion, in his own books. On Human Nature (1978) and Consilience (1998) treat religion as a product of evolution which science can study. The Creation (2006) is the “letters to a pastor” book Brand refers to.
Religion and the environment. Lynn White Jr., “The Historical Roots of Our Ecologic Crisis” (Science, 1967) blamed Christian dominion over nature. The counter-movement is “creation care” in evangelical circles. Wilson’s strategy in The Creation, calling biodiversity “creation”, comes from that bridge.
Elaine Pagels, Revelations (2012). A scholar’s reading of the Book of Revelation as a coded political text of its time, the opposite of Wilson’s brisk dismissal of it†.

Where others push back

The decline thesis. Globally religion is not declining: the number of religious people is growing, driven by population changes (Pew’s projections, from memory). The US is an exception, not the rule.
The attribution. Stories might be a symptom of tribalism, not a cause. People may cling to creation narratives because they belong to the group, not because the story matters.
Respect for believers. Critics of the “paranoid schizophrenic” remark said it trades goodwill built in The Creation for a joke.

What he leaves out. The positive social roles of religious groups (charity, community), and non-Western traditions where creation stories are less literal. Also, how a scientific origin story would itself work as a group marker (it already does, in some atheist and science-enthusiast communities).

Check yourself

What mechanism does Wilson give for why creation stories provoke hostility?
What does Pew’s latest data say about his forecast of a decline in religion?
Why might he still call himself an ally of evangelicals on conservation?

11. Microbes, the unseen frontier, and never retiring

In short. Wilson calls microbes the dark matter of the biological world. A gram of soil holds a billion bacteria of thousands of species, all unknown, and our lives depend on them. Technology now lets us read a bacterial genome in hours. He says medicine should build up our natural microbial flora, as conservation builds up ecosystems, and he advises never retiring.

🎙 In the talk. Brand reminds him that in Naturalist he wrote that if he were starting again he would study microbial ecology, and asks why. Wilson: microorganisms, “bacteria and archaea”, are “the dark matter of the biological world”. One gram of garden soil holds “one billion bacteria” of five or six thousand species, “all of them unknowns”, and a ton may hold “three million species”. “our lives depend on them” and “everything would die if the bacteria died”. We now have the technology: you can “see an entire bacterium genome in just a couple of hours”. He describes a recent conference in Washington where molecular biologists talked about evolutionary biology and ecology, and the shift from thinking the field ‘for dummies’. One speaker described our co-evolution with bacterial flora: the “500 or more bacteria in the mouth and esophagus” protect us from pathogens. Another group, working on antibiotic resistance: “an arms race that we may lose”. He draws the ecological moral: “disease as being an invasion of our ecosystem of alien species”, so medicine should “build up” our natural microbial flora, as conservation builds up natural ecosystems.

On personal energy: asked for advice for staying active, he gives a list (“genes few alcoholic beverages love the outdoors”) and then “never retire”. Find work you have “always wanted to do” and keep going. Older people have “a lot of canny savvy street smart experience to draw on”. He mentions a book “on why we desperately need you” for young scientists, and a trip to Vanuatu at 80, where he was “treated with great courtesy” because so few people there live to 65. The last words are Brand’s: “there’s a great deal of science yet to come”.

📚 Deeper theory

Microbial dark matter. The idea comes from the gap between what grows in a dish (about 1% of microbes) and what DNA surveys find. Environmental sequencing (Pace; Venter’s 2004 Sargasso Sea sample) and later single-cell methods have kept enlarging the tree of life. A 2016 tree (Hug et al., Nature Microbiology) showed that most known diversity is bacterial, and much of it is uncultured†. The count of species per gram of soil ranges widely depending on the species definition for bacteria, so Wilson’s figures should be treated as order-of-magnitude.
The human microbiome. Lederberg’s idea of a human ‘superorganism’; the Human Microbiome Project’s first results were published in June 2012. Later work on the gut-brain axis, faecal transplants for C. difficile, and microbiome links to disease grew fast, with many weak or confounded findings (from memory). The “ten trillion bacteria” figure Brand mentions is a rough estimate. A 2016 re-estimate (Sender, Fuchs, Milo) put the ratio of bacterial to human cells close to 1:1, rather than 10:1†.
Antibiotic resistance and ecology. Alexander Fleming warned about resistance in 1945 (Nobel lecture). The WHO and others now treat resistance as a major global threat. The ‘ecological medicine’ idea that Wilson picks up, treating pathogens as invasive species and microbiota as a protective community, has parallels in phage therapy, probiotics, and microbiome-sparing antibiotics. Blaser’s Missing Microbes (2014) is the popular case.
Disease as invasion. The comparison with invasive species goes back to Elton (1958), who treated both as ecological events. It is apt for disease outbreaks in new hosts, but it misses that many pathogens are native opportunists that become harmful when the community is disturbed.
Wilson as a microbe convert. Wilson’s life work was ants and island biogeography, so his turn to microbes in 1994 and again here is an example of the ‘next frontier’ habit, and a reminder of Brand’s place in the story: the Long Now interest in slow, long-lived systems (soil, forests, seeds).
Ageing and work. The “never retire” advice matches findings that continued purposeful activity is linked with longevity and health (for example the ‘Blue Zones’ reports and many cohort studies); the causes are hard to separate from the fact that healthier people keep working†. Wilson himself kept publishing until his death in 2021 at 92†.
Respect for older people. His anecdote about Vanuatu is a small field note: the value placed on elders is higher where fewer people reach old age.

📚 Roots: Biodiversity, ant societies and the microbial frontier

The problem. How many species do we need? Why are some invaders successful? What do we not know about life?

Charles Elton, The Ecology of Invasions by Animals and Plants (1958). Diverse communities resist invaders; the idea behind Wilson’s answer on supercolonies.
Robert MacArthur and E. O. Wilson, The Theory of Island Biogeography (1967). The species-area law: larger, closer islands hold more species. Brand’s “which basically revolutionized ecology population biology”. It applies equally to habitat fragments, which is why extinction follows when habitat is cut.
Robert May (1973) versus the “insurance” view. Diversity does not automatically give stability in simple models (May). Later field work (Tilman; Yachi and Loreau 1999, “the insurance hypothesis”) found that diverse systems are more stable on average, because different species cover for each other. Wilson’s “we need lots of species for insurance” is the insurance idea†.
Wilson’s biodiversity books: Biophilia (1984), The Diversity of Life (1992), Half-Earth (2016). Biophilia argues that we have an inborn affinity for living things and landscapes. The habitat preferences he describes belong here.
Ant societies as models. Supercolonies and invasive ants: the Argentine ant (Linepithema humile); work by Neil Tsutsui, Andrew Suarez, David Holway and others on how a genetic bottleneck in introduced populations left nests unable to tell each other apart†. Ross and Keller (2002, fire ant Gp-9 gene) linked a single gene to whether a colony has one queen or many. This is the “single mutation” Wilson alludes to; the mechanism, as I recall it, is that workers carrying one allele kill queens without it, which is not quite “silencing” recognition†.
Microbial ecology. Carl Woese (1977) found the third domain of life (Archaea). Environmental sequencing (Pace; later Venter’s Sargasso Sea study, 2004) showed most microbes had never been cultured. Joshua Lederberg (2001) called the human body a “superorganism” of human and microbial cells. The Human Microbiome Project published its first large results in June 2012, two months after this talk†. Martin Blaser, Missing Microbes (2014) made the ecological argument against overusing antibiotics, close to Wilson’s closing point.

Where others push back

‘Everything would die’ is an overstatement in detail, but the point stands that nutrient cycles depend on microbes.
Hype in microbiome science. Many claims about the gut and disease are correlation without proof of cause.
Retirement advice is not universal. Many people cannot afford, or are not able, to keep working; he is speaking from a position where work is chosen, not forced.

What he leaves out. The commercial side (microbiome start-ups), conservation of microbial diversity, and the question of whether microbial communities show the same group-selection issues his talk is about (they do: biofilms and microbial cooperation are a leading test-bed for social evolution theory; Kümmerli, West, Griffin; from memory).

Check yourself

Why does Wilson call microbes “dark matter”?
What does he mean by treating disease as an “invasion” of our ecosystem, and where does the analogy break?
How might microbial cooperation provide evidence for or against group selection?

Q&A

(Add questions and answers here only if worth keeping.)

About the source and how this page was made

Speaker: Edward O. Wilson (talk + conversation with Stewart Brand)

Event: Long Now Foundation seminar, co-presented with the Exploratorium, Palace of Fine Arts Theatre, San Francisco

Date: 2012-04-20 (Long Now seminar page; the talk itself says Wilson is 82 and the book is “current”)

Uploaded: 2020-05-02

Length: 1:32:10 (transcript end; Long Now podcast listings show 1h32m on one platform)

Channel: The Long Now Foundation

Video: The Social Conquest of Earth | Edward O. Wilson

One thing to know first. This talk was given in the middle of one of the most bitter fights in modern evolutionary biology. Wilson’s central technical claim (kin selection theory is “mathematically incorrect”, and multi-level selection replaces it) was rejected by most specialists in social evolution, and 137 of them signed a rebuttal in 2011. The talk mentions the fight, but says that nobody had answered the mathematics. That is not accurate (see idea 5). Read the talk as a bold, popular, partly contested synthesis, not as the settled view of the field.

Speakers

Label Who Role
Host A Rob Semper (inferred; the captions give “flexors zipper Rob”, and Wilson thanks “Rob”) Exploratorium. Introduces the evening. Not heard clearly in the transcript
Brand Stewart Brand (names himself) Long Now Foundation co-founder. Introduces Wilson, then interviews him and reads questions from cards
Wilson Edward O. Wilson (named by Brand; “Edward o Wilson”) Harvard entomologist and biologist, age 82. Gives the 49-minute talk, then answers
Audience Several people, read out by Brand Names in captions (“Jack Sawyer”, “Bruce Miller”, “Byron”, “William Petry”) are not reliable
”Paul” Paul Gauguin, not a speaker The painter whose 1898 canvas gives the talk its three questions

Confidence. Wilson and Brand: high (named in the talk, content matches the Long Now listing and The Social Conquest of Earth). Rob Semper: medium (from the Long Now listing that an Exploratorium executive introduced; the transcript name is garbled). The audience names: low. The transcript has no speaker labels, so which turns are Brand and which are Wilson is judged from the content: Brand asks the questions and makes the wry remarks, Wilson gives the long answers.

Caption warnings. The captions are automatic and often wrong. Corrections used on this page:

“you sociality / youth sociology / own sociality / you social” = eusociality.

“Ken’s selection / Ken selection / pen selection” = kin selection.

“Martin Novak and Karina carnita” = Martin Nowak and Corina Tarnita. “Bert holdover” = Bert Hölldobler. “the super organism” = the book The Superorganism (Hölldobler and Wilson, 2009).

“Charvet” = Chauvet cave. “Arta pittacus” = Ardipithecus. “Hammond lines” = hominid lines. “Homeland level of intelligence” = hominid-level.

“paleo Dixie Opera illa trop try” is a garbled list of Paleozoic insect orders. Do not trust any name in it.

“Morkul” and “retro dicta” are caption noise. “Retro-addictive test” = a retrodictive (after-the-fact) test.

“medical branch of Ted” = probably TEDMED, held in Washington in April 2012†.

“Vanuatu used to be new forever Jesus” = Vanuatu, formerly the New Hebrides.

“Anatoli different worldview” and similar fragments are noise. The meaning is “an entirely different worldview”.

Factual slips and shaky claims (details in the deep dives):

Gauguin. Wilson says Gauguin intended the 1898 painting “to be his last work before he committed suicide”. Gauguin did try to kill himself with arsenic after finishing it, but he survived and died in 1903. His own account of this is from his letters (from memory; widely reported, not checked here).

“Introduced the term biodiversity”. Brand says this about Wilson in the introduction. The word was coined by Walter G. Rosen for a 1986 forum; Wilson popularised it by editing the proceedings (BioDiversity, 1988)†.

“Early Mesozoic … 150 million years ago” for the ants. The early Mesozoic is about 250 to 200 million years ago. Ant fossils and molecular dates put ants at roughly 140 to 168 million years†, which is late Jurassic or early Cretaceous.

Haplodiploidy. Wilson says workers “do not have a father but they have a grandfather”. Workers are female and have a father. It is males that have no father (they come from unfertilised eggs). The point he wants is that full sisters share 75% of their genes by descent. Either a caption error or a slip.

Cave art at 32,000 years showing “raids”. Wilson says the art shows killings by spear and vengeance raids. Scenes of group fighting are rare in Palaeolithic art. Clear battle scenes in Spanish rock art are mostly Mesolithic or later, thousands of years after Chauvet. Treat this as an overreach†.

Ardipithecus “completely bipedal” with chimp-sized brain. The 4.4-million-year-old Ardipithecus ramidus was bipedal on the ground but kept a grasping big toe and climbed; its bipedalism is debated†.

Controlled fire at campsites by 2 million years. Secure evidence for controlled fire is much younger (a few hundred thousand to about one million years); claims for earlier fire are contested†.

“No one has addressed the mathematical argument”. Several published replies did, including a 137-author letter in Nature in 2011 (see idea 5).

“Evangelicals … 40% of Americans”. Surveys that count evangelical Protestants by denomination give a much lower figure (about a quarter, from memory). A figure near 40% matches “born-again” self-description in polls. Treat as loose.

Kin selection explained by “collateral relatives”. Hamilton’s theory covers all relatedness, including offspring, and the maths works on genetic similarity, not on any family label. Wilson’s restatement is a simplification.

🎙 marks what the speaker says. 📚 marks added background and the wider debate, not the speaker’s claim. † marks background from general knowledge that was not checked against a source.

Talks

The Real Creation Story — E. O. Wilson, The Social Conquest of Earth (Long Now × Exploratorium, 20 Apr 2012)

Summary

Map — talking points

Putting it together

Part 1 — Before you read

Key terms

People named (a reading list)

Wilson’s own frame

The current debate (live search, Nov 2026)

The ideas

1. “A real creation story”: why science, not religion or philosophy

2. Why eusociality is so rare: the nest, the maze and the spring

📚 Roots: The altruism problem: from Darwin to Hamilton’s rule

3. The hominin path: down from the trees, meat, fire and the camp

📚 Roots: Hominin evolution: the steps Wilson summarises

4. Tribal instincts, war and “the signature behaviour”

📚 Roots: Tribal psychology and the question of war

5. Multi-level selection against kin selection: the technical heart

📚 Roots: Levels of selection: the long quarrel about groups

6. A permanently conflicted species: instability, ethics and the future of violence

📚 Roots: Sociobiology and its politics

7. Ants as the model: superorganisms, supercolonies and distributed intelligence

📚 Roots: Major transitions and the superorganism

8. Humanity’s future biology: homogenisation, diversity and the limits of “progress”

9. Beauty, habitat and arts: the sensory shortcuts

10. Religion, creation stories and the future of belief

📚 Roots: Creation stories, religion and science

11. Microbes, the unseen frontier, and never retiring

📚 Roots: Biodiversity, ant societies and the microbial frontier

Q&A

Table of Contents

Label	Who	Role
Host A	Rob Semper (inferred; the captions give “flexors zipper Rob”, and Wilson thanks “Rob”)	Exploratorium. Introduces the evening. Not heard clearly in the transcript
Brand	Stewart Brand (names himself)	Long Now Foundation co-founder. Introduces Wilson, then interviews him and reads questions from cards
Wilson	Edward O. Wilson (named by Brand; “Edward o Wilson”)	Harvard entomologist and biologist, age 82. Gives the 49-minute talk, then answers
Audience	Several people, read out by Brand	Names in captions (“Jack Sawyer”, “Bruce Miller”, “Byron”, “William Petry”) are not reliable
”Paul”	Paul Gauguin, not a speaker	The painter whose 1898 canvas gives the talk its three questions