Philosophy — Noozify Original — October 20, 2026

'The Selfish Gene' at 50: How 'Meme' Outlived Its Own Theory

By Adam Cannizzaro

Watch the story in half a minute

Fifty years ago this autumn, Oxford University Press published the first book by a 35-year-old Oxford zoologist named Richard Dawkins. Near the end of its last chapter he ran into a vocabulary problem. He needed a name for "a unit of cultural transmission, or a unit of imitation." "Mimeme" had a suitable Greek root, but he wanted "a monosyllable that sounds a bit like 'gene'," so he clipped it. The new word, he added, should rhyme with "cream," and his examples were "tunes, ideas, catch-phrases, clothes fashions, ways of making pots or of building arches." Half a century later, meme is used far more often than any other coinage in The Selfish Gene, mostly to describe a picture of a cat with words on it. That is less a betrayal of the theory than the cleanest test it ever got.

The book was not, in the first place, about culture. Building on the work of W. D. Hamilton, George Williams, John Maynard Smith, and Robert Trivers, it argued that natural selection is best understood from the gene's point of view. Genes are copied down the generations, and the ones that build bodies good at copying them are the ones that last. Dawkins put it with maximum provocation in his preface, calling animals, people included, "survival machines—robot vehicles blindly programmed to preserve the selfish molecules known as genes." Matt Ridley, revisiting the book for Nature in 2016, called it "a sparkling best-seller and a scientific game-changer", and noted that the philosopher Daniel Dennett had called it "philosophy at its best." Peter Medawar's review in The Spectator in January 1977 found it "learned, witty and very well written". In June the publisher marked the anniversary with a fifth, 50th-anniversary edition, re-typeset, with a new epilogue by Dawkins and a new appendix on the 1930s work of the British zoologist A. G. Lowndes. By the publisher's count the book has sold millions of copies in more than 30 languages.

The meme arrived in chapter 11, "Memes: the new replicators," and its job was modest. Dawkins wanted to show that genes are one example of a broader class, the replicator, and that anything which copies itself with occasional errors could in principle drive a Darwinian process. Culture looked like a candidate. Memes, he wrote, spread "by leaping from brain to brain via a process which, in the broad sense, can be called imitation," and he quoted his colleague N. K. Humphrey's sharper version, "When you plant a fertile meme in my mind you literally parasitize my brain." Revising the book in 1989, he lowered expectations further. His "designs on human culture were modest almost to vanishing point," he wrote in a new endnote, and his purpose had been "to cut the gene down to size, rather than to sculpt a grand theory of human culture." In the same note he observed, with evident pleasure, that "the word meme seems to be turning out to be a good meme."

Others were less modest. In the 1990s a field calling itself memetics tried to make the analogy a science. Its high-water mark was probably the psychologist Susan Blackmore's The Meme Machine, published by Oxford University Press in 1999 with a foreword in which Dawkins wrote that any theory "deserves to be given its best shot." The field got a journal of its own, the Journal of Memetics, which ran from 1997 to 2005. Its final volume carried something close to an obituary. Bruce Edmonds of the Centre for Policy Modelling at Manchester Metropolitan University wrote that the journal had ceased "due to a lack of quality submissions," and that memetics had failed because it "has not provided any extra explanatory or predictive power beyond that available without the gene-meme analogy."

The sharpest objection had come five years earlier from the cognitive scientist Dan Sperber. In his chapter for the 2000 collection Darwinizing Culture, he pointed to a basic difference between genes and ideas. A gene is copied almost perfectly from parent to child, letter for letter. A story, a tune, or a habit is not copied that way. Each person who picks it up rebuilds it in their own head, and the new version only resembles the one they heard. Sperber's phrase was that cultural things are "re-produced," made again and again, but "not reproduced in the sense of being copied from one another." His example was laughter. A child who starts laughing is not copying anyone's laugh. Evolution by natural selection depends on faithful copies, and something that has to be rebuilt from scratch in every mind it reaches does not copy faithfully enough to evolve the way genes do.

The word, meanwhile, had found a better habitat. In 1990 Mike Godwin, a fixture of Usenet newsgroups and the Well, grew tired of online arguments that predictably ended with someone being compared to the Nazis, and set out on what he called "a project in memetic engineering." He wrote a rule designed to spread on its own, Godwin's Law of Nazi Analogies, which holds that "as an online discussion grows longer, the probability of a comparison involving Nazis or Hitler approaches one." He seeded it wherever a gratuitous Nazi reference appeared, and, as he recounted in a 1994 Wired column titled "Meme, Counter-meme," "the counter-meme was reproducing on its own!" It mutated too, sprouting corollaries he never wrote. Godwin used Dawkins's word as Dawkins meant it, and in doing so introduced it to a vast new population.

What that population did next was less faithful. The term "image macro" was coined on the Something Awful forums in 2004 for captioned pictures, LOLcats followed around 2006, and by the early 2010s a meme was, for most people, a template that anyone could caption and pass along. The dictionaries followed. Dictionary.com still gives Dawkins's sense first, then adds a cultural item "spread via the internet and often altered in a creative or humorous way," along with a verb, illustrated by the phrase "cute cats that get memed."

The printed record shows the switch. We counted the lowercase plural memes in Google Books' English-language corpus, using the plural because the singular is muddied in older books by the French même. In 1976, the year the book came out, it appeared about 0.02 times per million words. By 2000, after a decade of memetics, the rate was 0.16, and by 2010 it was 0.33. Then the internet sense took over. Between 2015 and 2022, the latest year in the data, the rate tripled to 0.85 per million words, about 37 times what it was the year Dawkins coined the word.

Dawkins noticed. In June 2013 he joined Saatchi & Saatchi's New Directors' Showcase at the Cannes advertising festival, a staged performance about memes that, Wired guessed, was meant to become one. In the piece, as Wired described it, Dawkins explained that an internet meme is "a hijacking of the original idea," altered deliberately by human creativity rather than mutating by random change and spreading by something like Darwinian selection. In an interview alongside it he was more relaxed. "The meaning is not that far away from the original," he said. "It's anything that goes viral." His original chapter, he noted, had already reached for the metaphor of a virus.

Both readings hold, and the tension between them is the most interesting thing about the word. The deliberate remix that Dawkins called a hijacking is roughly what Sperber described, something rebuilt in each mind rather than copied whole, and it helps explain why the strict science of memes never produced results. Yet the word itself behaved like a textbook meme. It was copied, varied, and selected, and the variant that suited its new environment crowded out the original sense. A replicator, in Dawkins's account, does not need to serve its carriers or remember where it came from. It only needs to get copied.

The book's other famous word had a rougher journey. "Selfish" was a metaphor about which unit natural selection acts on, but the title invited readers to take it as a claim about how people behave or ought to. The philosopher Mary Midgley opened her 1979 attack in Philosophy with a sentence that is still quoted, "Genes cannot be selfish or unselfish, any more than atoms can be jealous, elephants abstract or biscuits teleological." Dawkins answered her in the same journal in 1981, and the argument over what the metaphor licensed ran for decades.

On the title, at least, Dawkins came round. In his introduction to the 30th-anniversary edition he recalled that Tom Maschler of the publisher Jonathan Cape, shown the partly finished book in 1975, had liked it but not its name. "Selfish," Maschler said, was a "down word," so why not The Immortal Gene? "I now think Maschler may have been right," Dawkins wrote, because many critics "prefer to read a book by title only." The word to stress, he insisted, is "gene." He even asked readers to "mentally delete" one line from his first chapter, "Let us try to teach generosity and altruism, because we are born selfish," which he called misleading. In a 2008 essay in The Threepenny Review that retells the Maschler episode, the critic P. N. Furbank called the difficulties of evolutionary biology "above all, problems of language." The book's own closing line, rarely quoted by its detractors, is "We, alone on earth, can rebel against the tyranny of the selfish replicators." Ideas that travel on a catchy label tend to shed their fine print, as anyone who has pinned thesis, antithesis, and synthesis on Hegel can attest.

The science underneath has aged better than the title. Its central fight, over whether selection acts on groups or on genes, flared again in 2010, when E. O. Wilson and the mathematicians Martin Nowak and Corina Tarnita argued in Nature that inclusive fitness theory, the gene's-eye account of why animals help their relatives, should give way to multilevel selection. That is the idea that natural selection works on whole groups as well as on individuals, so that a group of cooperators can outlast a group of selfish members even though selfishness pays inside each group. Reviewing Wilson's 2012 book The Social Conquest of Earth for Prospect, Dawkins counted more than 140 evolutionary biologists who had objected to the paper, and closed by borrowing Dorothy Parker's line that it was not a book to be tossed lightly aside but "thrown with great force." Wilson replied in the same pages that inclusive fitness was "an unattainable phantom measure." That argument is not over, but Ridley's assessment was that the gene-centred view had become "central both to evolutionary theorizing and to lay commentaries on natural history." One throwaway remark in the book even seeded a theory of "selfish DNA," which two pairs of scientists set out in Nature in 1980, crediting Dawkins as their inspiration.

The anniversary edition adds a fitting coda. According to the Seoul Economic Daily's report on the new Korean translation, the fresh epilogue argues that half a century of genomics has left the book's central thesis untouched, and it quotes a poem ChatGPT wrote on the book's ideas. A machine that learned to write from human text, riffing on a book about copying with variation, is about as on-the-nose as a meme gets.

The gene's-eye view asked readers to stop treating the organism as the point and to see it as a vehicle. The word meme invites the same view of itself. Dawkins built it to carry an argument about genes, and the word outlived the trip, passed along by people who use it to label jokes and have never needed to know where it came from. That is exactly what he said a successful replicator does. "The word meme seems to be turning out to be a good meme," he wrote in 1989. He undersold it.