12.11.2026
Dancing the word
Jack Conrad reviews Chris Knight and Jerome Lewis The revolutionary origins of language New Haven CT 2026, pp384, hbk £25
First a confession: I know and very much admire Chris Knight. For many years now we have welcomed him, and other members of the Radical Anthropology Group - eg, Camilla Power and Ian Watts - to Communist University. Their contributions have always been well-argued, thought-provoking and, in scientific terms, cutting-edge.
We have long wanted to include Jerome Lewis amongst our regular CU speakers. But over the summer months he is always away, deep in Congo’s vast tropical jungle, visiting the BaYaka hunter-gatherers. He is, though - and this matters - no disinterested outside observer. Yet another ‘objective’, here today, gone tomorrow academic, who sees the BaYaka as mere objects of study. Jerome Lewis is deeply committed to the BaYaka people and their communist way of life. His son, Nando, underwent their none too demanding male initiation ceremony (fittingly the Knight-Lewis book is dedicated to the BaYaka).
That said, I have listened to, questioned and learnt much from Jerome at RAG’s meetings (when they were just down the road from me in Camden Town). So I write as someone who is familiar with, and is already thoroughly convinced by, the basic argument of The revolutionary origins of language: language is unique to human beings, must necessarily have had a Darwinian - ie, evolutionary - background, but cannot be accounted for by natural selection and gradual adaptation.
Revolutionary origins follows Decoding Chomsky. This 2016 book saw Chris Knight publicly enter the minefield of academic linguistics with a bold, frontal assault on what had been the dominant paradigm. Established by Noam Chomsky with Syntactic structures (1957), his ‘revolution in linguistics’ discounted all notions of Darwinian evolution and language as a social phenomenon.
Yes, humans have a natural facility for language. But how come? According to Chomsky, it is because of our unique neural architecture - effectively a language acquisition device lodged in our brain. Brought about by a mysterious accident, in a single individual, this macromutation, ie, saltation, happened some 70-50,000 years ago. At first, language was purely internal. The single individual talked to themselves, no-one else. Only later - presumably because of the advantage that this conferred - were words used to communicate with others: a useful byproduct, but nothing more.
Essentially language is equated with a computer programme. Hence the forlorn search by Chomsky and his acolytes for language’s ‘deep structure’ and building a ‘universal grammar’ machine: everyday English could thereby be translated into battlefield instructions for missiles and bombs. You can see why the whole project was funded and promoted by the Pentagon. Needless to say, Chomskyism is nowadays largely viewed as a failure. LLMs and AI do not work according to its principles. Instead, the likes of Chat GPT and Gemini rely on statistical probability, neural networks and massive amounts of training data.
With Revolutionary origins, Knight and Lewis set out to provide an alternative paradigm. A hugely ambitious project. Language, as they say, is the most difficult question in science. Certainly it is the most important: after all, the prime focus of any human science should be humanity itself.
Alternative
The explanation for language cannot just rely on our large, 1,200-1,350cc brains. Everyone knows that, secure in a loving, domestic, human environment, the common chimpanzee (brain size: 380cc) can be taught sign-language: eg, ‘Hug’, ‘Play’, ‘Listen’, Drink’. Washo acquired 350 signs. She even combined them to describe new things: ‘Water bird’. Nim Chimpsky came out with the longest recorded chimpanzee sentence: “Give orange me, give, eat orange, me eat orange, give me, eat orange, give me, you.” Note, the complete lack of human syntax. Despite that there is a string of distinct words. Yet, left to their own devices, chimps almost instantly abandon such an essentially cooperative form of communication. They revert to standard primate vocalisation and dominance displays.
In the wild it is fierce competition that rules, not least over food and sex. Communication is essentially honest, because it is costly. It can therefore be trusted in a chimp world, where the distrust of other chimps is the norm. Calls are instinctive and can be physically draining, because they involve the whole body, as in a pant-hoot: ‘Woo-ooo-woo-ooo-WOO-ooo-WOOOOOOO!’ It unmistakably announces food, location and status to other chimps. There is though no thought-through subtlety, no minutia involved. Such involuntary calls are often referred to as analogue, or graded forms of communication by evolutionary linguistics. Humans still communicate in such ways, as, for example, when we sob, scream, uncontrollably laugh or cry out in pain: ‘Ouch!’ or ‘Aaaargh!’. But that is not language.
Chimps, Pan troglodytes, are intelligent and highly political animals. Males form complex coalitions, so do females. However, there is a steep male-dominated hierarchy and life for those below is often hugely stressful. For newly mature females this is especially the case. To avoid incest they migrate away from their biological kin in the hope of being accepted into a neighbouring group. They are resented, sometimes driven away, by established females, but welcomed by randy males. Ovulation is visibly signalled with sexual swellings - a trigger for an alpha male to violently monopolise and guard her from other males. What can only be called rape is commonplace.
Mothers are always single-mothers. All care and provisioning relies on her. Males supply sperm, nothing more. And her young offspring are always at risk from both newly risen alpha males and rival females alike. Babies will never be handed to another female for momentary care. There is not the least possibility that they can be trusted. Infanticide is all too frequent. Best hope lies in working up the hierarchy by courting allies through grooming, sex and raising strong, potentially alpha sons.
It is not just a matter of biology. Environmental factors count too. Knight and Lewis use the closely related bonobos to make the point. The common chimpanzee lives alongside highly aggressive gorillas and have to range far and wide to obtain food. Gorillas have voracious appetites and hoover up most of the leafy, so-called fallback, food.
Separated by some 1.5 million years of evolution with the formation of the uncrossable Congo river, bonobos have a much easier time. No gorillas, abundant food. At the risk of gross simplification, the result is that the Pan paniscus hierarchy to the south of the river is female-dominated. It is females, helped along with lots of bonding G-G sex, who form the winning coalitions.
At some point, therefore, females must have overthrown the male-dominated hierarchy of the common Pan ancestor. This allowing the evolution of a distinct, slightly smaller, more friendly subspecies. Eg, incoming female bonobos are welcomed by established female bonobos.
Growing brains
Not that brain size is unimportant, when it comes to language. Far from it. Knight and Lewis trace the increase in the brain size of our immediate ancestors. Homo habilis, the first of the Homo lineage, lived some 2.4‑1.56 million years ago, and had a brain which ranged from 500-650cc. Homo habilis made stone tools and skilfully butchered carcases. But there exists no evidence of fire‑maintaining, let alone fire‑making.
It was Homo erectus, beginning about two million years ago, which broke through the ‘grey ceiling’. Early Homo erectus had a brain size of around 800cc. However, late Homo erectus took this up to 900cc - much bigger than Homo habilis. The burden for mothers must have been correspondingly much bigger too. Knight and Lewis turn to the ‘grandmother thesis’, first proposed by Kristen Hawkes and her colleagues in the late 1990s (they wanted to explain how the menopause evolved).
There is a shift from patrilocal to matrilocal living: instead of moving away from their kin group, daughters stay put with their mums. It is they, the grandmothers, who help out with the extra care and provisioning needed when it comes to the prolonged pregnancy and childrearing required for big brains. This too would allow Homo erectus females to form powerful coalitions. As a sub-species Homo erectus regularly made fire (the earliest reliable evidence dating back ~400,000 years). Their women therefore not only cooked: they probably chorused and collectively danced too.
So far, so just about Darwinian … and Homo erectus successfully spread out from their southern and eastern African homeland into Asia, reaching China and even Java.
Political strategy
We Homo sapiens took things still further. Males could be and were recruited into the work of provisioning females, their offspring … and their extended families.
Drawing on the archaeological record, fossil evidence, the insights of indigenous hunter-gatherers and the theorisation provided by fellow anthropologists - crucially Sarah Hrdy, Kristen Hawkes and Camilla Power - our authors, Knight and Lewis, argue that female-centred coalitions, coalitions which actively sought to recruit sons and younger brothers, overthrew alpha male domination: the human revolution.
Collective laughter, play, ritualised symbolism and, if need be, a little bit of force, established a new social order. But this was no re-run of Congo’s Pan paniscus revolution. This female-driven revolution was militantly egalitarian. Something brilliantly achieved by mapping social relations - and therefore sexual relations - onto the light and dark phases on the moon.
Notice that the lunar month is 29.5 days. Today a woman’s menstrual cycle, is, on average 28 days. That near match cannot be mere coincidence. It is surely significant. Knight and Lewis convincingly show that achieving, or simulating, synchrony around the phases of the moon played a key role in counterdomination and women collectively seizing control over their sexual availability to the opposite sex.
Menstruation signalled unavailability. The first word, say Knight and Lewis, is ‘no’. In return for men providing meat, they were given a ‘yes’. All females were expected to abide by the rule and show solidarity - dalliance with some new-found beau being akin to scabbing. Rape becomes absolutely taboo.
Dark moon is when lions hunt because of their excellent night-vision. Full moon is when we humans see better and therefore can safely hunt. Lions largely do not bother under bright conditions, because they can easily be spotted and are either gingerly skirted or warned off.
The militant egalitarianism of original communism continues to exist to this very day. Lewis and his accounts of the BaYaka show that in no uncertain terms: there is collective childcare - everyone has more than one mother - something which goes hand-in-hand with equal shares for all, personal autonomy and the active, often theatrical subversion of male arrogance. Any hunter who comes back to camp preening and priding himself about his prowess invites jokes, mimicry and takedowns … especially from the older women. The hunter is expected to self-reflect and quickly join in the general merriment.
Each month, as the moon begins to wax - and conditions become favourable for men to hunt - the women announce their coming ‘sex strike’. Dancing together, they sing out their message loud and clear, rhythmically stamp their feet and tease the men with bawdy jokes. Collective female power is thereby visibly asserted. Bodies are liberally painted with bright red ochre as a stand-in for menstrual blood.
The everyday world is reversed. The community symbolically returns to a parallel, a deeper world of creation, when humans, animals, everything was mutable. To this day the ‘Eland bull dance’ is still performed by hunter-gatherers in southern Africa. Women present themselves as the ‘wrong sex’, the ‘wrong species’, at the ‘wrong time’, with sticks standing in for horns and penises. Woe betide the foolish male who dares mock the ‘picket line’. It is sacred.
The conservative nature of symbolic culture, evidence from ancient rock paintings and the discovery of traces of red ochre in the archaeological record indicate that this, and similar dances, date back at least 170,000 years. Knight and Lewis suggest that it would have taken many millennia to fully consolidate the human revolution. The crucial marker being the use of red ochre. Beginning some 500,000 years ago, it becomes ubiquitous throughout Africa some 300,000 years later.
Camilla Power coined the term, ‘Female Cosmetic Coalition’ - there being a close symbolic association between red ochre and the blood of animals killed in the hunt. Essentially, what the women are saying is something like this: ‘We’re showing the red signal of the sex strike; now you go out and bring back the red blood of the game animal.’
Here in the crucible of symbols and an artificially created virtual reality are the conditions needed for language and what we understand as human consciousness. Words are cheap, they are digital, but they rely on metaphor and therefore on collective trust. Once trust is firmly established, the doors of language are flung wide open.
Words then come at the flood. All manner of things are named, including things that only exist in the imagination: tomorrow, the world to come, trickster gods, animal spirits, dead ancestors who come back to life. One word is joined together with another. Grammar and syntax spontaneously form around accepted usage. Fantastic stories get told and retold and humans collectively cross over into the realm of culture.
Weapon for today
For dogmatic ‘Marxists’, Knight and Lewis can only be right if they are out to prove Marx right, Engels right … both of them right. Nothing could be more foolish, nothing more mistaken. Like Marx and Engels, Knight and Lewis are guided by science: finding the best available evidence, establishing the best theory, the best explanation. If that contradicts, corrects, what Engels happened to say at this or that date - for example, in his famous 1876 essay, The part played by labour in the transition from ape to man - so be it.
Engels, of course, centred bipedalism, free hands and tool-making in this account. Essentially he echoed Benjamin Franklin and his Homo faber (admittedly a high achievement of the 18th century Enlightenment). Despite that the Grantite Revolutionary Communist Party automatically, unthinkingly, defends the 1876 Engels essay as a “masterpiece of dialectics”. There is even the claim that, if Engels were alive today, he would “rewrite” The part played by labour in the transition from ape to man “with fundamentally the same view” (communist.red/engels-and-human-development). An obvious absurdity.
Today, Engels, if he were alive, would be looking forward to celebrating his 206th birthday on the 28th of this month. But that is not, of course, my point. No, in my opinion, Engels would take an entirely different approach … because he would be guided by the science.
And, basing themselves on the best science, Knight and Lewis emphasise sexual reproduction, female-centred coalitions and their revolutionary overthrow of alpha male domination. Tools, though undoubtedly important, constitute something of a footnote. The indisputable fact of the matter is that technological development during the Palaeolithic was remarkably slow. Only with the Neolithic counterrevolution did things really take off in this respect: kiln pottery, polished stone weapons, sickles to harvest domesticated plants, ropes, pens and halters to control and manage domesticated herds (something conjoined with what Engels later called the “world-historic defeat of the female sex”).
If the Knight-Lewis theory can be disproved through some startlingly new discovery in the fossil, archaeological … or extra-terrestrial record, again, so be it. They will then, if they are to remain genuine scientists, and not collapse into mere ideology, have to suitably modify their theory. That or simply go back to the drawing board. Eg, if something like Arthur C Clarke’s entirely fictional black monolith is discovered on the moon and can reliably be dated back 300,000 years, then our views on the human revolution and the origins of symbolic culture, militant egalitarianism and language might perhaps be in need of urgent revision.
But it is more than stupid to rely on the word of Marx and Engels. After all, they could only but rely on the severely limited and often shockingly racist anthropology available to them during their day. That said, interestingly, because of their rigorous method, because of their ever-expanding world view, because of their accumulated body of theory, Marx and Engels were able to sift through the dross.
Certainly with Lewis Henry Morgan’s Ancient society (1877) they discovered 22-carat gold. Marx eagerly devoured his book in 1880 and after his death left behind his so-called Ethnographic notebooks. Engels dutifully read them in 1884 - and Morgan for the first time. Writing to Karl Kautsky over in Zurich he said that Morgan’s “definitive book” is “as definitive as Darwin’s was in the case of biology” (February 16 1884). Over the course of the next few months he produced his celebrated Origin of the family, private property and the state. A creative homage to Marx’s notes and the best available anthropology. Hence, in fact, a great deal, if not all, of what the late Marx-Engels team had to say about pre-class social formations has withstood the test of time: eg, original communism, matrilineal decent, matrilocal residence, no bonds of marriage, bride service, collective childcare, etc.
Sadly, pathetically, much of what passes itself off as contemporary ‘Marxism’ still relies on the means of labour - ie, tools - to the neglect of social relations. Certainly because there exists a widespread economistic fear of talking about female coalitions, sexual politics, menstruation and symbolic blood, the work of radical anthropologists such as Knight and Lewis usually goes studiously ignored.
For example, in his People’s history of the world (1999), Chris Harman, a former Socialist Worker editor, completely skips over the human revolution. His account of ‘before class’ is entirely based on a banal biological gradualism and technology:
humans were genetically programmed for extreme flexibility in response to the world around them - by being able to use their hands to hold and shape objects, being able to use their voices to communicate [language] with each other, being able to investigate, study and generalise about the world around them and being able, through long years of child-rearing, to pass on their skills and learning. All this required the growth of big brains (p4).
Big brains and language are assumed, not explained. Not surprisingly Harman devotes the same number of pages to the invention of the vacuum cleaner as he does to the rise of women’s oppression: three pages each.
Conclusion
If I have a complaint, and I do, the Knight-Lewis book is incomplete. There is much more that needs to be said, but goes unsaid. Too much is short-hand. Too much is left incomplete. Too much needs further elaboration. Maybe I am being patronising, but I am not sure the average reader will be able to fill in the gaps. After all, we are dealing with popular science.
The revolutionary origins of language is a medium-sized book which surely needs a big-sized book to do the arguments full justice by providing more background, more facts, more theory and more polemics.
There are also far too many typos and proof-reading errors. A prestigious publishing house such as Yale really ought to go to the modest expense of employing skilled proof-readers. Even a quick AI scan would help. A niggle doubtless, but it is still a pity. I very much hope such problems can be ironed-out with the paperback edition.
Nevertheless, this is a groundbreaking book. Nothing can take away from that. Genuine Marxists will certainly greet it as a triumph of historical materialism and a vindication of the dialectical leap.
Buy it. Read it. Recommend it.