Couldn't all four characteristics co-evolve concurrently? I think it's much more likely than sequential evolution, in part because each of them likely evolved over a long time.
The non-specific attempts to order these specific capacities in the hominin lineage is fraught. None of these models is or can be correct. Henrich can be more correct than Harari but that's really it.
There were obviously rudimentary versions of all four of these capability in the 6-8 mya split with chimps. The extent to which one of these distinctly ran ahead and then pulled up the other is not clear at all. It is evident the anatomy and physiology for human-level intelligence long preceded demonstrated levels of human sociocultural complexity. There were also different ancient human groups (some say species) which then later recombined. The exact cultural consequences of this are unknown and not a part of these models but based on the ancient DNA seem very important for the emergence of modern humans.
Yes, I think few things are purely "causal" since most observables in complex systems arise from multiple, interacting, mutually-re-enforcing processes.
It's just hard escape the framework of single-agent, sequential, one-way causality, since that's how we interact with most objects around us.
No, it doesn't. Moreover, it ignores a century of research in paleoanthropology, primatology (and zoology in general), linguistics, genetics and evolutionary theory.
(I know I should be more specific, but I have a bit of emergency situation - we have to permanently move to another town on a short notice - so I can't go into detail even though I'd love to. Sorry.)
Sapiens was such an utterly shallow book I don't understand the perceived need to refute it. Refutation is for original works of scholarship, not for books that take the scientific work of others and retell it for a popular audience.
Harari writes exactly the kind of books that flatter the elites and persuade the general public that the world as it is is mostly as it should be.
It is a whig theory of history, but with a patina of universalism, even though it pays very little attention to anything outside the European cultural system as it has developed over the last 300 years.
I haven't a clue as to the specific veracity of this claim, but its shape fits a very big hole in our general reckoning, so i feel somewhat justified in assuming it's probably nailing Harari pretty hard...
If you're interested to see how far we can reasonably push that sort of heuristic, you might like my piece, The Shadow of Aristocracy, from a couple of months ago
I gave it a very quick high-level scan about halfway through, and I think we basically agree. I may quibble on some of the wording, mostly because of my own theory of narcissism, but yes, I like where the argument is going. I will take some more time to read it in more detail, but there's a lot of similarity to how I think about how social systems work.
The way i roll, i'd be happy to consider it as a starting point for something more collectively expressed - if whatever doesn't survive other switched-on perspectives is trimmed away and it becomes something a bunch of folks can stand behind, i feel like maybe this angle could be a proper red pill (unlike that manosphere bullshit). Breach pedagogy is the aim; jailbreaking minds.
I think the timing on fire throws a monkey wrench in the copying story.
Signs of domestication of fire go back a half million or even a full million years ago.
Chimps can’t control, let alone start fire, Nor can chimps that are good at copying.
To make fire, our ancestors needed brains that were better at abstract thought (aka imagination.) The ability to think ahead to future needs and conditions and to combine long series of cumulative steps.
With abstract thought comes the ability to revisit past actions, to “read” others minds, and, since they can also read yours, the ability to bridge minds around common beliefs and norms. This allows people to be better at copying at training and at cultural accumulation. Eventually it leads to language and everything gets better still.
Fire also gives us huge advantages in caloric absorption via cooking , better heating (expanded range), and predator prevention. Cooking further leads to cooperative units to delay and protect food until it is cooperatively cooked.
I agree that cooperation, intelligence (abstract thought), fire, language, and culture all reinforce each other.
But you don’t get fire without abstract thinking well beyond what chimps can do, and we have had fire since well before Homo sapiens.
This is an amalgamation of the extensive writings and research of Wrangham and Suddendorf.
The interesting part of the Cousins et al 2025 paper ("Structured coalescent model") that you mentioned was not only about the ancient population bottleneck, but about very deep human history - splitting into two branches between 1.5 and 1.8 mln years ago and then reconnecting some 300 kY ago.
These big demographic changes were likely connected with major techno-cultural steps: the first one approximately coincides with the spread of Acheulean stone technology from Africa to Eurasia, and early ancestors of modern humans likely resided in Eurasia since then without interruption (that was "branch A" in Cousins paper, that included also Neanderthals and Denisovans). The second (reconnection) event coincided with the spread of Levallois technology both in Eurasia and some parts of Africa.
These major technological stages (especially Levallois one) already required deep "culture" - these were highly non-trivial, sophisticated techniques that needed intergenerational learning, transmission and refinement.
I know its not your main point here Kirill, but your mention of a specific tool-making tradition (i.e. a culturally-mediated behavior) already being established almost 2mya (and for that to happen surely there were other contemporary traditions just as deeply established that didn't leave archaeological residue) highlights the disproportionate preponderance of cultural evolutionary literature that seeks only to explain what's gone on in the last ~100k or ~40k or ~10k years, implying (without explicitly stating) that's when human culture became important. The climate chart by Richerson cited above is just one of many examples to be found once one starts looking for them.
Analyses looking at the most recent 1-5% of Homo history are of course interesting (and that's by far where the most data is) but alarms go off when I see extrapolations being made from them purporting to explain the nature of the dynamics of the entire evolutionary trajectory of the lineage's culture and other distinctive behavioral propensities and traits. I observe that when the field wears its 'evolutionary anthropology' hat, it takes its explanatory target to be the entire 2-3m year Homo span, but then when it puts on the cultural evolutionist hat to get specific about culture's role in that trajectory, it so often invokes data with a recency bias while it strives to present clean cause-and-effect narratives.
This is very much angels dancing. There is no way to distinguish the four factors because there is no way to talk about any of them without invoking all of them.
Well, if we look at the reasonable extremes it's possible to argue for, and internalise the dynamics of each proposed explanation, that gives us a possibility space and a set of intuitions to navigate it with - we arrive at a fuzzier but more nuanced understanding, laced with threads of load-bearing probability; a map of attractor states for the dynamical system.
Understanding is an endless road, paved with overconfident declarations of arrival.
Epicycles and dancing angels are not more nuanced understandings of the world, since they don't represent anything in the world. What they represent instead is overproduced elites earning their keep. Current notions of “culture", “intelligence", “cooperation” are not much different. Recall the grandiose theorizing around the "Cog" project at MIT; then compare it to what actually happened. We are still trying to square the circle; meaning our experts are still selling us smaller bits of paper and a better kind of glue.
Well yeah, but i think our referents differ a bit here - i'm including what you rightfully dismiss as mapped negative space, along with all the worthwhile stepping stones akin to Rutherford's model of the atom. It's all very well to stand above everything and sneer at how dumb some of us have been, that's totally valid IMO as one mode of recognition among many - but the deeper view accepts all the causality involved and looks for whatever value can be salvaged from each perspective, if any.
One thing that such an approach impresses one with, is the notorious fallibility of 'experts' who have often operated devoid of wider context deep in a silo, or even, as has been quite predominant in the field of economics, orienting themselves to a supposed level dictated by a heavy thumb on the scale, courtesy of the ruling class and their ecosystem of camouflaged assumptions - i hear ya, and I'm in absolute solidarity against that fucking cancer.
Yes, clearly we co-evolved with our milieu and the idea that the only driver is some random mutation in DNA is just a superstitious cult.
Also cool: Holocene warming was unusual in that past closing of Panama Isthmus (1 My), interglacials were a rapid rise to 300 ppm of CO2, quickly followed by steady decrease to increasingly lows (180 ppm in latest glacial max). Instead, at 7 ky, our interglacial interrupted the decrease with a further rise by 20 ppm. And the driver is: human land use. Basically our ancestors burning forests for agriculture.
See also how the little ice age was caused by the collapse of populations and agriculture past European invasions of the new world.
So: big climate swings --> species that's able to adapt faster --> the species burns stuff and stabilises the climate.
Yes, we are (so far) stabilising the climate by preventing the end of the current interglacial - and this looked like it was going to get out of control but then we also switched to electricity.
Intelligence damping the signal that created it, basically.
That's a very interesting take. Humans influenced climate much earlier than it is assumed generally. One way was through killing most of the megafauna on different continents. It had some significant environmental changes in both directions, hard to tell what was the net effect.
Wouldn't cooperation be pretty hardwired into biology?
How would multicellular organisms have evolved to begin with, otherwise?
Don't social entities exist as social super organisms?
Government certainly functions as a form of nervous system, executive and regulatory, while money and banking would seem somewhat analogous to blood and the circulation system.
Currently our financial sector enables a lot of freeloading, but that is because we haven't yet come to understand it has to be a utility and not a player. Like government.
>> Currently our financial sector enables a lot of freeloading, but that is because we haven't yet come to understand it has to be a utility and not a player. Like government.
This may well be a symptom for the current decline of the Western world. The "successful transition to finance capitalism (s. eg. Bas van Bavel “The Invisible Hand? How Market Economies Have Emerged and Declined Since AD 500” (2016) and also Piketty "CCapital in the 21st Century (2013) - and of-course "End Times" by our much adored Prof Peter Turchin)
l'll raise a quibble. I suspect we move too quickly to 'intelligence'. It's perhaps a modern conceit to talk about the brain being an information processor. Consider the brain (and other bits and pieces) to be a signal processor. A baby chimp does not have to 'understand' how to open hard shells, he has to replicate the signals he receives from watching others and then practice how he can achieve the sweet nut meat by strengthening the the signals that lead to a favourable outcome.
Over evolutionary time the signal processing ability is selected for and this leads to Culture, Cooperation, Language and Intelligence.
Intelligence did not arrive last, after language. Intelligence as schema — structured expectation, role, sequence, consequence — was already there, and language was an upgrade to it.Music, dance, ritual and mime are that cognition performed, not a warm-up for it. Mithen’s Hmmmmm and Donald’s mimetic stage carry a cultural loop without propositions. Deacon’s hierarchical association fits the same place: a prefrontal capacity co-evolved long before grammar. What language added was bandwidth, so an accumulation already underway could become propositional, external and compounding. It did not switch thought on.Harari is wrong to treat a late linguistic mutation as the origin of the mind. The cultural-evolution sequence is wrong in the other direction, in making intelligence the last trait to evolve. Schema is the earlier loop.I make that case here as three inheritance loops — genetic, neural, cultural — with language as the upgrade that took the third from idle to compounding:
This just seems like more speculation. Large cooperative groups and inter group cooperation as early as is asserted here needs some kind of evidence. It flattens the reality of cooperative societies of by avoiding various kind of stress that threaten them. They evolve various complex ways to deal with not only free riders but all kinds of violation of cultural norms.Ways to suppress powerful or persuasive individuals and so on.The least he could do is define what “large” means in this context.
John Cacioppo argued that the stress of loneliness evolved as an adaptation that kept early humans embedded in cooperative social groups. One interpretation is that this sociality made expansion from African forests into savannah environments possible. Another is that climate change first created the savannah environment, and cooperation then evolved as an adaptation to the challenges of life there.
If Cacioppo is broadly correct, would this imply that a significant increase in cooperation preceded advanced social learning and cumulative cultural evolution? Or do you see cooperation, social learning, and cultural evolution as coevolving from the beginning?
Great article One can agree that four evolutionary traits may have an overlapping, the order 4-3-2-1 still holds. Its intensity of these traits which matters. Harari , of course , doesn’t merit discussion.
That GRIP ice core is often used to show Holocene stability and the Pleistocene's huge temperature fluctuations. It is showing temperature variation in one place - the Greenland ice sheet during one ice age cycle. Fluctuations in Africa could well have been more modest, but still very significant, and still likely to have been a significant factor.
Having just now read Darshana Narayanan's review and critique, it makes me glad I never wasted my time reading Harari. It sounds like he is offering a new techno-utopian mythology as master narrative and religious vision. It's frightening that business elites and political leaders may be basing important decisions on this misinformation.
Couldn't all four characteristics co-evolve concurrently? I think it's much more likely than sequential evolution, in part because each of them likely evolved over a long time.
The non-specific attempts to order these specific capacities in the hominin lineage is fraught. None of these models is or can be correct. Henrich can be more correct than Harari but that's really it.
There were obviously rudimentary versions of all four of these capability in the 6-8 mya split with chimps. The extent to which one of these distinctly ran ahead and then pulled up the other is not clear at all. It is evident the anatomy and physiology for human-level intelligence long preceded demonstrated levels of human sociocultural complexity. There were also different ancient human groups (some say species) which then later recombined. The exact cultural consequences of this are unknown and not a part of these models but based on the ancient DNA seem very important for the emergence of modern humans.
Yes, I think few things are purely "causal" since most observables in complex systems arise from multiple, interacting, mutually-re-enforcing processes.
It's just hard escape the framework of single-agent, sequential, one-way causality, since that's how we interact with most objects around us.
The linked text by Joseph Heath explains why this is unlikely.
https://josephheath.substack.com/p/harari-vs-henrich
No, it doesn't. Moreover, it ignores a century of research in paleoanthropology, primatology (and zoology in general), linguistics, genetics and evolutionary theory.
(I know I should be more specific, but I have a bit of emergency situation - we have to permanently move to another town on a short notice - so I can't go into detail even though I'd love to. Sorry.)
Insigntful! Thanks for weighting in!
Sapiens was such an utterly shallow book I don't understand the perceived need to refute it. Refutation is for original works of scholarship, not for books that take the scientific work of others and retell it for a popular audience.
Harari writes exactly the kind of books that flatter the elites and persuade the general public that the world as it is is mostly as it should be.
It is a whig theory of history, but with a patina of universalism, even though it pays very little attention to anything outside the European cultural system as it has developed over the last 300 years.
I haven't a clue as to the specific veracity of this claim, but its shape fits a very big hole in our general reckoning, so i feel somewhat justified in assuming it's probably nailing Harari pretty hard...
If you're interested to see how far we can reasonably push that sort of heuristic, you might like my piece, The Shadow of Aristocracy, from a couple of months ago
Link?
Sorry, i go on a bit - if it's TL;DR you could throw it at an LLM to see WTF I'm on about
https://agnosticfundamentalism.substack.com/p/the-shadow-of-aristocracy
I gave it a very quick high-level scan about halfway through, and I think we basically agree. I may quibble on some of the wording, mostly because of my own theory of narcissism, but yes, I like where the argument is going. I will take some more time to read it in more detail, but there's a lot of similarity to how I think about how social systems work.
The way i roll, i'd be happy to consider it as a starting point for something more collectively expressed - if whatever doesn't survive other switched-on perspectives is trimmed away and it becomes something a bunch of folks can stand behind, i feel like maybe this angle could be a proper red pill (unlike that manosphere bullshit). Breach pedagogy is the aim; jailbreaking minds.
I think the timing on fire throws a monkey wrench in the copying story.
Signs of domestication of fire go back a half million or even a full million years ago.
Chimps can’t control, let alone start fire, Nor can chimps that are good at copying.
To make fire, our ancestors needed brains that were better at abstract thought (aka imagination.) The ability to think ahead to future needs and conditions and to combine long series of cumulative steps.
With abstract thought comes the ability to revisit past actions, to “read” others minds, and, since they can also read yours, the ability to bridge minds around common beliefs and norms. This allows people to be better at copying at training and at cultural accumulation. Eventually it leads to language and everything gets better still.
Fire also gives us huge advantages in caloric absorption via cooking , better heating (expanded range), and predator prevention. Cooking further leads to cooperative units to delay and protect food until it is cooperatively cooked.
I agree that cooperation, intelligence (abstract thought), fire, language, and culture all reinforce each other.
But you don’t get fire without abstract thinking well beyond what chimps can do, and we have had fire since well before Homo sapiens.
This is an amalgamation of the extensive writings and research of Wrangham and Suddendorf.
The interesting part of the Cousins et al 2025 paper ("Structured coalescent model") that you mentioned was not only about the ancient population bottleneck, but about very deep human history - splitting into two branches between 1.5 and 1.8 mln years ago and then reconnecting some 300 kY ago.
These big demographic changes were likely connected with major techno-cultural steps: the first one approximately coincides with the spread of Acheulean stone technology from Africa to Eurasia, and early ancestors of modern humans likely resided in Eurasia since then without interruption (that was "branch A" in Cousins paper, that included also Neanderthals and Denisovans). The second (reconnection) event coincided with the spread of Levallois technology both in Eurasia and some parts of Africa.
These major technological stages (especially Levallois one) already required deep "culture" - these were highly non-trivial, sophisticated techniques that needed intergenerational learning, transmission and refinement.
I commented on this in the post https://kirillpankratov.substack.com/p/two-early-humankinds
I know its not your main point here Kirill, but your mention of a specific tool-making tradition (i.e. a culturally-mediated behavior) already being established almost 2mya (and for that to happen surely there were other contemporary traditions just as deeply established that didn't leave archaeological residue) highlights the disproportionate preponderance of cultural evolutionary literature that seeks only to explain what's gone on in the last ~100k or ~40k or ~10k years, implying (without explicitly stating) that's when human culture became important. The climate chart by Richerson cited above is just one of many examples to be found once one starts looking for them.
Analyses looking at the most recent 1-5% of Homo history are of course interesting (and that's by far where the most data is) but alarms go off when I see extrapolations being made from them purporting to explain the nature of the dynamics of the entire evolutionary trajectory of the lineage's culture and other distinctive behavioral propensities and traits. I observe that when the field wears its 'evolutionary anthropology' hat, it takes its explanatory target to be the entire 2-3m year Homo span, but then when it puts on the cultural evolutionist hat to get specific about culture's role in that trajectory, it so often invokes data with a recency bias while it strives to present clean cause-and-effect narratives.
This is very much angels dancing. There is no way to distinguish the four factors because there is no way to talk about any of them without invoking all of them.
Well, if we look at the reasonable extremes it's possible to argue for, and internalise the dynamics of each proposed explanation, that gives us a possibility space and a set of intuitions to navigate it with - we arrive at a fuzzier but more nuanced understanding, laced with threads of load-bearing probability; a map of attractor states for the dynamical system.
Understanding is an endless road, paved with overconfident declarations of arrival.
Epicycles and dancing angels are not more nuanced understandings of the world, since they don't represent anything in the world. What they represent instead is overproduced elites earning their keep. Current notions of “culture", “intelligence", “cooperation” are not much different. Recall the grandiose theorizing around the "Cog" project at MIT; then compare it to what actually happened. We are still trying to square the circle; meaning our experts are still selling us smaller bits of paper and a better kind of glue.
Well yeah, but i think our referents differ a bit here - i'm including what you rightfully dismiss as mapped negative space, along with all the worthwhile stepping stones akin to Rutherford's model of the atom. It's all very well to stand above everything and sneer at how dumb some of us have been, that's totally valid IMO as one mode of recognition among many - but the deeper view accepts all the causality involved and looks for whatever value can be salvaged from each perspective, if any.
One thing that such an approach impresses one with, is the notorious fallibility of 'experts' who have often operated devoid of wider context deep in a silo, or even, as has been quite predominant in the field of economics, orienting themselves to a supposed level dictated by a heavy thumb on the scale, courtesy of the ruling class and their ecosystem of camouflaged assumptions - i hear ya, and I'm in absolute solidarity against that fucking cancer.
Heartened by the generosity of spirit revealed in your words.
Yes, clearly we co-evolved with our milieu and the idea that the only driver is some random mutation in DNA is just a superstitious cult.
Also cool: Holocene warming was unusual in that past closing of Panama Isthmus (1 My), interglacials were a rapid rise to 300 ppm of CO2, quickly followed by steady decrease to increasingly lows (180 ppm in latest glacial max). Instead, at 7 ky, our interglacial interrupted the decrease with a further rise by 20 ppm. And the driver is: human land use. Basically our ancestors burning forests for agriculture.
See also how the little ice age was caused by the collapse of populations and agriculture past European invasions of the new world.
So: big climate swings --> species that's able to adapt faster --> the species burns stuff and stabilises the climate.
Yes, we are (so far) stabilising the climate by preventing the end of the current interglacial - and this looked like it was going to get out of control but then we also switched to electricity.
Intelligence damping the signal that created it, basically.
That's a very interesting take. Humans influenced climate much earlier than it is assumed generally. One way was through killing most of the megafauna on different continents. It had some significant environmental changes in both directions, hard to tell what was the net effect.
Wouldn't cooperation be pretty hardwired into biology?
How would multicellular organisms have evolved to begin with, otherwise?
Don't social entities exist as social super organisms?
Government certainly functions as a form of nervous system, executive and regulatory, while money and banking would seem somewhat analogous to blood and the circulation system.
Currently our financial sector enables a lot of freeloading, but that is because we haven't yet come to understand it has to be a utility and not a player. Like government.
Evolution is still at work.
>> Currently our financial sector enables a lot of freeloading, but that is because we haven't yet come to understand it has to be a utility and not a player. Like government.
This may well be a symptom for the current decline of the Western world. The "successful transition to finance capitalism (s. eg. Bas van Bavel “The Invisible Hand? How Market Economies Have Emerged and Declined Since AD 500” (2016) and also Piketty "CCapital in the 21st Century (2013) - and of-course "End Times" by our much adored Prof Peter Turchin)
l'll raise a quibble. I suspect we move too quickly to 'intelligence'. It's perhaps a modern conceit to talk about the brain being an information processor. Consider the brain (and other bits and pieces) to be a signal processor. A baby chimp does not have to 'understand' how to open hard shells, he has to replicate the signals he receives from watching others and then practice how he can achieve the sweet nut meat by strengthening the the signals that lead to a favourable outcome.
Over evolutionary time the signal processing ability is selected for and this leads to Culture, Cooperation, Language and Intelligence.
Intelligence did not arrive last, after language. Intelligence as schema — structured expectation, role, sequence, consequence — was already there, and language was an upgrade to it.Music, dance, ritual and mime are that cognition performed, not a warm-up for it. Mithen’s Hmmmmm and Donald’s mimetic stage carry a cultural loop without propositions. Deacon’s hierarchical association fits the same place: a prefrontal capacity co-evolved long before grammar. What language added was bandwidth, so an accumulation already underway could become propositional, external and compounding. It did not switch thought on.Harari is wrong to treat a late linguistic mutation as the origin of the mind. The cultural-evolution sequence is wrong in the other direction, in making intelligence the last trait to evolve. Schema is the earlier loop.I make that case here as three inheritance loops — genetic, neural, cultural — with language as the upgrade that took the third from idle to compounding:
https://theintegraloflife.substack.com/p/three-loops-of-evolution
This just seems like more speculation. Large cooperative groups and inter group cooperation as early as is asserted here needs some kind of evidence. It flattens the reality of cooperative societies of by avoiding various kind of stress that threaten them. They evolve various complex ways to deal with not only free riders but all kinds of violation of cultural norms.Ways to suppress powerful or persuasive individuals and so on.The least he could do is define what “large” means in this context.
John Cacioppo argued that the stress of loneliness evolved as an adaptation that kept early humans embedded in cooperative social groups. One interpretation is that this sociality made expansion from African forests into savannah environments possible. Another is that climate change first created the savannah environment, and cooperation then evolved as an adaptation to the challenges of life there.
If Cacioppo is broadly correct, would this imply that a significant increase in cooperation preceded advanced social learning and cumulative cultural evolution? Or do you see cooperation, social learning, and cultural evolution as coevolving from the beginning?
Watching science try to wrestle with how and when non-intelligent mechanisms brought about extreme intelligence is popcorn worthy.
Great article One can agree that four evolutionary traits may have an overlapping, the order 4-3-2-1 still holds. Its intensity of these traits which matters. Harari , of course , doesn’t merit discussion.
That GRIP ice core is often used to show Holocene stability and the Pleistocene's huge temperature fluctuations. It is showing temperature variation in one place - the Greenland ice sheet during one ice age cycle. Fluctuations in Africa could well have been more modest, but still very significant, and still likely to have been a significant factor.
Having just now read Darshana Narayanan's review and critique, it makes me glad I never wasted my time reading Harari. It sounds like he is offering a new techno-utopian mythology as master narrative and religious vision. It's frightening that business elites and political leaders may be basing important decisions on this misinformation.