SCIENCE AT THE SHINE DOME canberra 6 - 8 may 2009
Symposium: Evolution of the universe, the planets, life and thought
Friday, 8 May 2009
Professor Kim Sterelny
School of History, Philosophy, Political Science and International Relations, Victoria University of Wellington, New Zealand
Kim Sterelny studied philosophy at Sydney University in the early Pleistocene, and has subsequently taught it at various Australian and New Zealand universities. He is currently based at the Australian National University, but still has a foothold at Victoria University of Wellington. His interests have always been at the interface of philosophy and the natural science; in the last decade, his focus has shifted to the connections between philosophy and the life sciences. He has a recent book on cognitive evolution (Thought in a
hostile world) and is gestating another. In addition to philosophy, Kim spends
his time eating curries, drinking red wine, bushwalking snorkelling
and bird watching.
When apes go weird, or why humans are so strange
(Click on image for a larger version)
Kim Sterelny: Welcome to the anticlimax and the end of the classy PowerPoint displays. I am a philosopher and, in general, philosophers are pretty technophobic and low-tech. I am at one end of the technophobic low-tech spectrum, but you will get a few pictures. Those who have seen my talks before will realise that I am gradually evolving towards science norms in this regard, although I have a long way to go.
(Click on image for a larger version)
I will begin by talking about two ideas, which I think in the last 10 to 20 years have framed a lot of thought about human evolution and human evolutionary history; they have been elaborated in a particular way, which I will call the standard model. I think the framing ideas are right, but I think their elaboration in the standard model is, at best, a pretty serious oversimplification. So I will begin by briefly explaining these framing ideas and what lies at the back of them. Then I will talk you through the standard model, being a little critical as I go. Then, in roughly the second half of the talk, I will to try to introduce something of a variant of the standard model – a kind of alternative that I think avoids at least some of the oversimplifications of that model – and leave you with a few thoughts about what kind of big picture of the evolution of human behaviour and cognition you get at the end of this story.
(Click on image for a larger version)
I will begin with an insight of Jared Diamond’s. In The Third Chimpanzee, which he wrote about 15-or-so years ago – this is something that I was very struck with at the time – he pointed out that, if a Martian zoologist had visited the Earth three or four million years ago and done a formal survey, he would have noticed a few hairy, chimp-like critters wandering around east Africa and not taken much notice of them. They would have gone down as a relatively minor element of a relatively minor lineage in a very diverse chunk of the world. As we all know, over the next three million years, something happened: minor fauna changed in very dramatic ways, expanding geographically and ecologically. There was a massive explosion in tool use. This chimp-like critter became extraordinarily cooperative for a mammal and extraordinarily cooperative for a primate. It changed its sexual, social and family behaviour in striking ways. I can recognise my daughter, for example, which my ancestor of a few million years ago probably could not do – male chimps do not have much chance of recognising their daughters.
So something happened and, equally important, it seemed to happen only to us; it happened only to our ancestors. You do not see a similar kind of extraordinary evolutionary change in other lineages. This all happened quickly. One of the reasons for thinking that something striking is going on here is the rapidity of the change. For those of you who remember a little of the recent history, when Alan Wilson and his collaborators came out with relatively recent dates for the hominid-chimp split – dates of four million-or-so years ago – they were very controversial and, in part, that was because people did not think that was a long enough time for the evolution of such dramatic differences between our lineage and the chimp lineage.
(Click on image for a larger version)
The second of the framing ideas has been the discovery – again, over the last 20-or-so years – that cognitive competences that we take to be pretty ordinary are, in fact, extraordinarily difficult. This came first with Noam Chomsky and his pointing out and elaborating the idea that language is an extraordinarily subtle, complex, intricate and surprising communication system; it became elaborated through cognitive psychology and artificial intelligence (AI). Artificial intelligence turned out to be far harder than anyone thought in the sixties. In the sixties, people in AI were confidently predicting – not just to get the grant money – that, in 10 to 20 years, they would be able to build machines that could genuinely think. In the sixties they were confidently claiming that, within 10 years, they would be building chess-playing computers that would beat any human. It took them 20 years longer than they had expected, and chess is an easy case because it is such a clearly defined microworld. It turned out to be much harder to build computers that were socially smart, such as ones that could chat a girl up for you and do things like that – I am still waiting for that machine!
Being a human, being able to survive and act adaptively in a human environment, turns out to depend on a whole range of skills, a whole range of cognitive competences, which are high-information-load competences, which depend on being able to access lots of information and use that information well. I think both of those ideas are right: something extraordinary happened to drive human evolution; and ordinary human life or ordinary human skills are highload skills.
(Click on image for a larger version)
A kind of standard story has developed about how all of this fits together. Part of the standard story is a selective story: a story about the selective environment. That goes by the name of the ‘social intelligence hypothesis’ or the ‘Machiavellian intelligence hypothesis’. The idea is that there is a kind of feedback loop between social complexity and intelligence. Positive feedback has driven the evolution of intelligence in the human lineage, mostly because you want to get the benefits of cooperation without being cheated. On various standard analyses, it turns out that is a difficult task. You have to be smart to both cooperate and not be ripped off – and, if possible, to do a bit of ripping off yourself. But, if everyone is under selection for being smart in that way, the social world gets more and more complicated: you have to be smarter and smarter to avoid being ripped off. So there is a positive feedback relationship between individual intelligence and social complexity.
The second idea is that high-information load is managed by innate, evolved and specialised cognitive mechanisms. That is known in the trade as a ‘massive modularity hypothesis’, and I will come to that shortly. First, I will speak a little about the Machiavellian intelligence hypothesis.
(Click on image for a larger version)
This is one of my favourite book covers. About 10 years ago, a series of volumes came out, both in the primatological literature and the human cognition literature, exploring this hypothesis. Roughly, it goes this way.
(Click on image for a larger version)
You would need some kind of internal feedback-driven model of human cognitive and behavioural intelligence because it was fast, extensive, dramatic and unique. If it had been the result of external forcing, through some sort of deterioration with increased seasonality, increased climatic fluctuations or something like that, you might see a human response to that which required greater human intelligence, but you would see parallel responses in many other lineages – and, arguably, you do not. So the model you need has to be an internalist feedback-driven model.
(Click on image for a larger version)
The most plausible candidate for that is the hominin’s dilemma, which is the dilemma of cooperating without being cheated. Cooperating, when it works, is an enormously profitable adaptation. Hominins acting together can do all sorts of things that they cannot do alone. They are safer from predators, they can extract resources from their environment, they can manage risk in their environment and they can manage fluctuation. There are all sorts of advantages to be got from cooperation; it is too profitable to opt out of once others have opted in, but it is risky because of the threat of defection. Many benefits of cooperation do not depend on everyone cooperating wholeheartedly. Collective defence is far more effective than individual defence. But, typically, collective defence will not depend on every single agent being prepared to put their body on the line, if necessary; it is much nicer if the others will do it.
So you want to be in a system with collective defence. You would rather be in a system of collective defence where others carry most of the burden, but you certainly do not want to be in a system of collective defence where you are doing the collective defending and others are not. So there is this risk of defection coupled with the extraordinary profits to be got from successful cooperation. So you have to be a cooperator, but you have to be a vigilant cooperator – a cooperator that minimally does not get ripped off and where you are not paying more than your fair share. So that selects vigilance, which selects for intelligence. But, of course, the same thing is happening with every agent in the population, so the population gets smarter.
(Click on image for a larger version)
Because the population gets smarter, the requirements for vigilance become greater; hence, you have to be smarter still to be a successfully vigilant cooperator. So the feedback crank turns and turns and you go from a last common ancestor, which was presumptively about as smart as a current chimp, to us.
(Click on image for a larger version)
I am going to come back to that picture because I think, although there is something right about it, there is something importantly wrong about it too. But, before I do that, I want to say a little about the second of the framing ideas and how it has been elaborated: the massive modularity picture.
As I have mentioned, there is ordinary human action: making and using tools and discussing things and coordinating with people. One of the things that we are incredibly good at is tracking what other people intend to do in ordinary social interaction. So we are socially and cognitively aware of not just ourselves and our own intentions but also others and their intentions. That is a very important aspect of a successfully coordinated social life. We use language and we are aware of the norms that govern our local community. I did not have to be told explicitly that, to give this talk, I should turn up wearing some clothes, even if it was warm enough not to do so. It is the same for a whole range of other perfectly conceivable activities that I could do and that many primates do but which I would never think of doing in public. I will leave it to your imagination to think of what some of those might be. All of that we, or at least almost all of us, do effortlessly; those of us who do not do such things stand out like a sore thumb and, unless they are extraordinarily gifted in some other aspect of their life, have extraordinary poor prognoses in human life.
(Click on image for a larger version)
So how do we manage to do all that? On the standard story, our minds are ensembles of special purpose devices. Over hominin history, we have been faced repeatedly with a certain welldefined set of problems – of social interaction, of language use, of awareness of the norms of our local community, of how to master and use simple tools and so on. As a consequence, we have developed special cognitive adaptations, each of which suits us for one or another of this class of recurrent problems. This picture does not deny the importance of learning. All of this depends on learning, but it does say that we have solutions to these problems partly preinstalled in us and the rest is done by highly channelled and canalised learning. So, yes, we have to learn the particular language that we speak, but we know a lot about language when we are born – a lot about language is wired into our genes – and we also know the kinds of things that we have to learn in learning a language. On this picture, what is true of language is true of these other standard features of human competence.
(Click on image for a larger version)
So on this picture, famously summarised as the ‘Swiss army knife model of the mind’, the distinctive features of human intelligence or our kind of distinctive capacity to act on the world does not depend on our having a particularly powerful general purpose learning machine; it depends on us having an ensemble of special purpose capacities. As I have got here, this idea has now sort of become common cultural currency – common enough to be exploited in various visual jokes. Mostly they are not very funny. This is one that I ripped off the internet.
(Click on image for a larger version)
It is about as funny or unfunny as any of them, but the point is that the joke only makes sense against a background of assumptions about localisation and being innately equipped to think in various ways and so on. So some version or other of massive modularity has become as much part of our contemporary cultural currency as progressivist images of evolution of the kind that Lindell was talking about earlier.
(Click on image for a larger version)
What does this picture depend on? It depends on a tacit assumption of stability. If the informational demands to which humans must respond remain stable over evolutionary time – noone has any idea about how long evolutionary time is, but presumably we are talking of at least tens of thousands of years here – they form a fixed target on to which selection can lock and preassemble in our heads large chunks of the information that we need. So the kind of tacit background assumption in all of this is that the Pleistocene, the period in which the distinctive human head was assembled by evolution, was an environment in which the informational challenges that our ancestors faced were serious and difficult but stable, recurring and relatively discrete.
(Click on image for a larger version)
I think sometimes that assumption is pretty decent. If you think, for example, of one of the experimentally better-established sorts of modules or specialisations, kind of a folk-physics module, there is nothing implausible about the idea. The physical properties of rocks, sticks, bones and things of that kind do not change over time, and their importance to humans has not changed over time. It is important for us now, as it was important for our ancestors, to understand the kinds of mechanical interactions between middle-size objects. So important, discrete, stable: it would not surprise me in the least if we had an evolved specialisation for dealing with physical interaction, with tool making and using. Chimpanzees apparently do not have such evolved specialisation. That is not surprising, as tool use is less central to their life than ours. However, if we are competent only when and because we have these partial solutions preinstalled, genuinely novel environments should pose an extremely difficult challenge to us – and one reason why I think this picture cannot be right is that they do not.
(Click on image for a larger version)
There is a kind of LAX paradox. If any of you have ever been through LAX, then first you have my commiserations: it is hell on Earth. It is a very unusual and new form of hell on earth – it is not a Pleistocene hell on Earth – yet remarkably few of us die while traversing that chunk of the world. It is a particularly vivid example of something that is surely obvious. We now live in a world massively transformed from that of our ancestors. That transformation does cause some dislocation in our behaviour. Medical literature is full of diseases of affluence, obesity and stuff like that. We continue liking fat, oil, booze and stuff like that, even though that may be a result of behaviour that was adaptive in the Pleistocene no longer being adaptive. So there is some problem of adaptive lag. But we are not shattered by adaptive lag; we are not rendered hopelessly incompetent in the face of novelty. To the contrary: people manage novel environments with remarkable success and that is why there are so many of us.
So the massive modularity picture of how we deal with highload information, high-information-load problems, has to be at best incomplete. I think that is all it is. I do think there are some modules as they say, but it is not the right general explanation of how our lineage responded to the problems of increasing informational demand.
For the rest of this talk I want to move from the standard model to developing an alternative to the standard model. I will begin by tweaking the standard model’s picture of the selective environment and then I will talk about an alternative model or an alternative picture of how we cope with high-information-load problems.
(Click on image for a larger version)
What is wrong with the Machiavellian model? First, it overstates the problem of keeping track of cheats. In small-scale social environments – it may not be true in contemporary environments – everyone knows who the cheats are. Everyone who lives in a village knows who you can trust and who you cannot. They know who the dangerous drunks are. They know who will keep their promise when they say they will turn up and help and they know who cannot be relied on. All of that kind of information is common knowledge in small-scale social worlds. Knowing what to do about dangerous drunks might be a different problem, but identifying them is not so hard. So I think it overstates the problem of vigilance of monitoring and it understates the problems of coordination. I think the problems of coordination are very serious indeed.
(Click on image for a larger version)
So I do think the general picture is right: managing cooperation was the selective driver of human intelligence. We are smart because we are cooperative and we are cooperative because we are smart.
The big-picture change in hominid ecology over the last 2½ million years is that we went from being food of predators to chasing predators away from their food and taking the food from them. In general, except in the case of fishing, we do not predate much on predators, but we have displaced them.
(Click on image for a larger version)
This is a Cape buffalo. They weigh about 1,000 kilos and are famously ornery and dangerous animals. About 200,000 years ago, things like that started to turn up in human middens on a regular basis. Think of the cognitive difficulties of killing something like that, when you are armed only with a sharp stick or a stick with a sharp piece of stone at the end. We did not have distance-killing technology 200,000 years ago. It is still not exactly clear when bows and spear throwers were invented, but there is certainly no sign of them in the record until about 35,000-or-so years ago. We were killing things like that much deeper in time and regularly.
How did we do it? Certainly we did it with technology – noone was doing it with their bare hands and teeth – but we did it also with cooperation and knowledge. To kill something like that with reasonable safety, you had to know a lot about the animal, you had to know a lot about your environment, you had to master the technology that you had and you had to do it in a group; it was a cooperative enterprise. That not only imposes considerable cognitive demands of both on-the-spot cooperation – of actually organising the hunt and making sure that you do not make mistakes – but also involves accumulating information over time and over the generations. No one individual would have figured out how to safely kill something like that, or what its habits were or any of the informational resources you needed to be a successful forager 200,000 years ago, equipped with basically fairly simple lithic technology. Accumulating those cognitive resources would have been a piecemeal multigenerational activity. I cannot prove that directly, of course, but I would still bet my house on it being true.
(Click on image for a larger version)
Cooperative foraging depended on cooperation, on information technology and on increasingly informationintensive extraction of resources from the environment. Cross-generational cultural learning and on-the-spot coordination were crucial for the kinds of lives that our ancestors lived.
(Click on image for a larger version)
So, instead of thinking of there being a sort of vigilance-intelligence feedback loop here, the model is a sort of life-history ecological-information feedback loop. Human life-history changes depended on increasingly successful knowledge-dependent profitable foraging. As foraging became more profitable, you could support people acquiring the skills longer. As they acquired the skills longer, their ultimate levels of expertise became greater; they became increasingly more effective and increasingly successful foragers at their peak.
Do not worry too much about the numbers up there. There is some quantitative information on this but for only a very small number of societies. That is from a central American foraging culture and there are some similar numbers from a few others. So there is a little bit of quantitative information about how long it takes adolescents to start to be able to feed themselves in foraging societies and at what age they peak. It seems roughly as though mostly you are not feeding yourself until you are about 18, you are not at your peak until you are about 30 and you are kind of not in life-profit until you are about 40 or maybe even a bit older.
(Click on image for a larger version)
So it is a very different picture of how the feedback loop works, but the core idea is similar: the selective driver of human evolution is the problem of managing cooperation.
(Click on image for a larger version)
So, instead of the Machiavellian intelligence hypothesis, we have a social foraging hypothesis. The core idea is the same – we are smart because we are cooperative and vice versa – but the main challenge is different. It is the challenge of assembling and using the informational resources that you need to extract highly valued and heavily dependent resources from the environment in a collaborative and cooperative way.
What about novelty? Novelty is pervasive in human life. The reason we can cope with novelty now is that we have always had to cope with novelty. In general, human environments have never been stable over long periods of time, and that is mostly because we change our own environments: we change them when we move, when we change our economic style and when various forms of social change take place, such as with family structure and size changes and so on. So human environments have always been in flux – and they have been in flux in ways that matter because humans affect their own social, physical and biological environments. Think of the different pathogens that people face when they move from sedentary lifestyles to non-sedentary ones, and vice versa. That is just one example of the many ways that humans’ own interactions with their world change the world they are in, in ways that matter for fitness. So genes have never been able to predict the environments into which we have been born.
(Click on image for a larger version)
So, instead of basically having prewired heads, we evolve mechanisms that enable us to accumulate information and use that information cross-generationally. The key idea I want to get across here is that these mechanisms involve not just internal changes to people’s heads – things like language and the ability to learn by imitation, learning how to do something from seeing someone else to do it; so there were internal changes to human heads which enable us to accumulate information – but also our organising the developmental environments of the next generation. It is the organising of that developmental environment that is important for the evolution of our capacity to accumulate and use information cross-generationally.
(Click on image for a larger version)
Roughly speaking, if you go back 2½ million years ago, there was a bit of social transmission; but it was social transmission as a side effect. If you are adults using a digging stick to extract buried food and you organise your foraging lives around doing that, the kids with you will explore their world in differential way: they will play with discarded digging sticks and see the results of broken digging sticks and so on. So, as a side effect of your economic activities, their trial and error learning will be tweaked in various ways. That is probably all you need for some simple but reliable transmission of information about simple tools.
(Click on image for a larger version)
Who knows when? Maybe something like a million years after that you got a gradual transition to a capacity to innovate. If you look at human technological evolution, mostly all we can see is stone technology – so we are seeing a very small sample – but it is amazingly conservative for a long period of time. Humans or hominins have been using stone technology for about 2½ million years and nothing that looks like even rapid change, in the sense of changing at the hundred-year scale, is visible until about 200,000 years ago. So for about 2.3 million years there is astonishing conservatism in the technological record. I think what has happened is that cultural learning existed, but for a long period it was very low fidelity. It was low fidelity both because the learning environments of children were not organised and because children did not have distinctive cognitive adaptations for cultural learning. At some stage – I suspect it was somewhere around a million to half a million years ago – that began to change and there was both the evolution of individual cognitive capacity and increasingly organised and adaptive developmental environments, until we have got to something like now -
(Click on image for a larger version)
‘now’ as in a couple of hundred thousand years ago – where the capacity to innovate is coupled with the reliable capacity to preserve innovation. You can think that the technological record was conservative either because our ancestors were too dumb ever to come up with a new idea or because they came up with new ideas reasonably often just by trial and error but those new ideas did not preserve; there was not sufficiently hi-fi cultural transmission to the next generation.
(Click on image for a larger version)
So, about 2½ million years ago and for the next couple of million years, humans made artefacts which looked like that. They are examples of Oldowan technology. It began about 2½ million years ago and has continued more or less until now, but it gradually became supplemented by things that look a bit fancier.
(Click on image for a larger version)
So, about 1.8 million years ago, bifacial hand axes or Acheulean hand axes started to appear in the record. This is a late one; the early ones do not look quite as nice and as fancy and symmetrical as that. So at this stage clearly there is some capacity to innovate linked with some capacity to preserve innovation. About a couple of hundred thousand years ago – the dating here is pretty open textured – you started to get fancier lithic technology.
(Click on image for a larger version)
These are microliths that quite likely are arrowheads or spearheads. When you get these, you get a variety of other stone implements as well. So the tool kit is getting larger and it is getting fancier. You are starting to get distinctive regional styles, which are not explained just by the local availability of raw materials. So there is a transition in human life: a transition in the capacity to come up with relevant information and reliably transmit that information to the next generation.
(Click on image for a larger version)
The key hypothesis is that that transition depends on both organised learning environments and individual cognitive adaptations. So what kind of learning environments do I have in mind? I think something that exists in the modern world is a good model for thinking about the evolution of this – namely, apprentice learning.
(Click on image for a larger version)
Apprentice learning is learning by doing. Noone acquires trade skills just by looking at a skilled tradesman. Noone becomes a carpenter just by looking at carpenters. Very few people become carpenters just by individual trial and error either. The acquisition of skills of these kinds is a hybrid process. It depends on trial and error exploration but in highly organised learning environments in which, typically, you have access to raw materials, including raw materials which are partly processed; you have access to appropriate models and examples, some of which are complete, some of which are incomplete, some of which have gone well and some of which have gone badly; you have access to tools; you have access to many examples of people using those tools – and, of course, in modern cases, you get at least some explicit teaching and instruction as well. Of course, in the Palaeolithic, there were not medieval craft guilds. But, if you look at the ethnography of forager societies, craft transmission in such societies often bears quite striking resemblances to apprenticeship transitions.
(Click on image for a larger version)
So, on this hypothesis, children are active explorers of their world, but they explore a world seeded with informational resources from and organised by the parental generation, and that makes it possible for them to reliably acquire learning-dependent complex competences.
(Click on image for a larger version)
I like this picture because it shows the social nature of learning in an apprentice workshop. That was taken in Australia in the 1960s.
(Click on image for a larger version)
On this picture, humans are strange, in part, because we have new heads; we have heads with new devices in them. But we are also strange because we have evolved the habit of building environments that scaffold the development of distinctive human skills. Those new developmental environments, modelled by the apprentice learning model, are as essential to the distinctive human competences as the intrinsic cognitive changes that have taken place in our heads.
(Click on image for a larger version)
In some ways, science is an example like this. Quine once said that science is just organised common-sense. It is no such thing. Science, in lots of ways, is an utterly atypical product of human cognition. But it is typical in other ways: it is essentially collective, it is cumulative and it is dependent on an informational world that we ourselves have made. In that way, and only in that way, science is typical of what makes us so strange a primate.
Discussion
Jenny Graves: Thank you, Kim. Are there any questions?
Question: Thanks for presenting a very lovely argument. I wonder whether you could elaborate one thing that you brought up at the beginning of the talk, which was why this happened to us and did not happen to other species.
Kim Sterelny: I wish I knew the answer to that better than I do. My best guess is that it has something to do with the fact that there was an external triggering event; that our ancestors lived in environments which became increasingly seasonal and increasingly open. Basically, instead of living in forests, we began to live in woodlands and savannahs because our world changed around us. That selected for a collective defence of some kind. We became much more vulnerable to predation than we had been and that our relatives who continued to live in forests were. But that is a best guess, I am afraid. I think there had to be some external trigger of one kind or another. But, once that external trigger took place, it was the internal dynamics that drove most of the rest of the process.
Question: So it is possible that we were just the first species to be put in such an external trigger.
Kim Sterelny: It is possible. But, of course, once it happens to us, we change the world in various ways which makes it very difficult for any other species to take a similar evolutionary trajectory. We have pre-empted it. We did not pre-empt it right at the beginning. Neanderthals and many other hominids were also evolving – but chimps are fresh out of luck now.
Question: I was going to ask about Neanderthals actually, but you have just mentioned them.
Kim Sterelny: My best guess is that Neanderthals were just unlucky. There is a lot of stuff in the palaeo-archaeological literature that looks for some kind of intrinsic cognitive difference between us and the Neanderthals that explains why we survived and they did not. But, if you look at human or sapiens history, many isolated sapiens populations have gone culturally extinct. Cultural extinction in their case need not mean biological extinction, because they could interbreed with their cultural replacements. In most cases they became extinct because they were unlucky in various ways. Jared Diamond’s Guns, germs, and steel is a kind of nice model of how accidental environmental differences can ramify in their importance in interactions of this kind, and I think a similar story is likely in the Neanderthal case. They were unlucky in the kinds of economics and ecology that constrained the kinds of technologies and social life that could be supported by the environments in which they found themselves; hence, in interaction with sapiens, we eroded their environments more than they eroded ours.
Question: There was a fascinating 1960s-something sci-fi book called Babel-17 by Delany, I think, which explored the consequences for a modern human of not being able to say the word ‘I’. At what stage does self-awareness come in?
Kim Sterelny: There is very interesting work in paleoanthropology on what they call the establishment of ‘symbolically marked cultural groups’. That is where they are looking for ochre, decorations and things of that kind, which are plausibly interpretable as signals that ‘we are aware of who we are; we are aware of ourselves as a distinctive group different from you lot.’ Obviously, a lot of conjecture is involved in any of this because the physical record, especially when you go back more than 30,000-or-so years, is very fragmentary indeed. But, even on the most generous kinds of interpretations, which are thoughts that ochre and things of that kind are archaeological markers of this, it does not go back all that far: about 120,000-or-so years ago in Africa. So, if that is the right archaeological signature of the self-awareness or self-identity phenomenon, it is quite shallow.
This is a collective self-identity example rather than an individual self-identity example, so it is much, much, much harder to think of what might be a plausible archaeological signature of self-awareness in some strong awareness of yourself as a narrative being over time. Thomas Suddendorf, in Queensland, argues that what he calls ‘mental time travel’ is relatively recent and is basically linked to planned behaviour. You need to be able to be aware of yourself as a narrative being over time to do things which will not get anything like immediate reward. If that is right – again there is archaeological signature of tools being made long before they will be needed and stuff like that around 150,000 to 200,000 years ago.
Question: Tell us how they killed the buffalo with sticks 200,000 years ago.
Kim Sterelny: With spears.
Question: Digging pits perhaps?
Kim Sterelny: They might have done it that way. They certainly had spears – heavy javelin-style spears. I am guessing. For somewhat later things, there is smoking-gun evidence. There are animals with spear tips in their bones; as far as I am aware, we do not have anything like that for buffalo. But it is known that heavy javelin-style spears were used about 300,000 years ago.
