1 Building Cumulative Culture
1.1 Methodological Preliminaries
This book will present an account of the origins, elaboration and interaction of two very distinctive features of our lineage: our dependence on cooperation and our dependence on culture. In these respects, as in many others, we are very different from other primates. In my view, those differences emerged by positive feedback amongst a number of initially smaller divergences from the great ape stock. These include bipedal locomotion, improved causal and social reasoning, reproductive cooperation, increasing dependence on tools, changes in diet and foraging style. Interaction between these initially small departures from great ape lifeways drove an evolutionary trajectory that took us ever further from our great ape relatives. In contrast to many others, I do not see the distinctive features of human life as grounded in a key, difference-making innovation: not language, not understanding others; not a distinctive kind of cognitive flexibility (Mithen 1996, Deacon 1997, Tomasello 2014). I will begin with some methodological remarks, and some cautions about the empirical foundations of the analysis before ending this section with a brief sketch of my general framework. The account is largely in the form of a narrative, and it begins in our deep past, shortly after the human lineage (collectively known as the hominins1) diverged from its sister lineage, the lineage whose living descendants are the two chimp species. The narrative is
1 A note on terminology: I will use “hominin” for any member of this lineage. I will use “human” as an informal term for the late, large-brained members of this lineage.
The Pleistocene Social Contract. Kim Sterelny, Oxford University Press (2021). © Oxford University Press. DOI: 10.1093/oso/9780197531389.003.0001
developed as an explanation, not a mere chronicle: it is full of causal claims about the factors that drove the human evolutionary trajectory, and about their relative importance. However, as the narrative begins in the deep past, 6+ mya (mya = millions of years ago), it is reasonable to wonder whether any attempt to reconstruct the behaviours and ways of life of such long-extinct hominins is just baseless speculation. Indeed, there is a standard sneer that evolutionary narratives, especially human evolutionary narratives, are “just-so” stories: easy to produce, sometimes sounding plausible, and impossible to test (Gould and Lewontin 1978).
This suspicion is not entirely groundless. Traces of past lives disappear; the more distant in time those lives, the more those traces are lost. Those that remain must be interpreted through the lens of theory, and these are themselves controversial. Even so, the evolutionary narrative produced here is, I claim, detailed, coherent and empirically grounded. It identifies multiple, causally interconnected strands, linking foraging strategies, social structure, life history, reproductive strategies and intergenerational cultural learning.2 What is said about each of these constrains what can be said about the others. Coherence, the mutual fit between these separate but causally interconnected elements, is a significant constraint on a complex evolutionary narrative, as any account of the rise of the hominins must be. Moreover, the narrative is empirically grounded at many points: if it is right, it predicts patterns in the traces left by our ancestors. For example, one causal hypothesis links a social change—increased social connection between residential groups—to a more reliable preservation of culturally transmitted information. If that hypothesis is right, signs of that social change should covary with signs of more reliable preservation and expansion of informational capital. In both cases, such traces will often
2 A second note on terminology: some authors use “cultural learning” as a term for some advanced species of social learning. I use “cultural learning” and “social learning” as equivalents.
be faint or ambiguous. But they are not entirely missing. To recycle an analogy I have used elsewhere, in sending an agent under cover, an intelligence service has to construct a “legend”: an innocent fake past of the agent. The construction of a legend becomes more difficult, as it becomes more complex and grounded. The more complex the legend, the more detail the agent has to remember, and the more difficult it is to retain coherence. Likewise, the more points at which the legend is vulnerable to external checks, the more difficult it is to construct. Contrary to the sneer, rich coherent narratives which make frequent contact with the data are not easy to construct. The narrative has gaps and is hostage to future discovery, but it is not a just-so story.3
In general, the empirical foundations of the explanatory framework developed here reflect consensus views in archaeology and palaeoanthropology. However, there are important exceptions to that claim. The picture of human evolution presented here, with its interaction between culture and cooperation, depends on three controversial claims about the past. The first is that our ancestors were competent, cooperative hunters in the distant past: probably as early as 1.8 mya. The second claim is not controversial as such; it is just not discussed. It is the idea that the character of human cooperation changed, between about 150 kya and 100 kya (kya = thousands of years ago) from cooperation as collective action to cooperation through reciprocation, through the exchange of aid. As I see it, these economic changes were initially much more marked in Africa; an economy of reciprocation developed later in the rest of the world. The third is that the threat of intergroup violence did not play a major role in structuring human social life until the Holocene, about 12 kya. I will explain my stance on these issues as they become relevant, but be warned.
3 For a more extensive version of this argument, see (Currie and Sterelny 2017); for an extended defence of the historical sciences, see (Currie 2018).
While the narrative presented here is not a mere speculation, there is no denying that there are important gaps in the historical record. That is especially true of the earlier hominins. Even so, the empirical record is somewhat richer than one might suppose. For example, there is a much richer fossil record of human evolution than there is of chimp evolution, and that itself tells us something important about the different distributions and habitats of the two lineages. Hominin fossils tell us a good deal about hominin lifeways: about their diet; about their physical capacities (for example, the hands, feet and shoulders of obligately bipedal hominins differ from those hominins who still spent a good deal of time in trees); about their habitat preferences (if fossils are found where the hominin died); about their success in dispersing over landmasses; about their life history (teeth sometimes indicate age at sexual maturity). Isotope data can tell us something about diet and even movement patterns (Lugli, Cipriani et al. 2019). Fossils even sometimes hint at social organization (for example, when fossils show evidence of surviving injury or illness that would have required care from others;4 conversely, evidence of violent death). In addition to the fossils themselves, from a bit earlier than 3 mya, we begin to have evidence of their material culture, and of the uses of that culture. That is most directly from the debris they left behind them (their middens, camp sites, work sites), but also from tools with their wear patterns. Unfortunately, this evidence is limited in three important ways. First, unless by amazing fluke, we will find evidence only of common products and activities, and the deeper in time, the more we see only the commonest phenomena. Second, sites disappear or degrade over time, but not equally. There are biases: from example, many Pleistocene sea-edge sites are now under water. Third, with rare exceptions, we see only hard material
4 There is evidence of care as far back as around 1.5 mya, from an individual whose diseases would have left him unable to forage or defend himself (Spikins, Rutherford et al. 2010).
technology and hard detritus. That said, stone technology is both especially revealing and important. Working stone is unforgiving and dangerous, as striking stone can send sharp flakes flying off in unpredictable directions. But it also gives the stone worker access to other materials. Wood, hide and fibre can be worked with stone, whereas soft materials cannot be used to transform other soft materials. Lithics are a keystone technology.
Three-million-year-old rubbish dumps do tell us about threemillion-years-gone lives, but only with the help of theories and models which tell us what the traces from the past are traces of (Currie 2018). Some of these theories are very specific: for example, about the chemical and physical differences between domestic and wild fires. Others are much more general: about the economics of foraging, or the costs and benefits of different ways of managing risk. The role of theories and models in interpreting traces expands our evidential base: current (and near-past) observations become relevant, because those theories must themselves be tested and calibrated. Some of these observations are from experiments: what can a certain kind of stone tool cut, and what does the edge look like afterwards? What does a bone look like after it is first butchered and then gnawed by hyenas, rather than vice versa? Can a thrown wooden, untipped spear penetrate zebra hide, and from what distance (Churchill 1993, Churchill and Rhodes 2009, Salem and Churchill 2016)? But others are from observations of ethnographically known foragers and other small scale societies. That is not because (say) the San of Southern Africa or the Hadza of East Africa are living fossils of Pleistocene lifeways. Rather, as Frank Marlowe says in discussing the Hadza (Marlowe 2010), it is because information about such peoples tests and calibrates our general models of forager economics and forager ecology. In building these general models, the diversity of forager experience is important. Robert Kelly’s superb survey emphasizes that diversity (Kelly 2013), and it is critical in showing the response of forager economic and social organisation to environmental variation. In turn, those general
models guide our interpretation of the traces they have left. For example, we know from ethnography that in many environments most hunts fail, even for the most skilled hunters. In those environments, gathered resources were critical, probably implying some form of division of labour. Likewise, we can estimate the size range of ancient forager bands from the ecological economics of ethnographically known bands (Kelly 2013). We cannot automatically assume that the forager economies of the Pliocene and Pleistocene were like those of the recent past. But if we do think they were different, we need to identify a factor (or factors) that makes them different.
In short: the attempt to understand the lifeways of our ancestors and the ways those lifeways served as foundation and springboard to ours is challenging. But it is not hopeless.
This essay builds on my previous work on the evolution of human social life, and on the cognitive capacities that support that life (or those lives). In particular, it relies upon a view of human cognition that emphasizes its plasticity. We have the capacity to acquire new skills for which we do not have specific genetic preparation, and we can re-purpose existing cognitive circuits to new tasks. This adaptive plasticity allowed our ancestors to acquire novel capacities, and sometimes these were important enough to reshape ancestral niches; stone tool making and fire control were probably examples of niche-changing new skills. Once lifeways change, so too do the selective forces acting on those hominins, ultimately changing their genetic makeup. This process has iterated through hominin evolution. So while I agree that gene-culture coevolution has built human biology, and in ways relevant to human cognition, change typically began with behavioural innovations that were expressions of our adaptive plasticity. In West-Eberhart’s phrasing, genes are the followers, not the leaders, of phenotypic change. While not yet orthodoxy, this view is no longer heterodox. See, for example, (Heyes 2018), (Anderson 2014). On the basis of this assumption about hominin plasticity, I develop two novel
claims: one about cooperation; the other about culture, and their coevolution.
The book itself is organized around a distinctive four-phase model of the emergence of the extraordinary forms of cooperation on which contemporary life depends. I take the first phase to be the suppression of the dominance hierarchy that early hominins would have inherited in some form as the great ape pattern. This suppression allowed a form of foraging that depended on collective action to stabilize and expand. Here the argument partially converges with that of Michael Tomasello, who also argues that collective action with immediate returns played a foundational role in hominin evolution (Tomasello, Melis et al. 2012, Tomasello 2016). The next phase is the transition from a foraging economy based on immediate return mutualism to one in which direct and indirect reciprocation was increasingly important. While still very profitable, the stability of this form of cooperation depended on new cultural and cognitive tools. The third phase is an expansion of the social and spatial scale of cooperation, as residential groups became networked with one another forming larger communities, although the second and third phase are separated more analytically than temporally. However, as chapter 3 shows, an expansion of the social and spatial scale of cooperation (or even passive tolerance of those outside the immediate circle of daily interaction) poses novel problems. That is especially true if the ancestral atom of hominin social organization even roughly resembled the chimp residential group, with its extreme unease with outsiders. Human residential groups are open rather than closed, and as hominin residential groups gradually became connected in larger networks, those networks made it possible, in some cases at least, to solve cooperation and collective action problems at larger scales than the forager band. There are many uncertainties about the baseline and timing of this transition, but in chapter 3 I offer an incremental account of this transformation. Neither the second nor the third phase have been the focus of much attention. In contrast, there has been much
attention on the fourth transition, the (re-)-establishment of hierarchical societies, and the continued (even expanded) role of cooperation despite inequality. This attention is no surprize, for as well shall see in 2.1, the coexistence of serious inequality and continued cooperation is puzzling. I offer my own account, drawing on ideas from Ray Kelly, Robert Kelly and Brian Haydon,5 but differing from them in important ways too. The account I offer places considerable weight on the prior evolution of a hierarchy of esteem, and with it, the oblique intergenerational transmission of the norms and mores of a community. Moreover, I also emphasize the limited opportunities of collective action from below, and give an account of why those opportunities are so limited.
This analysis of the growth and transformation of cooperation is tied to an account of culture and cultural evolution. This account builds on the analysis of The Evolved Apprentice (Sterelny 2012). I am part of the almost universal consensus in agreeing that in our lineage, cultural learning has become cumulative, and that is a critical difference between late hominins and almost all the cultural learning of almost all other animals. I also accept the consensus that one form of cumulative culture, incremental improvement of an existing capacity, depends on the high fidelity transmission of information across the generations. However, I depart from that consensus in two, linked ways. I do not think high fidelity cultural learning depends on a specific cognitive adaptation, and I think its importance to cumulative culture has been over-stated. So, first, I doubt that cumulative culture depends on a specific cognitive adaptation, a cognitive breakthrough that made cumulative culture possible. For in my view, high fidelity transmission across the generations does not depend on high fidelity individual learning episodes. Consider, for example, a sub-adult learning to make an adhesive, one which can only be made through a precisely ordered sequence, and then learning to use that adhesive to attach a
5 See (Kelly 2000, Kelly 2013, Hayden 2014, Hayden 2018).
point to a shaft. That juvenile is likely to have many opportunities to watch expert performance and demonstration, and to eavesdrop on peers. For in forager life, much takes place in the open in public view (Hewlett, Hudson et al. 2019). The sub-adult is likely to have many opportunities to experiment, supplementing socially sourced information with trial and error. Even if there is a lot of noise in the flow of information in a specific learning episode, intergenerational transmission can be high fidelity, if the incoming generation has the capacity to detect their own errors (from signals from the world, or from their elders), and the motivation and the support to locate and correct them. So one important claim is that there are many roads to high fidelity. That being so, it follows that cumulative culture via incremental improvement does not depend on the prior evolution of some specific cognitive adaptation for cultural learning. I do not think accurate imitation learning and/ or collective intentionality are essential prerequisites for accumulating cultural knowledge across a series of generations. Very likely, as cultural learning became increasingly central to hominin lives, gene-culture coevolution tuned human minds to those new demands. We became better at cultural learning. Even so, culture and cumulative culture became important in our lineage before those cognitive changes, whatever they might be.
Second, the importance of high fidelity transmission to cumulative culture has been over-stated; it is essential for only one form of cumulative culture. For cumulative culture has often been described by analogy with the genetic evolution of complex adaptation, and is hence seen only as the incremental improvement of existing capacities (Tomasello 1999, Tennie, Braun et al. 2016, Tennie, Premo et al. 2017, Tennie, Hopper et al. forthcoming). This is too narrow a characterization, and one that has led to an excessive focus on imitation as the critical cognitive capacity on which cumulative culture rests. Indeed, despite their focus on imitation, Michael Tomasello and Claudio Tennie have introduced a vivid metaphor for an alternative and much broader conception of
cumulative cultural evolution, through their idea of a “zone of latent solutions (ZLS)”.
An ability is within this zone, if an agent could acquire it fairly readily by individual learning, even if de facto it is learned socially. For Pleistocene hominins, the location of nearby waterholes was surely within their zone, if even most learned those locations as juveniles accompanying their mother. Obviously, an ability that could only be learned socially is outside the zone. So we can think of the difference between ordinary cultural learning and cumulative cultural learning, as this: cumulative cultural learning enables agents to gain capacities that clearly lie outside their ZLS, whereas ordinary cultural learning, of the kind widespread across animal lineages, only enables those animals to find solutions within that zone, though perhaps more quickly or with less risk. Importantly, the concepts of cumulative culture as (i) building capacities by incremental improvement, and (ii) enabling agents to acquire capacities outside their ZLS are not equivalent. While most abilities/ technologies that are the result of incremental improvement are probably outside an agent’s ZLS, the converse is not true. We have many capacities that we can acquire only through the greater efficiency of cultural learning. This is most importantly exemplified in forager’s famously extensive natural history knowledge of their patch. Very likely, any single item in their information store could be individually learned, but not the whole of it. Likewise, cultural learning enables agents to combine and cross-fertilize information streams from different domains (Muthukrishna and Henrich 2016): in finding out about the world, the division of labour is a powerful tool. Cumulative cultural learning combines (i) increases in the bandwidth of learning, (ii) incremental improvement and (iii) novel recombination, and this is another reason to reject the idea that cumulative cultural learning is grounded in some single, specific cognitive adaptation.
In the rest of this book, I sketch a dynamic in which the expansion of cultural learning, including cumulative cultural learning
in this broader sense, in our lineage supports the expansion and transformation of cooperation in our lineage. Moreover, these new forms of cooperation, in turn, make cultural learning more powerful, more pervasive in its effects on hominin lifeways. Culture and cooperation evolve together.
In very brief summary then, the argument of this book will depend (i) on the idea that large game hunting was important early in our evolutionary career, but that inter-group violence became a serious threat only late in that career; (ii) on a four-stage model of the emergence of our distinctive form of cooperation, and (iii) on the claim that cultural learning, including cumulative cultural learning, became important early, but without depending on special purpose cognitive adaptations. OK. You have been warned about what lies ahead, and what is new and controversial. Let’s go.
1.2 Culture and Cooperation
When agents act together, there are often synergies. Their collective output is greater than the sum of their individual outputs, were they to act as lone wolves. In such circumstances, there is a cooperation dividend. That dividend can be the result of collective action: a group of musk oxen standing together, acting in concert, can see off a wolf pack attack that would be extremely dangerous to any individual ox; a group of wolves can kill an ox that would be fairly safe against any individual wolf. It can be the result of complementarity and the division of labour. Eusocial insect colonies exemplify the power of collective action, but there is also a good deal of pooled individual action, mediated by the division of labour and caste differentiation. In many forager societies, the sexual division of labour is collectively profitable, as it enables the different sexes to specialize on different targets in different places, with appropriate tools and skills (O’Connell 2006). If both sexes targeted the same resources, they would deplete more rapidly, and other resources
would go untouched. Cooperation can manage risk, by protecting against unpredictable fluctuations in the environment. If we allow your people to forage in our territory in your drought years, in return for our right to forage in yours in our droughts, we are both buffered against environmental risk. If you share with me when I am sick or injured, and I share with you when you have a similar misfortune, we are both protected. An explanation of any particular form of cooperation has to show why it is profitable, and how that form of cooperation could emerge incrementally from a noncooperative or less cooperative predecessor. This is the generationof-benefit problem.
An explanation of cooperation also has to explain why cooperation is stable. Notoriously, in many cooperative interactions, while all co-operators are better off if they all cooperate than they would be if none of them cooperate, each agent is better off if he/ she does not cooperate, and everyone else does. For acting cooperatively often has costs: risk, energy, opportunity costs. Yet the profit of the cooperative interaction often does not require that every agent that stands to gain also needs to pay their full share of those costs. Indeed, in collective actions, the profitable outcome is more robust if there is some redundancy; if not every wolf has to be at the right place at the right moment. Otherwise collection action is profitable only if coordination is perfect (Birch 2012). This creates the famous evolutionary temptation to defect or free-ride; to accept help when I am injured, and then stint on helping others. This is the distribution-of-benefit problem (this way of framing the issues is from (Calcott 2008)). Once agents begin to cheat, others do too, and cooperation erodes. Benefit must be distributed in ways that incentivize further cooperation. These acidic effects of uncontrolled free-riding on cooperation make contemporary societies deeply puzzling. Almost all of us live in large scale social worlds where we depend on cooperation for virtually every necessity of life (Seabright 2010). Yet these are also highly unequal worlds in which tiny elites skim off a huge slice of the social surplus. Paradoxically,
contemporary society seems to combine extensive cooperation with rampant freeriding; a puzzle to which we return in 4.2–4.4.
Despite spectacular cooperation failures, contemporary humans have clearly solved both problems, though almost always in partial and error-prone ways. In comparison to most mammals and most primates, we are spectacularly cooperative. This is true of the deep history of our lineage; we have been obligate co-operators for hundreds of thousands of years; perhaps millions of years. This book is about the role of hominin culture in the solution of both the generation and the distribution problem. We have become more culturedependent as our cooperation has become more pervasive, and as the imprint of culture has increased, we have become more cooperative. Hominins are both extremely cooperative primates and encultured primates.
Contemporary humans are encultured in a very rich sense. Our sense of our own identity is complexly entangled with the history, the mores, the legends and the collective experience of the social groups in which we are embedded. We do not just live in specific communities, we consciously identify with the communities of which we are a part, and advertise that identification in dress, accent and other insignias. I do not deny the importance of that rich sense of culture,6 but here I will mostly be writing of culture in a more mundane sense. What we know, what we believe, what we can do and even to a considerable extent what we want, is learned from other hominins. Over the approximately seven million years of hominin history,7 cultural learning has had an increasing imprint
6 This rich sense of culture is often linked to a further claim: a conception of a community’s culture as a cohesive, integrated system. The extent to which the elements of a culture are linked is an empirical question, but I am persuaded by Sperber’s arguments that the more extreme versions of holistic views are untenable, as they have no adequate account of either variation within the one community, nor of the fact that cultures can change piecemeal (Sperber 1996). This becomes relevant again in 4.4.
7 There is actually a good deal of variation in molecular clock estimates of the divergence of the human and chimp lineages; some are as deep as 12 mya; some as recent as 5 mya. For a brief introduction to these complexities, see (Jensen-Seaman and Hooper-Boyd 2013). For a more extensive but exceptionally clear introduction to the use of molecular methods to estimate the dates of evolutionary divergences, see Bromham,
on hominin minds. In that sense, we have become increasingly encultured.
When the more expert aid the less expert, cultural learning is information sharing, and sharing information, like sharing any other resource, is a form of cooperation. Early in hominin evolution, much social learning was probably just a by-product of adult activities: informed action in the world created public information, and the less informed took advantage of that information. A juvenile with her mother will have many opportunities to see what her mother identifies as food, what she takes to be dangerous, who she identifies as an ally and so on. One major change in the hominin lineage took place as the informed began to actively facilitate this uptake, sharing what they know; perhaps a little under 2 mya (Hiscock 2014). On this picture, the growth of cultural learning in our lineage is both the expansion of a form of cooperation and an amplifier of cooperation. Shared information and shared technique, over deep time, has made collective action and other forms of more directly material cooperation more powerful, and it has also provided tools that help control the threat of free-riding. These material benefits of information sharing, once established, came to select for the cognitive capacities and social interactions that made cultural learning more reliable.
I have claimed that hominins diverged from other great apes via positive feedback loops between initially much smaller differences. One of those loops was between information-sharing and other forms of cooperation. An expansion of one is apt to open up space for the other. Thus reproductive cooperation, an important and perhaps early evolving element of hominin reproductive strategies, opens up more space for social learning by very young infants, by exposing them to more sources of information (Burkart, Hrdy et al. 2009). Bipedalism increases the benefits of reproductive
L. (2016). An Introduction to Molecular Evolution and Phylogenetics. Oxford, Oxford University Press, chapters 13 and 14.
cooperation, for once hominins walked upright, their infants could no longer ride safely and conveniently on their backs. Some form of creching would be very beneficial (especially, of course, as hominin infants became ever more immature at birth).8 The same is true of cooperation in the acquisition of resources. Perhaps sometime in the late Pliocene, near Pliocene/Pleistocene boundary, hominins gradually evolved a new lifeway centred on collaborative foragers targeting high value resources (Thompson, Carvalho et al. 2019). These resources are typically hard to find, heavily defended, or both. Animals defend aggressively; have vigorous and well-honed escape routines; hide, are camouflaged; rest in inaccessible places.
Plants defend themselves mechanically (with thorns and shells) and chemically. Harvesting such resources depends on a blend of cooperation, technology and expertise.
To pick just one example, in seasonal, subtropical areas many plants survive the dry by growing underground storage organs (tubers, corms, collectively, “USO”s). These are potentially rich sources of energy, full of carbohydrates (Laden and Wrangham 2005, Wrangham, Cheney et al. 2009). To harvest them, foragers have to recognize the right plant, which is not easy, since in the dry many of these plants are unobtrusive, nondescript stalks. They have to be dug out, which requires a robust, sharp digging stick. Once extracted, many USOs require processing to be made edible. In the right environments, they are available in quantity. But they are demanding targets. Even more obviously, the same is true of meat, brains and marrow from medium and large game. Harvesting game, either by direct hunting or by pirating the kills of other predators, requires a weapons technology, rich information about the habitat and the habits of the target animal, and, at least until the relatively recent invention of projectile technology, cooperation between the foragers. Yet efficient harvesting of resources, together
8 Here is just one place where our ignorance of soft material technology is troubling: we have no idea when baby slings were invented.
with an effective collective response to predators, makes a slower life history possible, giving juveniles more time to acquire skills and information, and adults more time to hone their skills and add to their information store through experience (Kaplan, Hooper et al. 2009). There is positive feedback between the efficient capture of difficult but rich resources, cultural learning and a slow life history. Adult profit supports juvenile skill acquisition, in turn making their future activities as adults profitable.
This general picture is not controversial. There are vigorous debates on detail: for example, the role and relative importance of hunting vs gathering and scavenging; the relative importance of size and skill; the cognitive prerequisites of cultural learning. There are debates about exactly when cultural learning became essential (Corbey, Jagich et al. 2016, Tennie, Braun et al. 2016), linked to debates about when hominins became dependent on stone tools (Shea 2017). But few would disagree that by the mid-Pleistocene, hominins depended on some combination of technology, cooperation and expertise. These all depend in various ways on cultural learning. This chapter is mostly about how and why cultural learning came to have such a massive footprint in our history, and on the effect of cultural learning in amplifying the benefits of cooperation. The next chapter is focussed on distribution; on the control of freeriding and the increasing difficulty of that control through the second transition. To put these projects into context, what follows is a brief overview of hominin history.
1.3 The Prehistory of an Unusual Ape
Identifiable hominin fossils date back into the Miocene and through the Pliocene, but the main morphological changes that make us look very different from other great apes—bipedalism and encephalization—seem to have mostly taken place from the end of the Pliocene and through the Pleistocene (Table 1.1).
Table 1.1 Navigating the Epochs
Epoch Dates before Present
Miocene
Pliocene
Pleistocene
Holocene
23.03–5.3 mya
5.33–2.58 mya
2.58 mya–11,700 ya
11,700 ya–present
Before giving a thumbnail sketch of hominin history from our divergence from the chimp lineage (probably at about 7 mya), a couple of initial cautions are in order. Much is not known. Especially for the first 4 million years of this period, fossils are few, and many are fragmentary. Only at the very end is there archaeological material: that is, evidence from the products of hominin action (tools, middens and the like), and even that is controversial. So all of what follows remains somewhat conjectural; some is very conjectural. Second, in the technical paleoanthropological literature there is a lot of debate about taxonomy: about assigning fossils to specific species, and, often, defending claims about species identity and difference. I do not place much weight on specific species identification in the Miocene, Pliocene and much of the Pleistocene: fossils are often fragmentary, and they are typically so rare that we have little evidence of the natural variation within these species. So when I use specific species names like habilus, erectus and Heidelbergensis, I use these to indicate a shift in morphology and probably behaviour over time. These are successively larger and more encephalized hominins (that is, the ratio of brain size to body size has increased). They are successively somewhat more similar to very recent humans. But we do not know how many biological species of hominins there were, say, a million years ago: we do not know whether there were reproductively isolated but morphologically similar lineages existing at the same time. Moreover, evidence of some gene flow between Anatomically Modern Humans
(AMHs), Denisovans and Neanderthals suggests that there can be gene flow between morphologically distinguishable lineages.
One obvious and notorious difference between humans and other great apes is our relatively and absolutely large brain. There are uncertainties about the history of hominin brain expansion, but Table 1.2 offers a rough guide.
This table illustrates a general trend, but it should be read with a lot of caution. For one thing, body size also matters: as bodies get larger, brains get larger too. For another, in some cases, especially with erectus, there is a large variation both regionally but also temporally, with later erectines typically larger-brained: see (Klein 2009) pp 306–307. Finally, it can be argued that neural density is critical, and this is not constant over mammals or even primates: see (Herculano-Houzel 2016). That said, the earliest hominins (ardipithicenes and australopithecines) had approximately chimp-sized brains; they seemed to live in somewhat more open and seasonal habitats than the pan species. At least some were habitually bipedal, but it is less clear when they become obligately bipedal. Some of these early hominins seem to have retained
Table 1.2 Hominin Brains over Time
Species Brain Volume in cubic centimetres
Australopithecines
Homo habilis
Homo erectus
Homo Heidelbergensis
Upper Palaeolithic Sapiens
European and West Asian Classic Neanderthals
Reference
434–530 (Klein 2009) p 198
Approx. 650 (Gamble, Dunbar et al. 2014) p 99
Average 950 (Rightmire 2013)
Average 1230 cc (Klein 2009, Rightmire 2013) (Rightmire 2013)
Average 1577 (+/- 135 (Klein 2009) p 308
Average 1435 (+/- 184) (Klein 2009) p 309
