An influential version of social theory is the ‘Machiavellian Intelligence’ hypothesis (Byrne and Whiten 1988; Whiten and Byrne 1997). Social interactions and relationships are not only complex but also constantly changing and therefore require fast parallel processing (Barton and Dunbar 1997). The similarity with Niccolò Machiavelli (1469-1527), the devious adviser of sixteenth–century Italian princes, is that much of social life is a question of outwitting others, plotting and scheming, entering into alliances and breaking them again. All this requires a lot of brain power to remember who is who, and who has done what to whom, as well as to think up ever more crafty wiles, and to double bluff the crafty wiles of your rivals – leading to a spiralling arms race.
‘Arms races’ are common in biology, as when predators evolve to run ever faster to catch their faster prey, or parasites evolve to outwit the immune systems of their hosts. The notion that some kind of spiralling or self–catalytic process is involved certainly suits what Christopher Wills (1993) calls ‘the runaway brain’, and this idea is common among theories that relate language evolution to brain size. These take the social function of the brain a step further but I will leave them until the next chapter. In general, social theories of the evolution of intelligence have been very successful over the past ten years. They have shifted the balance from male–dominated technological explanations to those that appreciate the complexity of social life. Research on the topic is growing but many questions remain. Why, for example, was there pressure for such a great improvement in social skills? Competition within the species is implied but why should just this species and no others take this expensive route? I also wonder how much our peculiar abilities to do mathematics, program computers, paint pictures or build cathedrals actually come down to social skills. Many people think the social theories are the best we have, but the question of brain size is far from resolved. No one knows for sure how and why we got our enormous brains.
Did memes drive brain size?
I am going to propose an entirely new theory based on memetics. In summary it is this. The turning point in our evolutionary history was when we began to imitate each other. From this point on a second replicator, the meme, came into play. Memes changed the environment in which genes were selected, and the direction of change was determined by the outcome of memetic selection. So the selection pressures which
produced the massive increase in brain size were initiated and driven by memes.
I shall explore this new theory in two ways, first by rather speculatively looking at our origins again, and then by examining in more detail the processes of memetic driving.
The turning point was when early hominids began to imitate each other. The origins of imitation itself are lost in our far past, but the selective (genetic) advantage of imitation is no mystery. Imitation may be very difficult to do but is certainly a ‘good trick’ if you can acquire it. If your neighbour has learned something really useful – like which foods to eat and which to avoid, or how to get inside a prickly pine cone – it may pay (in biological terms) to copy him. You can then avoid all the slow and potentially dangerous process of trying out new foods for yourself. This is only worthwhile if the environment does not change too fast, a factor that can be modelled mathematically. Richerson and Boyd (1992) have shown the conditions under which natural selection might favour more reliance on social learning (including imitation) rather than individual learning. Economists have devised models of how optimisers (who incur the cost of making a decision) can coexist with imitators (who avoid the cost but make inferior decisions) and studied how fads and fashions result when large numbers of people all imitate each other (Bikhchandani
et al.
1992; Conlisk 1980). Indeed, fads and fashions have been associated with imitation ever since Charles Mackay (1841) blamed such ‘extraordinary popular delusions’ as the South Sea Bubble and the Dutch seventeenth–century mania for tulips on ‘popular imitativeness’.
But why did generalised imitation apparently evolve only once? We know from studies of other animals, already discussed, that social learning is fairly common in the animal kingdom but true imitation is very rare. Why should it have arisen in early hominids rather than any other kind of animal?
I suggested that imitation requires three skills: making decisions about what to imitate, complex transformations from one point of view to another, and the production of matching bodily actions. These basic skills, or at least the beginnings of them, are available in many primates and were probably available to our ancestors of 5 million years ago. Primates have good motor control and hand co–ordination, and good general intelligence which would enable them to classify actions and decide what to imitate. Some of them can imagine events and manipulate them mentally, as is shown by their use of insight to solve such problems as reaching food with sticks or by piling up boxes, and, most notably, they have Machiavellian Intelligence and the beginnings of a theory of mind.
The connection between advanced social skills (or Machiavellian Intelligence) and imitation is this. To indulge in deception, pretence and social manipulation you need to be able to put yourself in another’s shoes; to take the other’s point of view; to imagine what it would be like to be that other. This is precisely what you need to be able to imitate someone else. In both cases you must be able to transform what you see someone else doing into what you would have to do to achieve the same end, and vice versa. Finally, our ancestors used reciprocal altruism, that is, helping others who will later help you in return. As we shall see (
Chapter 12
), a common strategy in reciprocation is to copy what the other person does; cooperating if they cooperate and refusing if they do not. With all these prior skills imitation is not such a huge leap for evolution to make.
I suggest that the social skills others have singled out as directly responsible for our large brain were in fact responsible for the prior step of acquiring imitation. As soon as our ancestors crossed the threshold into true imitation a second replicator was unwittingly unleashed. Only then did the memetic pressure for increasing brain size begin.
When was this turning point? The first signs of obvious imitation are the stone tools made by
Homo habilis
2.5 million years ago. We modern people are not natural stone flakers, and experiments trying to find out how it was originally done have shown that making stone tools is a fine art and not easily learned by trial and error. Almost certainly the skill of making stone tools spread through early peoples by imitation. This certainty is much increased in the later archaeological record which shows styles of tools, pots, jewellery and other cultural artefacts spreading through different cultures at different times.
Imitation could have begun earlier. Perhaps people imitated ways of making baskets, wooden scrapers or knives, baby slings, or other useful artefacts that would not have survived the way stone tools do. So let us imagine a very early culture of
H. habilis,
using simple stone tools to cut up and skin game or shape wood, and inventing and copying a few other simple artefacts.
As the new skills begin to spread it becomes more and more important to be able to acquire them. And how do you acquire them? – by imitation of course. Thus, being a good imitator becomes increasingly important. Not only that, but it becomes important to imitate the right people and the right things. In such decisions we would expect simple heuristics, or rules of thumb, to be used. ‘Imitate the most successful people’ might be one; but now there are memes this does not just mean imitating those with the most food or the strongest muscles, it means imitating the people who have the most impressive tools, the brightest clothes, or the
newest skills. What this amounts to is ‘imitate the best imitator’. As a consequence, whatever is deemed best spreads fastest.
Another important decision is whom to mate with, and again the answer should be the best imitators, because they will provide you with children who are more likely to be good imitators. All this pressure for better imitation creates more people who are good at spreading memes – whether the memes are ways of making tools, rituals, clothes or whatever. As imitation improves, more new skills are invented and spread, and these in turn create more pressure to be able to copy them. And so it goes on. In a few million years, not only have the memes changed out of all recognition but the genes have been forced into creating brains capable of spreading them – big brains.
That is the story in a nutshell, but I now want to unpack it and take it one step at a time, looking more closely at the mechanisms involved.
The first step we might call ‘selection for imitation’. Let us assume, echoing Darwin’s original argument, that there is some genetic variation in people’s ability to imitate. Some people quickly pick up the new technology of stone flaking, while others do not. Who will do better? The better imitators of course. If stone tools help with food processing, then better toolmakers will eat better and their children will eat better. So far, the same argument could be applied equally well to having strong hands for making the tools. But the difference is this – imitation is a general skill. The good imitators would also be good at copying ways of making wooden scrapers or baskets, or plaiting their hair, or carrying piles of leaves or fruits, or making warm clothes, or any skill that helps survival and can be stolen from someone else. Genes for being a good imitator will begin to spread in the gene pool. Now the environment in which the genes are selected begins to change. If you are absolutely hopeless at imitation, you, and therefore your offspring, will be at a disadvantage in a way you never would have been a few thousand years earlier. The new selection pressure begins with this step.
The next step we might call ‘selection for imitating the imitators’. Whom does it pay to imitate? The good imitators of course. Imagine a woman who is especially skilled at copying the latest ways of picking inaccessible fruits or carrying them back to the family, or a man especially good at copying the best toolmaker. If you are an inferior imitator it will still pay you to copy the best imitators. They will have acquired the most useful skills and you now need those skills. During the last millennium you did not. When no one had clothes there was no competitive edge to having them, but now they have been invented you will be less protected from cold and injury and less likely to survive than people who do have
them. Now that carrying–baskets have been invented you will get less of the best fruit if you cannot make one. Genes for imitating the best imitators will increase in the gene pool.
Note that this is an escalating process. A male robin can only get a bigger territory in predetermined ways, for example, by singing well – and there is a limit to how brilliant any robin song can be. But a male
Homo erectus
might get power and influence, and come to be copied, by wearing more impressive clothes, lighting bigger and better fires to cook the meat – or to scare the people who have not yet mastered fire – having the sharpest tools, and so on. There is no theoretical limit to this process or to the direction it may take. Selection pressures on the genes will be influenced by whichever memes happen to proliferate. Memes evolve as memes build on memes; new tools emerge; new clothes are made; new ways of doing things are invented. As these memes spread the most successful people are those who can acquire the currently most important memes. Genes for being able to copy the best memes, and genes for copying the people who have the best memes, will be more successful than other genes.
But which are the best memes? ‘Best’ means, initially at least, ‘best for the genes’. People who copy survival–related memes will fare better than people who copy irrelevant memes. But it cannot always be obvious which these memes are. The genes set us up with preferences that reflect their interests. So we like cool drinks and sweet foods, and enjoy sex, for example. These things feel ‘best’ to us because they were best for the genes of our ancestors. But memes can change faster than human genes, so the genes will not be able to track them effectively. The best the system can do is probably to evolve heuristics such as ‘copy the most obvious memes’ or ‘copy the most popular memes’ or ‘copy memes to do with food, sex and winning battles’. We will look at the effect of such heuristics in modern society later. In ancient hominid society such heuristics would initially have helped individuals survive and spread their genes, but then increasingly would allow the memes to outwit the genes. Any meme that
looked
popular, sexy, or very obvious would spread in the meme pool and thus change the selection pressures on the genes.
The third step could be called ‘selection for mating with the imitators’. In our imaginary society it would pay to mate with the same people you want to copy. If you mate with the best imitators, then your offspring are more likely to be good imitators and so to acquire all the things that have become important in this newly emerging culture. It is this conjunction that drives the process on – first it pays to copy the best imitators because they will have the most useful skills, next it pays to mate with them so that
your children can also get these skills. But the heuristics for choosing what to imitate are only rough–and-ready guidelines and the memes are beginning to proliferate beyond purely survival–related skills. For example, as memes for singing appeared, the best imitators would begin to sing best, singing would be perceived as important, and so copying singing would come to have survival value. In this way, the specific nature of the memes of the time would determine which genes were more successful. The memes began to force the hand of the genes.
There is a fourth and final step that might increase the process again – though it is not necessary to the explanation. We can call this ‘sexual selection for imitation’. Sexual selection, first described by Darwin and much argued about ever since, is a well–recognised, if controversial, process in biology (see Cronin 1991 for a review and Fisher 1930). Particularly interesting cases involve runaway sexual selection, in which elaborate but otherwise useless structures, such as the peacock’s fantastic tail, are selected for by generations of peahens choosing males with fancier tails. Once the process has begun it can incur enormous costs for the male but it works for the following reason. A female who chooses a male with a good tail will have sons with good tails who will attract mates with choices like hers. She will therefore have more grandchildren. The reason it is the females who do the selecting is the imbalance between the sexes in parental investment. Male birds can potentially have vast numbers of offspring but females are constrained to producing only a few eggs a year or, in the case of humans, a few children in a lifetime. So females cannot greatly increase the number of children they have. They can, however, increase the number of their descendants in future generations by choosing mates who will give them ‘sexy sons’ who will have many offspring. With lots of females all going for the same males, this process rapidly escalates until the costs become too great.