the practices which define of their respective roles. Practices, no less than bundles of information and instructions passed from mind to mind, fulfil the two necessary conditions for them to act as replicators. So it can accordingly be said – to go back to the threefold distinction as I put it at the very beginning of this chapter – that as organisms we are machines for replicating the genes in our bodies, as organisms with minds we are machines for replicating the traits in our cultures, and as organisms with minds occupying and performing roles we are machines for replicating the practices which define those roles and the groups, communities, institutions and societies constituted by them.
Since evolution, whether natural, cultural or social, is not proceeding towards any predetermined final state but only away from what may, for the moment, be a more or less stable equilibrium, it will never be any more possible for sociologists to predict the future of institutions and societies than for anthropologists to predict the future of cultures or biologists to predict the future of species. In the words of the American demographer Joel E. Cohen’s only half-joking Law of Prediction, ‘The more confidence someone places in an unconditional prediction of what will happen in human affairs, the less confidence we should place in that prediction.’5 The problem is not just the incalculability of the consequences of the interaction of an enormous multiplicity of separate events. It’s also that, as the philosopher Karl Popper has argued to particular effect, to predict the future state of human societies would involve, among other things, predicting the future of sociological knowledge itself, and there is no way in which we can claim already to know what we have yet to discover. Critics of sociology sometimes argue that because sociologists can’t predict how future societies will evolve it isn’t really a science at all. But then they will have to say the same about biology and its inability to predict the future evolution of species. If what distinguishes science from non-science is that its conclusions are prescriptive for all observers in accordance with the strength of evidence which they can all go and check for themselves, there is no argument whatever for dismissing explanations which can be tested only with hindsight as ‘unscientific’. Sherlock Holmes can’t predict the clues which will enable him to solve the crime; but when he follows up the clues which do indeed solve it, his solution is no less ‘scientific’ than if he had conducted a laboratory experiment whose outcome he had specified in advance.
On the other hand, it would obviously be a mistake to argue that human social behaviour isn’t predictable at all. How else, for a start, do advertisers grow rich? We are successfully predicting each other’s social behaviour every day of the week, and the continuance of the cultures and societies to which we belong depends on our ability to do so. If you and I are introduced to each other on a social occasion, I am at least as sure that if I hold out my hand you will shake it as I am that if I depress the accelerator pedal of my car it will start to go faster. We wouldn’t be the very social animal that we are unless we could rely on each other’s responses to each other’s behaviour for most of the time. When somebody’s social behaviour is totally and consistently unpredictable, we can tell at once that we are confronted with one of Aristotle’s wild beasts or gods. A society in which nobody’s behaviour was predictable wouldn’t be a society at all.
But wait a minute. Suppose that in order to justify what I’ve said in the preceding paragraph I am rash enough to bet you $100 that if I hold out my hand to Joe Soap, whom I’ve never previously met, he will shake it as our respective roles and the conventions of our common culture dictate. Your ploy is obvious. All you have to do is take Joe on one side and offer him $50 (or, if he is the kind of organism with a complex mind who turns out to be a really tough bargainer, $99) to keep his hands to his sides. This isn’t as stupid an example as it looks. It brings out just as clearly as a more serious-looking example would do the implications for the scientific study of human social behaviour of the familiar fact that most of it is a matter of purposes and goals and self-conscious decisions to pursue them. From this, some sociologists have concluded not merely that predictions about human behaviour can be overturned in ways that predictions about inanimate objects can’t, but also that the only way to explain human behaviour is for the observer to reproduce in his or her own mind what is going on in the minds of the people whose purposes and goals are dictating their behaviour. This second conclusion, however, is right in one sense but wrong in another. It’s right in the sense that for me to explain what you’re doing, I do have to know what you are doing. If I think you’re really trying to throttle your little schoolfellow when it’s only a game you’re playing, my research project about the social behaviour of young adolescents in educational institutions isn’t going to get very far, just as if I think you really believe that the spirits of your ancestors can somehow influence what happens in your own life when you’re only performing what you know to be a purely symbolic ritual at their gravesides, I shan’t be a very good sociologist of religion. But it’s wrong in the sense that it mustn’t be supposed that this makes the explanation of behaviour into an exercise of a quite different kind. It doesn’t. The question ‘what made you decide to pursue your chosen objective and act accordingly?’ can be addressed by the same methods, and the answer assessed by the same criteria, as the question ‘what made you respond instinctively to what you heard and saw in the way that you did?’ The fact of our self-awareness of our acquired and imposed behaviour doesn’t affect one way or the other the validity of the explanation of the behaviour of which our behaving selves are aware. What matters is that the researcher who is doing the explaining should know what’s going on – that is, should have identified the intention which makes the action what it is before going on to identify the motive which lies behind it and the environmental conditions which have brought that rather than another motive into play. The !Kung San of the Kalahari Desert are as aware as are the professors and graduate students who study them of the function of meat-sharing in reinforcing their social ties. But the function would be the same even if they weren’t.
Many people find this way of looking at human behaviour counter-intuitive because it seems more natural to look for the reasons which we have for our decisions than for external influences which we are able, if we so choose, to resist or ignore. But the antithesis is a false one. The concepts of social selection and environmental pressure are not in contradiction with the concepts of individual decision and rational choice. On the contrary, there is every reason to suppose that the human mind has been programmed by natural selection to calculate the trade-off between the costs and the benefits of one course of action rather than another. But although our imposed as well as our acquired behaviour is therefore a ‘matter of choice’ – the only thing we all have to do is die, and as Wittgenstein said, death isn’t an event in life – to say so explains neither the cause of the choice (and thereby the behaviour) nor its consequences. The question ‘with what conscious purpose in mind was this mutation in social behaviour introduced?’ is quite compatible with, but leaves still to be answered, the question ‘how did this mutation affect the subsequent evolution of the society in which it occurred?’ Let me give an example from military history. The rulers and generals of seventeenth-century Europe who first introduced infantry drill into the training of their previously undisciplined recruits had a clear idea of what they wished it to achieve and of how it would serve their interests, both personal and patriotic, if it did. What’s more, they had an evident inkling of the biological as well as sociological reasons for why they were right: as the famous Maréchal de Saxe, among others, was aware, men marching in step in close formation to the sound of music respond instinctively in a way which makes them more effective on the field of battle. But although the innovators succeeded in their aim – which is more than most innovators do – it’s not their desire to win wars and battles which explains their success. To explain that, and the consequent changes in how European wars were fought, it has to be shown why they were right – which means showing what competitive advantage was conferred by the adoption of this novel set of practices on the soldiers trained in it, the armies manned by those soldiers, and the states whose armies they were.6
‘Then if sociology can explain why people choose between alternative patterns of social behaviour in the way that they do, does this not amount to a claim that sociology is a predictive science after all?’ No, not truly predictive. A prediction, to deserve the name, has to be more than a guess