I should teach the world that the creation of a scientific theory depends upon imagination, and upon empirical tests of some sort. Science is both a creative activity, and a critical or sceptical activity.

Tests can never prove that a theory is correct, but they can sometimes show that it is wrong. Hence, any scientific theory is provisional, because it is always possible that a new theory comes along that makes better sense of the phenomena. The reasons for emphasising these points are the widespread views of science, on the one hand as the mere accumulation of facts or data, and on the other hand as proving that certain theories are true.

Variation in individuals, together with natural selection - the neo-Darwinian synthesis - suffice, in principle, to account for the evolution of species, and hence for the evolution of human beings. This is important, as a great idea. But it also implies that all human beings are descended from ancestors in common, and so its moral and political implications are tremendous. We need to remember our common humanity.

Philip Johnson-Laird is author of The Computer and the Mind: An Introduction to Cognitive Science (buy this book from Amazon (UK) or Amazon (USA)), and Mental Models: Towards a Cognitive Science of Language, Inference and Consciousness (buy this book from Amazon (UK) or Amazon (USA)). See his website.