Hold that thought, and allow me to return to my second singularity that helps define today as a Promethean moment: the coming into being of a new kind of science. The phrase is Stephen Wolfram’s, but I want to stretch it further than he did, because I think he only got halfway there (Wolfram 2002). He was really describing an old new kind of science. Let me explain. We can make a strong argument that the sort of science we do when we simulate, particularly when we simulate complex systems, is sufficiently different to normative science that we may call it a new kind of science. The argument (and Wolfram is by no means the first to make it) goes roughly that science first went through an observational stage, then an experimental stage and now is going through a simulation stage. These stages parallel developments in the tools and technologies we use, in particular mathematics, and also roughly track the sorts of phenomena that are understandable by different stages. This argument is countered by the logical positivist idea that experimental science is the real dinkum science, preceded in a developmental sense by observational science, and extended, reluctantly and in special cases, to simulation. And the most typical special case is complex systems. But in this view, experimental science remains the core of science, and we complex systems scientists remain beyond the pale with our success judged by how well we can eventually bring our problems into the core.

Our program, by these lights, is to make the complex simple enough to allow it to be handled eventually by normative science. Our simulation is a means to an end. Now Wolfram and others challenge this, arguing that we have been seduced by the success of experimental science at picking the low hanging fruit on the tree of knowledge. The real problems and real knowledge hang much higher up, forever beyond its reach. And if we condemn science to be what the logical positivists call normative, we condemn ourselves to ignorance. To that point I am in full agreement with the Wolfram view, but I put a different interpretation on the history of science, one which is based on thinking of the science enterprise itself as a complex adaptive system, and not necessarily as some sort of progression through stages: this is now normative, later that is. In this view, which I have elaborated elsewhere, I use the same evidence as a Wolfram might, but I do not come to a view that we are changing what we mean by normative science, that is changing in earlier times from observational to experimental and now to simulation. Instead I am rather more interested in just what science is.

My argument is that science is not a process—a method—for finding the truth, and hence does not really have stages (Bradbury 1999). It is rather a body of knowledge about the world and a recipe for growing that body of knowledge and a bunch of people busying themselves with that knowledge and that recipe all mixed up together. But it is important to realise that science is neither just the algorithm nor the output, but the whole lot including the people. It is also important to realise that we can push this analogy too far, that science is a special mix. Each part can change the other. The knowledge can change the recipe; the recipe can change the people.

Thus science is not a process for finding the truth or anything else. It is not a process at all, but a system, a complex adaptive system. It is a complex adaptive system made up of two interacting complex adaptive systems—the scientists and the knowledge system, the former a social system, the latter a playground of memes—with the interaction strongly mediated by the recipe. Now I need to make it clear that this recipe for growing the knowledge is not what is often called the scientific method or what we called earlier experimental or normative science—the approach used by most scientists since the seventeenth century. It is instead a recipe for discovering the scientific method, and all the other methods that achieve the same ends, which are to increase the body of scientific knowledge, to improve the recipe and to change the people.

What sets it apart from other such systems, such as art, religion or politics, is that the idea of belief is deliberately and persistently winnowed out of science; despite the persistent human tendency to keep putting it back in. Each of those other human enterprises has at its base a kernel of some fundamental and fundamentally unprovable, and hence illogical belief, whether it be belief in a god, an aesthetic or a manifesto, on which it then constructs the logic of itself. Even mathematics has this character, and is thus not scientific.