杏吧原创

COMPLEXITY is not just a word for messy stuff. It is a useful label for a set of interesting approaches to understanding it. From climate systems to the question of quite how it is we are conscious of them, we encounter phenomena on a daily basis characterised by large numbers of simple interactions between, say, gas molecules or neurons, that collectively produce behaviour of a different order.

The banking crisis of 2008 showed once more that treating complex systems as if they were determined by simple sums and averages of their parts is a recipe for misunderstanding 鈥� or worse. A working understanding of this among legislators and regulators might even have helped make world economic recession less likely.

John Miller, a professor of economics and social science at Carnegie Mellon University and a fellow of the Santa Fe Institute for the study of complexity, seems ideally placed to explain things, at least to the business community.

A Crude Look at the Whole presents most of the starter-level examples of complexity. Social insects are an illustration crucial to Miller鈥檚 economic focus. He gives an engaging account of how the simple behaviour of individual ants produces something that looks a bit like intelligence in the nest as a whole. (Although the report of a wave of army ants 60 feet wide and 3 feet thick seems to betray a scarcity of editor drones.)

But the fundamental difficulty of a simple account of this field is shown by the lack of interesting and challenging mathematics. Probably wisely, Miller makes no attempt to convey the fun you can have defining the states of a system as points in a 鈥減hase space鈥� with 7 billion dimensions. But that鈥檚 just one of a number of options for modelling the world economy as a collection of individual agents 鈥� an interesting task that seems central to Miller鈥檚 economic concerns.

Nor is there any real discussion of the implications for the philosophy of science. In 1972, the physicist Philip Anderson wrote that 鈥渕ore is different鈥�, leading to much discussion of the idea that the behaviour of complex systems is 鈥渆mergent鈥� from that of their components. That discussion has yet to reach a conclusion about what kind of knowledge, if any, emergence represents. And Miller may have overreached for layperson鈥檚 language when he writes that 鈥渨ater and stones, then, provide an existence proof that cleverness is not restricted to smart things鈥� 鈥� unless he actually does embrace the panpsychic view that all stuff thinks.

In such a field, politics is never far away. There are many undercurrents of conservatism here, for example, when Miller writes that an understanding of the complex behaviour of social networks could help a state that wanted to prevent 鈥渞abble-rousers鈥� gaining support. Such an understanding is of even more use to agitators for whom inventing things that the state does not know how to respond to is much to the point.

Miller鈥檚 conclusion is that 鈥渢he various proofs, observations, and conclusions that form each thread are beginning to fade into a deeper understanding as we look, in perfect silence, at the complexity that abounds鈥�. More research is clearly required to get there in an accessible way or, perhaps, at all.

This article appeared in print under the headline 鈥淲hat will emerge鈥︹��