Critical Mass: how one thing leads to another by Philip Ball

Philip Ball has written excellent and (relatively) short books on subjects as diverse as the nature of water and the history of colour in art. Now he returns, at much greater length, to a favourite theme: the nature of complexity, and the way the complexity of human society is affected by chance and uncertainty. Is there, he asks, a "physics of society"?

Such a big theme demands a big book, but size isn't everything. Although there is a great deal of intriguing information buried here, it isn't always easy to find. The author seems to have been overcome by the desire to share everything he knows, instead of cutting away the excess foliage and leaving the beauty of the blooms.

This is frustrating, because the good stuff really is good. It seems there are situations where people behave like the particles loved by physicists, and the insights derived from computer models can, for example, provide an understanding of crowd behaviour. If applied correctly, it could prevent disasters like Hillsborough in 1989.

Ball also convinces us that there are no economic cycles, but goes on to labour the point. He discusses how standards (like the Qwerty keyboard) become established, and uses the Unix operating system for computers as an example, but curiously without mentioning the success of Linux. Some of the best material concerns the way alliances form between states, and how the Second World War groupings were, if not inevitable, a likely outcome of the state of Europe after 1918.

It would, of course, be more useful if the physics of society could predict things in advance. Here Ball misses a trick. He notes how the "tit-for-tat" policies of the Cold War ratcheted up tension between the superpowers; but fails to point out that its dramatic end was a result of the same tit-for-tat policies in a mutually reinforcing de-escalation. All that was needed was someone to take the first step down. And this was a prediction of the war games beloved by the Pentagon.

I particularly liked Ball's discussion of the relationship between crime, punishment and social deprivation, and his analogies with phase transitions in fluids. This contains material likely to annoy both liberals and conservatives, but will make them think. This clear argument draws on Ball's own work on the way gases and liquids inter-convert, so that he not only knows what is important but what can be left out.

Would that the rest of the book were so cogent; and the turgid text isn't helped by muddy illustrations. This is all the more bizarre because in The Self-Made Tapestry, on a related theme, Ball offered us crisp text and a wealth of clear illustrations. You can get that in OUP paperback for half the price of Critical Mass - which is what I would urge you to do.

The reviewer's book 'Deep Simplicity' is published by Allen Lane