Once upon a time SF crackled with wish-fulfilment: free power, unlimited technological expansion, and loud snapping noises as mere physical laws were foolish enough to impede the Indomitable Human Spirit. Later it turned out that real-world constraints made for better stories – since they offered, out there in the vacuum of space, something for the human spirit to push against. (As in

Tau Zero, where the hated barriers of relativity are used as a staircase going endlessly up.) Then the more pessimistic aspects of current scientific philosophy started to have their effect, in particular that gloomy arrow of entropy which through epochs of dissipation points only to a final chucking-out time. After which, if you'll let me oversimplify wildly, everybody buggered off to read fantasy instead.Paul Davies's book is a synopsis of more recent changes in the mood of scientific thought. No real synthesis has been reached, mostly because all the new insights involve nested levels of fascinating complexity; but these days there's a cheerier arrow of time in the air, an arrow pointing in the direction of higher organization.

The book's predecessors include Ilya Prigogine's

Order out of Chaos, a knotty work with several examples of high-level patterns emerging from unstable chemical systems; James Gleick'sChaos, a pop-science survey of fractals, turbulence, strange attractors and other currently "sexy" notions (OK and well illustrated, though occasionally larded with facile journalese. You know: "For all his wild mop of hair, crazed eyes and addiction to monkey brains, Popplestein seemed just a normal Californian. Little did colleagues know that secretly he was thinking about ...chaos!"); and finally Douglas Hofstadter's efforts inGödel, Escher, BachandMetamagical Themasto model complicated systems and show that the trendy holistic and execrated reductionist viewpoints are complementary.What was the more pessimistic world-view missing out? It was in its way a product of wishful thinking about how complex the universe is. Eugene Wigner wrote a famous paper called "On the Unreasonable Effectiveness of Mathematics" and marvelled that so much of nature could be modelled by simple functions. Unfortunately this was largely a side-effect of the unspoken scientific consensus that physics is deterministic if you ignore those horrid quantum effects (Davies gives an elegant mathematical example which even on paper is guaranteed to "go unpredictable"), that nonlinear equations are a bit beyond the pale (all too much natural behaviour proves to be nonlinear), and that to study nonequilibrium systems is to make life needlessly complex. The word to delete is "needlessly".

In the famous words of sociology, we are all nonequilibrium systems. Likewise the Earth, the Sun, and (from start to entropic finish) the universe. As Prigogine showed, a million gloomily intuitive rules about entropy and decay fail to apply on the local scale once you take that "needlessly complex" view of things. There is much more here than I can cover; despite the dry expository style I was left with a frothy feeling of wanting to go away and write SF stories. Optimistic ones. Try it.

Only in one area does the book fall down badly. Although by now we've all seen too many pictures of the Mandelbrot set, there's plenty more scope for attractive illustrations. The author must feel a sick envy of Gleick's beautifully produced book: it looks as though Davies provided

veryrough sketches as guides for an illustrator, and Unwin gleefully decided to save a few bob by using the roughs, thus enabling them to keep the paperback price down to ... well, actually, £5.95.