ÁñÁ«ÊÓƵ

John Le Carre's novel The Honorable Schoolboy features a literary agent who wears a beard in order to create the impression that he might write the books he in fact merely peddles. John Brockman needs no such props. He runs a successful literary agency in New York, and this is the tenth book he has written or edited.

The Third Culture rests on a simple thesis: that while people doing complex or original work in literature or other branches of the humanities rarely feel compelled to explain what they are doing simply to the wider world, many scientists - biologists, cosmologists, or experts on computing and artificial intelligence - do feel the call and can explain superbly. And because these are the fields where the real progress is being made in understanding ourselves and the universe, the message they are sending deserves the widest possible audience.

Brockman's title refers, of course, to C. P. Snow and "The Two Cultures", a book warning of the dangerous gap in understanding between people educated in the arts and the sciences. This thesis was always wrong. There are plenty of scientists with a deep and wide understanding of the arts: the problem is that most arts people are too ill-educated or unimaginative to find out about science. But as this book shows, Snow was right about one thing - there are scientists in abundance, Snow's "third culture", who are good at closing the communications gap by conveying the excitement of the fast-moving subjects in which they work.

But what really distinguishes Brockman's book is his approach to their work. He has simply asked 23 leading communicators of exciting new knowledge to write about their own field - and then asked them about each other as well. So as well as reading Lynn Margulis on Gaia, you can read Daniel Dennett on Margulis, Paul Davies on Murray Gell-Mann or Alan Guth on Martin Rees. The only glaring omission is Jim Lovelock, whose absence leaves Margulis defending the Gaia thesis, but who still rates eight index entries because of the frequency with which others cite his work. The result could be used by the less committed as a bluffer's guide to new science. If nothing else, it will save you having to read a lot of books by some uncompromising authors.

However, it also emerges in the course of The Third Culture that many of the best-publicised controversies now running in science are less vital than one might have thought. The argument between Richard Dawkins and Steven Jay Gould about the importance of gene-level natural selection, turns out to be in essence a private quarrel between evolutionists who agree on many of the fundamentals.

Another useful task performed by The Third Culture is to dispel any lingering idea that science is not a human activity like any other, full of the same punch-ups and backbiting you find in orchestras, political parties or companies. The personalised arguments between Margulis and Dawkins, for example, would be at home in an episode of Drop the Dead Donkey. Steve Jones writes off Stephen Jay Gould as "a snail geneticist gone to the bad" (the joke being that this is also Jones's field) and says that he is "constantly in danger of a triumph of form over content". But the best fun in the book is probably in the later stages where it moves away from biology and into cosmology, non-linearity and advanced computing.

If you want to understand inflationary and alternative universes, for example, you could not do better than to read Alan Guth and Martin Rees, although in both cases the commentaries on them are less informative than others in the book. In our lifetimes, the origin and structure of the universe have left the world of speculation and turned into an empirical science, and these authors fully convey the excitement that this step brings with it.

Even more valuable, perhaps, is the material on the work of the non-linear gurus like Doyne Farmer and Stuart Kauffman. If they are right, their work is even more significant than the reassessment and broadening of Darwinism now under way. And even if one divides the claims by ten to allow for the natural enthusiasm of scientists working in the early days of an important discipline, something interesting must be going on. Kauffman's big insight is the natural emergence of order from chaos when simple rules operate on messy starting conditions. The result could be galaxies and solar systems, but with other raw materials could be corporations or businesses.

For example, Kauffman and a colleague have a patent for a system whereby billions or trillions of genes new to biology would be let loose in an organism: he is confident that "we're going to be able to make hundreds of new drugs" as the underlying order-creating machinery takes over. "Within five years, I hope we'll be able to make vaccines to treat almost any disease you want, and do it rapidly." One only hopes that the University of Pennsylvania, where he is professor of biochemistry, has a good patent lawyer.

The word most overused by the complexity theorists is "emergent". Any property of anything that arises from a broader underlying pattern is to be labelled emergent. An even more uncompromising approach to the potential of emergent properties is taken by Doyne Farmer, as befits the leading light behind a project to enrich science by building into a shoe a computer for predicting the outcome of roulette wheels on clandestine casino trips. He has obtained large sums from the Swiss Bank Corporation for a firm, the Prediction Company, which aims to spot underlying patterns behind currency fluctuations and other market movements.

If his organisation, the Santa Fe Institute, suddenly supplants Wellcome, the Ford Foundation and the others in the league tables of big research spenders, you will know that he was right. However, it could just be that like Brian Goodwin, the UK's most prestigious nonlinearity enthusiast, he has caught what Steve Jones terms "something in the air" at Santa Fe and become "a mystic". The Third Culture is likely to sell well, although it could have been improved by making the contributors say much more about each other and a little less about themselves. It deserves a place in this summer's holiday reading for the discerning beach-goer, and may well lead you on to understand many other good things underlying Brockman's emergent whole.

Martin Ince is deputy editor of The THES.