Earthlings in ant country

This is one of the literary oddities of the year. It is an ingenious and creative tour of some fascinating intellectual frontiers that could also function as a super-version of one of those bluffer's guides that save you from having to plough through a lot of hard stuff. But its authors' undoubted talents and enthusiasm are unable to match the vastness of the task they take on.

Ian Stewart is a mathematics professor with a gift for the public communication of science, and Jack Cohen is a biologist who, the dust-jacket assures us, makes his living partly by a consultancy that provides "credible creatures and ecologies" for science-fiction writers. This is a promising combination to explore consciousness, the nature of minds, the evolution of awareness and similar fashionable topics. Part of their approach involves an imaginary dialogue between members of a race that lives in octuple rather than binary groupings, written in a style that suggests that these earthlings have a liking for the fiction of Terry Pratchett - who appears on the back cover praising the book.

Large parts of the book are devoted to a praiseworthy attempt to make us think about the mind and other matters at the correct scale, instead of being sucked into the reductionism that insists we start at the quantum level and work up from there. The latter approach, as the authors see it, leads to attempts such as those of Roger Penrose and Stuart Hameroff to solve the so-called "hard problem" of how humans became conscious with a quantum explanation. As authors such as Patricia Churchland have pointed out, an impressive-sounding problem can have a simple solution: consider the way that natural selection produces new species.

According to Stewart and Cohen, you could describe a T-shirt by reference to parasites in cotton plants, the physics of cotton thread, the mechanical engineering of looms and a host of other considerations; but this level of detail would hide the truth about the garment instead of revealing it. Although they do not use the term, the authors are really arguing against what biologists jokingly call "physics envy", the desire - seen at its most extreme in the Human Genome Project - to emulate the big-budget, high-prestige science of the space and particle physics communities. They argue that the Human Genome Project could be a lot less informative than its supporters make out. After all, we know all about the genetics of HIV - and we also know that changing human behaviour is the key to halting the spread of AIDS.

The authors' case is that the most interesting things happen in a messy place called "ant country". If you think you do not know where that is, look around, because you already live there. It is the highly textured bit of territory, criss-crossed with paths like those of ants, between the things we can chart by arguing from the top down and those we can get at from the bottom up. While in principle it may be possible to predict the course of the universe from its basic laws, the sheer density of tracks in ant country means that in practice it will never be possible.

This concept comes very close to the idea of "contingency" explored at length by Stephen Jay Gould in Wonderful Life. His idea is that things happen all the time in the universe that make real prediction impossible: think of the role that stray asteroids and climate change are now thought to have in the development of Homo sapiens. Gould has told this story better, but Stewart and Cohen diversify it well in an idiosyncratic way.

One of their handy innovations (in chapter one, which also awards a Nobel prize to Maurice Wilkes, rather than the actual winner Maurice Wilkinson) is the "grim sower", alongside the better-known grim reaper. This concept derives from the fact that most animals and plants reproduce by the billion, in such overwhelming numbers that all the offspring cannot possibly survive. The average female frog lays 10,000 eggs, but for numbers to stay constant, 9,998 of them cannot reproduce, leaving two to replace the parents. For cod, the prospects are poorer still - the female produces about 40 million eggs, 39,999,998 of them unwanted if numbers are to remain stable (which suggests to this reviewer that over-fished populations in the North Sea would replace themselves in a few years if left alone).

Stewart and Cohen seem here to have reinvented the K and r classification of species, without saying so. K types, such as people or dolphins, have few offspring and look after them carefully, while the r types, such as cod, produce millions and do not worry too much if most of them die, as inevitably happens.

In the same spirit, Figments of Reality gets excited about the reductionist idea that there is a very close connection between a plant or animal and the genes that underlie it. In fact, the connections are subtle and have to do with the complexities of developmental biology, the environment in which the animal grows up, and other factors. There is often enthusiasm, for example among animal breeders and biotechnologists, for getting the "best" genes in a species. But the "best" gene today can be a duffer tomorrow. The genes that make a frog prosper in wet conditions can be fatal in a dry spell. Sexual reproduction is a useful device here, since it allows the genes that might have been useless in earlier conditions to reappear in case they are needed; the disadvantage is that the same hidden genes can cause disease or malformation when they emerge. Anyone seriously interested in this subject will already have read The Extended Phenotype by Richard Dawkins, published as long ago as 1982; but those coming to it for the first time can get the gist in a few pages from Stewart and Cohen.

However, they are at their best in discussions of brain function, the mind and consciousness, in which fancy theories such as the quantum approach to consciousness are stood up mainly in order to be shot down. Their preference is for "emergence", the view that the appearance of conscious behaviour from the complex interaction of our minds and their environment does not require a special explanation but is instead a feature that can arise much as species arise from natural selection. In other words, the "hard problem" of consciousness may have a rather painless solution.

Anyone who feels like a canter through such subjects and a lot more (free will, the origin of language and the like), imaginatively presented with few errors, will enjoy this book. Although its style is a bit twee for my taste, I learnt more than I expected from the dialogue between the octuple beings, who include a character called Liar to Children (a teacher) and a Liar to Adults (politician/priest). The book's very ambition and quirkiness mean that it will never displace more stolid works on the undergraduate syllabus. However, it should be highly welcome to curious minds, whether they belong to students or the general reader.

Martin Ince is deputy editor, The THES.