The insect in all of us

All of us "higher animals" begin life as a single cell, which by successively dividing, eventually gives rise to the complex form we call an adult. Because all of our cells descend from the same single "Eve" cell, every one of the billions and billions of cells in the adult body contains precisely the same set of genetic instructions. The cells in your teeth carry the genes for making hair, or fingernails, eyes, a heart, bones, skin, and even a brain. And yet, teeth tend only to be found in jaws, and brains only in heads. How individual cells know what to become and how they assemble themselves spontaneously into an immensely complicated and functioning set of components called an animal is the riddle of development or "ontogeny". It is a riddle that has exercised thinkers since at least Aristotle and is made all the more amazing when it is realised that the cells accomplish their precise feat without help from any central authority - there is no developmental map reader telling which cells to become what, where to go, and in what proportions.

Surely then development qualifies as simply a miracle. Perhaps, but scientists shy from miracles just as the church often shies from science - in both cases someone can get put out of a job. In place of miracles, scientists attempt "material" explanations for things and explanations of development are among the most vexing a scientist can attempt, simultaneously mixing fundamental simplicities with overwhelming complexity. Rudolf Raff's iconoclastic new book The Shape of Life shows why. Animals as diverse as worms, the insects, and mammals, and representing half a billion years of evolution all share a small number of highly conserved genes - the Hox gene cluster - that determine basic body plans and "north-south" axes of the body: the same starting point yields startlingly different results. Kafka's Gregor Samsa who "awoke one morning to find himself transformed into a gigantic insect" is not just a bad dream.

Cunning scientists can exploit the simplicity of the starting conditions to reveal fundamental routes of ontogeny. Tweak a single Hox gene in the embryo of a developing fruitfly and you can get its legs to grow out of its head. Tweak a different gene and you can get wings to grow in the wrong place, or get eyes to turn into wings. These sorts of experiments are beginning to reveal how it is that genes for, say, eyes, turn themselves off when in wing cells and vice versa. If the estate agent's mantra is "location, location, location", ontogeny's mantra may be "position, position, position". At the risk of gross simplification, genes may sense gradients and concentrations of the products produced by other genes.

The strength of a gradient or the concentration of some gene product instructs some other gene to produce its product or not, and how much of it to make. Lewis Wolpert has called this "the French flag problem": a line of otherwise identical cells could spontaneously form a blue, white, and red pattern by sensing their position in the line. In some unknown and undoubtedly hideously complex way all of these various pieces of positional information could orchestrate the production of a body. Development may seem miraculous but in fact could proceed from what may turn out to be comprehensible first principles.

The study of development has long been dominated by grand themes with equally grand names. The great German developmental biologist Ernst Haeckel coined the phrase "ontogeny recapitulates phylogeny" to describe the apparent trend for animals during their embryonic development to trace out the history of evolution. Thus, even advanced creatures like us vertebrates resemble tadpoles at an early stage; we have all heard stories of babies born with fish's "gills", and late in foetal life humans are said to resemble some of our monkey ancestors. From this sweeping principle (in fact a mere description of some events in embryogenesis and development) have sprung such additional principles as "heterochrony" and "developmental constraint". All three principles serve to put limits on the directions and ways that organisms can evolve, and are a favourite weapon of those biological scientists who prefer to see the limits of adaptation and natural selection as opposed to its creativity.

Thus, if you have to resemble a tadpole on your way to becoming a chicken there are only so many ways of becoming a chicken. In four calm chapters in the middle of Raff's fine book, these orthodoxies are laid bare as either wrong or simplistic, opening up a more optimistic and positivistic view of animal evolution. Despite sharing basic plans for development, animals have come up with a bewildering variety of ways of becoming different and adapted to their environments. Raff is more sardonic: "an accounting of ... evolution that began as ineluctable has ended up as ineffable". This view of evolution as not just a tinkerer but as a force that can produce new and varied forms slams head-on into an anti-Darwinian view of animal development that Raff calls structuralism and which I call Platonism.

Structuralists believe that the many and varied forms of animals (from worms to spiders to butterflies to frogs, birds, and mammals) arise as inevitable consequences of the physical properties of chemicals diffusing and reacting in cells. Because the physical laws are finite, so is the range of possible forms. Evolution, if allowed to be re-run, will tend always to produce rabbits, frogs, eagles, and the other forms we see, not as a consequence of adaptation to particular niches but because these morphologies arise spontaneously from the physical properties of our particular chemistry. Individual rabbits are, in effect, merely the worldly manifestation of ideal - Platonic - forms that exist deep in the physical laws of nature. The structuralist-adaptationist debate will no doubt carry on. The adaptationist Raff gains ironic effect from his understatement that "we can't be sure that we know the limits of morphology". From structuralist first principles, square bacteria should be impossible. But a bacterium called Haloarcula or the "salt box" has been found in the Sinai, and it is "astonishingly square", resembling a miniature postage stamp. Triangular bacteria have also been discovered.

Raff has produced a sober, literate and authoritative account of ontogeny that positions the field squarely to be challenged, refined, and advanced by careful scientific observation unburdened by the dead hand of century-old orthodoxy. Perhaps Raff intended a double entendre in his title: his book not only reveals the fundamental shapes common to most animals, but also reveals a scientific field of life (developmental biology) in better shape than ever.

Mark Pagel is senior research fellow, department of zoology, University of Oxford.