Evolution of the auto pilot

When a jumbo jet finds its way to the right landing strip in thick fog, this is hailed as a marvel of modern engineering. The control mechanisms of airliners are among the most complex and subtle machines ever designed by human beings. And yet every member of the animal world carries within less than a couple of fistfuls of that porridge-like organ called the brain machinery that is several orders of magnitude more competent and more complex than the control machinery of the jumbo jet. Alain Berthoz's The Brain's Sense of Movement has been written to address and explain those structures of the brain that guide the living animal through a complex world and are responsible for survival. This is a welcome and salutary text, written by an eminent French neuroscientist and translated into English without loss of brilliance.

Berthoz argues that the brain has evolved primarily to allow its owner to know its own body unconsciously, and direct it as subtly or as brutally as is necessary through a world that contains food or poison, allies or predators, earth, fire and water. Just observe an object within your reach and touch it first gently, and then sufficiently to make it move. Now do it all again with your eyes closed and with the head held at a different angle. It turns out not to be too much of a problem. While we tend to take such abilities for granted, the fact that they exist at all is a major clue to the raison d'être of the brain.

Despite this, the trend in recent years has been to relate brain structure to thinking, cognitive abilities and consciousness. Refreshingly, Berthoz and his colleagues have achieved their understanding of the brain by working on the way in which sensory modalities are coordinated not only to control the thousands of muscles that cause movement, but also to discover how the inner planning that unleashes or inhibits movement is linked to these senses. In particular, his work has centred on the vestibular sense, that is, our inner sense of balance generated by the semicircular canals situated just behind our ears. The fact that a tennis player can execute a subtle drop shot while sprinting towards the ball is largely due to what the signals from this little organ tell the brain about how the body is moving with respect to the ball, which the eyes hold in a steadfast gaze. And it can all happen in a fraction of a second. To designers of computer control systems for aeroplanes, this seems nothing short of miraculous.

Primarily, this is a book for anyone working in the neurosciences. But given the burgeoning interest in how the brain works, it also offers a palatable alternative for those tiring of the plethora of "mystery-and-wonder" books about the brain that have recently hit the popular science shelves of high-street booksellers. I hope that this book does make it to the high street, as Berthoz's skilful way of treading between neural tourism (X is connected to Y) and function (this is what X and Y get up to) can be appreciated by many.

He chooses a sensible track, starting with the motor theories of perception, fashionable in the 1940s, and advances using his own observations about the vestibular sense. The heart of the matter lies in the sources of coherence in perception (seeing a joined-up world) despite the movement and position of the eyes, head and body. The need for, and the role of, memory in achieving this seamless sense of existence is carefully explained. Happily, Berthoz retains his scientific integrity by resisting the temptation to dumb down his narrative. Some may therefore find the book hard going, but my advice would be to persevere as the result is satisfying.

Berthoz is harsh (to my mind correctly) on the computer metaphor of the brain. He sees it as a new form of dualism that pessimistically creates a gap between mechanism and behaviour. This is certainly true of the past 50 years of artificial intelligence, where creating behaviour (eg playing chess) by algorithmic hook or crook never told us a great deal about how the brain achieves such levels of competence. His preferred metaphor is the Bergsonian notion of an orchestra and its conductor (the neural brain) generating a symphony (conscious life). Of course this is precisely the agenda for those computational neuroscientists who use computers to create virtual neural brains that begin to tell us something about the symphonies of conscious life. While Berthoz does not speak to computational neuroscientists, they would gain by heeding his words.

Igor Aleksander is professor of neural systems engineering, Imperial College, London.