The Physiology of Truth: Neuroscience and Human Knowledge

  • Jean-Pierre Changeux
(transl. Malcolm DeBevoise) Belknap Press: 2004. 288pp. $45, £29.95, €41.50 0674012836 | ISBN: 0-674-01283-6

Can neurophysiology cast any light on the human condition? Books that set themselves this ambition, and there are plenty, are invariably disappointing. The problem is not that we lack information at the neuronal level — a great deal is known about cell receptors, neurotransmitters, re-entrant connections and so on. Rather, the difficulty lies in relating this microscopic knowledge to higher human faculties such as thought, emotion and consciousness.

To get round this, popular-science books by the likes of Francis Crick, Joseph LeDoux or Antonio Damasio typically have the following trajectory. We start with a few chapters on the neuronal nitty-gritty. But then the gears surreptitiously change, and we switch to speculation about the mind's higher powers. However, any serious theorizing at this level tends to be ‘boxological’, rather than physiological — we are given flowcharts connecting posited brain modules, but there is no bottom-up, cell-level account of how these modules might work.

Perhaps this is unsurprising, given the kind of evidence that is currently available about the large-scale operations of the mind. In recent years, functional-imaging data have been added to findings from studies of brain lesions. But even these new data are at too gross a scale: it is like trying to figure out how a computer works by noting when different bits get hot and what goes wrong when certain parts are broken. With luck, this might give us some idea of where certain operations are located, but it is not going to tell us about the mechanisms that make them possible.

Jean-Pierre Changeux's credentials as a neurophysiologist are outstanding. He has been director of the Unit for Molecular Biology at the Pasteur Institute in Paris for more than 30 years, where he has played a prominent role in understanding allosteric proteins and their relevance to neurotransmitter reception. Nor is Changeux any stranger to popular science writing — his Neuronal Man (1983) and subsequent book-length dialogues with other prominent French intellectuals have been great successes in his native country and elsewhere. Nevertheless, his new book, The Physiology of Truth, suffers from the typical flaws of the genre. Initial chapters concentrate on neurophysiological signalling and modulation, but by the end the topics are knowledge, culture and the history of science. Interesting points are made at both levels, yet the initial neuronal material seems to shed little light on the later large-scale issues.

Still, the book does have the virtue of suggesting a possible, deeper explanation of why the micro–macro gap may be so hard to bridge. Throughout the book Changeux emphasizes the plasticity of the brain. Significant neuronal variation can be found among even simple organisms such as water fleas, and the brains of monozygotic human twins often exhibit striking differences. Changeux sees this variability as a result of epigenetic selection — the genes provide a general ‘envelope’ for brain development, but the details depend largely on the selective favouring of some spontaneously formed synaptic connections over others during development.

If this neural darwinism is right, then perhaps it is inevitable that any attempt to identify the neurophysiological mechanisms behind higher cognitive faculties will end in failure. Maybe there simply aren't any such mechanisms to be found. This is the view of ‘functionalists’ in cognitive science. They believe in large-scale patterns in human thinking of the kind portrayed in the familiar flowcharts, but they deny that there are any uniform physiological mechanisms to explain those regularities. Not that they assume any kind of spooky magic; rather, they argue that different mechanisms will underpin the regularities in different people.

From the functionalist point of view, asking about ‘the physiological mechanism’ responsible for scientific reasoning — to take a topic from the end of Changeux's book — is like asking for ‘the low-level explanation’ of why all word-processing programs work in roughly the same way. In truth, there isn't any one such explanation. Different programmers use different tricks, subject only to the constraint that their programs end up doing what word-processors have to do. Similarly, neural darwinism may ensure that our brains use different tricks to achieve roughly the same ends, subject only to the constraint that we all end up getting around in the world reasonably well.

Changeux has plenty to say about neural darwinism, and touches on functionalism in passing, but he doesn't quite spell out the connection between them. Still, his book presents a more satisfying picture of the brain than most of its competitors in this crowded market. On standard accounts, it can simply seem frustrating that we never get any bottom-up explanations of higher cognitive functions. If the structure of the brain is laid down by a definite genetic plan, then why can't we find out about the underlying mechanisms? Changeux's book fails to identify any such mechanisms too, but at least he gives us some insight into why the search for them may be doomed to permanent frustration.