book reviews: the trouble with neuroscience

newscientist | NO CREVICE of the human experience is safe. Our deepest fears and desires, our pasts and our futures – all have been revealed, and all in the form of colourful images that look like lava bubbling under the skull. 

That, at least, is the popular conception of neuroscience – and it's worth big money. The US and the European Union are throwing billions of dollars at two new projects to map the human brain. Yet there is also a growing anxiety that many of neuroscience's findings don't stand up to scrutiny. It's not just sensational headlines reporting a "dark patch" in a psychopath's brain, there are now serious concerns that some of the methods themselves are flawed.

The intrepid outsider needs expert guidance through this rocky terrain – and there's no better place to start than Brainwashed by Sally Satel and Scott O. Lilienfeld. Satel, a practising psychiatrist, and Lilienfeld, a clinical psychologist, are terrific sherpas. They are clear-sighted, considered and forgiving of the novice's ignorance.

Their first stop is the fMRI scan – a staple of much brain research. Worryingly, the statistical techniques used to construct the images sometimes create a mirage of activity where none should exist. They have a telling example: one research team watching a salmon in an fMRI scanner as images of human faces were flashed at it saw its brain spark into life in certain shots – even though it was dead.

Such fishy results are troubling enough, but even legitimate scans can be problematic. As the authors point out, brain images should be used only alongside other kinds of evidence. But all too often they are given the final say on human behaviour. A common pitfall, assert Satel and Lilienfeld, is "neurodeterminism" – the idea that a murderer, say, had been cursed with a brain defect that destroyed their sense of morality.