'My brain made me do it' – a legitimate defence?

Neuroscience is becoming increasingly relevant to the law. Can brain scans ever prove whether you are guilty or innocent?

NEUROSCIENTISTS seek to understand how the brain underpins our behaviour, thoughts and feelings. Given that the law is also concerned with human behaviour, albeit for quite different reasons, it is hardly surprising that remarkable advances in our understanding of the brain have led many to believe that neuroscience is becoming increasingly relevant to the law.

In the US, a number of universities teach courses on the interface between neuroscience and the law, and the Chicago-based MacArthur Foundation has invested several million dollars to fund research in this area. In the UK, the Royal Society has just published a report on neuroscience and the law.

Some argue that neuroscience has already cast doubt on the idea of free will, and therefore raises questions about the legitimacy of punishing people for actions over which they had no control. In the US there is a steady increase in defence attorneys seeking to introduce neuroscientific evidence. So is "my brain made me do it" a legitimate defence in a criminal trial?

There will surely be cases where such evidence is relevant. Most countries specify an age of criminal responsibility somewhere between 6 and 16; in England and Wales it is 10. Brain imaging studies have shown that the brain continues to develop throughout adolescence, with the prefrontal cortex, implicated in impulse control and decision-making, not reaching maturity until 20 or so. Such studies have also shown that there are huge individual differences. It is hard to believe that all 10-year-olds should be held fully responsible when they break the law.

Few situations are so clear-cut, however. For example, brain imaging has shown that there are often differences between the brains of people categorised as psychopaths or with antisocial personality disorder (who are disproportionately likely to commit violent crimes) and the more law-abiding majority. But it needs to be stressed that these are average differences only - it is not possible to diagnose individual criminal psychopathy on the strength of a brain scan.

In any case, if psychopaths think, feel and behave differently from others, then of course these differences will be reflected in their brains. That is hardly sufficient to establish that the atypical nature of their brain was the cause of their behaviour.

Similarly, men who were abused as children and who also carry a particular version of a gene called monoamine oxidase-A (MAOA) are more likely to behave violently. But not all men with the gene and a history of childhood abuse commit violent crimes; in one study fewer than a third had been convicted of one by the age of 26 (Science, vol 297, p 851). Should a convicted criminal's sentence be reduced on the strength of a genetic screen?

Rather than such evidence serving to reduce a convicted criminal's sentence, one could argue that it might be used to increase it, or at least influence decisions about release or parole. Such decisions involve assessing the risk of reoffending, and such risk assessment is notoriously imprecise.

Unsurprisingly, those who have to make these decisions err on the side of caution. In 2003 the UK introduced indeterminate sentencing, which allowed judges not only to set a minimum prison sentence but also to require convicts to satisfy the authorities that they would not pose any threat if released. The provisions were initially designed to detain a small number of exceptionally dangerous criminals, but by March 2011 there were 6550 people in prison under these terms. Even if indeterminate sentences were abolished, the problem of risk assessment would not go away. It seems at least possible that neuroscientific or genetic evidence might be able to reduce the risk of getting these decisions wrong.

The law is also concerned with establishing whether people are telling the truth. Is the witness who claims to have seen the defendant at the scene of the crime lying? Is the defendant who protests his innocence actually guilty? It is widely recognised that polygraphs are not reliable enough to be used in a court of law. Might fMRI brain scans, which can detect changes in brain activity when people perform a particular task, do a better job? Several experiments have shown differences in activity in certain regions of the brain when people are asked to answer some questions truthfully and others with a lie.

Two companies, Cephos and No Lie MRI, have already been set up in the US to commercialise these discoveries. Neither has yet succeeded in persuading a court to accept their evidence - and there are good reasons to remain sceptical. At best fMRI might sometimes be able to detect a difference between a witness who is telling the truth and another who deliberately lies. But there is good experimental evidence to suggest that brain imaging will not distinguish between a reliable witness and one who is honestly mistaken. By the same token it seems probable that defendants who have repeatedly protested their innocence under persistent questioning might end up believing they are telling the truth even if they are not. There is also evidence that people can be taught to fool the machine.

There's no doubt neuroscience will provide some startling revelations about human behaviour in the coming years, but we must not get ahead of ourselves. At this point, our priority needs to be to make sure that advances that may affect the law are communicated properly to legal professionals so that when it becomes appropriate, neuroscience is used in court to the benefit of all involved.

Nicholas Mackintosh is professor emeritus in the department of experimental psychology at the University of Cambridge and chair of the working group for the Royal Society's report Brain Waves Module 4: Neuroscience and the law

http://www.newscientist.com/