Simon Baron-Cohen, the eminent researcher into autism, has set himself a mission. He wishes to convince us to jettison talk of "evil", and focus instead on the concept of empathy deficiency. For him, this represents a shift from a position that is woolly and permeated by theological assumptions towards something much more objective and scientific. In a book that is partly a popular science treatise and partly a self-help manual (urging us to immerse our interpersonal problems with our spouses or colleagues in the "universal solvent" of empathy), he interweaves life stories and clinical evidence in an engaging and informative manner.

His chief goal is to work out how people can act with extreme cruelty towards others. In the first sentence of his book, he recalls a conversation with his father when he was seven years of age. It was a conversation that every parent - and especially Jewish ones - must dread: Nazis had turned Jews into lampshades and bars of soap. How could this be possible? How could seemingly respectable people treat other human beings as mere objects? Indeed, as every child rapidly learns, human history is packed with examples of cruelty, often carried out simply for the pleasure of seeing others suffer.

From this early exposure to the fact of human viciousness, Baron-Cohen went on to study people who have "zero degrees of empathy", including psychopaths, those with borderline personality disorders and extreme narcissists. He admits that not everyone who possesses a low degree of empathy acts in a cruel way (observing, for instance, that some people with autism are exceptionally talented and respond by withdrawing from society rather than attacking it), but their lives are certainly blighted by their inability to empathise with others.

Most of us would agree with Baron-Cohen that labelling cruel people "evil" does not get us very far. It is, he correctly observes, actually no argument at all. But does the concept of "empathy erosion" have a better explanatory power?

³ÉÈËVRÊÓƵ

Baron-Cohen argues that empathy has two main components: an ability to identify what another person might be thinking or feeling and a willingness to respond to those thoughts and feelings in appropriate ways. He is most convincing when he addresses the first of these requirements. Along with his colleagues, he has devised an Empathy Quotient; by analysing people's responses to a series of questions, it enables him to rank women higher on the empathy scale than men, and students in the humanities higher than those in the sciences.

It comes as no surprise that Baron-Cohen, a specialist in developmental psychopathology, offers an intriguing discussion of what he calls the "empathy circuit" in the brain. Through functional magnetic resonance imaging (fMRI), he concludes that there are at least 10 interconnected brain regions involved in empathy, with neural activity significantly correlated to individual levels of compassion. In this way, he helpfully explains, the idea of empathy varies "like a dimmer control" on a light.

³ÉÈËVRÊÓƵ

Crucially, however, he is no biological determinist, but repeatedly insists that environmental factors and genes are constantly interacting. He is particularly impressed with the work of child psychiatrist John Bowlby, whose attachment theory explored the effect of parental care (or lack of it) on the child's subsequent emotional life. Parental affection is likened to an "internal pot of gold", enabling the child to develop into a responsible, resilient and empathetic adult. Rejection by carers has the opposite, highly destructive effect.

When Baron-Cohen turns to the second part of his definition of empathy, he is less convincing. His chief problem is that the concept of empathy is as normative as the doctrine of evil. It requires not only an acknowledgement of other people's thoughts and feelings, but also an "appropriate response". But how do we decide what is the correct response? Is it to be defined by the other person? By the empathiser? By a third party? Baron-Cohen cannot tell us.

What may seem like an appropriate response to help alleviate the sufferings of an oppressed individual or group could be regarded as empathetic or patronising; it may be praised as thoughtful or condemned as unfairly partial and spiteful. Who is to say? Like the jurist who argued that he could not define obscenity but knows it when he sees it, so too we are led to assume that - just by observing - we know when people are empathetic and when they are indifferent or unkind. Sometimes that may be true (people turned into lampshades); but it is not always so clear cut (freedom fighters). As a result, replacing "evil" with the label "non-empathetic" may be pleasingly secular but is less helpful than it seems.

Baron-Cohen concludes by saying that "unlike religion, empathy cannot, by definition, oppress anyone". The trick lies in that phrase "by definition". From whose standpoint is oppression to be judged? Who decides what is the correct response to another person's thoughts and feelings? He is grappling with one of the most important questions for our times, and although his answers are partial, they are sophisticated. The debate will certainly continue.

³ÉÈËVRÊÓƵ