One morning in January 1993, a terrible fire broke out in a house in the French region of Gex, near the Swiss border. The house was occupied by Dr Jean-Claude Romand, a medical researcher who worked in Geneva at the World Health Organisation; he lived there with his wife, Florence, and their two young children. The fire brigade was called, but it was too late for Mme Romand and the children. Dr Romand survived the fire in a coma; it was two days before he regained consciousness. When he did so, it was to be told that the police had discovered that his wife and children were already dead when the fire broke out. Furthermore, on the same night, someone had broken into the house of his parents some miles away and shot both of them dead.
At first, the police assumed that the Romand family had been the victims of a feud. Then routine inquiries established that Dr Romand did not work at the WHO, nor was he a doctor. For 18 years, he had perpetrated "a monstrous deception". Starting as a reclusive medical student at Lyon University, where he failed to take his second-year exams, he managed to persuade everyone he knew that he was in practice, with a university post at Dijon and then a prestigious job in Geneva. In order to fund this deception, and raise a family in the provincial, bourgeois manner, Romand showed considerable creative ability. The Gex region has a large population of French professionals who work across the frontier in Switzerland and pay no French tax. Romand persuaded his wife's family and several of his friends to give him all their savings so that he could "invest the money at 18 per cent in a Geneva investment bank". Without exception, they fell for it.
So every day, Dr Romand kissed his wife and family and "drove off to work" in his Mercedes. He crossed the frontier with all the genuine commuters and passed the time in Geneva in the public library at the WHO, or sitting in a supermarket car park, or walking in the woods. Then he came home after a hard day at the office "inventing a new cure for cancer" or working on arteriosclerosis, his "main field". He was a friend of national figures such as the then prime minister, Laurent Fabius, and the political media hero Dr Bernard Kouchner. His children, who were very proud of him, went to the local private school; he and his family took holidays abroad; his wife enrolled in dance classes and worked as a charity volunteer. Dr Romand was a pillar of the Gex community, which supported his Catholic parish, and often took a principled stand in local politics. He even became a supporter of animal rights. All went well until the inevitable day when one of his trusting investors wanted her money back urgently and he couldn't deliver. So, rather than confess the truth, Romand destroyed the entire lie, which meant destroying his wife, his children, his parents and himself. By chance, he survived the fire, to be charged with five murders and receive a life sentence with a minimum term of 22 years. The Adversary, published to great sensation in France last year, is an account by the novelist Emmanuel Carrere of his own fascination with the Romand case.
One can see why a novelist might be drawn to this story. There are the technical details of deceit: how do you deal with telephone calls at work (answering service); why did no one ask for his or her money back (he left his victims with plenty more and they did not want to earn lower interest rates); how did he deal with medical questions (he knew a great deal about medicine). Then there is the psychological angle: how does it feel to tell a huge, sustained lie? Is it just like telling a little one (the cheque's in the post, this review's already been e-mailed), but worse? Or does the strain involve out-of-the-ordinary powers? Is it possible to live like this and remain entirely sane?
Carrere - who, unfortunately, has not been well served by this translation, which fails to capture his dry, throwaway style - considers some of these questions and has made a thorough inquiry into the daily problems and the slow growth of the deception, moving from the point where it would have been possible to make a fresh start to the point where multiple murder seemed to be the "logical" solution. He is good at describing the "vast beach of dead and empty time" where Romand passed the 18 years of his lie. He also provides one key to Romand's behaviour: that he became incapable of distinguishing between fact and fantasy and was deceiving himself even when he believed he was telling the truth. But the author's explanation breaks down on the sordid facts of the case. Romand's children adored him. He loved them. Then he shot them, one by one, in cold blood, having tricked them into looking the other way. He beat his trusting wife to death with a rolling pin, after she finally challenged one of his lies. He told his mistress to close her eyes while he gave her a necklace, and then attempted to strangle her with a plastic cord. Every evening, he would call his parents to wish them goodnight; but when he shot his mother, she was facing him, and so she probably died knowing what sort of man her son really was. He carefully cleaned the murder weapons after he had committed his crimes.
When confronted with the truth, Romand wriggled and lied for as long as he could, like any petty conman. In court, he appeared to be far more preoccupied with his own distress than with the suffering he had caused others. And the prosecution produced evidence suggesting that his suicide attempt - for which he used out-of-date Nembutal and waited until the dustbin lorry was on its early morning round before setting fire to the house - was as false as everything else.
There is a tradition in French criminal justice that before condemning a man, the court must understand him. This is an admirably humane practice, but it only postpones judgement; it does not replace it. Hence the life sentence. On the other hand, the French intellectual fascination with criminal behaviour, with crime as a political or social statement, which stands in a solid, national tradition that stretches back to Genet and Camus, none the less postpones judgement for ever. And it is frequently based on a false premise: that the greater the crime, the greater the interest. The trouble with the Romand case is that so long as he was a brilliantly successful liar he was fascinating, but once he was trapped into the brutal resolution of his problems he became merely repulsive. The most celebrated recent example of a French intellectual ending up looking the fool over a cause celebre was that of Marguerite Duras. She intervened in the Villemin case in 1985, where a mother was accused of drowning her infant son, announcing that her concern was entirely for the mother, and that for "humiliated", working-class women isolated in villages in the Vosges to drown their infants was certainly "sublime". Some years later, the mother in question established her innocence.
Carrere has not committed a Duras. He is too aware of the ambivalent role he himself is playing in Romand's inability (or refusal) to take responsibility for his acts. But he still proves incapable of exercising judgement. And he still flirts with the idea that "crimes so outrageous . . . confer tragic stature". He cannot understand the popular reaction to such crimes, or the view that, in failing to kill himself, Romand destroyed any sympathy that his final predicament might have aroused.
Today, "the doctor" is a celebrity prisoner who still invents interesting lies and has started to take refuge in a superficial religious conviction. And this is perhaps the most repulsive thing about him. A hardened woman reporter at the trial, both progressive and anti-clerical, said to Carrere that if "the little creep" thought he could expiate his crimes simply by saying the rosary for the next 20 years, she would be in favour of reinstating the death penalty.
Patrick Marnham's most recent book is The Death of Jean Moulin (John Murray, £20)