Symposium by Lorentz Fellow Richard Gill
On 22 March, five days after his final attendance at the Lucia de B case, Professor of Statistics, Richard Gill, will be appointed Distinguished Lorentz Fellow 2010-2011. To mark the award, Gill is organising a breakfast symposium on the relations between the media, science and law.
‘I am very proud,' says Richard Gill, unlocking his office door, 'and the University can also be proud.' He is evidently pleased with his appointment, which carries a prize of € 10,000. The press release mentions that 'Gill is a strong advocate of the responsible use of statistics in legal cases.' Gill has played a major role in a number of cases, including that of Lucia de B, in which he argued that the statistical analysis on which much of the evidence was based was not reliable.
It is pure coincidence that the symposium is due to take place so soon after the final court session on Lucia de B, but according to Gill the case is a perfect illustration of the intended focus of the symposium. Gill: ‘For several years I've been chewing over the idea of there being three different interests at play in legal cases, namely the media, the scientific community and the law courts. Each group has its own interests and acts accordingly, which is fine. But the main thing is that all three should work in harmony with one another. If two parties work together against the third, there are bound to be problems.'
Lucia de B
Nurse Lucia de B was sentenced to life imprisonment and detention in a mental hospital for seven murders and three attempted murders. It soon became apparent that a great number of errors had been made in the course of the trial. Gill: ‘One doctor stated, for example, that one of Lucia's suspected victims had suffered an unnatural death. Later, it appeared that the doctor had thought this because the patient had been a patient of Lucia de B. In a normal situation he would have regarded it as a 'natural death'. The judge adopted the opinion expressed by the doctor and used it as 'medical evidence'. This poses an enormous problem. Gill: 'What is the truth and who determines this? There is an extremely fine dividing line between the thinking of a professional and that of a lay person. The symposium will also address this issue.'
It was philosopher of science Ton Derksen who demonstrated in his book on the Lucia affair that the staff of the judiciary used statistics, although they claimed not to. Gill: 'They did use statistics, only they didn't call it by that name. They saw it as evidence. However, the evidence was derived from statistical data that was wrongly interpreted by a bunch of amateurs. Major fundamental errors were made by people who had not been trained as statisticians. And they were taken seriously by the judiciary, and also by the media. Lucia's fate was sealed on the basis of incorrect observations.'
System under pressure
Initially, Gill was not at all interested in the Lucia de B case. It was only when his wife mentioned that statistics were being used as evidence, that he started to look at the case in more depth. 'I read the book written by Ton Derksen about it. The book made it clear that my field, statistics, was being abused by the judiciary. I was livid. It's not my concern whether a person is guilty or innocent or is wrongly imprisoned; for me, what's important is that it doesn't happen based on lies about my field.' Gill decided to take action. He wrote an article about the case in Natureand started a petition, and the issue became news. This was the impetus for the media to take up the case again. Gill: 'The case was reopened because we put pressure on the system. Even if the judiciary will never admit to it.'
The Lucia de B case shows clearly how important each of the three groups involved is. Gill: ‘You see that it goes wrong if the legal people and the experts get together and try to work out how it all fits together. It is basically the media that has made sure that Lucie will now have a fair trial, and that more attention is being paid to forensic science.' It is an interesting phenomenon that Gill, in his year as Lorentz Fellow, would like to research further. 'I want to work on forensic and technical statistics; there's a lot of room for improvement there. The communication between science and arts people also needs to be improved. And legal experts also have to understand the information provided on the basis of statistics. The symposium is the first step on the road.'