Yes, I got it wrong – and then an innocent’ man was jailed for life.
Fourteen years ago David Canter helped to convict a man of murdering his pregnant wife. The psychologist explains why he now rejects crucial evidence of a fake suicide note.
In 1992 I was asked to assist a murder inquiry by commenting on the authorship of a suicide note in the handwriting of Paula Gilfoyle. I had provided guidance to a number of other big investigations.
The police officer leading the inquiry presented the suicide note to me with the fascinating thesis that it had been dictated to Paula by her husband, Eddie, who was believed then to have encouraged Paula to put her neck in a noose while he stood behind her, after which he killed her by lifting her legs to hang her.
I was told that it was believed that Eddie had murdered Paula, who was eight and a half months pregnant, after tricking her into writing the suicide note on the pretext that he needed it for a paramedical evening class he was taking. His motive, in the crime novel tradition, was hinted at being an affair that he was having.
I have always been fascinated by the psychological implications of the written word, having even studied my own correspondence with my childhood friend, the film director Mike Leigh, when I was a student in the 1960s. In the context of crime the written word can often be regarded as the crime scene, especially when I have given guidance on threat and extortion, for which the text provides the criminal act.
So I worked with some postgraduate research assistants to consider how we might analyse Paula’s suicide note to determine if it did indicate her intention to kill herself. It is important to emphasise that from the start that I did not think we could establish authorship categorically but needed to consider the note in its whole context.
Part of my report was the result of comparing the suicide note with other written material we had of Eddie’s and of Paula’s. The approach that we took to analysing all this material was highly numerical; the proportions of personal pronouns, connectives, sentence lengths and so on.
But we had no firm basis in previous research to work from. No one has been able to provide a truly reliable, objective, numerically based procedure for determining authorship with small examples of text. There is a tremendous difficulty in doing this because we all change how we speak and write depending on the circumstances of our communication. In fact, a moment’s thought about the nature of communication reveals that we must change it in accordance with the situation.
The written material we were working with was a great mixture of jottings left in notebooks, letters to friends, postcards and a series of notes that Paula wrote to Eddie about their relationship. So comparison of the text across them was very problematic. However, I came to the conclusion that the suicide note was not typical of Paula’s style, nor was it typical of Eddie’s and taking all the circumstances and note into account formed the opinion that Paula “had not intended to take her own life”.
My report was never presented to the court but apparently had an influence behind the scenes. But I had always been curious about how a pregnant woman would write a suicide note under dictation from her husband with whom she had had a strained relationship, and then put her head in a noose with him standing behind her. So when the opportunity arose a few years later for me to talk to Eddie and his family (an option I had been denied as a prosecution expert) I jumped at it. The picture that emerged from these discussions was much less clear than the original story. There was no strong evidence that Eddie had ever dictated a suicide note, or even claimed that he was doing a course as a paramedic, just hearsay from friends of Paula, which was never presented to the court.
By the time I met Eddie I had been able to oversee a variety of other studies on text and authorship and had formed the view that the purely numerical approach could be very misleading. I had developed an interest and understanding of the need to reveal the implicit and explicit narrative in any material, possibly closer to an English Literature approach than a purely statistical one.
This set me to consider the unfolding story within the correspondence between Eddie and Paula that culminated in the suicide note. To my surprise I realised that no one had ever done that. The documentary evidence of the relationship between these two people leading up to Paula’s death had never been reviewed in sequence, either in the original court case or at Eddie’s first appeal. All the focus was on circumstantial evidence from apparent witnesses.
Eddie and Paula had been on different shifts so left a dozen or more crucial notes for each other. When I read these through in sequence a very clear narrative emerged. This showed Paula’s disquiet about her pregnancy and relationship to Eddie for which the suicide note was a natural ending.
Furthermore, the related research, which has recently been supported even more strongly in a major doctorate that I have supervised, emphasised that the point of suicide notes is often to excuse people who take their own lives. The notes typically take the form of an apology to exonerate close relatives. Fewer than half of those who take their own life leave suicide notes; some studies put the figure as low as 15 per cent. Notes are usually left because they know how hurtful and unacceptable the death will be to others. This is exactly what Paula’s suicide note is about. With a bitter irony, it is mainly devoted to telling Eddie not to blame himself.
In addition, we have also found that in about a tenth of suicides there are no obvious prior indicators. This means that the judges’ claims in both the original trial and two appeals, that Paula’s demeanour in contact with others, was clear evidence that she did not intend to kill herself, completely ignored what many now recognise, even with the recent suicides around Bridgend. People can present a happy face to the world while still intending to kill themselves.
All these matters and other factors, most notably the demanding creative imagination, beyond Eddie’s abilities, that would be needed to invent the sequence of notes and letters changed my opinion. I formed the view that my original analysis had been too greatly, if inadvertently, influenced by the story that the police had originally given to me. The reliance on number crunching was also misleading. I therefore wrote a further report for Eddie’s second appeal which argued that there was a psychological logic that made it very likely that Paula had taken her own life.
David Canter is Professor of Psychology at the University of Liverpool
Source: The Times 25th February 2008