Forget finger prints and DNA, Sheena Hastings meets the forensic scientist for whom a regional accent or an odd turn of phrase can prove key evidence.
ANYONE with Peter French’s special talents is of course going to suffer requests to perform a party trick. You can imagine him being backed into a corner by tiresome folk saying: “Go on, then, where do you think I’m from...?”
It’s difficult not to give in to temptation – and I don’t. After listening to very little speech from me – we’re there, after all, to talk about him – the affable Prof bangs the nail deftly on the head.
He reckons I come from Greater Manchester... middle-class... educated at a girls’ grammar school (“a repository for nice girls”, as he quaintly puts it)...but with speech that now has strong enough Yorkshire tinges to suggest I’ve been here rather a long time. Correct on all counts.
His day job may not be as glamorous as television’s CSI, where crime scene investigators deal in certainties and all cases seem to be cleared up in a trice, but Prof French’s expertise and eminence in the field of speech analysis is an important component of many court cases.
In 20-odd years of specialist consultancy in forensic speech science he has been called to all points of the compass to act as an expert witness, and his evidence has helped to convict drugs barons, terrorists, fraudsters and murderers as well as some of the hoaxers who waste emergency services’ time by claiming to have committed or know about series crimes.
Well known cases where his finely-tuned ear, scientific analysis and mightily experienced and respected professional judgement have been used either by the prosecution or defence – and sometimes as an independent source of evidence consulted by judges – include the International War Crimes Tribunal trial of former Yugoslav President Slobodan Milosevic and the murder of Liverpool youngster Rhys Jones, who was caught in the crossfire between rival gangs.
He was also been an expert witness in the murder trial of David Bieber, killer of West Yorkshire police officer Ian Broadhurst, as well as the belated conviction of Yorkshire Ripper hoaxer John Humble, nicknamed “Wearside Jack” by police, and Who Wants To Be A Millionaire? fraudster Major Charles Ingram.
He is one of a handful of such expert forensic speech analysts in the UK, and his York-based team analyse speech recordings and prepare reports for 500-600 legal cases a year.
Ten years ago the majority were drugs related; today much more time is spent on analysis of voices recorded and used as evidence in terrorism cases. He’s just back from a jihad trial in Limerick.
“Who’d have thought it?,” he says. “But now there are so many al-Qaida related cases that people almost don’t remember them from one week to the next.”
Thanks to television detective shows that glamorise the work of forensic investigators, hundreds of hopeful young people write to Peter French saying “I want to be like you.”
They flock to university courses with no hope of a job at the end. It’s a small and rarefied field, and one that fascinates partly because most of us have very little understanding of how very personal and particular our voices are – and although no expert analysis by a speech scientist or other witness can be 100 per cent certain, language analysis has been a major component at many a trial.
French grew up in Consett, County Durham (you can still hear it in his voice), and after trying a few manual jobs, studied English and education at Leeds University with a view to teaching. The teaching never happened. Instead he went into a few years of postgraduate studies which embraced researching a dialect map of Europe and child language development, tracking youngsters from 15 months to seven-years-old.
A lecturing job at the College of Ripon and York St John (now York St John University) brought French into contact with the late, great dialectologist Stanley Ellis, who taught at Leeds University and was one of this country’s leading experts in forensic speech analysis. He advised West Yorkshire Police during the hunt for the Yorkshire Ripper, the killer of 13 women in northern towns and cities between 1975 and 1980.
A turning point in French’s career came one evening when he and Ellis were sitting eating fish and chips on the front at Scarborough. They were there to talk at the conference of the Yorkshire Dialect Society.
“Stanley was involved at that time in what we now call forensic speech science, mainly comparing voices in criminal recordings with those of known suspects. There wasn’t much of that kind of work around at the time, as people didn’t record much back then.
“There was a feeling in academic circles that it wasn’t the thing to be doing - it was getting your hands dirty. There was also a view that it wasn’t politically a very correct thing to be acting for the police against criminals. The ethos then was very different.
“Anyway, Stanley said, as though we all didn’t already know, ‘I’m involved in this forensic work...would you like to join me?’ I said ‘I’ve been waiting and thought you’d never ask!’ I did my first case as (his) understudy and thought ‘This is absolutely fantastic’.”
French learned from the master then came out from under his wing. As forensic speech science techniques developed, helped by advances in acoustic methods which paint visual pictures of voice patterns and can isolate individual elements as well as screen out ambient noise, he raised the profile and status of the profession by founding an international professional association.
A watershed moment came in 1987, with the passing of the Police and Criminal Evidence Act. Among other things it decreed that every interview of a suspect had to be recorded, to counteract possible claims that written notes had been doctored.
“This meant that everyone gave a voice sample by virtue of being interviewed. It was as though there had been this enormous thunderstorm and it was raining cassette tapes overnight.”
At that point French gave up university work to concentrate on forensics. In the 80s, he says voice analysis was done by ear, listening to how vowels and consonants were formed, the intonation (melody) of speech, voice quality etc. What he brought to the mix by the end of that decade was the acoustic component – using sound spectrograms or pictures of speech, and using them to match pattens while also extracting mathematical measurements of voice signal.
More recent times have brought ASR – automatic speech recognition, and electronic network frequency analysis, which can authenticate and date a voice recording by analysing ambient electrical “hum”.
But still human judgement will always be crucial, says Prof French, and a joint auditory and acoustic analysis of a voice sample will take around 15 hours. What this great labour aims to do is to find what is distinctive or unusual about a suspect’s voice and compare those characteristics to the criminal voice recording.
The expert witness only ever deals in a percentage likelihood of two recordings matching each other. Despite CSI, forensics are not about certainty.
In the Milosevic case he was asked to report on the likelihood of voice recordings having been tampered with. In the Rhys Jones case his expertise was used to analyse recordings taken by “bugs” in suspects’ homes, on which they discussed false alibis.
In the case of the murder of Charlene Ellis and Letisha Shakespeare, innocent victims of drive-by shootings by gangs in Birmingham 10 years ago, he was asked to give an opinion on “ear witness” evidence of one potentially incriminating phrase heard by a witness who had not seen the speaker’s face.
An increasing mass of evidence in recent years has come from mobile phones, with some perpetrators deliberately recording their crimes and victims of violent crimes sometimes calling 999 while the attack is about to happen or ongoing.
“People do record their own rapes, muggings or murders or they call 999 then the line is open, with every noise recorded,” says the professor. “The voice of the perpetrator is often heard. It’s worrying how many people record a crime to show their friends. There are several rape cases a year where this has happened.”
What about the emotional wear and tear of hearing such barbaric crimes over and over again as they are analysed? “There was a time when I felt contaminated... but you just develop self-protective strategies.”
Prof French believes that both the Interception of Communications Act and the Regulation of Investigatory Powers Act – which prohibit use of telephone intercept evidence in court – should be repealed.
“We are the only country in the world that doesn’t allow them in evidence, and the fact that there has been an intercept (under a warrant issued by the Home Secretary) can’t even be mentioned. I don’t want them used because it would mean more work for people like me, but because the police would be able to bring charges earlier in cases that currently take many months to put together. They could maybe be over in a matter of weeks.”
Peter French Career Profile
Professor Peter French did a degree in English and education at Leeds University, but swapped a love of literature for language, with further degrees in linguistics and phonetics and a PhD in analysis of recorded conversation.
Research fellowships and lecturing at Birmingham then York St John University followed, before increasing quantities of expert witness work led him to set up his own forensic speech and acoustics laboratory, JP French Associates, in York.
His team is involved in 500-600 court cases a year in this country and abroad, where voice recordings of one kind or another are submitted as evidence in legal proceedings.
Peter French teaches the MSc course in forensic speech science at York University, where he holds an honorary professorship. He is also the founder of The International Association For Forensic Phonetics and Acoustics.