The language of crime
Dr Andrea Nini talks about using forensic linguistics to study the infamous Jack the Ripper letters.
Dr Andrea Nini, a lecturer in Linguistics and English Language at The University of Manchester and an expert in forensic linguistics, began his research when he stumbled upon the letters by chance and became fascinated by them.
He analysed over 200 letters claiming to be from the perpetrator of the Whitechapel murders of 1888, fitting his work around his PhD and other commitments post-PhD whenever he could.
The police received about 130 of these letters immediately during and after the murders. The rest of the letters continued to be sent for around a decade afterwards. There has been much speculation over whether any of them were written by the real killer - who was never caught - or if they were all written by hoaxers, or even journalists looking to create some 'fake news' to help sell newspapers.
By analysing the use of certain words and phrases across the entire set of letters, he found that two letters that were among the first to be received by the police - the Dear Boss letter and the Saucy Jacky postcard - were written by the same person. His study also suggested that these two texts were linked to another letter sent later on, the Moab and Midian letter.
This possible connection to a third letter is all the more fascinating considering that the Dear Boss and Saucy Jacky texts were the first to be made public - prompting a flood of copycat letters - and the Moab and Midian letter arrived later.
Hoax or fake news?
However, this third letter is a mystery in itself. One of the Central News Agency's journalists, Tom Bulling, forwarded a copy of the letter and the original envelope to the police. The letter has since been lost, making Moab and Midian the only Jack the Ripper letter that no one other than one or two people has ever seen.
"He didn't say why he didn't forward the original one, and it is incredible that nobody asked. And, to this day, we have no idea why he did such a thing," Dr Nini says.
The release of Dr Nini's research prompted some news outlets to jump to the conclusion that Tom Bulling authored all three of the connected letters to generate business for his employer - something that Dr Nini is keen to correct.
"We need historical evidence that the Moab and Midian letter was a hoax, written and fabricated by Tom Bulling," he explains. "At that point we can say, ok, possibly the other two letters are likely to have been written by Tom Bulling, because of the linguistic evidence. But that's beyond anything that I can do."
Considering the notoriety of the murders, it is perhaps surprising that Dr Nini's research represents the first time that a professional linguistic analysis of the letters has been carried out to find out if more than one was written by the same person.
"Not many people know that these letters exist," Dr Nini explains. "Everybody knows about Jack the Ripper to some extent, but not many people know that the name actually came from one of these letters."
Indeed, the murders are so famous that an entire tourism industry has sprung up around them and books continue to be published about the murders each year even now. In this context, it is little wonder that the media has been so interested in Dr Nini's research, with outlets covering the story including the Times, BBC News, the Daily Mail, the Express, Fox News, Forensic Magazine and IFL Science. He has also discussed the research on BBC Radio Manchester and talkRADIO.
"I thought that that would happen," he says. "My other research on the Bixby letter also received a lot of attention, and I thought that if you combine Jack the Ripper with forensic linguistics, there is bound to be a lot of media attention. So I was expecting it."
The world of forensic linguistics
Forensic linguistics, or linguistics as applied to legal problems, is a small but growing field of practice. The 2017 TV show Manhunt: Unabomber recently renewed public interest in the discipline over two decades after the events of the show - a series of bombings in the US - actually took place.
At the time, the police were at a loss for leads until the bomber started sending letters to the press. When the bomber's brother recognised the writing in his 'manifesto', FBI agent James Fitzgerald and colleagues used a linguistic analysis of these letters to put together the evidence needed to obtain a search warrant, culminating in the arrest of the culprit in 1996.
“Not many people are aware of what forensic linguists can do for them, so they don't request help from us even though they might need it.”
The TV show has generated mostly positive reviews, but Dr Nini says that more could be done to promote the use of forensic linguistics within the police and the legal sector.
"Not many people are aware of what forensic linguists can do for them, so they don't request help from us even though they might need it," he explains.
A handful of cases in the UK have been solved with the help of forensic linguistics, particularly those where the murderer sent text messages 'from the victim' to reassure their family that all was well. In some of these cases, forensic linguists were able to compare the text messages with those actually sent by the victim and determine whether they had really sent them by analysing the style in which the messages were written.
Still, the practice of analysing language for clues to solve criminal cases is not currently widespread. Yet Dr Nini is optimistic.
"It's very likely that it'll be a matter of time, because people are not hand-writing any more, and they're not writing documents on their own with typewriters or anything like that," he says.
"The work of forensic handwriting experts and forensic document examiners hasn't disappeared, but it's probably becoming less and less frequent - and part of that is being replaced by forensic linguistics."
Is it likely that the increasing use of very short forms to convey information - including both text messages and posts on social media platforms like Twitter - presents a challenge to those looking to analyse text-based evidence for clues to authorship?
"The shorter the text, the less likely it is that there is evidence in it," Dr Nini says. However, he notes that the letters he analysed in his own research were short, too - some of them were only 50 words long.
"That's another contribution of the Jack the Ripper study; testing and applying methods used to analyse short texts on this case. It's an addition to that body of research," he explains.
Teaching forensic linguistics
Dr Nini is also passing on his own expertise in forensic linguistics. As well as conducting his own research and fulfilling requests for help as an expert witness in criminal and civil cases, he has launched and currently leads the Forensic Linguistics course unit on The University of Manchester's undergraduate and postgraduate linguistics courses.
The unit is incredibly popular, with around 100 students taking it in its first year. Dr Nini uses a hands-on approach to his teaching, getting his students to work on a different forensic linguistics case each week.
His undergraduate students have tackled challenges ranging from whether it's possible to tell if a statement to the police has been tampered with to cracking a code used by prisoners to communicate with the outside. Master's students are set more complex problems based on real-life cases such as the authorship of internet posts claiming to be from a time traveller called John Titor, and letters sent by the Zodiac killer.
"They have to find their own creative solution to each case," Dr Nini says. "Overall, it's been great. I've learned a lot from engaging with the students as well, which is fantastic."
Now that the Jack the Ripper study is over, Dr Nini is embarking on his most ambitious project to date: a study that will attempt to establish the extent to which each person uses a different version of their language - known as their idiolect - that is unique to them, as is believed to be the case by many linguists, but has not yet been proven.
He will attempt to find a mathematical basis for this intriguing theory, borrowing methods and theories from physics and computer science. Interestingly, the idea for this research came out of his work on the Jack the Ripper letters, in which he discovered that there was one particular combination of four words virtually unique to only two of the letters.
"How is that even possible?" he asks. "It is amazing how even though we all pretty much use the same words, we combine them differently. The probabilities of finding shared word combinations sometimes are so low that is shocking, almost like probabilities of DNA matches. And we're talking about language, and I think that's extremely fascinating."