✕ CLOSE Online Special City News Entrepreneurship Environment Factcheck Everything Woman Home Front Islamic Forum Life Xtra Property Travel & Leisure Viewpoint Vox Pop Women In Business Art and Ideas Bookshelf Labour Law Letters
Click Here To Listen To Trust Radio Live

How “Language Police” detect identities of anonymous authors

The article I shared here a few weeks ago about how people’s peculiar language usage creates a stylistic footprint, which can set them apart from others and be used to unmask their identities, resonated with more people than I expected. Many readers asked if I could share more insights into this area, called forensic linguistics, so I’ve dug up a great, down-to-earth article that lays it bare. Originally titled “How ‘Word Cops’ Unmask Anonymous Authors,” it was written by John Donovan and published on September 14, 2018 on the HowStuffWorks.com site. Enjoy:

When you think of forensics, you probably think of bones in a shallow grave or fingerprints at a crime scene. And that’s understandable. We can thank TV dramas like “CSI” for that.

But forensics – “relating to or dealing with the application of scientific knowledge to legal problems” (thanks Merriam-Webster, for that)- is way more than femurs and fingerprints. There are forensic psychologists, for example, and forensic engineers. The American Academy of Forensic Scientists includes both forensic pathologists, the ones who deal with the injured or dead, and forensic accountants. Any field where science and scientific knowledge can be applied to the law – whether it’s actually solving crimes or simply providing information for use in something as mundane as, say, a pending contract – is, by definition, forensic.

SPONSOR AD

Even experts in something as sleepy-sounding as linguistics – the study of language and how it’s used and structured – can be considered forensic scientists.

“A lot of the work, quite frankly, is nerdy,” says James R. Fitzgerald, a forensic linguist and 20-year veteran of the FBI who, proving not all linguistics work is sleepy or for nerds, helped break open the case of Unabomber Ted Kaczynski in 1996. “You’re sitting and counting the number of dots in multiple points of ellipses … which actually helped me solve a case once. And you’re doing research on the internet and various corpora, about the frequency of certain punctuation features, or alternate spellings of certain words.

“A lot of people can be amateur linguists. But it really takes someone studying the field, knowing the field, having a true appreciation of language usage, that I believe makes the best forensic linguist. It’s not a part-time gig. It has to be taken as a full-time, serious profession.”

Forensic linguists, perhaps because of their innate nerdiness, may never get their own TV show. But these scientists – and they are scientists, first and foremost – can help solve mysteries and crime in their own way.

What forensic linguists do

In early September 2018, the New York Times ran an anonymous opinion piece that was authored, the paper said, by a senior official in the administration of President Donald Trump. The essay was damning in its portrayal of a chaotic White House and an out-of-control president. Immediately, people around the country – especially, perhaps tellingly, inside the White House – began to wonder: Who wrote it?

That question is squarely in the forensic linguist’s wheelhouse, a task that Fitzgerald calls “authorial attribution analysis:” Figuring out who wrote something, whether that’s a scathing op-ed, a ransom note, a society-shaking manifesto, a threatening email or a blatantly one-sided and perhaps unfair pizza review on Yelp.

Putting a specific name to a specific grouping of words is not always easy or possible, especially if you have only one writing sample to work with. Still …

“You can learn a lot from looking at one document – what’s called linguistic demographic profiling,” says Robert Leonard, the director of the graduate program in forensic linguistics at Hofstra University and, with Fitzgerald, co-director of the school’s Institute for Forensic Linguistics, Threat Assessment and Strategic Analysis. “You can learn an awful lot about people, without comparing it to anything, because actually you’re comparing it to all the other documents you and all other linguists have studied since the beginning of time.”

What can you discover?

“We can see whether somebody is very well educated, perhaps trained in the social sciences, how old they are, perhaps, from the phraseology they use, and many, many other things,” Leonard says. “What work experience they have, where they live … “

Somewhat more difficult to determine, Leonard says, is a person’s gender. Changing gender roles have something to do with that.

Profiling a writer from a single document can immediately help narrow a pool of suspected authors (to, for an example, a well-educated middle-aged person from the Midwest who speaks English as a first language). A trained linguist can help unmask an unknown author further by working with several writings, then comparing the document in question to others whose authors are known.

How forensic linguistics works

People’s language, either spoken or written, differs in many ways, for many reasons that can include (but are not nearly limited to) their education, their surroundings, their age, their mood and their intended audience.

Those differences, according to Carole Chaski – she’s the executive director of the Institute for Linguistic Evidence, a non-profit scientific research organization, and the chief executive officer of forensic linguistics consulting firm ALIAS Technology – fall into a few categories. From a paper she wrote in the Journal of Law and Policy:

In linguistic theory, language is divided into levels for analytical purposes. These levels are sound, word, and word combinations. These levels, respectively, are analyzed in phonetics and phonology; morphology and the lexicon; syntax; semantics and pragmatics; and prosody.

Included in those categories are things like punctuation and spelling, too. Forensic linguists, then, in examining writing, look scientifically at everything from the whole to how a certain sentence is structured to the use of a period or a question mark or an apostrophe. They look at articles (“a” and “the”). They look for “markedness,” a linguistic term that refers to how a certain word or phrase differs from the norm.

“When I actually sit down with these documents, what I look for are indicators of lexical features; what are the words individually being used, are they unusual, are they distinctive, are they rare? Are they unique?” says Fitzgerald, who profiled his life and career in a three-book series “A Journey to the Center of the Mind.” “What are some stylistic features of the author? How does he or she use punctuation, how do they format their communication – and that’s important, too. Do they use semicolons, do they use em dashes, do they use en dashes …

“I look for mistakes of course, and are they forced mistakes, are they mistakes in an attempt to disguise the actual identity of the author, or do they appear to be natural mistakes? As I’ve always said, it’s much easier for an anonymous author to dumb down than to smart up.”

In trying to identify the author of an anonymous writing, misdirection is a common tactic. In the New York Times op-ed, the word “lodestar” was seized upon by several armchair forensic linguists. It’s a word used often by Vice President Mike Pence.

The general consensus, though, is that the word was placed in the opinion piece to throw linguists off the trail of the real author.

Becoming a word sleuth

Anonymity works sometimes, but forensic linguists are a pesky lot. An English professor at Vassar, doing some linguistics sleuthing, correctly exposed Newsweek columnist Joe Klein as the previously anonymous author of a novel on Bill Clinton’s first presidential run. “Harry Potter” author J.K. Rowling was outed as the writer of a detective novel published under the nom de plume Robert Galbraith.

Identifying Kaczynski as the author of an anonymous manifesto mailed to news outlets helped lead to his conviction. That remains Fitzgerald’s most famous case. He and Leonard worked on the JonBenét Ramsey case, too, helping in 2006 to rule out someone who falsely confessed to killing the young Colorado beauty contestant in 1996.

Most of forensic linguistics isn’t that high profile, though. It’s painstaking work, done outside the limelight in front of a computer.

“I guess all forensic sciences suffer from what’s called, ‘The CSI Syndrome,’ where everything is tied up neatly in a bow in an hour,” says Leonard who – by the way – had a previous career as an original member of the rock group Sha Na Na and performed at Woodstock in 1969, just before Jimi Hendrix. “What I look for in students is [that they are], of course, highly intelligent and able to go through language data in infinitesimal detail. To not only know about how language works in terms of languages and dialects and interaction among people in different social groups in different languages, … but also to be able to focus on the most minute of data points, and understand the context in which it occurs.”

 

Join Daily Trust WhatsApp Community For Quick Access To News and Happenings Around You.