Computerized Text Analysis  Detect Killer Instincts In Psychopaths
Scientists at Cornell University, New York have now learnt that computerized text analysis could possibly reveal how psychopathic killers make identifiable word choices which are beyond conscious control when talking about their crimes. Anthony Hopkins as Hannnibal. Reuters

The next time you're in the mood to narrate a murder-mystery, be wary of the words you use. Scientists and language researchers from Cornell University, NY, suggest that psychopaths are more likely to use certain word patterns than others, thereby revealing their killer instincts.

The team of researchers have discovered that a computerized text analysis could reveal how psychopathic killers make identifiable word choices (these are beyond conscious control) when talking about their crimes. The results of the study, published in the journal Legal and Criminological Psychology, could lead to new tools for diagnosis and treatment. The findings could also have implications on law enforcement and social media.

Given the significance of social media communication in today's reality, Jeff Hancock, a Cornell professor of computing and information science at the University of British Columbia, explains that the ability to identify criminal patterns via word therapy and analysis could have broader implications in helping clinicians identify people in need of treatment, as well as aiding law enforcement officials in tracking suspects across Facebook and Twitter.

Our paper is the first to show that you can use automated tools to detect the distinct speech patterns of psychopaths, said Hancock, who was also the lead researcher for the study, This can be valuable to clinical psychologists because the approach to treatment of psychopaths can be very different.

While admitting that previous work did look at how psychopaths use language, Hancock explained that his team's research matched personalities to words that psychopathic criminals use, all of which reflect selfishness, detachment from their crimes and emotional flatness.

Hancock and his team analyzed stories narrated by 14 psychopathic male murderers held in Canadian prisons and compared them with 38 convicted murderers who were not diagnosed as psychopathic. Each subject was asked to describe his crime in detail. Their stories were taped, transcribed and subjected to computer analysis.

Psychopaths used more conjunctions like because, since or so that, implying that the crime had to be done to obtain a particular goal. They used twice as many words relating to physical needs, such as food, sex or money; non-psychopaths used more words about social needs, including family, religion and spirituality. Unveiling their predatory nature in their own words, psychopaths often included details of what they had to eat on the day of their crime.

Psychopaths were more likely to use the past tense as a prologue which suggests detachment from their crimes, explained the researchers. According to the researchers, psychopaths also tended to be less fluent in their speech, using more ums and uhs.

The exact reason for this is not clear but researchers speculate that in framing the words as they do, psychopaths try hard to make a positive impression. That requires them to use more mental efforts to frame their story.

Cornell Computer and Information Science Professor, Jeff Hancock, will deliver his talks about social media, law enforcement and broader implications of his new research into computerized text analysis and the minds of psychopathic murderers on October 17 at the next Cornell media luncheon.