Linguist jumps out of skin over "sorting head out"
« previous post | next post »
An article today’s NY Times and another in WalesOnline tell us about a linguist in Wales who was praised for discovering that a murderer — who had been having an affair with his victim — unconsciously revealed his identity as the writer of a fake text message that included either the phrases, “need to sort my head out” and “sorting my life out” (according to WalesOnline) or “sorted her life out” and “head sorted out” (according to the NY Times). Regardless of which quotation is accurate (assuming that one of them got it right), this can sound like something from an ill-conceived TV cop show and hardly something that would cause the linguist to “jump out of” his skin.
If we are to rely on these press reports of this linguist’s work, writing “sort X out” along with his use of periods for commas by themselves are not sufficient clues for nabbing a forger/murder. I don’t know of any court of law that would rely on this evidence alone. But wait, maybe the court didn’t have to rely on it. WalesOnline also tells us: “The police then seized one letter [from the suspect] saying “she [the murdered woman] is not getting away with my life” and another letter in which the suspect requested a gun. In contrast, the Times reports that the police found a five-page letter the suspect had written about his desire to kill himself and the victim. Whichever report is accurate (or even if both are), these letters seems like more convincing evidence than the "sorting out" phrase.
If these science writers told even part of this story accurately, the police were putting far too much faith on the alleged smoking gun of “sorting X out” plus the allegedly odd punctuation. Forensic linguists have a much larger tool kit to use. (see here and here). It includes comparing known samples with unknown samples in the same register (often difficult to find), knowing about regional and social variation, having a large enough sample to account for natural language variability, knowing with some degree of certainty that a given linguistic feature can be diagnostic of individual or group usage, and focusing on aspects of language that are used less consciously than vocabulary or punctuation (syntax comes to mind).
It’s possible, of course, that these science writers got it wrong. Science writers have been known to do this (see here).