Quantcast
Viewing all articles
Browse latest Browse all 1389

How Forensic Linguistics Helped Solve JK Rowling Mystery

Image may be NSFW.
Clik here to view.
The Cuckoo's Calling
Harry Potter author JK Rowling – recently fingered as the pen behind detective novel The Cuckoo’s Calling, published in April under the pseudonym Roger Galbraith – is understandably furious about being outed. But it might not have happened without some detective work by a computer in Pittsburgh, Pennsylvania, of all unlikely places. London Sunday Times reporter, Cal Flyn, who broke the story after receiving a Twitter tip-off, first approached Pittsburgh-based Patrick Juola of Duquesne University to assess whether the tweet naming Rowling as the author was from cloud-cuckoo land or not.

Joula is a professor in computer science and an expert in forensic linguistics who has helped develop “authorship attribution” software that identifies a writer’s unique voice within a text. The professor was given electronic copies of The Cuckoo’s Calling to compare against Rowling’s own The Casual Vacancy, Ruth Rendell’s The St Zita Society, PD James’ The Private Patient and Val McDermid’s The Wire in the Blood. He then used the program to break down Cuckoo into chunks of 1000 lines and compared each chunk individually against the other four books.

Joula explained that the key for comparison is not long, characteristic words, like a character’s name (the hero-detective in Cuckoo is Cormorant Strike, which was also the code name for a 2012 Sri Lankan joint military exercise). Rather, it’s the use of short words that provides the key clues: prepositions, articles and connecting words, which make patterns that are too difficult to mask, “because they’re so subconscious and the author won’t even think of changing these”.

The results of the tests were that JK Rowling was the most likely to be Cuckoo‘s author. This was enough evidence for Flynn to take to Rowling’s publisher, and Rowling subsequently confessed. In keeping with fiction’s classic mystery stories, she was the least likely suspect, until a computer algorithm with a detective’s nose was put on the case. – Jennifer Platt @jenniferdplatt

Book details


Viewing all articles
Browse latest Browse all 1389

Trending Articles