How Forensic Linguistics Works

TL;DR

Forensic linguistics uses statistical and linguistic analysis of writing style, word choice, and regional patterns to identify authors and solve crimes.

Key Points

• 1.Burrows' Delta identifies authors by comparing word-frequency distributions across large texts. When applied to J.K. Rowling's pseudonymous Robert Galbraith novels, the 500 most common words matched Rowling's known works far more closely than any other author's, exposing her identity.
• 2.The Jaccard coefficient matches short texts to authors using shared linguistic features. In the 2005 Jenny Nickel murder case, nine features — like spelling of 'myself' and abbreviations — showed the alleged runaway texts matched the 45-year-old boyfriend's writing, not Jenny's.
• 3.Regional vocabulary and deliberate misspellings can pinpoint a suspect's geography and education with no prior suspect. A single ransom note led linguist Roger Shuy to identify the kidnapper as a well-educated man from Akron, Ohio, using the hyper-regional term 'devil's strip' and suspiciously purposeful misspellings.
• 4.Sentence structure shifts in confessions can reveal police fabrication. Timothy Evans's wrongful execution case showed that incriminating sections of his confession used unnaturally precise timelines and complex clauses, unlike his natural run-on style, indicating those passages were fed to him or edited by police.