Linguistic Indicators of a False Confession

Abstract

To date, most investigations of false confessions have focused on their prevention rather than their identification. In this study we investigated whether certain linguistic variables might help to distinguish between false confessional statements and true accounts. Using a within-subjects design, we elicited both false confessional statements and true accounts from 85 participants. We examined these for the presence of nouns, verbs and adjectives. Additionally, participants completed the Depression, Anxiety and Stress Scale (DASS) to determine the relationship between noun, verb and adjective use and self-reported measures of state anxiety. Results showed that whereas nouns and verbs failed to discriminate between false confessions and true accounts, adjective use significantly decreased during false confessions. Anxiety was not associated with veracity. The current findings suggest that there are measurable linguistic differences between false confessions and true accounts that cannot be attributed to level of state anxiety.

Key words:

• adjectives
• anxiety
• DASS
• deception
• false confessions
• grammatical category
• lexical class
• linguistic markers of deception
• nouns
• parts-of-speech
• verbs

Notes

1. A search of key databases in psychology, linguistics, computational linguistics, communication studies and behavioural sciences (IEEE Explore, LLBA, JSTOR, OvidSP, ProQuest, PsycINFO, Scopus, and Web of Science), from 1985 to 2010, was conducted using combinations of several key words and key phrases: “deception”, “lying”, “confessions”, “false confessions”, “linguistics”, “language”.