Finding linguistic indicators of personality
This chapter looks at some of the scientific evidence for linguistic indicators of personality. Linguistic analysis models were applied to the small samples of texts to detect measurable patterns in their respective authors’ use of language. Linguistic dimensions associated with psychological conditions such as emotion were coded separately. A number of specific linguistic features emerged as especially sensitive to personality differences: social words, emotional words, first-person pronouns and present tense verbs. The language feature scores and personality dimension scores were then entered into mathematical analyses to find out whether consistent and systematic relationships emerged between certain linguistic markers and the personality profiles. Computational linguistics techniques were applied to authors’ texts, in the form of emails, to identify markers that can be used to predict author demographics and personality. Linguistic features that consistently exhibit strong statistical relationships with specific personality profiles can be further classified in terms of their relevance as personality markers.