Using computational text analysis tools to compare the lyrics of suicidal and non-suicidal songwriters. Submitted to Cognitive Science

Recent research in cognitive science and psycholinguistics has provided evidence that extreme behaviors such as suicide may be predicted by analyzing an individual’s writing style. This study seeks to extend research of this kind to a new and less traditional written genre: song lyrics. We applied the computational tools Coh-Metrix and Linguistic Inquiry and Word Count to a corpus of 35 songs each by 16 different songwriters in order to investigate how differences between suicidal and non-suicidal songwriters ’ lyrics reflect previous research on suicide. The results of our study suggest that suicidal songwriters use words of lower concreteness, fewer words, more future verbs, and fewer death-themed words in their lyrics. Overall, our results contrast notably with previous research, providing a new perspective on deriving psychological information from non-traditional texts