Telling Stories: Analyzing Text to Understand Personality, Social behavior, and Narratives

  • James Pennebaker | The University of Texas at Austin

Pronouns, prepositions, articles, and other short function words have traditionally been relegated to STOP word status in the AI world. Dozens of studies are summarized pointing to the value of function words in understanding people’s thinking styles, social interactions, emotional states, honesty, status, and a broad array of behaviors. Ultimately, function words can tell us how people are paying attention and how they are connecting with their audience. Recently, my lab has found reliable patterns of function word use that underlie stories and other types of narratives. The ways most stories are constructed reveal the changing social relationships among the author, the audience, and the story’s characters.

Speaker Details

Jamie Pennebaker is the Regents Centennial Professor of Psychology. Over his career, he has made significant contributions to social, clinical, cognitive, and health psychology. After discovering a method to help people cope with traumatic experiences through brief writing exercises, he built LIWC, a computerized text analysis program, that identified psychology-relevant language dimensions. In recent years, his research team has focused on ways to measure natural language use to better understand social and psychological processes. Author of over 300 scientific papers and 8 books, he is one of the most cited researchers in the social sciences. In addition to being continuously funded by federal grants since 1984, Pennebaker has received multiple research and teaching awards.

Series: Microsoft Research Talks