Sociologists have long hypothesized that the nature of the human self has evolved over time. From Durkheim to Foucault, social theorists argue that sociohistorical shifts ranging from the industrialization to the growth of psychotherapy have coincided with changes in the characteristics that distinguish people’s core modes of selfhood, rendering people more or less individualistic, calculating, expressive, or narcissistic. In this paper, I leverage computational techniques to assess whether people’s own self-understandings have evolved over time. Drawing on a novel corpus of over 25,000 English-language autobiographies published between 1770 and 1968, I explore how the ways in which autobiographers characterize themselves have evolved between the eighteenth and twentieth centuries. To gain insight into how the self-characterizations of members of different social groups may have converged or diverged over time, I also disaggregate trends in self-representations based on the gender, nationality, location of birth, and occupation of autobiographers. This research project will provide insight into how autobiographers understand and represent themselves in texts, and it will constitute the first computational analysis of a large-scale dataset of full-text, English-language autobiographies.

---