This study aims to shed light on conversations of surveillance over the past 40 years of American discourse, using a corpus of entries to the Congressional Record, mainstream and independent news sources, as well as movie scripts and reviews. These components were selected as representative of attitudes across American political, media, and pop cultures.
The personality, mood and tone markers found in each of these content types were then determined using the IBM Watson Personality Insights API. By comparing sentiment measurements across the corpus, we examine points of similarity and difference in surveillance discourse across present and past, cultural and countercultural, institutional and popular — watching surveillance, as it were, through assemblages of text and data.