Data sousveillance: An analysis of cultural and congressional surveillance rhetoric over time

Jacky Connolly, Sarah Hackney, Rajene Hardeman, Erin McCabe, Allison Nellis, Laurin Paradise, and Chris Alen Sula

About this Project:

We are a group of advanced digital humanities students and faculty from Pratt Institute’s School of Information. Over the course of the Spring 2016 semester, we designed a DH project in the cultural studies discipline for a conference at the Cultural Studies Association (CSA) in June. The theme of this year’s conference is Policing Crises Now, prompted by Policing the Crisis (1978). After discussing various topic prompts offered by the CSA, we chose to focus on policing in the context of surveillance.

Surveillance is the monitoring of a population, often for the purposes of data collection, control, or social manipulation. As the current media landscape increases its coverage of US government surveillance practices, questions arise about the intentions and values of this type of monitoring as it is happening now.

This study aims to shed light on the conversations happening around surveillance over the past 40 years of American discourse. Certain historically significant events have shaped our understanding of surveillance over the the past few generations, and we have gathered data around these time periods so that we might capture these shifts in discourse. Our data consists of Congressional records, mainstream, conservative, and alternative news sources, as well as pop cultural resources (relevant films about surveillance topics). We gathered a baseline corpus from the Congressional records to see if our other documents exhibit different traits through sentiment analysis.

Sentiment analysis is the use of text analysis to identify and extract subjective information from data. For this project, we used IBM Watson’s Personality Insights service, which determines the “personality characteristics” and worldviews of a text corpus. Review of congressional records provided us with seed terms for our surveillance context, which allowed for further review of independent, business, and conservative news sources, and popular culture (relevant film media and critical reviews). By comparing general and specific sentiment measurements across these various sources, we examine points of similarity and difference in attitude across the present and past, mainstream and political, institutional and popular with regards to surveillance. This project is an act of sousveillance: through data scraping, mining, and quantitative analysis, we are watching from beneath.