Digital methods in memory studies: A beginner’s guide to scalable reading of Twitter data

Abstract

This article makes a methodological contribution to the growing subfield of digital memory studies. It demonstrates a possible way forward for memory studies scholars who want to try out digital methods but also remain in conversation with the kinds of research traditionally produced within the field. The article revolves around a showcase of an analytical workflow for conducting a scalable reading of large quantities of tweets through access to the Twitter API. The article argues that using only basic computational approaches to social media data in combination with API access can drastically improve data collection practices and enrich analytical practices, producing results recognizable and compatible with existing research in memory studies. As a case study, the article uses a dataset of nearly 200,000 tweets collected around two events that prompted Twitter users to discuss the history of the American children’s television program Sesame Street. It does so to demonstrate: first, how a visualization focusing on chronology can help underpin arguments about heightened activity around certain events. Second, a close reading of selected tweets from these events can support claims of shared activity, even if no hashtags were used. And third, how using simple tools for distant reading makes it possible to converse with questions and issues about gatekeepers and connectivity already central within memory studies. Furthermore, the article demonstrates how the Twitter API supports a more systematical, large-scale collection of tweets than usually seen in memory studies, making researchers less dependent on the algorithmic bias that rules the search in the platform’s regular interface.