Toward Mainstreaming of Feminist (Counter)Publics? The Networked Structure of Feminist Activism on Twitter

Abstract

This study investigates the roles of feminist actors in the Twitter discourse about sexualized violence that came up during the confirmation of Brett Kavanaugh as a U.S. Supreme Court justice in October 2018. We theorize the discourse from the perspective of feminist counterpublics because it helps in understanding the extant power relations in public debate that are also manifest in digital networks. In this empirical study, we reconstruct the Twitter network discourse about the judge’s confirmation and examine the role of feminist actors and their counterpublic characteristics. Our findings indicate that the discourse network was highly polarized, but feminist voices were well integrated within the (progressive) center. Although feminists acted in the role of challengers seeking the attention of established political actors and news media, they did not remain isolated. Instead, they positioned themselves in the center of the progressive network and therefore influenced the general public discourse.

Keywords:

• Counterpublics
• feminist activism
• sexualized violence
• social network analysis
• Twitter

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Only actively tweeting accounts were considered as nodes. Accounts that were addressed only via @mentions were removed from the network.

2 The multilevel community detection algorithm is a modularity optimization algorithm (Blondel et al., 2008). Modularity optimization procedures are often used in network research because of their well-interpretable results and computational efficiency (Elgesem et al., 2015, p. 173; Stoltenberg et al., 2019, p. 130).

3 For manual coding, we selected the top 20% of the actors for each community. For smaller communities, we coded at least 30 actors (or all of the smallest ones). Therefore, we sorted the actors based on their total degree (in-degree + out-degree), i.e., the summation of the attention they receive as measured by the communicative relations targeting them (in-degree) plus their level of activity to connect to others (out-degree). Therefore, our selection of coded actors comprises the most prominent and the most active ones.

4 Intracoder reliability test results for each variable: actor type: Holsti = 0.97; position: Holsti = 0.97; feminist versus anti-feminist standpoints: Holsti = 0.97.

5 Each of these communities included at least 20 actors. Several smaller communities that consisted of fewer nodes were not considered for further analysis. The modularity value of 0.53 indicated a significant community structure. Note that the network figures were generated with the open-source program Gephi (Bastian et al., 2009).