In this month’s research spotlight, COSMOS highlights recent research that focuses on collective mobilization, in the form of political protests. Specifically, the studies titled, “A Computational Approach to Analyze Identity Formation: A Case Study of Brazil Insurrection” and “Role of Co-occurring Words in Mobilization in Brazilian Social Movements,” both studied two protest movements surrounding the 2022 Brazilian elections, one anti- and the other pro-government. These studies were published and presented recently at this year’s annual Americas Conference on Information Sciences (AMCIS 2024) in late August, which took place from 15 to 17 August in Salt Lake City, Utah.

Protests represent a coming together of individuals to speak out and to respond to an event, particularly a political one; this phenomenon is known as collective action. There are various theories behind what social elements bring together such individuals to enact collective action, and each of these studies investigated different characteristics of the protests in January 2023 that took place following the Brazilian 2022 election results. Leading up to 2023, supporters of the losing candidate, Bolsonaro, claimed the election had been fraudulently stolen, eventually protesting publicly in January—thus, being anti-government protesters. In response, pro-government supporters of the winning candidate similarly organized to identify rioters and dispel misinformed claims.

“A Computational Approach to Analyze Identity Formation: A Case Study of Brazil Insurrection” examined how online discourse on Twitter and Instagram contributed to collective identity formation leading up to and during the Brazil insurrection. Using Moral Foundations Theory and Resource Mobilization Theory, researchers analyzed hashtags, content similarity, and network structure to track the development of community groups and their cohesion. They found that as the protests approached, network cohesion increased, with topics converging and identity formation intensifying. Resource mobilization patterns showed growing support, reflecting shared motivations and unity against a common adversary.

 “Role of Co-occurring Words in Mobilization in Brazilian Social Movements,” applied Diffusion of Innovations (DOI) theory to analyze Instagram discourse in three stages: initiation, amplification, and sustainment. Over 600 co-occurring word pairs were manually coded for their conative (action-oriented), referential (informational), and emotive (expressive) language use. Anti-government posts primarily employed conative language, peaking during amplification and correlating with the rapid escalation into the January 8, 2023, riots. Pro-government narratives, however, maintained a consistent use of referential language, focusing on policy details and rational arguments. The results highlight distinct patterns in discourse strategies between opposing groups, indicating how language use and collective identity shape the dynamics of protest mobilization and escalation.

These research studies show two different investigative methods for discovering the subprocesses behind collective action, such as collective identity formation and language use.