This study explores the dynamics of the Jordanian Twittersphere by analyzing a month's worth of tweets produced in Jordan. The research investigates how online communities form, the linguistic divisions within these communities, and the topics they engage with. Initially, the study hypothesized that the Jordanian Twittersphere would be divided into two main clusters—Arabic and English speakers—with distinct discussion topics. However, the reality proved to be more complex.
Data collection:
The data collection process faced challenges, particularly in accurately identifying tweets produced in Jordan due to limitations in Twitter's API and users' inconsistent location reporting. We overcame these obstacles by refining the sampling approach and leveraging geographic filtering techniques. The final dataset included 380,635 tweets from 5,346 unique users, with the majority of content being in Arabic.
The analysis:
The analysis employed network and content analysis to uncover patterns in retweet communities, language distributions, and thematic discussions. Using the Louvain algorithm, the study identified 642 distinct communities, most of which were small and highly localized. Content analysis revealed that most retweet communities were linguistically homogeneous, with Arabic-dominated clusters being the largest. Similarity metrics showed that tweets within the same community were generally more thematically cohesive than tweets across different communities.
We found political, entertainment, and activist networks, with discussions centered on regional and international issues. Notably, Pro-Palestinian activism, solidarity with Rached Ghannouchi, and discussion of the Turkish elections at the time emerged as major themes. The study also highlighted that different communities had discussed similar themes, but used different framings. So for instance, Arabic content on Palestine appealed to religious discourse and group identity, whereas that in English appealed to international organizations, and used terminology that highlighted violations of international law. This exploratory study thus led us to questions about how the discourse on Palestine changes when a local versus an international crowd is being addressed.
We also found that small clusters of minority groups were Tweeting about their own local politics at home. Future research could expand on these findings by examining how connected these actors are with users in their home countries, and how minority groups political discourse varies across regions.
