Broadcast information diffusion processes on social media networks: exogenous events lead to more integrated public discourse

Understanding information diffusion is vital to explaining the good, bad, and ugly impacts of social media. Two types of processes govern information diffusion: broadcasting and viral spread. Viral spreading is when a message is diffused by peer-to-peer social connections, whereas broadcasting is characterized by influences that can come from outside of the peer-to-peer social network. How these processes shape public discourse is not well understood. Using a simulation study and real-world Twitter data (10,155 users, 18,000,929 tweets) gathered during 2020, we show that broadcast spreading is associated with more integrated discourse networks compared to viral spreading. Moreover, discourse oscillates between extended periods of segregation and punctuated periods of integration. These results defy simple interpretations of good or bad, and instead suggest that information diffusion dynamics on social media have the capacity to disrupt or amplify both prosocial and antisocial content.