New research sheds light on how Donald Trump’s offline rhetoric might have mobilized online political discussions related to the storming of the U.S. Capitol on January 6, 2021. The findings have been published in the journal Frontiers in Sociology.
“Early in 2020, my team and I started monitoring the COVID-19 pandemic on social media, as well as the 2020 presidential election. During the pandemic, we noted a large number of groups created narratives focusing on public insecurity and employed different strategies to engage with target audiences and/or via distinctive targeted messages. We thought the pandemic ‘model’ might be applicable to the 2020 election, as COVID-19 continued to spread,” said study author Claire Seungeun Lee (@drclaireselee), an assistant professor and Donahue Ethics Fellow at the University of Massachusetts Lowell.
“At the same time, we noticed the contentious atmosphere surrounding the presidential election; specifically, that far-right groups were openly supporting President Trump. In this regard, we began collecting data from November 2020 onwards in order to observe the message and engagement of different far-right groups in the upcoming election. During this period, we witnessed how the two presidential candidates constructed their political campaigns in very different ways (e.g., messages, audiences, targets).”
“The January 6, 2021 insurrection emerged as an event during our initial stage of the data collection,” Lee explained. “We quickly recognized the significance of the event, which prompted us to refocus our research on communication during the incident. Even though the event was unfortunate, it provided us with a firsthand perspective on political communication. Most importantly, since the data collection program was in place prior to the event, we were able to capture data that some users would later delete and therefore be unable to be collected during a later collection process.”
The researchers used language processing algorithms to compare the linguistic similarity of three of Trump’s speeches to tweets that used the hashtags “#trump2020,” “#MAGA2020,” or “#QAnon.” The three speeches included a speech delivered on January 4, a speech at the Capitol on January 6, and a subsequent speech on January 6 as the Capitol was under assault. Lee and her colleagues then conducted thematic analyses of tweets with the highest similarity to Trump’s language.
“The unique approach of our study is the use of a mixed-methods approach of quantitative and qualitative thematic analyses,” Lee said. “Using this approach, we were able to combine offline speech information with Twitter data during key speeches leading up to the date of the insurrection; exploring the link between Trump’s offline speeches and QAnon-related hashtags across a three-day timeframe.”
Several themes emerged based on tweets that were similar to Trump’s January 4 speech: declaring Trump the “greatest president,” criticizing Vice President Mike Pence, and calling for actionable support for Trump. “More importantly, a well-known powerful slogan emerged – ‘Storm the Capitol,’ which has later become synonymous with the riotous actions of January 6,” the researchers noted.
Themes based on tweets that were similar to Trump’s January 6 speech included criticism of the media and social media companies, complaints about illegal votes and the “stolen” election, and references to Pence’s (dis)loyalty. “In the January 6 rally speech, [Trump] said, ‘They defrauded us out of a win in Georgia’. This sentiment was picked up online and protracted in Twitter,” the researchers said.
Two main themes emerged from tweets similar to Trump’s speech in the wake of the riot: references to the “stolen” election and references to the rioters dispersing from the Capitol grounds. “While the theme of stolen election was present in the earlier speech made at Trump’s rally on January 6, the theme of dispersal became more relevant after the January 6 (Evening) speech, in particular with #MAGA2020 and #QAnon,” the researchers explained.
“Our analysis shows a link between online political participation and offline political speech,” Lee told PsyPost. “This research illuminates this phenomenon and offers policy implications for the role of online messaging as a tool of political mobilization, as well as the transference of information into online messaging.”
But “due to the large-scale data and a particular focus of the study, we only could explore and analyze a very small portion of the data,” Lee noted. “We are planning to look at other far-right group discourses around the election and pandemic. In doing so, we want to find out whether different groups and/or their messages are interrelated to each other in ways we have yet to explore.”
“Social media has created another dimension in communication for which we can study other aspects of message transference, online proliferation, survival rate, as well as sentiment and emotion,” Lee added. “This field of study is in its infancy and continues to evolve as a new frontier in the study of communication.”
“Computer platforms are very good at capturing and recording this information, but we as researchers need to continue to develop accurate analysis methods to ensure the integrity of our results. There is no denying that human interaction is very complex so we need to ensure we use analysis methods provide the best possible accuracy in our findings. This is an area researchers need to continuously reassess and refine.”
The study, “Storm the Capitol: Linking Offline Political Speech and Online Twitter Extra-Representational Participation on QAnon and the January 6 Insurrection“, was authored by Claire Seungeun Lee, Juan Merizalde, John D. Colautti, Jisun An, and Haewoon Kwak.