Researchers at the Georgia Institute of Technology looked at nearly 29 million tweets surrounding four recent Black Lives Matter (BLM) events to identify the social media patterns of its activists. They found that the community is unlike many other social movements because of its ability to bond over the course of many months. More than a third of those who participated via social media for the first time continued their participation during the next BLM event. The study also found being from a state with historically high rates of black victimization due to police violence, especially in the South and Midwest, was linked to peoples' response on social media even if the state wasn't the scene of the current BLM event. The researchers also found a change in tone on Twitter the day before protests sprung up around the nation.
The Georgia Tech study is among the first to examine the online language of the BLM movement. It collected tweets surrounding four major events: the shooting of Michael Brown in Ferguson, Missouri (August 2014); the decision not to indict Ferguson police officer Darren Wilson (November 2014); protests and the shooting deaths of two New York City police officers (December 2014); and the death of Freddie Gray while in custody of Baltimore police officers (April 2015).
The team also plugged in data of more than 10,000 deaths due to police shootings since 2000. From that, it outlined which states, based on their African-American population, had the highest rates of deaths. Missouri, Kansas, Oklahoma and New Mexico were among the highest in the mainland United States.
"These were the same states where we found very negative language on Twitter during Black Lives Matter events," said Munmun De Choudhury, the Georgia Tech assistant professor who led the study. "Other areas, especially in the South and Midwest, used words about death to express their feelings. Potentially, Twitter users in states with historically higher rates of fatal police shootings of blacks tend to interact with others to seek and provide psychosocial support around issues of racial inequality."
However, much of that negativity disappeared on the immediate eve of protests. Language changed. The researchers noticed that tweets about death, anger and first person singular nouns transformed into more posts about family, friends and first person plural pronouns.
"We saw more sadness but lower anger and anxiety the day before protests," said Benjamin Sugar, a Georgia Tech master's student who co-authored the study. "It doesn't mean they weren't angry or anxious. But in 140 characters, people showed thoughts of moving forward and making a difference."
BLM was born on Facebook in 2013 following the acquittal of George Zimmerman in the shooting death of Florida's Trayvon Martin. During each subsequent event, the movement has contradicted typical Twitter stereotypes. The study found that on average 36 percent of first-time BLM users participated again.
"Other hashtags movements have helped spread awareness about important issues, but many of them quickly lose their momentum," said Shagun Jhaver, a Georgia Tech Ph.D. student and co-author. "The Black Lives Matter movement realizes it's part of a long-term social transformation and shows continual engagement. And it continues despite having no formal hierarchal structure."
The research, "Social Media Participation in an Activist Movement for Racial Equality," won best paper when it was presented last month at the International AAAI Conference on Web and Social Media in Germany. It was funded, in part, by the National Institutes of Health (#1R01GM11269701). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Institutes of Health.