We probably could have predicted Gakirah Barnes’s death based on the stories she left on Twitter. By the time she was 17, the self-identified gang member allegedly shot or killed 20 people. Gakirah would say things on social media like:

Vernon is a well-known street on the Southside of Chicago, and Gakirah is identifying it as the boundary between her and any rival crew or clique. Her handle, @TyquanAssassin, is a moniker she adopted after her good friend Tyquan Tyler was killed in 2013. The devil and gun emoji indicate that crossing Vernon could lead to any transgressing rival being shot.

Chicago saw a 58 percent increase in homicides between 2014 and 2016, effectively negating two-thirds of the decrease in homicides the city had experienced since the early 1990s. Some city leaders, including Superintendent Eddie Johnson of the Chicago Police Department, have suggested that social media posts like Gakirah’s might be contributing to the surge. Their diagnosis may sound to some like an attempt to duck responsibility for the failure of the local law enforcement system to interdict more illegal firearms or do more to stop repeat shooters before they injure or kill again.

But having studied the phenomenon – known in the academic community as internet or cyberbanging – I can tell you that the frequency with which young people use platforms like Facebook, Twitter, and Instagram to hurl insults, taunt enemies, and brag about violent acts is playing a meaningful role in fueling retaliatory efforts between gangs and cliques in marginalized neighborhoods. It also has significant implications for gun violence prevention.

For the past four years, I have examined the relationship between Twitter activity and gang violence among young people who live in Chicago. Because of the complexity of interpreting social media communication, it has been important to develop an interdisciplinary team. I work with social workers to accurately decode what teens are saying, and collaborate with data scientists to detect patterns in social media communication that may lead to gang violence. The process often feels like an archaeological dig, carefully combing through Twitter conversations, studying emojis and hashtags, videos and images, to figure out the cues that often end with gunfire erupting.

In our rush to understand whether social media causes violence, we often forget what brings young people to social media in the first place: connections to other young people. Those connections were apparent in Gakirah Barnes’s tweets, even as she was becoming notorious enough to be dubbed the “Teen Queen of Gangland Chicago” by the Daily Mail. Unique in her status as a female shooter for her gang, Gakirah kept up a feed full of threats and taunts to “opps,” or members of oppositional groups. But after analyzing her social media communication, I noticed that Gakirah and other users in her network also used Twitter to collectively cope with losing friends to shootings.

A pattern emerged in Gakirah’s tweets. First, there would be a post expressing general loss or grief. Then, it would be followed by a more aggressive or directly threatening tweet. For example, at 10:53 a.m. one day, Gakirah turned to Twitter to grieve the death of a friend.

Less than 10 minutes later, Gakirah posted a second message, vowing revenge. A scan of news reports from the time suggests that Gakirah’s friend may have died in a fatal confrontation with police, but it’s not law enforcement from whom she plans to exact retribution. Instead she sends notice that she will be hunting gang rivals, squaring the ledger of loss by targeting a mutual foe of her deceased friend.

Jeffrey Lane of Rutgers University has identified a code of the “digital street,” wherein informal rules established in a brick-and-mortar neighborhood dictate how young people communicate with each other online. As this pair of tweets from Gakirah shows, online expressions of trauma take what would have been a moment of private mourning and put it on cyber display. The code of the digital street then compels rivals to reply with posts disrespecting the dead. And from there, an obligation to strike back at the offending posters takes hold. One such “opp” replied to Gakirah’s tweet, baiting her to enter his terrority — so that then he would have a reason to retaliate against her crew. (Identifying information has been blocked to protect the identities of the users who are still living.)

Opp Response

Social media doesn’t allow for the opportunity to physically de-escalate an argument. Instead, it offers myriad ways to exacerbate a brewing conflict as opposing gangs or crews and friends and family take turns weighing in.

The dynamic poses challenges to existing approaches to violence interruption, which treat shootings like a communicable disease that spreads through face-to-face interactions and can be prevented by steering one of the parties toward peaceful alternatives to armed response. Those programs, developed before social media became part of daily life, don’t have the capacity to keep up with thousands of users hurling endless insults at each other.

Gakirah was shot on April 11, 2014, three blocks from her house. She died in the hospital. Her last original tweet (below) included her address. (TMB is an abbreviation for Trap Money Brothers or Boys).

After her death, many of Gakirah’s Twitter friends articulated deep pain. But there were also the too-familiar, unmistakable plans to retaliate, with other followers changing their handles to reflect intentions to avenge her killing. The future of gun violence prevention depends on a deeper understanding of how social media fills a need for disadvantaged communities hungry for connection, and when a hashtag or emoji is a signpost to the next exchange of gunfire. As research provides that understanding, the question will then become: What responsibility do tech companies bear for the shootings bred via their servers?

Dr. Desmond Patton is an assistant professor at Columbia University’s School of Social Work. He is also a faculty affiliate of the university’s Social Intervention Group and the Data Science Institute.