According to UvA researcher Petter Törnberg, the idea of the bubble on social media is outdated. Using a computer simulation, he found out that conflict on social media is what drives polarization. "On social media, it's not about arguments or opinions, but emotion and identity."
"It's pretty clear that social media contributes to increasing polarization," Petter Törnberg said. Törnberg is a researcher at the UvA with a background in computer science, but now works within digital geography. "Of course, we cannot compare our society to one without social media. But most scientists researching polarization and radicalization think that social media contribute to rising polarization."
Polarization is a "metacrisis," Törnberg says. As feelings between polarized groups become increasingly negative, social debate becomes difficult. It hangs over us like a dark cloud, making it harder for us as a society to deal with other crises such as the corona pandemic or climate change.
Echo on social media
In his new research, Törnberg examined how social media drives polarization. The term "echo chamber" has been a dominant concept in the literature since the 1990s, Törnberg explains. "In short, it says that social media causes us to isolate ourselves in groups of like-minded people. Your own opinion echoes around you, so to speak, and there is no exposure to other ideas. This causes polarization and makes our ideas more extreme."
But there does not appear to be much convincing evidence for this. Says Törnberg, "It starts with the fact that people usually don't feel like they are in an echo chamber." Research confirms that view, says Törnberg. "We can't find those echo chambers on mainstream social media, such as Twitter and Facebook. According to the idea of echo chambers, isolation causes polarization, and contact with "the other side" causes moderation. The latter doesn't seem to happen. Seeing a crazy tweet from a political opponent does not cause you to re-evaluate your opinion or like that person more. Instead, it evokes negative feelings and causes people to lash out at each other.
"Contact between dissenters on social media is often negative, while contact within a group is more positive. It is precisely the exposure to other groups combined with conflict that fuels that polarization." According to Törnberg, academics often view social media as a place where people exchange high-quality, rational arguments. "That's not what it looks like in practice, of course. Social media is more about conflict than rational debate. It's more about emotion and social identity than opinions and arguments."
In his research, Törnberg summarizes previous research on polarization with mathematical formulas. "In that model, I simulate people-actors in the model and their interactions with each other. Observations from previous real-world research are assumptions in my model. For example, I make the assumption that similar people influence each other more strongly and that because of social media, we have more interactions outside our local social network. With the model, you can experiment and see if you can mimic reality. For example, what happens if you change the possibilities to communicate between actors?"
Social media puts people in contact with others who are physically farther away from them but still belong to the same group. At the political level, this splits or "sorts" society into left and right. "This works like a vortex," says Törnberg. "The stronger the political identity becomes, the more issues become involved: what you eat, what car you drive and what your favorite sports team is. It is no longer about differing opinions, but about differences in social identity. All kinds of cultural issues thus become part of your political identity. This is consistent with earlier findings: You are influenced more by people you think are like yourself. Consequently, you gradually get an us vs them situation. It creates a vicious circle of conflict and polarization."
"In the model, you see that when actors communicate mainly locally, as it was before the advent of social media, that different clusters emerge. Within those clusters you still have people from the left and right, but they also belong to another group, for example, their soccer club." This then forms not just a political but also a local identity. "When you add long-distance communication to the model, those local clusters disappear. Here you see that vortex coming back that brings everything together. Within the two groups, people are increasingly the same."
Lord of the Flies
In his research, Törnberg proposes a new metaphor to replace the echo chamber idea. "We should see social media not as a collection of echo chambers or bubbles, but more like the island in Lord of the Flies." In that book, a group of British boys are stranded on a desert island. There, they soon form groups with their own identities. "Like the Internet, the island is a place where groups are pitted against each other and conflict reinforces group identities, which in turn increases conflict. So you end up in a vicious cycle."
But how, then, can polarization be reduced? "A society with smaller local clusters also has conflict, but they keep each other in balance because people have a common denominator. That disappears if all aspects of cultural life are divided into two groups; you are then left with just the opposing political identities."
Currently, social media is mostly concerned with getting readers to pay attention to content. The biggest conflicts get the most exposure. More attention needs to be paid to issues that people agree on or think similarly about between groups, for example, by getting political dissenters of the same religion to see that they share the same core values."