Rough cuts on the incredibly interesting implications of Facebook’s Reactions

Cross-posted to Citizen Technologist

How do we express ourselves in social media, and how does that make other people feel? These are two questions at the very heart of social media research including, of course, the ill-fated Facebook experiment. Facebook Reactions are fascinating because they are, even more explicitly than the Facebook experiment, an intervention into our emotional lives.

Clearly a co-branding opportunity was missed.

Let me be clear that I support Facebook’s desire to overcome the emotional stuntedness of the Like button (don’t even get me started on the emotional stuntedness of the Poke button). I support the steps the company has taken to expand the Like button’s emotional repertoire, particularly in light of the company’s obvious desire to maintain its original simplicity (contrast that with Slack’s more cluttered approach, or the company’s own feeling indicator). But as a choice about which emotional expressions and reactions to officially reward and sanction on Facebook, they are consequential. They explicitly present the company with the knotty challenge of determining the shape of Facebook’s emotional environment, and they have wide implications for the 1.04 billion of us who visit Facebook each day. Here are a few rough reactions to Facebook Reactions.

  • To some extent, Reactions track existing research about emotions that motivate sharing in social media, and the new buttons now allow us to reflect those emotions back to friends in their posts. We have buttons for enthusiasm, amusement and humor (Haha, Love), awe and inspiration (Wow), and anger (Angry). Interestingly, other high arousal emotions thought to motivate sharing—anxiety is a key one—are absent from Reactions, and one low arousal emotion theoretically less likely to motivate sharing, sadness, is present. This suggests either the research is incomplete, and people very often do express and react with sadness (more so than anxiety), or the company has culled the set of supported emotions by other criteria than popularity. I’m guessing it’s a combination of these: that people do frequently express sadness on Facebook and receive high engagement when they do, such as after a death. But that sadness also has perhaps the most glaring cognitive dissonance with the Like button, urging the company to split it off regardless of popularity. We simply cannot “like” our friends’ grief.
  • Facebook’s choices about which emotions to include and exclude will shape the platform’s emotional environment and, thus, users’ emotional experiences on it—just like the Facebook experiment did and just as Facebook’s design choices have always done. Sanctioning four positive emotions with Like, Love, Haha and Wow buttons as well as two negative emotions with Angry and Sad buttons means posts that stimulate those emotional reactions are likely to be better rewarded, while posts that do not are less rewarded. Now that we can explicitly reward angry posts from friends with the Angry button, will there be more anger shared? Almost certainly. Now that users no longer have to fight against the grain of the Like button to express grief and sadness, will we see more posts about grief and sadness? Likely. Facebook has encouraged the mix of emotions to stay positive overall, however, with a total of now four buttons to express positive reactions (even when, arguably, these are the reactions that fit best under the original Like button).
  • Facebook has always clearly wanted to avoid fostering a disparaging emotional environment, which is likely a reason it has fastidiously avoided a Dislike button. It will be interesting, thus, to see if people sometimes use the new buttons sarcastically or disparagingly, and I can imagine people trolling a company or politician with the Sad and Angry buttons (or even a sarcastic Wow). But these are more ambiguously disparaging than Dislike would have been. Responding with Sad or Angry, for example, may unintentionally invite a literal interpretation by the post’s author and, perceiving that they’ve made a friend feel sad or angry, might be empathetic in response and start a dialogue about why there was a negative reaction. This is much less confrontational than a Dislike button—and, arguably, superior design.
  • In Facebook’s announcement, the company hints at a very big question mark: how to rank posts in News Feed with the new information from these buttons. “Initially … if someone uses a Reaction, we will infer they want to see more of that type of post. … Over time we hope to learn how the different Reactions should be weighted differently by News Feed.” This is a thorny issue that puts Facebook’s role in shaping our emotional experiences into sharp relief. If sad posts make us feel sad, but in expressing that reaction, Facebook decides to show us more sad posts which further our sadness, is that a good thing? If we react with anger at others’ angry posts, is it a good thing that Facebook will show us more posts that perpetuate our anger? Beyond the personal implications, what of the political implications? Forget the filter bubble: Facebook’s ranking of News Feed could suck us into an emotional bubble.

Luckily, one thing the company arguably has been in rolling out these new buttons is careful, in a way they often are not. Facebook drew on its years-long relationship with Dacher Keltner, the Berkeley psychology professor who also consulted on Inside Out, to craft its Reactions, and tested the new buttons for a long time before rolling them out today. In addition, by adding a Wow button, Facebook may further the platform as a vehicle for spreading awe, an emotion Keltner’s own research suggests has particular benefits for health and well-being, beyond those of other positive emotions (though, be careful—a Wow reaction could also be the sign of an envy-inducing post).

But here, now, we are in the tricky space where Facebook is explicitly choosing the emotions we may feel and not feel as we interact with News Feed throughout the day. This is no less “manipulative” than the Facebook experiment, and by rolling out globally, arguably more consequential. Choices the company makes in ranking News Feed will now more explicitly affect the amount of amusement, love and awe we experience in our daily lives—but also the amount of anger and sadness. Facebook’s choices will have consequences because emotions have consequences.

Galen is a Ph.D. candidate studying social media behavior at the Berkeley School of Information, and is director of the Berkeley Center for Technology, Society & Policy.