Scientific American
Social media is changing the character of our political conversations. As many have pointed out, our attention is a scarce resource that politicians and journalists are constantly fighting to attract, and the online world has become a primary trigger of our moral outrage. These two ideas, it turns out, are fundamentally related. According to our forthcoming paper, words that appeal to one’s sense of right and wrong are particularly effective at capturing attention, which may help explain this new political reality.
It occurred to us that the way people scroll through their social media feeds is very similar to a classic method psychologists use to measure people’s ability to pay attention. When we mindlessly browse social media, we are rapidly presenting a stream of verbal stimuli to ourselves. Psychologists have been studying this issue in the lab for decades, displaying to subjects a rapid succession of words, one after another, in the blink of an eye. In the lab, people are asked to find a target word among a collection of other words. Once they find it, there’s a short window of time in which that word captures their attention. If there’s a second target word in that window, most people don’t even see it—almost as if they had blinked with their eyes open.
There is an exception: if the second target word is emotionally significant to the viewer, that person will see it. Some words are so important to us that they are able to capture our attention even when we are already paying attention to something else.
We thought that perhaps moral words might be special in just this way, grabbing our attention even when it was allocated elsewhere. Indeed, we have previously found that people seem highly attuned to moral stimuli in the world around them. This result could help explain why people share certain moral content online: the words that appeal to our morality break through the noise on social media.
To test this idea, we ran a pair of experiments in the lab, using a task that intentionally mimicked Twitter feeds. People were presented with a rapid stream of fictitious tweets with different types of words used for a hashtag. We found that moral words (such as crime, mercy, right), emotional words (such as afraid, love, weep) and moral-emotional words (such as abuse, honor, spite) captured more attention than neutral ones (such as coast, novel, maze). This was even true when people had been focused on a prior term. That is, these moral and emotional words broke through the noise and grabbed subjects’ attention.
We then examined how this finding relates to sharing on Twitter. We analyzed a large data set containing Twitter conversations about political topics (specifically, gun control, same-sex marriage and climate change). These conversations were part of a previous paper we published, showing that messages with moral-emotional words were 20 percent more likely to be shared by another person on Twitter. We then asked whether the words that captured attention in the lab were the same ones that went viral on Twitter.
We analyzed nearly 50,000 Tweets that used words we had studied in our experiments. Some of them strongly captured attention, as measured by our attention task in the lab. Others were hardly noticeable. Most importantly, attentional capture was a robust predictor of sharing messages with these same words online. Moral and emotional words (such as lewd, kill, evil, faith and sin) were the most captivating. When they were embedded in tweets, they were significantly more likely to be shared as compared with neutral words. This was particularly striking because our measure of attention was based on an entirely different group of people than those who retweeted the messages.
