“These aren’t people,” President Trump said Wednesday at a White House meeting with California officials. “These are animals.” In a tweet Friday morning, Trump specified that he was referring to MS-13 gang members and not “Immigrants, or Illegal Immigrants,” in general, and called this “a big difference.” Is it, though? Also, Trump has said many negative things about many groups; why should his choice of term here, regardless of whether it was applied narrowly to a murderous gang, be particularly troubling?
The answer to both questions lies in the distinction between dehumanization — seeing people as animals — and prejudice or dislike. The two are related, but dissimilar enough that different regions of the brain are active, depending on which judgment is being rendered. Dehumanization is a special evaluation of others, and it is consistently associated with hostility or antagonistic behavior. What’s more, people are adept at visiting the sins of another group’s extremists on the group as a whole. If you evaluate MS-13 gang members as not human, for example, you are likely to extend that dehumanization more generally to Central American immigrants and others who cross the border illegally — and to support hostile policies and behavior toward them.
In empirical studies on five continents involving tens of thousands of participants, my colleague Nour Kteily and I have used a provocative way to measure dehumanization: We show people the popular “ascent of man” diagram inspired by Charles Darwin, and we list groups below the diagram and ask people to place those groups along the “ascent of man” scale, from ape to advanced human, to indicate how “evolved and civilized” they judge each group to be. Levels of dehumanization using this scale can be chilling. For example, Hungarians rate the Roma minority group and Muslim refugees nearly 30 points below what they rate ethnic Hungarians on the 100-point scale; in the Czech Republic, the difference is 40 points. In Western Europe and the United States, levels of dehumanization, particularly of Muslim groups, are similarly telling.
What are the consequences of dehumanization? Historically, dehumanization has been associated with the darkest chapters in human history, from slavery and colonization to war and genocide. In the years leading up to the Holocaust, Jews and the Roma were collectively referred to as “Untermensch,” or “subhumans.” In the days before the Rwanda genocide, which claimed the lives of more than 500,000 Tutsis in less than 100 days, the Tutsis were referred to as “cockroaches” on the public radio station.
The intuitively plausible link between dehumanization and hostility is borne out, our research shows. In numerous contexts, the degree to which people dehumanize another group is strongly correlated with their endorsement of hostile policies and their willingness to act with hostility. For example, how strongly Americans opposed the 2015 Iran nuclear accord, and the degree to which they advocated war with Iran, was strongly correlated with how much they dehumanized Iranians. Levels of dehumanization also are correlated with the number of Muslim refugees that Spaniards, Greeks, Hungarians and Czechs think should be allowed into their countries, and how they think these refugees should be treated once they are there. For Hungarians and Czechs, dehumanization of the Roma is strongly associated with their support for particularly nasty policies, including that of providing Roma women birth control that has a high chance of permanent sterilization, a policy that was in place in the Czech Republic as recently as the 1990s.
In all of these studies, the predictive power of dehumanization remained even when prejudice was taken into account. That is, knowing how much people dislike another group tells you only so much about how they will behave toward that group; knowing how much they dehumanize the group allows you to predict their behavior better. We tested this dissociation between dehumanization and dislike directly in a neuroimaging study using fMRI scans. In this study, we provided participants with two different ways to rate groups: how cold they felt toward the groups, and how much they dehumanized the groups. We found that there was a complete dissociation in the brain when making these evaluations: The brain regions that were active when providing judgments of dislike were completely separate from those that were active when providing dehumanizing judgments.
The chief defense of Trump’s blatant dehumanization is that it was qualified. He was not referring to all immigrants but to a subset of extremists crossing the border through Mexico. However, we have found that this distinction becomes vague in the minds of Americans. When we ask Americans (or Spaniards) how much all Muslims are to blame for the acts of Muslim extremists, the average response is about 40 on a 100-point scale; at the same time, when it comes to the acts of white American and Christian extremists, people tend to hold all white Americans and Christians only 10 points responsible on this same scale. And the degree of collective blame was a good predictor of support for hostile policies toward Muslims. A spike in hate speech and hate crimes committed against innocent Muslims occurs in the aftermath of attacks by Muslim extremists.
Trump’s dehumanizing language reflects language that has been weaponized throughout the “illiberal democracies” in Eastern Europe. For example, Zsolt Bayer, the founder of Hungary’s ruling party and a close friend of Hungary’s current president, wrote in a 2013 editorial that a significant portion of the Roma “are not fit to live among human beings. … These people are animals and behave like animals. … Inarticulate sounds pour out of their bestial skulls. … These animals should not be allowed to exist. In no way. That needs to be solved — immediately and in any way.” Bayer’s quotes on Muslim refugees are similarly dehumanizing. He compared Muslim refugees to “lice” and used an analogy to illustrate their threat to Hungarians: “If I send you a nice big bowl of chocolates, and I inform you that I only poisoned ten of them, what would you do?” That analogy should sound familiar: It was adapted by Donald Trump Jr. and used on the campaign trail.
President Trump and many other politicians who use dehumanizing language might claim that their comments are limited to a nasty subset of another group, not the group as a whole, and that their words should not be interpreted as an endorsement of violence. But even if these claims may be rhetorically true, they are psychologically irrelevant. Dehumanization is a virulent psychological process that predicts hostility around the world. Playing with the language is inviting violence against anyone perceived to be associated with the dehumanized. I suppose that matters only if you think the targets of the violence are worthy of moral consideration.