The Weight of Our Words

Apr 13, 2018
Originally published on August 13, 2018 11:20 am

This week, we look at the language we use around race and religion, and what it says about the culture we live in.

In 2014, two shooting sprees occurred six months apart in busy American cities. They had uncanny similarities: in both events, multiple people were shot and killed, including two police officers. Both ended with suicides, and both involved ominous, anti-government messages left on social media. But one crime received nearly five times as much coverage as the other one. The key difference? One shooter, Ishmael Brinsley, was Muslim, and the other shooters, a white supremacist couple named Jerad and Amanda Miller, were not.

New research finds that print journalists are about four times more likely to report on terror attacks when the perpetrator is Muslim. In fact, according to Erin Kearns and colleagues at Georgia State University, "a perpetrator who is not Muslim would have to kill on average about seven more people to receive the same amount of coverage as a perpetrator who's Muslim."

But there are disturbing implications for the way Americans perceive Muslims, and the way Muslims perceive themselves. As we'll discover, reporters (and all of us, of course,) are battling psychological biases about members of groups who are unfamiliar.

"These biases are not just simply something that's in our head," social psychologist Muniba Saleem tells us. "But they are in fact affecting our behaviors and our public policy decisions towards Muslims."

We'll also look at how President Trump's rhetoric is changing the way Americans talk. It's based on the work of Chris Crandall, a researcher at the University of Kansas who studies prejudice. In 2016, he found that Trump's disparaging comments about Muslims, Mexicans, and even overweight people may have altered basic social norms.

"The election changed people's notion of what was tolerable," says Crandall.

This week's episode was produced by Rhaina Cohen, Lucy Perkins, and Maggie Penman. Our team includes Jenny Schmidt, Parth Shah, Thomas Lu and Laura Kwerel. Our supervising producer is Tara Boyle.

Copyright 2018 NPR. To see more, visit http://www.npr.org/.

SHANKAR VEDANTAM, HOST:

This is HIDDEN BRAIN. I'm Shankar Vedantam. There's a phrase politicians often use to describe certain acts of violence - radical Islamic terrorism.

(SOUNDBITE OF ARCHIVED RECORDING)

PRESIDENT DONALD TRUMP: Radical Islamic terrorists are determined to strike our homeland - as they did on 9/11, as they did from Boston to Orlando to San Bernardino.

(SOUNDBITE OF MUSIC)

VEDANTAM: In a speech in 2017, President Trump criticized the media for failing to report accurately on terrorism.

(SOUNDBITE OF ARCHIVED RECORDING)

TRUMP: It's gotten to a point where it's not even being reported. And in many cases, the very, very dishonest press doesn't want to report it.

VEDANTAM: Is the president correct? Has the media held back in its coverage? Is political correctness keeping us from grasping the true danger we face? This week, we step away from the politicians and the pundits to look at the empirical evidence - social science research into how the American media actually cover terrorist attacks. We will also look at what effect that has on our perceptions of terrorism and our attitudes toward the Muslim community. New research has found that there are indeed systematic biases in coverage - but not in the way President Trump suggests.

ERIN KEARNS: A perpetrator who is not Muslim would have to kill on average about seven more people to receive the same amount of coverage as a perpetrator who's Muslim.

VEDANTAM: In the second half of the show, we look at another side of the political correctness question - about where we draw the line between what's free speech and what's hate speech. Today on HIDDEN BRAIN - what our language says about our politics.

(SOUNDBITE OF MUSIC)

VEDANTAM: In 2014, two terrorist attacks occurred six months apart. They had eerie similarities. In both cases, two police officers were shot and killed. In both cases, a third victim was shot as well. Both ended with the perpetrator's killing themselves. In the aftermath of the attacks, investigators learned of criminal records, missed red flags and anti-government threats on social media. The first incident occurred in June 2014 in Las Vegas. It was carried out by a husband and wife team.

(SOUNDBITE OF ARCHIVED RECORDING)

UNIDENTIFIED REPORTER #1: We learned this afternoon the identities of a Las Vegas couple who ambushed and killed two police officers and gunned down a civilian who tried to stop them.

VEDANTAM: The couple was Jerad and Amanda Miller.

(SOUNDBITE OF ARCHIVED RECORDING)

UNIDENTIFIED REPORTER #2: Investigators described Jerad and Amanda Miller as anti-government and on a mission to kill police officers.

VEDANTAM: On the day of the shooting, Jerad posted this on Facebook - the dawn of a new day, may all our coming sacrifices be worth it.

(SOUNDBITE OF ARCHIVED RECORDING)

UNIDENTIFIED REPORTER #2: The Millers then covered the bodies with a swastika and a Gadsden flag, first used in the American Revolution, emblazoned with the words don't tread on me.

VEDANTAM: They left a note on the bodies of the slain officers. It read this is the beginning of a revolution. The couple then walked to a nearby Walmart and ordered customers out. When they were challenged by one person, they killed him too. The rampage ended when the couple were confronted by authorities. Amanda shot and killed her husband and then killed herself. That was the first attack.

(SOUNDBITE OF MUSIC)

VEDANTAM: The second one that we'll examine today occurred later the same year in late December. Before dawn on the morning of December 20, Ismaaiyl Brinsley let himself into his ex-girlfriend's apartment in a Baltimore suburb. He pointed a gun at his own head and threatened to kill himself. When she talked him out of it, he shot her instead. She was injured but survived. Ismaaiyl then got on a bus to New York where he shot and killed two police officers in Brooklyn. In a note he'd posted to social media earlier, he said he was intent on carrying out retribution for police killings of unarmed black men.

(SOUNDBITE OF ARCHIVED RECORDING)

UNIDENTIFIED REPORTER #3: Police say one of the last posts he put on social media was this. I always wanted to be known for doing something right, he said. But my past is stalking me, and my present is haunting me. The post followed with another ominous warning. I'm putting wings on pigs today, he wrote. They take one of ours. Let's take two of theirs.

KEARNS: On balance, these two incidents are fairly similar in that they've both killed two police officers. So we should expect more coverage because of the targets, because of the fatalities.

VEDANTAM: This is Erin Kearns. She's a criminologist at Georgia State University. Erin and her colleagues have been studying the way that the media cover terrorist attacks and the amount of coverage different incidents receive.

KEARNS: The Millers actually killed an additional person, so you'd expect that if anything that might have a little bit more coverage.

VEDANTAM: But that isn't what Erin and her co-authors found.

KEARNS: Ismaaiyl Brinsley received about 4 1/2 times more coverage than the Millers, and he was Muslim.

(SOUNDBITE OF MUSIC)

VEDANTAM: The fact that Ismaaiyl Brinsley's case received so much more coverage than Jerad and Amanda Miller's could be explained by a number of factors. Perhaps it had something to do with the fact that his attack was in New York City while the Millers were in Las Vegas. New York is a bigger city, the center of the news media. Concerns about Mayor Bill de Blasio's relationship with the New York Police Department may have added fuel to the fire. The story came at a time when race and policing was often in the news. But Erin Kearns and her fellow researchers have found there is another factor that determines which attacks catch the attention of the media and which don't.

KEARNS: When the perpetrator is Muslim, you can expect that attack to receive about 4 1/2 times more media coverage than if the perpetrator was not Muslim.

VEDANTAM: Put another way...

KEARNS: A perpetrator who is not Muslim would have to kill on average about seven more people to receive the same amount of coverage as a perpetrator who is Muslim.

VEDANTAM: Perhaps this is not a huge surprise to you. If you've watched coverage of a recent terrorist attack, you might have a sense that this was the case.

KEARNS: The results themselves I don't think were very surprising to myself and to my colleagues. The magnitude of the findings though is something that we were all taken a little bit aback by - that it's such a drastic difference in coverage when the perpetrator is Muslim.

VEDANTAM: Erin and her colleagues did not include broadcast journalism in their data set. In other words, this analysis does not encompass the dramatic coverage on cable news - coverage like this.

(SOUNDBITE OF NEWS MONTAGE)

UNIDENTIFIED REPORTER #4: U.S. officials believe Farook's wife, Tashfeen Malik, had been radicalized before stepping foot in the U.S., raising alarm bells about the fiance visa she came in on.

UNIDENTIFIED REPORTER #5: And as we've been reporting, law enforcement officials identifying this suspect as 29-year-old Omar Seddique Mateen, as we've said, a U.S. citizen.

UNIDENTIFIED REPORTER #6: We should also point out that the New York Daily News has been on a jihad against conservatives over the last couple of weeks.

VEDANTAM: The researchers looked mostly at print news sources, like The New York Times, The Washington Post, The Wall Street Journal, local papers from around the country, as well as cnn.com. Ironically, it's these mainstream media outlets that are routinely accused of political correctness in their coverage of terrorism. The researchers studied these sources for their coverage of terrorist incidents within the United States. They looked at a five-year period from 2011 to 2015. To identify cases to study, the researchers used a data set called a Global Terrorism Database.

KEARNS: The way that terrorism is defined within the Global Terrorism Database is talking about the threat or use of violence, the incident having a political, religious, social or economic motive, being committed by a non-combatant and with the goal being, you know, for fear, to coerce or intimidate a population.

VEDANTAM: Jerad and Amanda Miller were self-identified white supremacists. They were affiliated with a far-right group. They intentionally sought out and killed police officers and said it was the beginning of a revolution. This would fall cleanly onto the Global Terrorism Database's definition of terrorism. But Erin and her co-authors have noticed that the media, and in turn the public, do not apply the terrorism label evenly. That was the case with another attack that same year.

KEARNS: After the Frazier Glenn Miller attack in Kansas back in 2014...

(SOUNDBITE OF ARCHIVED RECORDING)

UNIDENTIFIED POLICE OFFICER: Four or five shots have been fired into the front door. There's a male with a shotgun.

KEARNS: ...Where Frazier Glenn Miller, who was a former leader of the Ku Klux Klan, killed three people at a synagogue, yelled Heil Hitler at the end of the attack, this wasn't described as terrorism very commonly in the media.

(SOUNDBITE OF ARCHIVED RECORDING)

UNIDENTIFIED REPORTER #7: There's a significant circumstantial case here pointing to a possible hate crime because Miller's history as an avowed white supremacist...

KEARNS: Which led one of my co-authors, Dr. Anthony Lemieux, to write a piece then about why we aren't applying this label to this particular attack even though it clearly fits within it.

(SOUNDBITE OF MUSIC)

VEDANTAM: To examine how people draw conclusions about which cases of terrorism should be labeled terrorism, the researchers conducted a study.

KEARNS: We have presented participants with real-life terrorist attacks, and what we've found is that when the perpetrator of those attacks was Muslim, people were much more likely to consider it to be terrorism than when the perpetrator was not Muslim. In those cases, people were more likely to say that perhaps it's a hate crime or not be sure how to classify it.

VEDANTAM: The problem with studying actual incidents of terrorism is that each one has many idiosyncratic features. It's difficult to tell whether differences in perception and coverage in cases such as the Jerad and Amanda Miller shooting and the Ismaaiyl Brinsley shooting are because of the identity of the perpetrator or some other factor. To address this problem, the researchers conducted an experiment. They controlled for all sorts of different factors. Volunteers were given descriptions of fake terrorist attacks, including the location, the number of casualties and other details. Holding everything constant, volunteers saw the cases differently when the perpetrator was Muslim.

KEARNS: What we found here, again, is that even if the target's the same, the weapon is the same, if the perpetrator is Muslim, the participants are much more likely to consider that to be terrorism.

VEDANTAM: The tendency not to give some cases of terrorism that label and to cover other cases more intensively shapes the way we all think about the phenomenon. Erin and her colleagues looked at how many terrorists attacks in the United States are actually carried out by Muslim extremists. The result might surprise you.

KEARNS: So if we look at these attacks in this five-year window, we see that only about 12 percent of them were perpetrated by Muslims whereas over 50 percent actually were perpetrated by some far-right cause. But most people don't perceive that as being what the actual threat is.

VEDANTAM: To be clear, that 12 percent number is disproportionate. Muslims account for just 1 percent of the U.S. population. But in a rational world, this should mean that 12 percent of the media's coverage of terrorism would be of terrorism committed by Muslims. When we come back, we'll hear about some of the psychological reasons this doesn't happen. Stay with us.

(SOUNDBITE OF MUSIC)

VEDANTAM: This is NPR.

(SOUNDBITE OF MUSIC)

VEDANTAM: This is HIDDEN BRAIN. I'm Shankar Vedantam. A recent study found that news coverage of terrorism disproportionately focuses on Muslim perpetrators. Research conducted by my next guest helps us understand why there's a gap between what we see in the media and reality.

MUNIBA SALEEM: My name is Muniba Saleem, and I am an assistant professor at the Department of Communication Studies at University of Michigan.

VEDANTAM: Muniba Saleem studies how the media cover terrorism. She's also interested in how the media can change the way we relate to one another. Her interest in these subjects goes back to her own childhood. Muniba, who is Muslim, was in high school in Ohio when the September 11 attacks occurred.

SALEEM: As even a high school student had to answer questions regarding, you know, why did this individual think that this was the right thing to do, or what does my faith say about these kinds of actions? And to be honest, at that point, I'm not sure if I even knew what to say or how to answer these kinds of questions. But I did have to answer them nonetheless.

VEDANTAM: As a teenager, she struggled to understand the shift in the way others thought of her.

SALEEM: The fact that I was a Muslim was enough to put me in the same category as these perpetrators who had committed these terrible crimes. And it didn't matter that I was a teenager or that I was an American or that I was Pakistani. What mattered was that Muslim identity. That seemed to kind of overencompass (ph) all of the other information that people had about me. And so I was asked questions regarding, you know, what is Islam? What does it say about these acts of terrorism? And do you believe in the kinds of things that these people did? Do you sympathize with what they have done? Do you hate Americans? Are you anti - I mean, all of these series of questions that I had honestly never faced before.

VEDANTAM: Muniba remembers one instance several years later where she was watching a play, and there was a joke about American politics.

SALEEM: The entire audience laughed in the theater, and I'd laughed as well, as did my friends and this couple that was sitting next to me. But then I noticed that they were very aware of the fact that I was laughing. And during the intermission, they made this comment where the lady turned around and she said, you know, it wouldn't hurt you to be more supportive of America. And I thought about why was I being pointed out for laughing as opposed to everybody else in the audience who also thought that that was a funny comment?

VEDANTAM: As a social psychologist today, Muniba Saleem understands many things she did not understand as a 15-year-old kid in Ohio. Terrorism doesn't just have physical consequences. It has a number of psychological effects. For one thing, simply reminding people of death, as September 11 certainly did, can change the way people think.

SALEEM: It did a lot of things. It increased, you know, this concept of mortality, which psychology shows that whenever that happens, we also then tend to have more in-group cohesion, which basically means that we tend to kind of stick to our in-groups. We tend to be more patriotic. We tend to support things that are of our nation a little bit more. So it provided a lot of those kinds of psychological effects as well.

VEDANTAM: Muniba also came to understand why many Muslims were seen with suspicion in post-9/11 America. It has to do with something called salience. When you look out at the world, certain details seem to pop out at you. They are more salient than others.

SALEEM: We generally like to think about ourselves and the groups to which we belong in a more positive manner because it makes us feel good about ourselves. So any group there we are not a part of, that's what's referred to as our out-groups. And so that information is always going to be examined a little bit more closely and scrutinized a little bit more closely. So what that means is as a woman, for me, the behavior of other women is perhaps, especially when it's negative, is not going to be as salient as the behavior of another man. So one of those key elements is the fact that for a lot of our American audience who are non-Muslim, that Muslim identity was salient. The second part, of course, is that the perpetrators had claimed that they were doing this in the name of Islam. And so that identity or that label became very salient in people's minds.

VEDANTAM: You might ask why some identities become more salient than others. For example, the majority of terrorist attacks are carried out by men. But while it's very common for Muniba to be questioned about why Muslims commit acts of terrorism, no one has ever asked me to explain why men become terrorists. Muniba says that has to do with who's in the majority and who's not.

SALEEM: Think about it as a jar of red marbles. If you only have a single blue marble within that jar, then wherever that blue marble is moving, that's going to become very salient to you because in that jar of red marbles, that blue marble sticks out a little bit more. And so I think there's a little bit of that happening in the way in which we oftentimes encounter news stories that are referring to racial and ethnic minorities.

VEDANTAM: Of course, the 9/11 attackers explicitly said they were acting in the name of all Muslims. They didn't say they were acting in the name of all men. But Muniba's point is that because Muslims are a small minority and men are not, the misdeeds of Muslims become salient to us.

(SOUNDBITE OF MUSIC)

VEDANTAM: This also holds true for other minorities and other crimes. The researcher Shanto Iyengar at Stanford University once conducted a study where he presented stories about local crime to volunteers. He found that when volunteers saw that an African-American perpetrator was responsible for a crime, volunteers tended to extrapolate the stigma of criminality to African-Americans as a group. When a white American was responsible for an identical crime, volunteers typically saw the criminal as being an individual - an aberration rather than the rule. Terrorist attacks also trigger another psychological phenomenon whereas normally we might see that a group of a billion people has enormous diversity, all kinds of different attitudes and political views, terrorism causes many Americans to view the entire Muslim world as fundamentally homogenous.

SALEEM: The idea behind that phenomena is actually referred to as out-group homogeneity, and all that is saying is that when the behavior or when the action is involving an out-group the group that you don't belong to, then you think that all individuals within that group are in fact represented by that particular idea or that particular attribute. So you think that the entire out-group is homogeneous, and they are all the same. They all talk the same. They all behave the same. So if one of them did something, then others must believe and act in the same way as well.

VEDANTAM: The way the media portray terrorism has serious effects not just on our perceptions of Muslims but on the public policies we support. In one study Muniba conducted, volunteers were randomly assigned to watch different clips of Muslims before answering a series of questions. Some volunteers watched news clips in which Muslims were represented as terrorists. Others saw neutral clips about Muslims. A third group saw positive news clips - clips that showed Muslims volunteering in their communities.

SALEEM: And immediately after being exposed to the video clips, we asked them what they think about Muslims in terms of how aggressive they are, how violent they are. But we also asked them whether they support various kinds of public policies that are targeting Muslims. And some of these were policies that were more for international countries - so military action in Muslim countries. And other policies were more domestic - so harsher civil restrictions for Muslim Americans.

And what we discovered is that participants who were in the negative news condition ended up thinking of Muslims as more aggressive, as more violent and subsequently supported policies that were targeting Muslims both domestically and internationally compared to those who were in the neutral or positive news condition. So continuous coverage of Muslims as terrorists is simply activating and strengthening these kinds of associations - ultimately facilitating us to think of Muslims in an aggressive manner and then support harmful behavior towards them.

VEDANTAM: What was striking about this paper was not just that people supported aggressive foreign policy interventions, but they actually supported more aggressive policies toward American citizens who happened to be Muslim.

SALEEM: That was actually fascinating for me to know - that a lot of Americans are not differentiating between Muslims who are living in other countries versus Muslim-Americans who are citizens and who are perhaps their neighbors and their friends. So we saw people supporting policies such as domestic surveillance without the consent of Muslim-Americans. We saw individuals supporting perhaps that Muslim-Americans should not be allowed to vote, that they should have separate and more thorough airport security lines, that they should be monitored, you know, by the government. Their phones should be tapped without their consent. All kinds of unconstitutional policies, we saw Americans supporting those.

VEDANTAM: So when I look at this body of research, what I'm seeing, when I step back and look at it, is a series of processes that in some ways are driven by, you know, fairly understandable and normal, you know, human mental processes - the idea that we see aberrational things or unusual things happening together. We draw correlations between them. We see patterns. And members of the news media, of course, are human beings. And so they gravitate toward certain stories and cover those stories more aggressively because those stories stick out in their minds. They're more salient.

As a result of that kind of media coverage, your research is finding that the attitudes of Americans themselves is changing in a way that becomes hostile toward people living in their communities, who are fellow American citizens who happen to be Muslim. And, you know, what is so troubling is that these relatively innocuous psychological biases can eventually have consequences that are anything but innocuous.

SALEEM: I think what's really important is to realize that these biases are not simply something that's in our head. But they are in fact affecting our behaviors and our public policy decisions towards Muslims - both in this country and outside. And also, the other side of things, which I've actually just recently started looking at, is how the same media representations are influencing Muslim-American youth and adolescents who are growing up in this country who identify both as Muslim and as American.

It's making them feel as if they cannot be both of those identities even though there is no reason why they should feel that way. They're having to feel as if they have to choose one over the other because they're being questioned about their American identity on account of being Muslim. They're having to answer questions about their loyalty, their patriotism as an American. And that is in fact affecting how they think about themselves psychologically, their physical health, their self-esteem and a host of other important consequences.

VEDANTAM: In some ways, these findings are disheartening. It does feel like a vicious cycle. I asked Muniba if she saw any way to combat the psychological biases that terrorism produces.

SALEEM: I think one of the greatest ways to break this cycle - and this is supported by both anecdotal evidence but also research - is contact. Increased contact with Muslims - of course, especially when it's positive - tends to decrease reliance on these kinds of biases and then subsequently, you know, support for these kinds of harmful actions. So time and time after again, we see that those Americans who have more contact with Muslims and more frequent contact tend to report less of these kinds of biases, tend to not support these kinds of policies as much. They tend to look at the media reports in a more critical manner, and they have a very good understanding that what they see in the media is not the entire story. So contact is definitely one of the most important things that we can talk about, and we can encourage individuals to pursue in order to break this vicious cycle.

(SOUNDBITE OF MUSIC)

VEDANTAM: When we come back, how we don't always mean what we say, especially when we're defending free speech. You're listening to HIDDEN BRAIN. This is NPR.

(SOUNDBITE OF MUSIC)

VEDANTAM: This is HIDDEN BRAIN. I'm Shankar Vedantam. In this segment, we're going to be talking about race and free speech. If you have small kids with you, please know that this segment includes a racial epithet.

You often hear a popular claim in the United States - we are too politically correct. Some pundits say that political correctness gets in the way of accurately reporting on the threat of terrorism.

(SOUNDBITE OF MONTAGE)

MICHAEL FLYNN: Political correctness kills. It causes - it will cause death.

JAMES WOOLSEY: You can't fight something effectively that you can't talk about.

UNIDENTIFIED PERSON #1: You know, we're now at the point where PC is silencing people, and it can have lethal consequences.

VEDANTAM: Others say political correctness infringes on free speech.

(SOUNDBITE OF NEWS MONTAGE)

UNIDENTIFIED PERSON #2: Is hate speech protected under the Constitution? In a word, yes.

UNIDENTIFIED PERSON #3: There's a fine line between free speech and hate speech.

UNIDENTIFIED PERSON #2: Are they not familiar with the First Amendment?

UNIDENTIFIED PERSON #4: But hold up, we can't have that conversation or any conversation until we both agree that we have the right to say what we believe and that we're not going to be punished for doing that. And I see...

VEDANTAM: After the violence in Charlottesville in the summer of 2017, these questions were top of mind.

(SOUNDBITE OF MONTAGE)

UNIDENTIFIED PROTESTERS: (Chanting) White lives matter. White lives matter.

UNIDENTIFIED REPORTER #8: We begin with the deadly chaos on the streets of Charlottesville, Va.

UNIDENTIFIED REPORTER #9: It was a weekend of street battles and stark displays of racism.

UNIDENTIFIED PROTESTERS: (Chanting) Jews will not replace us.

VEDANTAM: Photos of enraged white men using the Nazi salute and marching with torches shocked many Americans. And since then, many of us have been doing a lot of soul-searching. Who are we? What do we stand for? And as a nation, what do we tolerate?

(SOUNDBITE OF MUSIC)

VEDANTAM: The Constitution upholds the rights of Americans to say almost anything, no matter how distasteful, without censure from legislatures, the police or the courts. Our protections for speech, even hate-filled, vitriolic speech, go further than most nations. In the weeks following Charlottesville, many people made free speech arguments to defend the white supremacists who'd marched. When people defend free speech, what are they really defending? That's a question Chris Crandall has studied. He's a professor of psychology at the University of Kansas. We started by talking about an incident that happened a few years ago, before Charlottesville. In 2015, members of the Sigma Alpha Epsilon fraternity at the University of Oklahoma were caught on video singing a racist song.

(SOUNDBITE OF ARCHIVED RECORDING)

UNIDENTIFIED PEOPLE: (Singing) There will never be a (expletive) in SAE.

CHRIS CRANDALL: They were singing, there will never be a N-word in SAE. And they were singing happily and clapping along about the exclusion of African-Americans from the fraternity. The video of the song was put on Facebook, and it spread around the campus. And the two song leaders of the fraternity who were captured on video were expelled from the university, and the fraternity itself was shut down on the campus of the University of Oklahoma. And there was really a media firestorm about their singing the song.

VEDANTAM: In the aftermath of the incident, one of your graduate students noticed something unusual about the responses to the incident. What did he notice?

CRANDALL: Yeah, Mark White was following this on the Internet, and he looked at the tenor of the comments, and people were saying, well, this is simply free speech; these people have a right to say these things because Americans have the right to free speech. But underneath the surface, in the background or sometimes right at the forefront, it really looked like what people were doing was justifying the content of what was said, not the fact that they have the right to say something freely, but rather that they seemed to be justifying the racist speech itself by giving an account - hey, free speech - that allowed them to do that without punishment.

VEDANTAM: Now, it's worth pointing out, of course, that the First Amendment protects people against government intrusions on free speech. It doesn't actually prevent private companies or universities from deciding what is or isn't acceptable speech in the workplace or on a college campus. But setting aside that distinction, you conducted an experiment where you had volunteers listen to racially charged commentary, and you evaluated whether they would reach for free speech arguments to defend it. Walk me through the experiment.

CRANDALL: Yeah, so in one version of the experiment - and we did several of these - people read about somebody who wrote something on Facebook that was deeply racist. For example, a barista wrote on Facebook that the black customers were problematic, that they were noisy and rude and other racially stereotypic actions. And the people in the study read that the person had been fired for posting this racist speech. In the other condition, people read about a guy complaining about customers, but no racial information was given.

So the same behavior was described, but the racial element was eradicated. And so the question was, do racial attitudes of our participants determine how much they're going to defend the speaker? In the racialized condition, the more you had negative attitudes towards African-Americans, the stronger you endorsed free speech as a justification for why the person should have been able to say that without being fired. So that suggests that racial attitudes might be behind free speech defense.

But you might say quite easily, that's simply a correlation; I'm not impressed; maybe people who have negative racial attitudes are also libertarians, and they believe that free speech is super important, and they're just simply expressing that. The problem with that argument is when you remove the racial content from the story - so the - in the condition where the guy complains about customers, but there's no racial element - racial attitudes correlate zero with free speech defense. It seems that people pull them out and deploy them when they're appropriate. So people pull out free speech as a defense when they're defending racist speech but not when they're defending simply aggressive or negative speech.

(SOUNDBITE OF MUSIC)

VEDANTAM: One of the elements of the study that you conducted asked people about their attitudes about police officers. If I understand correctly, you were measuring controversial speech as it was directed towards a minority group versus controversial speech directed against police officers.

CRANDALL: Yeah. In the police version, we were interested in not only attitudes towards African-Americans - so if somebody said something racist - but we also thought that police might be the exact opposite. That is, people who might be very anti-African-American might be particularly pro-police. So we thought if we used the police, maybe you'd get the opposite effect. And we did. It wasn't as strong, but we did find that the more a person had negative attitudes towards African-Americans, the less they defended anti-police speech, although in the other condition, the more they defended the free speech of racist commenters.

VEDANTAM: So in other words, if the speech is basically anti-black, someone with high racial animus might say, that speech is permissible under First Amendment free speech grounds. But if the speech is anti-police, the person might say, I don't really agree with this person expressing these kinds of views. So the person is inconsistent in their defense of free speech.

CRANDALL: That's right. This is what we find - is that people are really inconsistent.

VEDANTAM: You also found in the experiment, Chris, that it wasn't just people with racial biases who were sometimes hypocritical when it came to free speech issues. You also found that people who were low in racial bias were also hypocritical, just in the opposite direction.

CRANDALL: Yes. People who were low in racial prejudice were just as inconsistent in applying the First Amendment as a defense than people who were high in racial prejudice. When they were asked to use the First Amendment to defend racist speech, they actually reduced their willingness to do it. They would defend the free speech of somebody who was complaining about their customers or police at an average level. But when it became racialized, the low-racist people actually walked away from the First Amendment defense. They went substantially lower.

This suggests that when people are using the free speech defense, they know what they're doing because the high racists use it in defense of racist speech, and the low racists drop it like a hot potato and say, no, no, no, no, no, we're not defending this speech with First Amendment. But they would defend a barista who's complaining about his or her customers or the police.

VEDANTAM: When you found people defending the racist speech on First Amendment grounds, on freedom-of-expression grounds, is this because people may have, at some level, felt bad about themselves, Chris? They felt bad that the person who has views similar to their own is being punished in some way, and they're trying to defend that person as a way to defend themselves?

CRANDALL: That's exactly what we thought was going on. And when we started showing these studies, everybody would nod when we'd say, and so maybe they're defending themselves. But we did several studies where we tried to find out if people's self-esteem was attacked by this, if they needed help to feel good about themselves, and we could never find evidence that people were defending themselves. What we found evidence for was that people were defending their right to say things. They were defending the speech so that their future speech would be protected. They wanted to create a world where this kind of prejudicial speech was acceptable for them to say, for others to say, in the future.

VEDANTAM: How would you respond to people who might say, you know, let's look under the hood of Chris Crandall's brain? Isn't it possible he's just, you know, a lefty academic who's coming up with research findings that endorse his own pre-existing views of how the world works?

CRANDALL: I think that some of that criticism would probably land. I am sort of a lefty academic. I'm a little bit more centrist, I suppose. But we were interested in studying prejudice because prejudice is a particular social problem. It shows up all around the country. It affects people's lives. And so we studied that.

But people would be inconsistent about, for example, being environmentalist and failing to recycle, or going too fast in the right-hand lane of a highway. People are inconsistent all the time, and they manage to get over that pretty easily, mostly by being unaware of it or not paying attention to it too much.

(SOUNDBITE OF MUSIC)

VEDANTAM: Not only are we inconsistent, we take cues from how other people behave. To explain this, we need to take what might seem like a bit of a detour to talk about a toy. If you're a parent, maybe your kid has one of those inflatable punching bag toys - a clown or a shark, or, like the one in this YouTube video, an inflatable Spider-Man.

(SOUNDBITE OF VIDEO)

UNIDENTIFIED CHILD: Whoa, he got me.

(SOUNDBITE OF THUMP)

VEDANTAM: These toys, which are known as Bobo dolls, were used in a famous psychology experiment in the 1960s. The researcher Albert Bandura had kids watch as a person repeatedly hit a Bobo doll with a hammer. I'll let researcher Chris Crandall pick up the story from here.

CRANDALL: Now, what Bandura had his children watch was somebody who came in, went and grabbed a hammer, straddled the Bobo doll so that it was lying on the ground and then hit it in the face with a hammer saying, punch him in the face, hit him with a hammer. Afterwards, sometimes what he called the model, the person doing the punching, was punished. Sometimes nothing happened at all. Sometimes they were rewarded. The children were then put in a room filled with toys and, over in the corner, a Bobo doll and, elsewhere in the room, a hammer. The kids who had seen the model either have no consequence at all or be rewarded quickly went over, grabbed the hammer, got the Bobo doll into the middle of the room and started immediately pounding it about the face with the hammer and repeating, punch him in the face, hit him with a hammer. The kids who had seen the model be punished did not do this at all.

But here's the trick. Bandura wisely said at the end of this, you know, if you saw something in that video that interested you, you should know you won't be punished for this. At this point, a majority of the kids get a look of glee on their face, run over, grab the Bobo doll, grab the hammer, and start beating it about the face and straddling it and doing what they'd seen the model do. The moral of the story is that we learn from what people do, and sometimes we don't do it. We may want to punch the Bobo doll, but we know that we'll be punished for it. If somebody comes in and says, it's OK, you can use that hammer and Bobo doll, they will gleefully run and grab the tools and start beating the face of the Bobo doll.

(SOUNDBITE OF MUSIC)

VEDANTAM: So what do Albert Bandura and the Bobo dolls have to do with free speech? The link here has to do with social norms and what's permissible and not permissible. Chris Crandall conducted a study in the lead up to the 2016 presidential election. He looked at both liberals and conservatives and asked them a series of questions.

CRANDALL: We were interested in what prejudices people had, and we also wanted to ask them what they thought it was OK to express in the world. So we asked them about Muslims, Canadians, blind people, immigrants, women. And we asked them, how do you feel about them on a zero-to-100 scale? And the lower the number, the less you like them. But other people, we asked them, what do you think is OK to express in the U.S.? So we were able to find out not only what people say they have as prejudices, but what they think is acceptable to express in the U.S.

VEDANTAM: You brought these same people back after the election, and you asked them the same questions. What did they say?

CRANDALL: For both Trump conservative supporters and Clinton, more liberal supporters, we found that they thought the nation had changed substantially in what it was OK to express as a prejudice. For the groups that Trump had actually targeted in his campaign, Muslims...

(SOUNDBITE OF ARCHIVED RECORDING)

TRUMP: I think Islam hates us.

CRANDALL: ...Mexicans...

(SOUNDBITE OF ARCHIVED RECORDING)

TRUMP: They're bringing drugs. They're bringing crime. They're rapists.

CRANDALL: ...Illegal immigrants...

(SOUNDBITE OF ARCHIVED RECORDING)

TRUMP: They beat us at the border. People are flowing through. Drugs are coming across, pouring across.

CRANDALL: ...Fat people...

(SOUNDBITE OF ARCHIVED RECORDING)

TRUMP: Only Rosie O'Donnell.

CRANDALL: ...There was a significant increase in how acceptable it was to express prejudice towards them. The interesting thing is there was no change in the acceptability of prejudice towards groups that Trump had not aimed his prejudice at - blind people, Canadians and so on.

VEDANTAM: What do you draw from the experiment, Chris? I mean, it seems remarkable that in a matter of weeks, you know, from stage one to stage two of your experiment, people's views of what they consider acceptable are changing.

CRANDALL: Yes. It was really only about two weeks, and we don't think that anybody changed their hearts and minds in that time. What we think is that the election of Donald Trump changed people's understanding of what America felt. The election of Donald Trump, despite all of his overt expressions of prejudice, meant that it must be OK in America to have these prejudices. And so their scores on the acceptability of prejudice for the Trump-targeted groups went up, but not the ones he didn't target. The interesting thing about this is that people's own prejudices did not go up following their sense that America accepted it. And in fact, to a small degree, our participants, both Trump supporters and Clinton supporters, said they went down a little bit in prejudice - just a little bit. And what we think is going on is that they looked around, saw the acceptability of prejudice, saw in America how much there really was, more than they thought. And they said to themselves, oh, I'm less prejudiced than I thought I was. I must be OK because there's a lot more prejudice out there than I thought.

VEDANTAM: Did it strike you as remarkable that things changed so quickly? I mean, I think most of us think about social norms as being relatively stable and enduring things. They're the things that are acceptable in a country or a society one day - are probably going to be the same things that are acceptable tomorrow or next week. And you're finding there was a dramatic change. Was it just the scale of the election and the fervor of the election that you think might've made the difference. Or do you think that actually social norms actually are very malleable, and they can change fairly quickly?

CRANDALL: You're right that most norms are stable, especially ones that have been around a long time and are as public as prejudice norms in the U.S. But this is a little bit different. First of all, the presidential election and coverage of Trump and prejudice was massive. The mass media and personal discussion over Trump and his discussion of racial groups, immigrant groups, fat people and so on was a huge intervention. Second of all, people were surprised that the American public approved of it by electing him. Now, of course, it wasn't the popular vote. But still, winning the presidency alone is likely to have an effect on social norms.

But one key difference with this social norm is that there is a large number of Americans who are suppressing their prejudice, who were holding it back, who wouldn't say what they really felt because they knew it would be punished. This is what is meant by political correctness. People have attitudes that are negative, but they don't say them out loud because they know that they're unpopular. The election of Trump removed the suppression. It was a key that opened up the floodgates just a little bit for people who had been suppressing their feelings a lot. And that's why the number of hate crimes seems to have jumped right after the election - not before when the speech was all there, but after when the nation seemed to approve of Trump's prejudices.

VEDANTAM: So it's almost, you know, a parallel - this eerie parallel to the Bandura study because what you're saying is it's not enough just to see the model strike the Bobo doll with the hammer. That isn't what actually causes the children to then imitate the adult. It's actually seeing what happens to the model after he or she does that to the Bobo doll. Does the model get rewarded? Does the model get punished? It's that action that then determines whether the children imitate the model or don't imitate the model.

CRANDALL: That's exactly right. It's not so much what's in your head and heart as it is you looking around and seeing what's acceptable, seeing what's OK, seeing what people will tolerate. And the election changed people's notion of what was tolerable.

VEDANTAM: I'm wondering if this research has any connection with some other work that you've done looking at the question of authenticity. You know, you found that we all crave politicians who are authentic, and certainly that was one of the big appeals of Donald Trump. But sometimes you've argued or your research has found that authenticity is a way to express more subtle feelings.

CRANDALL: Yeah. People said that Donald Trump was very authentic, and we wondered - there's really two ways that that could go. One is that we think he's authentic because he says things that are unpopular. And by saying things that are counter to the social norms, we see that he's revealing something about himself that's different from every other person. To say that you like ice cream is to reveal very little about yourself, but to say that I don't think ice cream is any good reveals quite a lot about you because it's so unusual and so atypical. With the case of prejudice, we wondered if maybe people thought he was authentic 'cause he was saying these horrible things that they thought were horrible. But holy cow, at least he says it.

(SOUNDBITE OF ARCHIVED RECORDING)

TRUMP: I don't frankly have time for total political correctness. And to be honest with you, this country doesn't have time either. This...

CRANDALL: That's one possibility. The other possibility is exactly the opposite. And that is, authenticity is really just code words for saying, you are saying the prejudice that I have. An authentic person is somebody who says what I feel when I can't say it myself.

(SOUNDBITE OF MUSIC)

VEDANTAM: Trump supporters who listen to this critique might say that they're being boxed into a corner. Just because they believe that Trump is authentic or they believe in free speech protections doesn't automatically mean that they're closet racists.

CRANDALL: That's a fair criticism. The reason we are interested in Trump is that we are interested in trying to explain how Americans were able to tolerate his racism. I think that most of us in the social science community were surprised that Trump was elected not because of his politics but because his racism, his attitudes towards immigrants, his anti-fat prejudice, his misogyny didn't seem to stop most Americans or many Americans from voting for him. And that's the question that we're trying to explain. It's not particularly a liberal or a conservative thing to be hypocritical. It's a very human thing to do that. But the public policy implications for prejudice is why we chose to look at Trump and authenticity in this way.

VEDANTAM: Where does the research go from here? And are you still trying to figure out ways to study if social norms are continuing to change as we move further and further from the 2016 election?

CRANDALL: We've been following some of the same people that we followed at the election. And what we have found is that the norms for prejudice are becoming more tolerant of prejudice even beyond what Trump targeted. So the bad news is that it seems that all prejudices are becoming somewhat more acceptable as the course goes on.

VEDANTAM: Chris Crandall is a psychologist at the University of Kansas. Chris, thank you so much for joining me today on HIDDEN BRAIN.

CRANDALL: Thank you, Shankar.

(SOUNDBITE OF MUSIC)

VEDANTAM: This week's episode was produced by Rhaina Cohen, Lucy Perkins and Maggie Penman. Our team includes Jenny Schmidt, Parth Shah, Thomas Lu and Laura Kwerel. Our supervising producer is Tara Boyle. Anya Grundmann is NPR's vice president for programming. For more HIDDEN BRAIN, you can follow the show on Facebook and Twitter, and make sure to subscribe to our podcast. I'm Shankar Vedantam. Thanks for listening to HIDDEN BRAIN. See you next week.

(SOUNDBITE OF MUSIC) Transcript provided by NPR, Copyright NPR.