Quotes of the Day

You're in a large group of people when someone makes an edgy remark about African Americans. You glance over at the African-American man sitting across from you to see how he responds. Everyone else does, too. It's natural.

"Think of the Oscar ceremony," says psychologist Benoit Monin. "Someone says something that has to do with race, and you pan to Samuel L. Jackson."

A new report published in the journal Psychological Science finds that when reacting to an ambiguous but potentially racist situation, non-blacks were much more likely to focus on the reaction of an African American than that of whites.

Jennifer Randall Crosby, an assistant professor of psychology at Agnes Scott College in Decatur, Georgia, used eye-tracking technology to follow the gazes of 25 non-black college students while they watched a video. The film depicted four men, three white and one black, all wearing headphones.

One group of participants was told that all of the men in the video could hear one another, while a second group was told that two of the men — the African American included — could not hear the discussion taking place.

In the video one of the white men makes a statement that could be viewed as raciallly offensive: "I think one problem with admissions is that too many qualified white students are not getting the spots they've earned," he says. "These students work hard all through school and then lose their spots to members of certain groups who have lower test scores, and come from less challenging environments. They get an unfair advantage." Crosby deliberately did not include the words "black" or "African American" in the conversation, in order to distinguish glances that could be based on simple association — looking at a black person when "black" is mentioned — from those looking for the African American's reaction.

As might be expected, participants in both groups watched the speaker more than any of the other three people in the film. When they believed that the African American man's headphone was off, they paid little attention to him. Yet participants who thought that everyone in the group could hear the conversation spent five times longer watching the black man in the video.

"We think that people are looking for cues," Crosby says. In ambiguous situations, we look to those most likely to be offended to define what discrimination is.

When Imus mouthed off last year about the women's basketball team from Rutgers, the media looked to African-American intellectuals and female cultural leaders to determine whether his remarks — referring to the young athletes as "nappy-headed hos" — were his standard brand of on-air provocation or if he had in fact crossed the line into racism.

In 2002, when then Senate Majority Leader Trent Lott suggested that if separatist candidate Strom Thurmond had been elected President in the 1948 election, "we wouldn't have had all these problems over all these years," we turned to prominent members of the black community to determine how far he had overstepped.

"There was a lot of conversation about whether he was a racist, or whether was he just ignorant," says Crosby. The ensuing cultural discourse, and subsequent condemnation of Lott's comments, fascinated Crosby, and prompted her research. "How do we figure out what is discrimination?" she asked. More blatant offenses or extreme examples such as hate crimes are easier to determine. Crosby, however, wanted to home in on the nuanced and ambiguous circumstances more common in everyday life.

In her latest research, currently under peer review, she asked students to fill out questionnaires about scenarios in which discrimination was possible but not explicit. A company remains open on Martin Luther King Day, or a police officer stops a black man whose clothing and hair match those of a crime suspect, for example. "They're ambiguous because we want more information," Crosby says. "If it's an ambulance company, you might want it to stay open," or if the person in question is actually the criminal, you would want him to be stopped.

Participants were asked to rate the discrimination on a scale from 1 to 9, where 1 represented no discrimination, and 9 definite discrimination. To gauge how people influence one another's views of discrimination, she made the questionnaires appear as if previous participants had filled out their answers on the same page. "When faced with responses attributed to a white individual, people averaged 4.4 when whites said the items weren't discrimination and 5.2 when whites said the items were discrimination. When the same responses were attributed to a black individual, the means were 3.3 and 6.1, respectively — a significant move from the baseline," she explains.

Consistent with her previous findings, non-blacks' assessments of the situations were strongly affected by whether African Americans had supposedly answered before them.

Crosby worries that this deference may mean we don't trust our own instincts when deciding what is offensive. As Monin, one of the study's co-authors, says, "That's great, of course, that downtrodden groups have a voice," but it also means that too often we may be leaving the responsibility for confronting discrimination in the hands of those discriminated against.

Crosby recalls an example of this from her undergraduate career at Stanford. The school's sports teams were called the Indians from 1930 to 1972, when the name was dropped because of protest from Native American students. Still, from time to time the former mascot would appear on t-shirts and paraphernalia — and each time it fell to Native American students to bring up their objections to the administration, Crosby says. "Why is it always their job?" she asks.

• Tiffany Sharples
• A new study shows that during a racially charged conversation, people will look for cues from people of the race being discussed.