Part 2 of 2 about the role of social context in how we see people
We all know that long-held cultural stereotypes influence how we evaluate and interact with others. But what about how we actually – visually – see people? New research, for the first time, shows that stereotypes change how we see seemingly objective characteristics such as race and gender and how those characteristics interact in the brain to categorize people.
“One fascinating and distinctive aspect of social perception is that a single face opens up the opportunity for multiple possible perceptions,” says Jon Freeman of Dartmouth College. When you look at a face, you instantly categorize it into a social group by sex, race, age, emotion, social status, and other factors. A growing body of research has found that not only does the brain simultaneously extract these categories but also that they impact each another in systematic and meaningful ways, as Freeman will discuss at the CNS annual meeting next month in Boston.
Gender, race, emotion, and other social categories are all interdependently linked through stereotypes, he says. Freeman and his team have used a computer mouse-tracking technique, paired with brain imaging data analysis, to show that these interlinked stereotypes may actually change how we visually see people at the neural level.
Past research in social psychology has shown that people, at least in the United States, have subtle (often unconscious) biases that link together certain characteristics – for example, seeing men angrier and women as happier and linking African Americans to men and Asians to women. So, when you perceive a male face, Freeman says, it will spontaneously activate stereotypes that are shared with other dimensions like race, which will bias your visual perception. “When looking at an African American face, for example, automatically activated stereotypes that are incidentally shared with the male category — such as aggression — may lead the face to look slightly more masculine.”
Using a computer mouse-tracking technique he helped develop, Freeman has tested these stereotype linkages. In his team’s experiments, participants see faces on a screen and click on, say, a male or female response or a black or white response to categorize what they see. The participants do not know that the researchers are recording their mouse movements, looking at the trajectory of their hand en route to a response.
“What we and others have shown for years now is that you can readily see subtle deflections in their mouse movements, indicating a tentative shift in perception,” Freeman says. For example, if there are subtle feminine cues in a man’s face, participants will gravitate a little bit to the female category before identifying the face as male. “They do not seem to be particularly conscious of this subtle shift in perception, but when we average across many trajectories, we observe their slight bias to partially interpret the face to be a bit closer to female,” he says.
They have shown these patterns time and time again with various visual cues. For example, in a recent study, when they showed low-status janitor attire around a white face versus high-status business attire, people would subtly deflect their mouse toward a black response, even though they ultimately categorized the face as white. The same exact face, Freeman says, can rapidly, within tenths of a second, trigger stereotypes, in this case stereotypical associations between race and low status in the United States. With this technique, they are now finding the same kinds of links between gender, race, emotion, and other social categories that have been shown in past research.
In new data, not yet published, Freeman’s team has found that the neural patterns of how we categorize people overlap to the same extent as people’s subjective categorizations of faces via the mouse-tracking studies. Each social category is distinctly represented in the brain through a neural pattern in the fusiform cortex, a region involved in face perception, and the orbitofrontal cortex, a region involved in activating stereotype knowledge and in generating expectations about what we see.
“We find that these social category representations seem to be fundamentally intertwined at the neural level in a way that corresponds with their subjective perceptions,” Freeman says. “As one example, the degree to which we find that mouse trajectories are slightly biased toward the angry category when categorizing men and toward the happy category when categorizing women, we find a corresponding degree of similarity between neural patterns in the fusiform and orbitofrontal cortex.”
Importantly, the work controls for overlapping visual features, to make sure the faces are not inherently angrier, happier, etc. “It shows these regions have representations of a face’s characteristics that are biased by the baggage we bring to the table, such as our stereotypic expectations,” Freeman says.
“These stereotypes may trickle down to affect basic visual perceptions of a face,” he says. They have not yet explored whether it is possible to override or control these perceptions. Ultimately, though, it will be an important issue to explore because if stereotypes can impact basic perceptions of faces, it is then a sustaining factor that influences our behavior. For example, if men look angrier and women look happier due to a stereotypical link, you might avoid men and further exacerbate stereotypical tendencies.
Freeman also points to other work out of his lab that looks not just at stereotypes but also at the role of natural context. In recently published research, for example, they found that how we visually process a scene (like the clothing example) also impacts how we process faces in that scene – for example seeing an Asian face in an Asian versus non-Asian setting. “We find that the orbitofrontal cortex is highly sensitive to the compatibility of facial cues and the surrounding social context,” Freeman explains.
More broadly, this work, along with other research being presented at the CNS symposium, is “showing that basic visual perceptions of other people are highly malleable and may be readily shaped by factors like context or stereotypes,” Freeman says. “What we take as fact from our basic visual perceptions of other people may instead be a subtle compromise between the visual information actually there and what our expectations are telling us should be there.”
Freeman will present his work as part of the mini-symposium Putting Person Perception In Context: Insights from Social Neuroscience on Sunday, April 6 at the CNS meeting in Boston. Read more about how race interacts with gender and emotion in part one of this story, featuring work by Jay Van Bavel.
-Lisa M.P. Munoz
Press can register now to attend the CNS annual meeting in Boston.