@PLOSNeuro #SfN14 highlights: Mind, Brain, and the Ethics of Intergroup Behavior
“Think of discrimination not as an act of harm against out-groups, but as an act of help towards in-groups.” – M. R. Banaji
How do we value and evaluate people? How do you know how talented, smart, or good someone is? Whether or not we realize it, we are constantly processing information and evaluating these questions in other people. Dr. Mahzarin Banaji of Harvard University explains how her research has identified these tendencies and driven an entire body of work on social cognition.
Social interactions are so fundamental to human life that we are able to identify and imitate behaviors (e.g., facial expressions) as early as a few days after birth. This behavior continues into toddlerhood, shaping the way we act and perceive ourselves as well as the world around us. Differences in the societies in which we live, and the people who surround us, engender a vast diversity of thoughts and mentalities. Harnessing such richness in insights and perspectives creates the best labs and the greatest research, believes Dr. Banaji.
The same fundamental tendencies that create diverse perspectives also cause inherent biases in the way we perceive ourselves and those similar to us, compared to people who are “different” in some way. Moreover, these biases only exist when evaluating social information. For example, probability estimation studies show that left lateralized regions of the inferior frontal gyrus and inferotemporal cortex consistently activate for non-social category judgments (e.g., deciding whether a car is more likely to run on gas or electricity). Conversely, once judgments contain social information (e.g., thinking about what other people are thinking or what action they are about to perform), activation occurs across the entire default social network of the brain, including the medial prefrontal cortex, posterior cingulate, bilateral temporoparietal junction and anterior temporal cortex. Furthermore, the activation of areas related to social cognition is selective for the judgment of humans; activation patterns differ when we try to predict similar actions in, for example, a robot. Dr. Banaji summarizes this concept: “Even when we engage the seemingly same process, the brain takes seriously the social nature of the object.”
For decades, explains Dr. Banaji, educational institutions exercised clear discrimination against women and people of certain ethnicities. She cites Dan Oren’s book Joining the club: A history of Jews and Yale, which outlines a clear policy of discrimination against Jews – epitomized, perhaps, by the rejection of an applicant for having “a Mediterranean nose” in the 1940s.
Our biased behavior reflects our sense of kinship towards others. Imaging studies demonstrate that judging people who share a similar mentality as our own (e.g., a liberal judging another liberal) draws on activation of the ventromedial prefrontal cortex – the same region that activates when we reflect on ourselves. In contrast, cogitating on people of dissimilar backgrounds (e.g., a liberal judging a conservative) recruits the dorsomedial prefrontal cortex. Dr. Banaji explains that these biases influence how we conduct interviews and perceive others in important contexts such as employment opportunities or during judicial proceedings, either negatively or favourably. She then recalls a personal incident where she refused a magazine interview based on a disagreement with its style of reporting – until she discovered that the reporter had studied at Yale University during the time she was a professor there. “We shared a zip code for about four years”; that was enough of a common ground to shift her perception and create a sense of kinship.
Social biases also arise based on physical traits. In the absence of other information, people consistently describe pictures of others as having an “honest smile”, “untrustworthy face”, or “incompetent” look. This phenomenon may relate to a person’s facial features. When observing a computer-generated face with wide-set eyes, people generally perceive the face as “smart”. However, moving the eyes closer and closer on the same face will result in the perception that the person becomes increasingly less intelligent. Dr. Banaji explains that these decisions are extremely rapid, occurring between 33–160 milliseconds (ms) after seeing a face. These judgements have real-world implications: for example, the perceived competence of a face can predict the margin of victory in an election.
Dr. Banaji then had the audience perform an experiment. As quickly as possible, we were to identify if a name was more male or female. Easy enough. We then had to identify whether an activity was career-related or home-based. The experiment became a little harder: “shout ‘left’ if the word is male or the activity is career-related, shout ‘right’ if the word is female or the activity is related to home”. We took about 640 ms to respond – not bad! Finally, the kicker: “shout ‘left’ if the word is male or the activity is related to home, shout ‘right’ if the word is female or the activity is career-related”. This was tough; in fact, the audience took an average of 1074 ms to respond. Research suggests that 75% of men and 80% of women show this effect. Apparently, women have no problem associating females with career-related words – the issue is associating males with words related to home.
Are we bound by our implicit biases?
Fortunately, our increasing awareness of such biases and prejudices has shifted the way in which we behave as a society. Today, Dr. Banaji proudly reports, Harvard admits students with greater equality in gender, race/ethnicity, and geographical backgrounds than many other institutions, and financially supports a large percentage of students who could not otherwise enrol. Similarly, the American Symphony Orchestra – once dominated by men – is deeply gender-integrated as a result of “blind auditions”, held to avoid gender biases.
To measure these subconscious social biases, Dr. Banaji — together with Anthony Greenwald and Brian Nosek — created the Implicit Association Test. The test relies on three established scientific concepts: associative learning, bounded rationality, and unconscious inference. With an estimated 17 million respondents to date, Dr. Banaji claims this is a robust measure of what people think but are unable to say – how much we attribute talent to others based on physical features or geographical location – and is now leveraged in the clinical setting. You can measure your own biases by taking the test here.
Take-home message
If our social biases are so implicit and prevalent, why seek a solution? Eliminating our own “blind spots”, or selfish perceptions, will enable us to evolve as a more harmonious and integrative global community. Dr. Banaji aims that by increasing awareness about our basic social tendencies, we will increasingly adopt values and behaviour that are in sync with our intentions. You can read more about Dr. Banaji’s work in her book Blindspot: Hidden Biases of Good People.
Also see my post-lecture Q&A with Dr. Banaji.
