Us vs Them: The biological basis of bias

The first step towards a healthy society is acknowledging that biases exist.

WrittenBy:Swetha Godavarthi
Article image
  • Share this article on whatsapp

(The final of two parts)


Support Independent Media

The media must be free and fair, uninfluenced by corporate or state interests. That's why you, the public, need to pay to keep news free.


As humans we carve out the world into social groups. In this part, we examine the biological basis of us (in-group) vs them (out-group). By studying patients with damage to specific regions of the brain and directly examining changes in brain activity in normal humans, neuroscientists are piecing together how “I”, “Us” and “Them” are represented in the brain, and thereby, unravel the biological basis of complex social behaviours.

“I” creates the “Us”

A prerequisite to creating a “Them” is to have an “Us”, for the person to identify with one group – a group that he or she shares some trait or belief with. This can be a basic label like gender, race, or colour of skin, extending to more abstract and complex societal labels such as nationality, caste, religious beliefs, and political ideologies. How do we figure out who are like us and who aren’t?

In order to find commonalities and/or differences with others, the brain has to firstly perceive its own identity – the “self”. We often introspect. Daydreaming about the future, focussing on our strengths for confidence (“I am good at my job!”), and maybe not as often on understanding our shortcomings (“I get angry easily.”)

This “thinking about the self” recruits a region at the very front of the brain, called medial prefrontal cortex (mPFC). People with damage to this region are unable to engage in such introspection and self-evaluation. In the social context, mPFC is important in understanding the beliefs, emotions and intentions of other people.

In an fMRI scanner, this brain region “lights up” when we hear, read or talk about groups we belong to.

fMRI is functional magnetic resonance imaging. The principle underlying fMRI is that oxygenated blood flow increases locally and temporarily in the active regions of the brain. Thus, recording blood flow while a person is performing a specific task/action can be used as a proxy for brain activity.

Indeed, people reading words describing “in-groups” (labels that they identify with) show a stronger mPFC activation, than while reading “out-group” words. It is also common for people to make decisions in a manner that benefits their group over others (e.g. the Tajfel experiment in Part I). This in-group favouritism is also mediated by mPFC.

Another example of the personal “self” extending into social “us” is seen in the activity of the anterior cingulate cortex (ACC). This region, located deep in the cortex, is activated when we receive painful stimuli, and is responsible for our emotional response to pain (fear or sadness).

Acceptance into a group (or social approval) is associated with greater activity in the ventral part of ACC. That is, vACC responds to social feedback about the self. It is also activated when we see others in a painful situation, an empathy-related response, especially for in-group members. We are better at feeling the suffering of people we consider to be like us.

Thus, neuroanatomically speaking, the in-group is represented similarly to the personal self. Just as we strive for our personal survival and success, this shared representation drives us to strive for the survival and success of our groups.

What about “Them”?!

Forming or becoming a member of a group not just involves identifying people who are more like us, but also excludes people unlike us, who get categorised as “them” (out-group).

We, and indeed all animals, engage daily in categorisation. Most of this classification is socially innocuous, but evolutionarily important. For example, you would not confuse a baby with a crawling toy baby. You just “know” that they are different! Our brains have gone to great lengths to reliably distinguish the animate (in-group) from the inanimate (out-group). When we are looking at other humans, or indeed other living creatures like fishes, birds, and reptiles, the sides of fusiform gyri (FG) in the brain “light-up”, while the middle of FG is activated when we look at objects like tools, furniture, or vehicles.

At a higher level, we readily distinguish our own species (in-group) from others (out-group). An example of this species-specific bias is our expertise in discriminating between human faces, but failure at telling primate faces apart. This distinction between human vs other animals (primates) is achieved by a network between posterior cingulate cortex and superior temporal sulcus (located right behind the temples in the temporal lobe).

Continuing with this theme of sequential categorisation, we break our own species into classes as well. Most of us use faces as the most important feature to identify individuals. The fusiform face area (FFA) (spanning the temporal lobe and the visual cortex) is responsible for facial recognition. But, we are not equally good at discriminating all faces. For instance, it is quite common to hear how Indians find East Asian faces very similar, and not surprisingly, I have heard my East Asian friends say that all Indian faces look the same! This is not bigotry, but a phenomenon called “own-race bias”.

FFA shows higher activity for own-race faces, making us adept at identifying faces matching our race, while lumping together faces from other races.

The categorisation of us and them not only affects our physical perception of the “other”, but also alters our emotional perception. Amygdala, an almond shaped structure in the temporal lobe, plays a key role in processing social emotions. Its increased activity is associated with negative emotional responses like fear, anxiety, and aggression. In a racial bias experiment, when white participants viewed brief, subconsciously presented pictures of black people, amygdala activity increased. Greater the race bias, greater was the activation, indicating an emotional bias against out-group members.

This emotional bias also affects our ability to empathise with members belonging to other groups. Earlier, I discussed about ACC, the region involved in pain perception and empathy-related response. Interestingly, response in ACC is much lower when we see members of other races in pain as opposed to our own kind, phenomenon called empathy-bias.

Our social identity also colours our perception of events and biases our world view. For example fans of Dortmund and Bayern Munich recollect the match contributions of their teams’ UEFA 2013 Champions League final very differently.

In an interesting study, boys were randomly assigned to two teams. They were then shown videos of members from their own team or from the other team performing rapid reaching actions. When asked to judge the speed of the reaching action, participants rated in-group members to be faster than out-group members performing identical actions. This bias was also encoded in the increased activity of inferior parietal lobe, a region associated with transforming observed actions to actions by the self.

Our brains have evolved to help us forge bonds and connect with each other. But an emergent and subconscious consequence of this is stereotyping of (eg. “they all look the same”), bias against (eg. “caste is a predictor of intelligence”), and apathy for (eg. “black men are more resilient to pain”) the out-group, a phenomenon called implicit bias. But these biases are not set in stone! They can be altered.

Let us go back to the facial own-race bias experiment. Is the bias hard-wired or just an outcome of familiarity? When people of different races are arbitrarily assigned to groups, FFA shows higher activity for in-group members, even after a brief but equivalent exposure to both in-group and out-group members. In other words, social identity assists in enhanced encoding of in-group member faces, overriding our implicit bias or our familiarity with them. We treat members of our group as “individuals”, and out-group members as a homogeneous mass.

Similarly, in the second racial bias experiment, when white participants are presented photographs of black people in a manner where they consciously perceive them, the amygdala activity reduces, with ACC and PFC showing greater activation. This indicates that reflective and controlled processing of social groups can regulate spontaneous/ automatic biases in the brain.

These are just a few of the brain regions involved in building our social self. Several other brain regions and their myriad connections and interactions, collectively creates the “social brain”, with the complexity of organisation of our social brain mirroring our complex social behaviour.

Studying kinship (in-group) and bias (out-group) emergence offers a means to use neuroscientific advances to bear on the very real problems of social inequality, discrimination, and intergroup conflict. The first step towards a healthy society is to acknowledge that biases, specifically implicit biases, exist. They are an evolutionary adaptation to facilitate complex social living. However, because these evolutionary biases are neural and importantly societal constructs, they can be regulated and altered by cognitive and environmental control.

On one hand, tolerance and acceptance of negative attitudes or explicit biases by a society reduces the need for an individual to exert conscious top-down cognitive control over implicit biases. On the other hand, direct exposure (through interactions and contact with out-group members) and indirect exposure (through reading about exemplars from out-groups) have been shown to help reduce the impact of our biases on our actions and interactions.

As the wise Atticus Finch tells his daughter: “… if you can learn a simple trick, Scout, you’ll get along a lot better with all kinds of folks. You never really understand a person until you consider things from his point of view […] until you climb into his skin and walk around in it.”

(Thanks to Ajit Ray for critical inputs and suggestions and Shruti Muralidhar for feedback.)

You may also like