This morning, I experienced an unsettling double-take on my walk to work: what appeared for a moment to be a baby left alone in a car turned out to be a large clown doll in a baby’s car seat. My brain had to use physical cues to determine whether I was seeing an inanimate doll, which would activate movement towards less clown-laden destinations, or a human being, which would activate social cognition and helping behavior. This basic ability to detect entities that can experience mental states—termed mind perception—allows us to see others as worthy of moral consideration, while failure to perceive minds in others has been theorized to facilitate prejudice and even torture (Harris & Fiske, 2011).

Humans rapidly make such distinctions between the mindless and the minded on a daily basis. In fact, the brain is able to distinguish between animate and inanimate faces within a few hundred milliseconds (Wheatley, Weinberg, Looser, Moran, & Hajcak, 2011). Bottom-up perceptual inputs have been found to play a key role in determining when we believe faces have minds: when presented with a continuum of morphs between human and inanimate faces, people perceive faces to have minds only past a categorical threshold biased towards the human end of the spectrum (Looser & Wheatley, 2010).

In a recent paper (Hackel, Looser, & Van Bavel, 2014), we tested whether top-down social motives can also shape the interpretation of perceptual cues indicating the presence or absence of a human mind. In particular, we examined the influence of group membership on mind perception. Group membership represents a strong social motive associated with biases in social perception and evaluation (Van Bavel et al., 2008), as in-group members are often considered more relevant targets for fulfilling core social needs like affiliation (Brewer, 1988).

In three experiments, we presented participants with a series of morphs between human and inanimate faces (i.e. dolls, statues), which were described as based on in-group or out-group models. Participants rated how much each face had a mind, allowing us to calculate the point of subject equality (PSE) for in-group and out-group faces—that is, the point along the morph spectrum at which people reliably perceived faces as having a mind—as a measure of mind perception threshold.

In the first experiment, participants were assigned to minimal groups, testing whether mere social categorization would influence the threshold for mind perception. Indeed, participants showed a more stringent mind perception threshold for out-group as opposed to in-group faces, requiring 67% humanness to interpret an out-group face as having a mind, but only 63% humanness to interpret an in-group face as having a mind.

We reasoned that intergroup bias in mind perception should depend on the strength of one’s identification with their group. That is, people should only show an intergroup bias in mind perception to the extent that they care about a particular group. In Experiment 2, only those NYU students who strongly identified with NYU showed a more stringent mind perception threshold for BU as opposed to NYU faces.  This finding suggests that the strength of particular identities can shape intergroup mind perception in a flexible manner.

Although people are often motivated to find in-group minds, this pattern may not always hold. For instance, soldiers on a battlefield would likely be motivated to detect enemy minds. We hypothesized that threatening out-groups might present motivationally relevant targets, as one may want to consider the out-group’s thoughts and plans. Indeed, our third experiment replicated the relationship between in-group identification and mind perception, but found that out-group threat was associated with a reversed effect: Democrats and Republicans who found the out-group threatening had more lenient thresholds for out-group mind perception. In settings of intergroup competition, people therefore may be particularly ready to track the minds of those perceived as a threat.

This work indicates that top-down motives may influence not only higher-level mind attribution, but also how we interpret bottom-up perceptual cues to the presence of a human mind. To the extent that mind perception serves as a building block for higher-level social cognition, this influence may have consequences for how we engage in social cognition and social behavior towards members of different groups. Finally, this work highlights the flexibility of social motives, revealing that threatening contexts may reverse an ordinary pattern of dehumanization and prompt readiness to perceive out-group minds. Future work can determine the extent to which threatening out-groups are perceived along different dimensions of mind, such as agency or experience, which are differentially associated with moral responsibility and moral care, respectively (Gray et al., 2007). In this manner, understanding how salient contexts flexibly shape mind perception can hopefully help understand and improve intergroup relations.


Hackel, L. M., Looser, C. E., & Van Bavel, J. J. (2014). Group membership alters the threshold for mind perception: The role of social identity, collective identification, and intergroup threat. Journal of Experimental Social Psychology52, 15-23.