Empathy is a complicated emotion, even for mice. On seeing another in pain, a mouse will act as if it itself is also hurting—much more, though, if it knows the first mouse. Capuchin monkeys will help out another monkey, without any reward, but only if they're on friendly terms. People also feel less empathy for those they dislike. But our species adds another layer of complication: We empathize more with people who are "like us" than with "them." This study, published this month, suggests that this Us-Them divide is more general—that the brain responds differently to any action performed by "one of us," not just to signs of trouble.

The study is part of a boom in research around the idea that people literally share feelings, at the physiological level: When I see you in pain, for example, neurons fire in my brain just as they would if I myself were in pain. It's an intriguing notion, not least because it offers a way to integrate aspects of human behavior that are usually looked at separately. Empathy is a social fact, arising out of people's relationships to each other; and it's a psychological experience for each of us; and it's a physiological phenomena in each empathizer's body. A model that connects those different levels would offer a more complete explanation. It would also, of course, offer ways to corroborate theories: It's great to be able to ask people if they feel empathy, but it's even better if you can measure it as well.

In a study published last year, for instance, Xiaojing Xu of Peking University and his colleagues used an MRI scanner to measure activity in the anterior cingulate cortex region of 33 college students as they watched films of other people getting poked in the cheek with either a Q-tip or a nasty sharp needle. A neural circuit involving the ACC is active when you feel pain in your own body, and when you see others being hurt (and, like a disgruntled monkey's, your circuit will respond less if you have a low opinion of the person being hurt, at least if you're male). In addition, Xu et al. found, their subjects' pain circuit responded much more when the injured party was of the same race. Chinese reacted more to photos of Chinese injury, and whites to white.

The new paper, by Jennifer Gutsell and Michael Inzlicht of the University of Toronto, didn't address empathy directly. They showed 30 white students a video of a college-age person drinking a glass of water, and then asked each volunteer to do the same. The students were hooked up to an electroencephalograph, which measured activity in the motor cortex area of their brains. Unsurprisingly, the motor cortex was most active when each volunteer performed his or her own act. It was also active, though, when each watched another white person drink. But if the drinker was black, South Asian or East Asian, the motor area didn't activate much at all.

Interestingly, all the students, as part of an intro psychology class, had completed something called the Symbolic Racism Scale, which is supposed to measure degrees of prejudice. When the researchers compared scores on that test with the EEG measures, they found a correlation: The more prejudiced a student appeared on the Racism Scale, the less motor activity was recorded when s/he watched non-whites.

Gutsell and Inzlicht theorize that there is indeed an Us-Them empathy gap, but that it's part of a more general phenomenon. People, they write, "mentally simulate" each other all the time, not just in pain and grief. But we only simulate people like us. We don't, they believe, "mentally simulate the actions of outgroups."

Does this mean racial preferences are somehow hardwired into the brain? Highly unlikely, for two reasons. First, there's that earlier research I mentioned, in which men empathized less with others if they believed those others had been unfair. That suggests that empathy is shaped by experience—that we learn who to "mentally simulate" and who not to.

Secondly, race figured in these studies because the investigators chose it, not because it's the only possible Us-Them division. Race is convenient for these experiments because everyone can see it, and most everyone can name his place on a racial map. It's not 100 percent reliable (two students Gutsall and Inzlicht considered "white" had to be dropped from their experiment because they self-identified as non-white) but it poses fewer pitfalls than religion, class, nationality or other groupings (an experiment based on religion, for instance, could raise questions about how the experimenters knew that students could tell Christians from Muslims).

In other words, race is used in "neuroscience of prejudice" experiments because it's handy, not because it's special. The real news here is the big impact of Us-Them divides on feelings we often assume are universal.

More on empathy research:

Frans de Waal's The Age of Empathy is a great introduction to the field (de Waal is one of the founders of the shared-neural-patterns model for empathy). Papers discussed in this post are here:

Gutsell, J., & Inzlicht, M. (2010). Empathy constrained: Prejudice predicts reduced mental simulation of actions during observation of outgroups Journal of Experimental Social Psychology DOI: 10.1016/j.jesp.2010.03.011

Xu, X., Zuo, X., Wang, X., & Han, S. (2009). Do You Feel My Pain? Racial Group Membership Modulates Empathic Neural Responses Journal of Neuroscience, 29 (26), 8525-8529 DOI: 10.1523/JNEUROSCI.2418-09.2009

Singer, T., Seymour, B., O'Doherty, J., Stephan, K., Dolan, R., & Frith, C. (2006). Empathic neural responses are modulated by the perceived fairness of others Nature, 439 (7075), 466-469 DOI: 10.1038/nature04271