For everyone interested in how their brain works, I'd suggest checking out a book coming out soon called Picturing Personhood, by MIT anthropologist Joseph Dumit. Dumit shows how easy it is for brain scans to become cultural Rorschach tests. Scans of mental activity, such as fMRI or PET, are basically complex graphs that represent the relationships of data gathered in very narrowly defined experiments and which are then statistically massaged with special-purpose software. But for most of us non-scientists (and even some scientists) it's easy to look at these images as objective snapshots of thought. As Dumit points out, it is even easier for us to impose what we want to believe about human nature on those pictures, getting a comforting feeling of certainty from our misconceptions about how neuroimaging really works. This neuro-Rorscach effect is only going to become more common. That's because neuroscientists are using their scanners to probe the social brain. Most people may not have a lot of preconceptions about how the cerebellum influences motor control, but when you get into the way we feel about one another, everybody's got an opinion. A case in point can be found in the new issue of Nature Neuroscience: Dartmouth scientists looked at how racism affects our ability to think clearly. This research builds on several years of work by social psychologists, They have found evidence of subtle effects of hidden racial biases with an experiment. they call the Implicit Association Test. The test is remarkably simple. You look at words and names flashing on a computer screen. In one version of the test, if you see a word, you press the left button if it's a positive word (beauty, for example), and if it's negative, you press the right one (filth). If a name appears, you press the left button if it sounds like a white name to you, and the right button if it sounds like a black name. In other words, you're pressing the same button for positive words and white names, and negative words and black names. The researchers then flip the test around, so that black names and positive words use the same button, and negative words and white names. They subjects have to hit the buttons as fast as they can, and the researchers then measure how long it takes for people to press a button. Tests like these reveal some surprising patterns. Some people show consistent differences in their button speed, depending on how the test is conducted. Some white people, for example, take longer in the black/positive-white/negative version than in the white/positive-black/negative version. Often, these people will claim not to be racist, and yet they show clear (but subtle) biases in the way they take the test. You can take an online version of the test here. Neuroscientists saw in this work an opportunity for an elegant use of their scanners. In 2000, Elizabeth Phelps of NYU and her colleagues decided to look for a signature in fMRI scans of these results on the IAT. They showed white people unfamiliar faces of whites and blacks. Whites who showed a big lag on the IAT tended to respond to unfamiliar black faces with higher activity in a region of the brain called the amygdala. This region is associated with fear and vigilance in uncertain situations. The smaller that IAT lag, the smaller difference in the response from the amygdala. And this pattern disappeared altogether when white people saw faces of black people who were either friends or celebrities--i.e., familiar. Now Dartmouth scientists have found another way that racism may affect the brain. The prefrontal cortex--the front third or so of the surface of the brain--manages the information processing that goes on in the brain. For one thing, it keeps us focused on mental tasks despite all the internal and external distractions we face. But many researchers have argued that this "cognitive control" isn't in infinite supply. Doing some intense cognitive control for one task may leave a person with less control over some other one. Racism, the Dartmouth team has concluded, can drain our cognitive control. The Dartmouth team first gave a group of white Dartmouth students the IAT test to measure hidden race bias. Then they showed the subjects pictures of black people as they scanned their brains with fMRI. While Phelps had focused her scanning on the amygdala, the Dartmouth team paid close attention to the prefrontal cortex. They found that people who showed strong racial bias had a strong response to the pictures in a patch of the brain just above the temples--the dorsolateral prefrontal cortex. Many researchers have shown that this region is crucial for cognitive control. What's most impressive about this result was that the Dartmouth team was able to draw a graph of the relationship--the stronger the bias, the stronger the cortex response. The Dartmouth team also ran another test. A black experimenter interviewed the subjects on some hot-button issues, such as racial profiling. The researchers hypothesized that the interview would demand a lot of cognitive control from people with strong racial bias. Immediately after the interview, the subjects had to play a game that also demands a lot of cognitive control. The game is known as a Stroop test. You see words that are in different colors, and you have to identify the color. The tricky part comes when the word is GREEN and the color is red. Now, the dorsolateral prefrontal cortex has to override the normal response to choose green and pick red. As they had predicted, the interview made people with a strong racial bias do worse on the Stroop test. And once again, their results form an impressive trend from low bias to high bias. So, what does this all mean? Here's one take, from Boston Globe, "To the litany of arguments against prejudice, scientists are now adding a new one: Racism can make you stupid." In other words, biology supports a political stance. But you could use the same sort of argument to make a very different (and very loathsome) argument: it's not racism that makes you stupid, but the social pressure to keep your racism bottled up. If a racist person was comfortable with his racism, then presumably he wouldn't have to use up his cognitive control suppressing stereotypes. If you want to elevate a brain scan to social commentary, the meaning of the scan itself becomes slippery. Is the scan I've pasted above your brain on racism, or liberal guilt? The very notion that racism can leave a biological signature can mean a lot of things, depending on your preconceptions. Is it therefore a pathology? Race bias is unconscious, after all--even people who say they're not racist can show these consistent lags on IAT. Does that exonerate people who commit hate crimes--it's just how their brains work? Or can we just all retrain our brains to bring an end to hatred? The Southern Poverty Law Center certainly seems to think so. Before we all start making grand statements about social policy, it's necessary to take the Dartmouth study with a few grains of salt. The Globe's use of the word "stupid" is actually way over the top. Taking an extra tenth of a second to answer a Stroop test question doesn't translate into imbecility. And some neuroscientists from the University of Michigan and Temple University wrote a commentary that will be appearing in the same issue of Nature Neuroscience, and they point out that even the IAT itself may not mean what some researchers claim for it. The black/positive lag may, for example, represent not a personal bias, but a knowledge of widespread stereotypes. And there has been precious little work done on the responses (both on the IAT and in the scanner) of black people. Still, the results are there. The lags are real, the trends in the brain scans relatively clear-cut. Something is going on. Perhaps before too long, we'll find out what.