Suppose 30 students are eating lunch in a school cafeteria. Twenty of them are white, and 10 are black. The 10 black students are sitting at one table, while the 20 white students are at two other tables. The question that will "naturally" strike most observers of this scene is, why are all the black kids sitting together?

Sociologists tell us this question is far more likely to arise than the question of why all the white kids are sitting together. This is an example of how whiteness works as an "unmarked category." That is, in this scenario the black kids are more likely to be perceived as black kids, while the white kids are more likely to be seen as just kids.

To see the hidden power of unmarked categories, consider the relationship between gender and health risk. Last week I wrote about a new study in the New England Journal of Medicine that evaluated the risk of death faced by people at different weight levels. I predicted that journalists would be fooled into buying the authors’ conclusion that higher-than-average weight was a significant risk factor, despite the fact that the study’s data contradicted this claim.

Sure enough, The Washington Post article titled "Just a Few Extra Pounds Could Mean Fewer Years, Study Finds" was typical of the coverage despite the fact that the study’s subjects carrying what the study defined as "a few extra pounds" actually had a lower death risk than anyone else.

But what’s particularly fascinating about the study’s data is the difference in mortality risk between men and women. According to my friends at the Harvard Medical School, the "ideally lean" weight for which all Americans should strive requires achieving a body-mass-index figure of 18.5 to 21.9. Meanwhile, these same experts assure us that the medical consequences of being "morbidly obese" _ of having a BMI of 40 or higher _ are almost too terrible to contemplate. (Only the fattest 2 percent of the U.S. population is at this weight level.)

Now let’s look at the study’s data. It turns out that the death rate among the study’s 6,992 "ideally lean" men was nearly 50 percent higher than among the 6,569 "morbidly obese" women! There were 1,513 and 981 deaths in the two groups, respectively. Note the percentage of people in each of these two categories who entered the study with a pre-existing chronic disease was almost identical, thus disposing of the argument that the death rate among "ideally lean" men was a product of men who were thin because they were sick.

In other words, in this study of more than a half-million Americans, "ideally lean" men were far more likely to die than the very fattest women _ women who on average weighed close to 300 pounds. And this same statistical pattern is reflected throughout the study: 42,173 men died, in comparison to 19,144 women.

Now in a superficial sense, gender wasn’t an unmarked category in this study. (After all, the researchers did divide their data according to sex.) But in the more profound cultural sense, the gender of the participants remains practically invisible.

The completely imaginary risk of carrying "a few extra pounds" generates front-page headlines, while the overwhelmingly real risk of being male is hardly ever mentioned. Is this because we can’t do anything about being male, while we can do something about being fat? In fact, both premises are false: While there is no known method short of stomach amputation for producing long-term weight loss, castration clearly extends life expectancy. So, in fact, there is a cure for the masculinity epidemic, although the cost of pursuing it seems rather high.

This is yet another example of how _ from Amber Alerts to anti-marijuana campaigns to the war on terrorism _ there seems to be little rational connection between the risks we most fear and the risks we actually face.

(Paul C. Campos is a law professor at the University of Colorado and can be reached at Paul.Campos(at)