This is taken from the book Drunk Tank Pink, which has been on my reading list for some time. There are numerous studies to corroborate these findings--that the labels or expectations we have when assessing people greatly influences what we see or don't see in them. -- Dallas
LONG AGO, HUMANS began labeling and cataloguing each other. Eventually, lighter-skinned humans became "whites," darker-skinned humans became "blacks," and people with intermediate skin tones became "yellow-," "red-," and "brown-skinned." These labels don't reflect reality faithfully, and if you lined up 1,000 randomly selected people from across the earth, none of them would share exactly the same skin tone. Of course, the continuity of skin tone hasn't stopped humans from assigning each other to discrete categories like "black" and "white" — categories that have no basis in biology but nonetheless go on to determine the social, political, and economic well-being of their members.
These racial labels impose boundaries and categories on an infinitely complex social world, but once in place these boundaries are very difficult to dissolve. People are apt to resolve racial ambiguity by resorting to racial labels. In a Stanford University study, an experimenter showed white students a picture of a young man whose facial features made it difficult to determine whether he was white or black. For half the students, the man was labeled "white," and for the other half he was labeled "black." The students were asked to draw the image in front of them as accurately as they could. To sweeten the deal, the student who created the most accurate drawing was promised a $20 cash prize.
The ones who were told that the man was black tended to exaggerate his "typically black" features, whereas those who were told he was white did the reverse, exaggerating his "typically white" features. Although the students were looking at exactly the same photograph, they perceived the image through a lens that was tinted with the racial label that the researcher provided earlier in the experiment. [continue]