My post below about Barack Obama and the ensuing comments led me to contemplate the swirling confusions surrounding the use of the term "black" and "African-American" in our language. Just the other day, my son, who has had the preferability of the term "African-American" drilled into him in school, asked me, about Ghana, "Aren't most people there African-American?"
"No," I said, "most of them are African-African."
On the other hand, Americans who actually have recently moved here from Africa, such as my Egyptian friends at the deli down the street or my economic mentor, Israel Kirzner, are never called "African-American."
So, is African-American really a racial classification? Well, it's not that either, because someone whose ancestory is mostly of non-African origin can still be "African-American."
A lot of the same problems surround the word "black" as well. One of the commenters in the thread above said "Obama is half-white and half-black." But, given that his father is from East Africa, his patrilineal heritage is probably a complicated mix of East Indian, Austronesian, Khoisan (see below) and "black" genes, where black is meant to indicate the fairly distinct genetic group that originated in the forests of Western Africa and spread south and east from there. "Blacks" in that sense are actually one of several genetically distinctive groups in Sub-Saharan Africa, others of which include the Khoisans mentioned above (think Nelson Mandela, who isn't genetically "black" at all!) and the Pygmys.