A lot has been made of Hillary's "fake southern accent" last weekend in Selma, at the remembrance of Bloody Sunday. Powerline and other rightwing blogs have posted comparative samples, like here. In fact, things are more complex than they first appeared, as laid out here, where LaShawn Barber reviews some of the political posts. (See also here.)
The same charge of "faking an accent" has been leveled at Obama in various places (including the last link above) and even Al Gore, here. For an upper middle class white girl from northern Illinois, who lived in Arkansas as a member of the real elite, even quoting a passage "in character" (as one blog puts it) like this seems like a bad move at best. But Obama, who worked as a community organizer in Chicago, surely knows the kind of English he was using very well. Even Al Gore is ultimately from Tennessee, although it's a tenuous connection for the son of a US senator. So, can the son of an African father style shift in Alabama? Or a rich white kid officially from Tennessee but mostly from Washington, D.C. when he talks in black churches? (And with Gore, unless I'm just not remembering, it's basically about moving from his normally very standard English to a southern-sounding variety, not using features particularly associated with African-American speech.)
What this calls to mind is not so much anything about linguistics per se, but more the considerable body of work in folklore on "authenticity", where that notion naturally turns out to be slippery. Regina Bendix's important book, In search of authenticity, argues that this search is central to the formation of folklore studies, and the sense of loss that comes with modernity.
Gotta run now, but will return to this issue later ...