As anyone who has dated a musician knows, discussing the actual act of making music can be an infuriating activity, the result usually being a hash of spacey poeticisms or technical minutiae. Here the term "muscle memory" is indicated: The memory is there, but mainly elsewhere, in the fingers or wherever. I've been guilty of this too--it's just that speaking about the "how" of music has a way of making one feel infantile and dumb. Maybe this is in part why the pleasures of criticism are ultimately, strangely, other than those of the music it refers to.

There are plenty of pieces in the new Granta music issue that explore such phenomenological territory, most of which fall flat because they do not overcome this very inarticulateness. Fortunately, there are also plenty of pieces that take a different tack, situating music instead in the social field, finding in it a sort of residue of the varied movements of economy and desire.

Nik Cohn's "Soljas," for example, charts the explosion of "bounce" hiphop: "It started in the late 1980s, a wild mix of rap and Mardi Gras Indian chants and second-line brass band bass patterns and polyrhythmic drumming and gospel call-and-response." When read next to "Frank [Sinatra]'s Place," the pieces illuminate how through two very different brands of gangsters--the New York mobster and the New Orleans playa--music is connected to such social realities as crime.

The exception is Greil Marcus' piece on Harry Smith's enormously influential Anthology of American Folk Music. Marcus is best when he's a little insane, his writing sweaty and almost coherent. But here he's in his donnish mode, his obsession too stale. It's also self-aggrandizing-- is his definition of a "secret history" one he himself uncovers? When presumably the very allure of the word "secret" is, as the work in this issue demonstrates, its implication of an irreducible multiplicity, then a narrative openness is as impossible as music itself.