Almost all of what I want to teach lately has to do with the sound of literature (and the sound of music itself in the case of my jazz course.) When I'm bringing in a poem sung, chanted, or recited on my ipod, it's not to entertain the students or supplement THE TEXT, the *real* poem on the page, but to study that performance in its own right. It's the concept behind Charles Bernstein's Close Listening, a book I used when I co-taught my seminar on poetry and performance. I have tons of Spanish poetry read aloud or sung--hours of it.
In my jazz course we've heard Langston Hughes reading with Mingus, Creeley with Steve Swallow et al, Kerouac making his own verbal music.
What I've found though is that we don't yet have a very good vocabulary for talking about the performance of language. This is another example of where teaching is out ahead of research. Instead of viewing teaching as the presentation or communication of the results of research or already existing knowledge, I view it as something more tentative that is paving the path for future projects.
Usually the argument is that students need teachers that are up-to-date with their fields, so their teachers should either be active researchers or recent PhDs. That's a valid argument. I would argue, though, that students need teachers who are going to do research in the future. I won't really know exactly where my ideas about poetry and performance are going to lead for a few more years.