The most intriguing entry, to me, in Sarah's collection of reviews, essays, and assorted correspondence was her email conversation with Dennis Lee and Robert Bringhurst when discussing the pitfalls and potentialities of "polyphonic poetry". I've always been fascinated by the subterranaean linkage of poetry with music. There are many structural similarities: organic ordering with choices of prominence, shading, repetition; thematic statement with variations; mood and pacing shifts; rhythmic play; contrasts in dynamics; and on and on. But music, not being burdened with having to deal with overt meaning, can nevertheless create differing emotions simultaneously with major and minor keys, instrumentation, through composition, infusing subtleties into all corners of a soundscape.
Sarah and Bringhurst discuss the latter's experiments with poetic polyphonic composition, and though the talk gets bogged down in the historical linguistic arguements over what the musical term has meant, I can understand the problems inherent in the experiments and sympathize with their similar passions in the process. (I was surprised neither brought up American classical composer Charles Ives' "carnival" symphony. I don't think that experiment works, but it's fascinating to think what may be done with a more polished and talented ordering in that line.)
What interests me much more, though, is something quite different, and which I found largely missing from the back-and-forth. Polyphonic experiment was often described as simultaneous voices, with a subordinate interest in consecutive polyphony. But there was no mention of polyphonic colouring with one voice simultaneously. Sarah rightly brought up Berryman's Dream Songs, but again, this just emphasizes that even multiple, highly contrasting voices have to be broken by sequence, no matter how wrenched, tight, and creative the syntax is (and Berryman was a groudbreaking syntactical genius).
The only way I can presently see clear to a breakthrough in one-voice-simultaneous polyphony is for heightened, contrasting, and multiple meanings of words themselves, rather than separate voices with different narratives, moods, and so on. Of course, poetry at its best has always called upon multiple meanings in its diction, but usually in the context of a single consistent voice with a particular story or thought. I'd welcome a correction on this, but if Sarah and Bringhurst are seeing such difficulty (and it is difficult) with multiple voice and consecutive voice polyphony, I can't imagine much historical evidence for single-voice-simultaneous polyphonic poetry. And I don't know how possible it is, never mind its effectiveness. In this as in other areas, music has the last note.