I'm reading Roberto Harrison's Os--slowly, because it's dense work that calls for lingering. So far I'm very impressed by:

1) The economy of the writing, which reminds me in an odd way of Zukofsky (though the tone, vocabulary and apparent goals are utterly different). Vivid words in compressed clusters that, through the overlap of connotations between them, convey something that would otherwise require extensive theoretical elaboration and a loss of concreteness. The relative distance of a word from the one next to it seems to be one parameter at work here.

(The charge of the language--the fact that something always seems at stake--makes this economy more powerful).

2) A really intelligent use of the line break; it's often hard to decide if a line is a discrete unit or not--whether the sentence (they're not really sentences, though they often act like them) is meant to continue uninterruptedly into the next line. When the latter seems to be the case, there's usually something about the next line that makes this conclusion uncertain. There's a wide variety of possibilities in here, and Harrison, in the first twenty pages of the book, explores a lot of them (I like the variety of his approaches--though the poems clearly "go together," they don't settle into a style).

3) I've been struggling, in my own writing, against a tendency to always conclude poems with gestures that sound like endings. Harrison is highly skilled at endings that neither seem to wrap things up nor seem simply to break off in mid-statement (this probably has something to do with the general syntax of the writing--not quite sentences, but not fragments either). That's worth further study.