Tom Beckett's got quite a few obsessions. How many? Only his sweetheart knows.
A few I've benefitted from: squirrels, poetry, poets, and, easily the largest one, poetics. When you re-read his interviews going back to those he did for The Difficulties, you'll see the man's obsessed with craft and somehow, in these interviews, he gets poets talking about theirs. An excerpt from his latest interview with Thomas Fink:
"""""""""" TB: Could you talk in a broader way about your process of composition? Where does the material for those first few drafts come from? How is it accumulated? Do any procedures come into play?
TF: I begin and develop poems in five or six different ways. I’ll talk about the most frequently used way. First, I accumulate 1 to 5 pages of lines in a notebook (or makeshift stapled “notebook” of used paper with one side blank). Some lines are usually free-writing, others immediate responses to passages from texts I’ve been reading, others interesting fragments or sentences from those texts. But I don’t copy the texts verbatim; I substitute other words for most of the text’s words through procedures like (and often not identical to) those that the Oulipo, Language Poets, New York School, and others have given us. One example of a procedure I don’t exactly use is N + 7; I find more flexible variations. Modifying Louis Zukofsky’s idea of homophonic translation, I write down bits of conversation in foreign languages that I hear on a train or in a restaurant. However, results of such procedures may not stand; they’re frequently transformed.
Once I feel I’ve accumulated enough writing, I first go to work on the least interesting parts that are my own free-writing, and I perform procedures on them to change some word-combinations to more felicitous ones. Then, still not knowing where the poem is heading, I revise the whole 1 to 5 pages in another notebook, getting rid of dead passages and doing whatever I can to make the language and emerging tropes or concepts more interesting. When I’m finally satisfied that all the lines are reasonably strong, the hard part arrives: I see which fragments and sentences connect with or play off of one another. It’s usually well under 50% of the lines/sentences, not surprisingly, and that portion is transferred into a third notebook, where I improve relationships between lines/sentences by adding or subtracting words, altering syntax, and adding further sentences in dialogue with ones already there. Finally, before moving to the “shaping” phase on the computer or, in some cases, creation of stanzas, I re-test the order of sentences/fragments and keep moving them around until I’m happy with the sequence as a poem. So even though constraints and procedures produce effects, those effects may be partially or wholly edited out."""""""""""""'
Fink clearly doesn't rush a poem to its "finish". Intuitive, exploratory, diligent, he's a sculptor, language chipped away at, shifted around, re-constituted if necessary (oh if Michelangelo could have similarly picked up the chunks of marble, when he needed to, knicking the stone floor beneath his feet), poem forming within the hands of his mind.