Some years ago, in Detroit, I knew a young woman, a fine soprano who often gave solo recitals or performed with the Detroit Symphony. Once, describing how she practiced a song, she said she just sang through the piece until she hit a snag, made a mistake. Having worked out that glitch, she would go back to sing through the song again. Now the once-snarled passage would be fine. Before long, however, the mistake would turn up again, later in the song – she always spoke of it as the same mistake turning up in a different, later passage. Over and over, she said, she went through this process, until she had pushed the mistake almost to the song's end and then could finally push it off the edge – as if over the cliff at the end of a promontory.

To me, this seemed like the way I often worked, writing poems – though excruciatingly reversed. I usually started with some notion of how a vague body of materials might be shaped into a poem. I would go through these rough materials again and again – not looking for a mistake, something wrong, but rather for something right, something surprising for its oddity or originality. Adding that into my rough hodgepodge, I'd go back through the materials over and over hoping for new surprises. Unlike my soprano friend, though, I had no suspicion where this might happen. She was aiming, after all, to reveal a fine, already existent structure. I was more like some prospector running tons of slushy mud through a washing pan looking for anything bright. I can take it for granted that this glittering particle – and any which might follow – will lead me into a solid pocket of discovery. That, I think, is because those bright glints are inklings of my own private thoughts and language rising (if I'm lucky) out of the slush of received thoughts and beliefs, of my own habitual, commonplace language.

I'm at odds, then, with writers who believe that their first thoughts are their best, those worth keeping. As for my first thoughts, they are usually things someone else told me I ought to (and had pretty damned well better) believe; they've been got up in the most politically acceptable, neutral language, devoid of character or personality. I found something similar in psychotherapy: mine often consisted of saying things over and over until I got them into my own language, where I could learn something about who I was, what I felt – as opposed to who my society and I hoped I was. This must be part of what James Baldwin meant in saying "The purpose of art is to lay bare the questions which have been hidden by the answers."

It's no surprise if a poem takes me eight or ten years. I don't, of course, keep drudging away at it day after day; I work until I feel it going dead under me, then put it away until I can come back almost as a stranger. That is, I let my unconscious handle it for a while, hoping for things my conscious mind may have resisted – perhaps for fear of showing myself in a bad light, perhaps for fear of saying things that will offend others and damage my standing in my community or the literary world. But, of course, those are the things that offer the chance to say something we don't hear every day, something that might enrich the atmosphere. (The trouble is that it offers the probability of offending our neighbors and our own high self-appraisals.) I usually work on at least half a dozen poems at a time, moving from one to the other until I feel I've done all I can with one of them – which, like a much-rehearsed song, I can then shove over to the end of my desk.

Each of these poems may have a pile of anywhere from ten to forty worksheets, some scattered about in various rooms, on various tablets, envelope backs, stick-it tabs. Some will show only an "a" changed to "the" which may well get changed back three or four times; some will show whole sections of the poem moved from one place to another or simply crossed out. Elsewhere, there will be long lists of synonyms, related terms, variant phrasings. Now that I work on a computer these smaller changes don't always get recorded – they're simply deleted or replaced on the screen. On the other hand, as I get older and have less immediate access to memory, thesaurus numbers are more and more in evidence.

My desk and work tables reflect – are, perhaps, an objective correlative for – the scramble of my mind and my manuscripts. Every surface is covered by odd scraps of paper, dictionaries for several languages, half-empty cups of cold coffee or tea, computer disks, clips and staplers, long-buried pencils and pens, temporarily abandoned projects, file folders, unmarked calendars. Of course I keep promising that I'll get back to tidy them up; frankly, I must admit that I hope that all this kafuffle never gets cleared up – not, at least, until the kafuffle on my worksheets produces one or two more sparkles.

Poetry Daily / Amazon.com

Search Poetry Daily / Amazon.com for other books: