Liberties

When I was young, I wondered what the essential ingredient in a successful lyric poem actually was. I had learned that a poem did not have to have meter and rhyme, that a poem could do without the first person, and that no topic was impossible to poetry. But when I was disappointed in a poem I could not say why it was lifeless. What, I wanted to know, gave satisfying poems their life-likeness and intensity?  At twelve I discovered that poems were not necessarily born as they appeared on the page, but often went through many drafts with seemingly unpredictable changes. I perceived, in astonishment, that poems grow and mutate like living things. But I could not determine the reasons behind the poet’s revisions, and I spent a good part of my fifteenth year puzzling over drafts of poems by Dylan Thomas, sent to me in microfiche from the University of Buffalo, which held the originals in enthralling number and variety. Already, Hopkins and Thomas had convinced me of the indispensable value of sound, so what I was trying to deduce from the drafts was the reason for revisions in sound as well as in every other layer of a poem, from plot to commas. I found that by imagining myself into the sensibility of the poet holding the pen (and often copying out each successive draft myself), I could ask “myself” why “I” preferred this word to that, this length of line to that, this tone to that. I was unhappy when I could not “figure out” the motive behind a given change, but by the time I had investigated all the Buffalo holdings (there were, as I recall, about thirty drafts of “Fern Hill” alone), I could guess some plausible motives for alterations: “I” was hunting down