A writer's correspondence is a component of the work. Not equivalent to it, and not to be interpreted as though the letters were drafts of it, but continuous with it: written in the same hand, out of the same preoccupations, often in the same months, and frequently superior to the published prose because there was nobody to please. A poet who spends thirty years considering a problem does not discontinue the consideration when he picks up a different kind of paper. To publish the poems and withhold the correspondence is to publish the conclusions and suppress the working, and the consequence is to make the writer look more finished than he was -- which is a distortion, and a flattering distortion that no serious reader wants. The objection that the correspondence was private mistakes the category. Everything a writer of that stature wrote was addressed to someone; the poems were addressed to strangers, which is a more demanding audience, not a more legitimate one. There is furthermore a practical consideration. A poem is frequently unintelligible without the occasion that generated it, and that occasion is in the correspondence and nowhere else. An editor who withholds the letters does not thereby leave the poems undisturbed; he leaves them more difficult to read, and subsequently furnishes the occasion himself in an annotation, which is identical information with the writer's voice extracted from it. The distortion is cumulative. A body of work with the letters removed acquires an appearance of inevitability that no working life has ever possessed.
[[PASSAGE B]]
The letters were addressed to a person. That is not a detail concerning their production; it is what they are. A letter derives its meaning from a relationship the reader of a printed edition does not stand in, and it presupposes things -- a quarrel three weeks old, a joke, a shared antipathy -- that the editor must either footnote into flatness or leave to be misinterpreted. What arrives on the page is not the letter but a transcript of one, and the difference constitutes the whole of what made it worth writing. Nor is the appeal to continuity as persuasive as it sounds. A writer picks up different paper precisely because he is undertaking something different: complaining, apologising, soliciting money, being funny for one reader who will understand. Publishing the working is a good reason where there is working to publish. Much correspondence is not working at all. Printing it tells us less about the poems than about our appetite for the poet. There is an additional cost that the case for publication rarely counts. A correspondence published entire changes how the subsequent generation writes letters, and what it changes them into is a form composed with a reader over the shoulder. That is no loss to scholarship, which will have its material either way. It is a loss to the thing being studied. The appetite is real and not disreputable; it is simply not equivalent to an interest in the poems.
