I do violence to these lines in detaching them from what precedes and follows them, but I do so for a reason. More often than not, the human pang in Stevens is secreted inconspicuously in the poem, instead of being announced in the title or in the opening lines. It is the usual, if mistaken, way of the commentators to begin at the beginning and take Stevens's metaphysical or epistemological prolegomena as the real subject of the poem, when in fact they are the late plural of the subject, whose early candor of desire reposes further down the page. And so I isolate what I take to be the psychological or human "beginning" of the poem, its point of origin in feeling...
This was extremely useful advice: Stevens was an acquired taste for me, and even now I find that I usually have to rearrange a Stevens poem before I can appreciate it or understand why it is put together the way it is. There is a peculiar similarity here with the business of appreciating mathematical proofs, which I was reminded of by something Gowers said in his recent (and very good) post on proving the unique prime factorization theorem:
I worry sometimes that accounts like this of how a proof might be discovered can be off-puttingly long. So it’s important to stress that the actual proofs are much much shorter. Here’s how the proof that the above thoughts lead to ends up. I’ll just do the uniqueness part, and I’ll write the whole thing in logical order, which is more or less the reverse of the order in which one discovers the steps.I am sympathetic to the idea that real appreciation is never without an element of reverse engineering, or of hypothetical intellectual history: the completed work in itself might offer some immediate delight, but to go any further you need to have some theory, accurate or not, of how it might have been arrived at, and why it was then organized as it was. (Notable exception here: Milton. But then I think of Paradise Lost as pure verbal texture a la Campion's songs.) I imagine that this is a more tenable procedure with a poem or a proof than it is with, e.g., a novel.
It has always struck me as a regrettable deficiency -- maybe a necessary one -- of education in math and physics that there is very little intellectual history, and I think it is fair to say that many practicing physicists are not connoisseurs by temperament; there is a widespread tendency to regard historical aspects as "impractical" -- which they are, for premeds and engineers -- though I have always found it invaluable to have a sense of how people have gone about doing things. (My views on pedagogy are too uninformed and too reactionary to discuss safely in public.)
[As an admin note I'll be traveling the two weeks after Thanksgiving: Boston Nov. 26-30, New Jersey Nov. 30-Dec 2, New York that Friday night and part/all of Saturday, San Francisco Dec. 6-8. It is only mildly indiscreet to note that this is all job-quest-related to some extent; the old talk has been revised and will have to be practiced and touched up repeatedly; it is against the culture of physics to read out any part of a talk, so that, in practice, one has to memorize the first three or four agonizing minutes until one hits one's stride.]