Poetic Data and the News from Poems: A For Better for Verse Memoir

At first, it was all about the homework. In a course introducing poetry to students who had little if any disciplined exposure to English metrics, we would spend a Thursday going over the nuts and bolts of scansion. I would set a passage from Keats or Dickinson or Herrick for them to triple-space, scan, and submit at our next meeting the following Tuesday. I would mark up the submissions with grim despatch for return on Thursday, when we would have the best-humored time we could going over a range of problems the exercise had disclosed. At the top of the range were the subtleties of verbal and vocal interpretation that scansion underscores and that prosodic vocabulary makes efficiently available for analytic consideration and debate—in short, the matter I actually wanted the course to engage. But teaching at that level would nearly always mean teaching less engaging things for a while first. We had to bushwhack our way up to the top, over awkwardnesses of the sort any new learner of an unfamiliar skill has to get past: how to work a trochaic or anapestic substitution, or balance the books with spondee-pyrrhic pairing—ropes that one learns to handle by practice.¹ Practice took the form of group work in class on given poems, and of the aforementioned homework; and right there was the rub. The lapse of a week between receiving an assignment and getting it back corrected, a long span by any pedagogical measure, can be aeonically long within the embattled field of undergraduate attention. For all but the most gifted or eager learners, the procedures of scansion were so non-intuitive, and the associated prosodical vocabulary so arcane, that the return of an assignment often found my charges discouragingly close to the point where they had been when the assignment was made to begin with. The time lag was killing us.

So it had gone with me, three decades more or less, plugging away with indifferent results at the transmission of skills that I persisted in teaching because even my mediocre rate of success fed cherished interlocking convictions: that scansion takes a reader further than any comparable practice can into the formal life of metered verse; and that the very clunkiness of scansion's apparatus bids fairest to free students of the complementary fetters they tend to drag across the classroom threshold: on one hand the despotic grip of school-bred hermeneutic expectation (libido interpretandi, as Augustine should have called what Whitman did call the lust to "get at the meaning of poems"),² and on the other hand the no-account slumber of a generalized expectation about verbal beauty that, having mumbled something about "sing-song" or "flow," lapses into aphasia. Scansion's wake-up call to the educable ear, its analytic mapping of structure and proportion, and its invitation to reconnect these formal first-derivative digital abstractions to the analog rhythms of breath and pulse, put students' birthright of what Blake called the "improvement of sensual enjoyment" back within intellectual reach.³ So I believe; so I believed; so I slogged on.

I doubt the word "digital" would have come so pat into those last sentences had not the practical frustration of a pedagogical conviction driven me a couple of years ago into the arms of Charles Babbage & Co. Enlightened souls within my university's computational wing had for some time practiced affirmative action among the humanities faculty, tempting us with offers of course relief and funded technical support to submit proposals for digital projects with an application to teaching. I had applauded this summons, but only from the sidelines. Then one day it occurred to me, having drummed out pentameters with digits left and right on more desktops than I could remember, that poets with good reason called versification "numbers," that meter was a counting device of a quite mechanical sort, and indeed that computation and prosody rode alike on a binary system with a remarkable capacity to extract the most complex functions from a sequence of the simplest elements: ones...