Francesco Giusti interviews Derek Attridge in the Los Angeles Review of Books. Their discussion uses Attridge's most recent book, The Work of Literature, as its starting point.
Caroline Levine, for instance, in her much talked-about Forms, understands forms as highly general tendencies (her subtitle and chapters name the categories she is interested in as “Whole, Rhythm, Hierarchy, Network”), a simplification that enables her to find formal similarities between works of literature and social or political realities. This is not to say that she doesn’t have some interesting analyses, but it seems to me that they owe very little to her overarching concept of form. Whatever is common to both poetic rhythms and institutional rhythms, literary hierarchies and sociopolitical hierarchies, is so general that it is of little analytical power. On the other hand, if we use the term for whatever purposes are called for by the particular analytical project we’re engaged in, as Jonathan Kramnick and Anahid Nersessian have recently suggested, it loses much of the explanatory capacity it has as a conceptual tool.
I agree with this criticism of Levine, and have made some of my own.
Attridge then goes on to say:
As a first corrective to the abstract notion of form, and restricting the discussion to literature, I would suggest that the term is best understood as a verb. Form, or forming, is something that happens in the experience of reading (or hearing) that constitutes the literary work as a literary work. Even the visual aspects of a literary work are part of that temporal experience. Secondly, as you suggest, form always has to be understood in relation to singularity: it is an aspect of what makes any literary work singular, and any given formal device may be used to very different effect in different works.
A bit later:
Formalism, then, is a name for the critical attention to these aspects of the literary experience, and I find it difficult to see how any responsible reading of a literary text can ignore them — though, clearly, a great deal of literary study over the past two or three decades has chosen to do so, preferring to treat the semantic dimension of the work as if it were the only thing of interest. To the historian or the sociologist, perhaps it is, but as a student of literature there is much more to take into account. My sense is that to look at meaning irrespective of these formal qualities is a sure way to misunderstand meaning.
While I'm sympathetic, I don't see that his appreciation of what form is and does leads him to an adequate descriptive practice. I've addressed this in a series of posts on a book Attridge coauthored with Henry Staton, The Craft of Poetry. They provide detailed accounts of a dozen or so poems. I add mostly descriptive commentary of a half dozen of their examples and have comments here and there on their method and on literary form in general