In Defense of Reason collects three separate volumes of Winters's criticism: Primitivism and Decadence, Maule's Curse, and The Anatomy of Nonsense. Although as a young man he had published some modernist-inspired poetry, he was famous as a critic for explaining what was wrong with the modernist poetry of, e.g., Eliot, Pound, H.D., Stevens, and Hart Crane (even though he was quite good friends with Crane, at least for a while).
Skimming through it again, I recalled why I hadn't liked it at that time and feel disinclined to revise my opinion upwards now. Here's a taste: "The doctrine of Emerson and Whitman, if really put into practice, should naturally lead to suicide: in the first place, if the impulses are indulged systematically and passionately, they can lead only to madness; in the second place, death, according to the doctrine, is not only a release from suffering but is also and inevitably the way to beatitude." In the margin of the library copy I am looking at, someone has pencilled a big "?". "?" indeed.
Winters thought that "every line or passage of good poetry, every good poetic phrase, communicates a certain quality of feeling as well as a certain paraphrasable content." It was that "paraphrasable content" that tended to be elusive in modernist poetry. Often it was not to be found; when it could be found, as when Winters found "the doctrine of Emerson and Whitman" in Crane, it was often worthy only of being condemned.
Winters had a term for a poetic phrase that seemed to have a "paraphrasable content" that it did not, upon examination, truly have: "pseudo-reference." He even anatomized the ways a phrase could promise such content and fail to deliver: "grammatical coherence in excess of, or in the absence of, rational coherence"; "Transference of Values from one field of experience to another and unrelated field"; "reference to a non-existent plot." A good creative writing assignment, I think, would be to ask student writers to create one example of each of Winters's varieties of "pseudo-reference." Taken together, they almost constitute a poetics. Ashbery's Girls on the Run, for instance, seems a sustained instance of references to a non-existent plot.
Winters died in 1968, before Ashbery became famous, but Winters's predilection for expecting poems to make clear propositions, which were then to be evaluated as propositions, suggests to me that he would have found Ashbery's ascendancy as exasperating as James Fenton did.
(Reviewing Ashbery's Selected Poems in 1985, Fenton wrote, ''"The critics always get everything wrong,'' Mr. Ashbery remarked in the interview just quoted. Yet one feels that the work is designed precisely in order to insure that they will not get it right. There is no 'getting it right.' Just as I would be arrested if I tried to find out what eminent thing my old friend was up to, so I shall be stopped at the gates if I try to penetrate an Ashbery poem. Look! The camera has already been activated and the dogs let loose in the grounds. Let's just scramble back over the electric fence.")
My catching-up-with-theory program was taken me to Jacques Rancière's Mute Speech, and Winters's approach to poetry seems very like that Rancière describes as prevailing before the Romantic era: "...it is not only a matter of pleasing by means of stories and discourses, but of educating minds, saving souls, defending the innocent, giving counsel to kings, exhorting the people, haranguing soldiers, or simply excelling in the sort of conversation that distinguishes men of wit. The system of poetic fiction is placed in the dependence of an ideal of efficacious speech, which in turn refers back to an art that is more than an art, that is, a manner of living, a manner of dealing with human and divine affairs: rhetoric." Which adds up--Winters seems to have really, really disliked Romanticism.