Wednesday, March 11, 2015

What Rhymes with Wallace?: Wallace Stevens and Rhyme

One of the unexpected pleasures of the recent Louisville Conference on Literature and Culture since 1900, for me, was running across Anthony Madrid in the corridors between sessions.  I hadn’t seen him for months, not since he pressed a collection of Edward Lear’s verse into my hands in a tiny used bookstore under the Metra tracks in Evanston, and was glad to let him pull me into a nearby room where he was about to join other members of the Wallace Stevens Society in a panel on Stevens and rhyme.

Madrid spoke first, beginning with a general pr├ęcis of his argument about the trajectory of rhyme in English verse.  I’d heard this before, when Madrid joined Don Share and Lea Graham on a panel on the poetry of Michael Robbins I chaired at the Midwest MLA a year ago, but it’s such an intriguing argument I was happy to hear it rehearsed again.  The gist of it is that after the Elizabethan period, whole categories of rhyme are, essentially, decommissioned from English verse, or become far less common (critics of Madrid’s theory love to find exceptions, but a full reading of his doctoral work in The Warrant for Rhyme reveals a strong case for a general trend of the kind he describes).  Rhymes that involve strong semantic links—semantic similarities, or opposites, or rhymes from the same semantic category—greatly diminish over the course of the seventeenth century.  So me/thee, mine/thine, he/she, berry/cherry, and the like become far less common.  Madrid makes much of this: the link between rhymes becomes less rational, he says, and more a matter of mystery, as if the poet wills the rhyming words to belong together for reasons unknowable to the intellect.

The anti-semantic nature of rhyme becomes a norm in the eighteenth century, and it is only with Thomas Percy’s 1765 Reliques of Ancient English Poetry, an antiquarian’s selection of old ballads and other poems, that the older style of rhyme begins to return for the Romantic era.  A Romantic like Byron, when he is serious, as in Childe Harold, rhymes like an eighteenth century poet, but when he’s comic, as in Don Juan or Beppo, he makes rhymes that go out of their way to draw attention to themselves, and appear as stunts (as Butler’s comic rhymes in Hubridas did).  The rhyme becomes something deliberately original, frame-breaking and winking.  And this sounds the death-knell for rhyme, since over the nineteenth century rhyme becomes less a holistic part of poems and more of an attention-grabbing device, until in the modern era it is all-but abandoned.

This brings us to Wallace Stevens, whose poetry only rarely rhymes—11 early poems are consistently rhymed, and after that rhyme occurs as an occasional grace, like alliteration.  When it does rhyme, it almost always uses the classical rhyme of the eighteenth century, which does not seek to draw attention to its cleverness (an exception being “├╝bermenschlichkeit/ word soon come right”).  There is rarely a semantic quality to the rhyme, and Madrid argues that this is aligned with Stevens’ belief in the mysterious powers of the imagination—the rhymes are convened only by the power of imagination, not on some rationally apprehensible basis.  Stevens rhymes less and less over time, too, as the poetry becomes less overtly musically-driven and more liturgical.

After Madrid, Eugene Vydrin of NYU took the stage to speak about rhyme and the adherence of the poem to reality in Stevens, and was followed by Joon Soo-Bong of Seoul National University, who spoke about schematic rhyme as essentially antithetical to Stevens’ aesthetic of flow.  All very interesting, but I was particularly fascinated by Roi Tartakovsky, who turned to empirical linguistic research to explain rhymes and other sonic effects in Stevens’ poetry. 

Tartakovsky noted that one of the most important recurring patterns in Stevens’ poetry is that of the recurrent pair of sonically similar words—as in the “sea” and “she” that come up at the start of the first and second stanzas of “The Idea of Order at Key West.”  Linguists have told us that acoustically similar words are actually less well reproduced in the memory than other words, and tend to become confused—but that rhyme words are, in fact, remembered well.  So Stevens works with a combination of words that aid in the recollection of argument or thesis, and words that actually blur rational or argumentative distinctions.  I’d be very interested in seeing a fully-worked out analysis of Stevens drawing on this insight.

Of course there was much more to see and hear (and eat and drink and argue over) at the conference, but I’m glad I was pulled into the little demimonde of the Stevensians, and I’m glad to see rhyme returning as a category of serious critical analysis.


  1. I lost an earlier post. I wonder if Stevens's "birthday books" to his wife to be survive. I expect they were mostly rhymed. Williams tried rhyme but abandoned it after Pound disdained it, even though Pound used it early on. Frost of course never abandoned it. Now that I think of it, Stevens must have abandoned rhyme and iambic pentameter close to each other, though the I5 persisted for a long time. Eliot, of course, used it early on, then less as time went.

  2. Gwynn, these are strange comments. Both Williams & Pound have early & late rhymed poems. Stevens of course never abandoned pentameter, or for that matter, rhyme. These are just facts, demonstrable by looking at poems.

  3. Bill Manhire3:07 AM

    One rhyming/assonantal effect I sometimes notice is a kind of semantic drift. For instance, in Plath’s “Daddy”, all the “black” vowel sounds – black itself (repeatedly), ghastly, Fascist, vampire, panzer-man, bastard, rack, fat, Meinkampf – leak back into the title, with the result that the word “daddy” is no longer happy and yappy, but totally terrifying.