On Translation: Rhyme and Form

An expansion for: Nine Maxims On Translation
E Bruce Brooks / University of Massachusetts / 5 Dec 2002

Rhyme and form are among the most aggravated issues in translation. The matter is further aggravated, for the aspiring accurate translator, by the decline in the prestige of rhyme and form in Western literature over the past century. It is almost a given with publishers and reviewers that rhyme in a poem (or translation) is "forced" and "artificial." That social fact does not solve the problem. It does make the solution to the problem difficult to apply. Rhyme and form are deeply connected; we here discuss them largely in terms of rhyme.

Somewhere in the 20th century, it became common practice to render Chinese poems (all of which rhyme) in unrhymed English. The prefaces of these versions were full of justification, typically bearing on the difficulty of using rhyme. Waley did perhaps more than any other translator to establish a metrically loose and unrhymed format for Chinese poetry in English. Here is the last paragraph of Waley's 1960 revised preface to his influential book One Hundred and Seventy Chinese Poems:

"Chinese poetry rhymes. At the time when these translations first appeared (1917), rhyme was considered the hall-mark of poetry, and there are still people who consider that a translator of poetry who does not use rhyme has not done his job. But rhymes are so scarce in English (as compared with Chinese) that a rhymed translation can only be a paraphrase, and is apt to fall back on feeble padding. On the whole however, people are used nowadays to poetry that does not rhyme, and I think lack of rhyme will not be generally felt as an obstacle."

In the same period, some translators of Japanese poems (none of which rhyme) produced rhymed versions in English, and turned such forms as the five-line tanka, with its 5-7-5-7-7 syllable count, into into English quatrains. The prefaces of these versions were also full of justification, some noting the difficulty of having readers accept an unrhymed short object as a poem, others simply saying "I like rhyme in a short poem." Here is H H Honda (1967) on the question of the form of the tanka:

"As for me, I prefer the quatrain as a pattern for Tanka's translation to the quintet. And if I have to give reasons, the chiefest is the latter's irregularities of beat which appear almost of themselves. In my opinion poetry must first appeal to the ear. In the quatrain the flow of rhythm, if the verse is well written, is not disturbed, whereas in the quintet, as I have already mentioned, hitches in the rhythm are unavoidable unless the measure be trochaic."

It seems that Honda is here accusing the tanka itself as being a bad form. The Japanese poets ancient and modern, he implies, have passively accepted an ugly and inadequate vehicle for the expression of their thoughts. This ignores centuries of literary criticism in Japanese, extolling the subtle skill with which Japanese poets from Nara to now have handled the possibilities latent in the asymmetrical tanka form. It ignores the fact that dynamic asymmetry has been expounded, at book length, as the soul of all Japanese aesthetics. Honda or no Honda, it would seem that Japanese poets and their readers, over the centuries, have not greatly chafed at asymmetry in general, or at the supposed "irregularities" of the tanka form in particular.

They have reveled in them.

One sometimes wished, in those years, that one could get the Chinese and Japanese translators to switch places, so that they might at once indulge their propensities and respect their originals.

One fact rarely mentioned in these apologetic prefaces is that a strongly metrical poem, especially when the metrical facts include couplet parallelism, tends to want rhyme, whereas a less strongly metrical poem tends to avoid rhyme. The lines of most Chinese poems, certainly those which set the standard for the tradition, have the same number of syllables, and the same number of stresses. They often involve close parallelism, noun matching noun and verb matching verb. And they also rhyme. Conversely, almost no Japanese poetic forms use lines of equal length, and Japanese is not itself a strongly stressed language. It comes as no great surprise, then, to learn that Japanese poetry does not rhyme.

As an implied general theory, this can be tested. Classical Latin poetry is quantitative (length-based) rather than accentual (stress-based), and its typical stanza forms are polymetric. It does not rhyme. On the other hand, as classical Latin mutates over the centuries into mediaeval Latin, we can see the poetry developing an accentual basis, and a tendency to parallelism. At the same period, we can also see it taking on rhyme as a standard feature (the student songs; the Dies Irae).

With all this in mind, it is instructive to watch Waley at work on Chinese poems. He speaks at length in his prefaces about Hopkins's "sprung rhythm" (partly to make clear that he invented it before Hopkins), but his own rhythm in translations might better be called "unsprung." It tends to avoid rendering parallel couplets as strongly parallel, or preserving the caesuras or the stress counts of single lines. To the extent that he does bring these features over, the result tends to ask for rhyme as a natural completion. This creates a problem, and Waley will sometimes address the problem by throwing in an awkward and clumsy line, to break up the texture of parallel and periodic diction in the original when it threatens to emerge too strongly in his translation. He sees the danger, and he heads off the poem's wish to complete itself on its own terms by further violating its terms. This is not so much a rendering of the original, as an attempt to keep the original at bay.

Waley was in contact with the literary tastes of his own day. His own day, as the contemporary playwright Somerset Maugham has noted, saw the decline of verse as a stage medium, and its replacement by conversational prose of crushing ordinarity. Observing all this, we can perhaps understand why Waley avoided the songforms (tsz), why he so violently disliked Li Bwo (his book on Li Bwo is The Monograph That Should Never Have Been Written), and why he concentrated most on the poets who would be least ill-served by an informal, conversational medium (the chatty and comfortable Bwo Jyw-yi, the voluble and unconventional Ywaen Mei). Chinese poetry, as Waley brings it over in quantity, is an old man's poetry: genial but toothless.

A general objection to Waley's procedure is that it works best for the least characteristic Chinese poets, is increasingly at a loss with the more characteristic Chinese poets, and is completely helpless with such things as the formal intricacies of the polymetric verse forms, in the hands of any poet at all. The cure, one would suppose, is for the translator to give up the struggle to cut down the Chinese original to the tastes of the Bloomsbury public, and to reproduce such features as rhyme and meter when they are present. Following this advice is technically difficult, but translation is acknowledged to be technically difficult. One who shrinks from that particular technical difficulty can always find fulfilment in some other virtuoso performance. Skiing. Recorder playing. Weekends at Bloomsbury, commenting adroitly on performances of Walton's "Facade."

Here are two test cases of rhyme and form in translation, for those viewers who may wish to pursue the subject a little further:

There is also an extensive consideration of rhyme and form among other features of a poetic original in the following series, which are mentioned in a different context later in the lecture:

The ultimate problem with rendering rhymed regular form as unrhymed polymetric form, and vice versa, is that it lies. It is no use acknowledging with one side of one's mouth that "Chinese poetry rhymes," and then presenting Chinese poetry out of the other side of one's mouth as though it did not rhyme, and did not need to rhyme, and is better off for not rhyming. What is true of the poem must also show up in the translation of the poem, or people are going to draw the wrong conclusions. Carl Sandburg, in a collection of his own poems for children, which he titled Wind Song, defended his own unrhymed style of poetry by citing what he thought was the example of "the great Chinese poets Li Bwo and Du Fu." Alas for Sandburg: his informants had lied to him. Neither Li Bwo nor Du Fu ever wrote an unrhymed poem in their lives.

When the major poets of one culture are being misled about the achievements of the major poets of another culture - not little crabby fine points, but the most basic and constitutive ones - it is surely time to call a halt. To quote Waley again, "Chinese poetry rhymes." There is the beginning of the matter, and there should be the end. Let the translators leave the rhymes where they found them, and where the original poets thought it good to put them.

And if they find themselves temperamentally ill at ease with the poets they have chosen to translate, well, there are plenty of other poets, and poetic cultures, left to choose from.


All lectures and abstracts posted on this site are Copyright © by their authors.

Back to Lecture Page

5 Dec 2002 / Contact The Project / Exit to Lectures Page