Lady Kasa: MYS 0598

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

We here compare rhymed and unrhymed translations of the same Japanese poem (Manyôshû #598). The original, let it be said at once, does not rhyme.

It does do certain other things. Poetry in the Nara period was largely in two modes, both defined by a texture of alternating lines of 5 and 7 syllables (end-stopped, in that no word, and no group of a noun plus its postposition, is ever broken across a line end). The 5-7 syllable couplet is the module. The poem is built of two or more such modules, capped by a concluding 7-syllable line. The "short song" (tanka), has the minimum two modules plus the final 7, or 5-7-5-7-7 in all. The "long song" (chôka) has three or more modules (one example in the Manyöshü anthology, by Hitomaru, reaches seventy-four modules) plus the same final 7. The tanka has a caesura after the second or third line; the latter (5-7-5 / 7-7) gives a greater sense of final repose. Japanese in the Nara period had only open syllables, and only eight vowels. Any poem of nine lines or longer could not help having at least two lines end in the same vowel. Poets did not avoid such lines, but as in classical Latin verse, they are ignored when they occur (like lines 2 and 4, which both end in -u in the poem below). They play no structural role in the poem.

This is one of twenty-four poems addressed to the young and eligible Ôtomo no Yakamochi, heir to the headship of the Ôtomo clan, by Lady Kasa, who was only one of his many female correspondents at this period, c735.

All twenty-four, along with some smaller groups by the same lady and from the same affair, are preserved in the anthology Manyôshû (MYS), which in its final form was edited by Yakamochi himself. In real life, he let the affair with Lady Kasa lapse after one or two encounters, but as an editor, he has preserved for us in great detail her feelings as she moves, poem by poem, from first happiness to neglect and illness. Our sample poem is from late in that progress.

MYS 0598
Tanka (five lines in 5 + 7 + 5 + 7 + 7 syllables,
no rhymes, division after line 2)

To object to rhyming quatrains would be to eliminate much of English poetry, not to mention Chinese poetry. Rhyming quatrains are obviously valid. The problem of the rhyming quatrain in the first version above is not that rhyming quatrains are bad. It is that the rhyming quatrain used by the translator has a different shape, and gives a different effect, than than the asymmetrical, onward-pressing, unrhymed form of the original Japanese poem. Honda's version also omits the key image of the stream (kawa) without (na-) any water (mizu > mi), which describes the poet's feeling of desolation and neglect, it transfers the poignant last line to the less climactic second position, and it loses the effect of the increasingly small time units in that line (his "by day and night" means merely "all the time").

The second version restores the key image of the waterless river, though as the proper name, which is correct, but in which readers will not perceive the meaning component which contributes to the poem. It also restores the climactic line to its proper position at the end, though with only two of its three time-units functioning to give a sense of urgency to the poem. The poet will soon die unless she is renewed by the attention (rainfall) of her lover. All these moves are certainly in the right direction, as far as they go.

Probably it is best to go all the way - to leave at least the most signficant of the poet's materials, and arrangements of those materials, where the poet left them. Anything added or taken away from that short list of features will distance the reader that much more from the expressive force of the original. The point of any translation is to convey the expressive force of the original. Otherwise, why bother?

The 5-7-5-7-7 tanka form was the only one available for a short Nara period poem. Its choice was thus somewhat inevitable. Still, that choice was made. The poet could instead have written a long poem, or invented a new form altogether; instead, she acquiesced in the standard of the day. The poem she made is intimately shaped to that form; it flows like a stream, in the bed that the form defines for it. Is there a test case for this statement? Yes: Nara poets did sometimes depart from tanka standard prosody, but they did so with expressive intent, not carelessly or arbitrarily (Brooks Embassy 272f). What Lady Kasa found effective, we also need to make effective. It is not our task to joggle her elbow, or invent new conventions for Nara poetry, or rewrite the history of the Japanese language. We start where Lady Kasa began, and we try not to stray from that point any more than we can help.

We will return to Lady Kasa later, and also to this poem, as one of a series of four whose interrelationships may concern a translator. To jump ahead to that series of notes, click here: To Lady Kasa's Series


From Yakamochi, Copyright © 1967, 1996 by E Bruce Brooks.

Back to "Rhyme and Form"

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