Translation Philosophies and Their Consequences: Part 2
In Part 2 I will compare some translations of the Japanese work “One Hundred Poets, One Poem Each”, the “Ogura Hyakunin Isshu”, sometimes shortened to simply the “Hyakunin Isshu”. The ‘Ogura’ is an anthology of 100 Tanka by 100 different poets. Tanka is the most significant and most enduring form of Japanese poetry. It has a written history of about 1400 years. Tanka is the basic form from which both Renga and Haiku have emerged. It is a formal tradition, retaining its syllabic structure over all these centuries. The syllabic structure is 5-7-5-7-7, for a total of 31 syllables. It is a form that is loved by all strata of Japanese society from the Imperial Court to the working peasant.
I have chosen this work to compare the consequences of differing translation philosophies for two reasons. First, it is one of the most widely read poetry collections in Japan. It is hugely popular. It is so popular that a card game has been created from the anthology which is often played during New Year celebrations. The card game requires that the Tanka of the anthology be memorized. The way the card game works is that each Tanka is divided into two; the first three lines or ‘ku’ of 5-7-5, and the concluding two lines, or ‘ku’, of 7-7. The cards with the concluding lines are laid out on a table, or the floor. The person leading the game then draws from the cards with the first three lines. The leader then reads the first three lines and the contestants then have to find the correct card that has the concluding two lines which match the three lines just read. I have heard that some people are so conversant with this collection that all they have to hear are two or three words and they know immediately what to look for.
In addition to this card game, the ‘One Hundred Poets’ has generated a great deal of art. First, there are the illustrations on the cards for the game. But the Tanka have also generated a lot of paintings based on the individual Tanka in the collection. There are art books devoted to these paintings.
Given all of this it is clear that the ‘Ogura’ anthology has had an enduring impact on Japanese culture and poetry.
The second reason I have chosen this work is that there are numerous translations into English available. And these different translations reflect, in a way that is very clear, the different philosophies of translation. Understanding this helps us to understand why the different translations are so different.
Let’s take a specific example. Here is Tanka 88 in three versions:
Because of one night,
A lovers’ nap on those reeds
Of Naniwa Bay,
Ought I, giving my body,
Love you always, do you think?
Translated by Tom Galt
I’ve seen thee but a few short hours;
As short, they seemed to me,
As bamboo reeds at Naniwa;
But tide-stakes in the sea
Can’t gauge my love for thee.
Translated by William Porter
For the sake of one night
on Naniwa Bay
short as the nodes
of a reed but at the root
what is left for me?
Like the wooden
out in the sea
must I, too,
wear myself out
pining for my love?
Translated by Peter McMillan
The first translation by Tom Galt is an example of literal, or word-for-word, translation philosophy in application. The syllabic structure of the Tanka, the 5-7-5-7-7 count, is retained. Interestingly, the Galt translation is the most explicitly erotic. By mentioning the body directly, Galt expresses the charged nature of this Tanka.
The third translation by Peter McMillan is an example of a meaning based translation. Notice that all connection to the formal constraints of Tanka are gone. The translation has eleven lines. The syllable count is 51. This is a free verse poem. Notice also the run-on from L6 to L7, a common free verse convention. Also in this version the erotic element is less explicit, the consequences of the one-night stand not as clear.
The second translation by William N. Porter takes a different approach. Recognizing that Tanka is formal verse, Porter translates the Tanka but instead of mimicking the formal constraints of the Japanese, Porter creates a formal structure that he believes is more resonant of English language poetry. Porter writes in his ‘Introduction’, “A tanka verse has five lines and thirty-one syllables, arranged thus: 5-7-5-7-7; as this is an unusual meter in our ears, I have adopted for the translation a five-lined verse of 8-6-8-6-6 meter, with the second, fourth, and fifth lines rhyming, in the hope of retaining at least some resemblance to the original form, while making the sound more familiar to English readers.”
Porter’s approach is to recognize that Tanka is formal verse and then translate the verse in such a manner as to align the translation with formal verse conventions as found in English. For Porter, who was working in 1909, this means an iambic line (hence the even numbered line count) and the use of rhyme. Porter retains the five-line structure, but curiously he reverses the relationship between the long and short lines of traditional Tanka. In traditional Tanka L2, 4 & 5 are long, 7-syllable, lines; while L1 & 3 are short, 5-syllable lines. Porter reverses this with L2, 4 & 5 being short 6-syllable lines, and L1 & 3 being longer, 8-syllable lines. I find this curious; but that’s the way he decided to shape his translations. The three short lines, L2, L4, and L5, are further marked by a shared end rhyme. Japanese poetry does not use rhyme as a structural element of its forms. But English poetry does and Porter thought that adding rhyme would indicate to the English reader, who is used to rhyme in formal verse, the formal nature of the original poetry. Interestingly, Porter is not the only translator to adopt this strategy. H. H. Honda published a translation in 1956 which transforms the Tanka into rhymed Quatrains. It has not been as successful as the Porter translation, so I’m just going to note it in passing.
Both the Galt and Porter translations give the reader a feel for the formal nature of the poems in the anthology. That is to say all the poems in these translations have the same form. For the Galt translation that means they all adhere to the 5-7-5-7-7 syllabic structure, unrhymed. For Porter that means that they all adhere to the 8-6-8-6-6 syllable structure with an end rhyme scheme.
For the McMillan translation all of the poems vary as to lineation and syllable count. No two consecutive poems have the same form. The consequence of this is that the reader is not able to compare the formal relationships among the poems. In a way, I would say that McMillan’s version is so far from the originals that they are, in a sense, different poems altogether. McCullough put it that for some a poem in a foreign language is an occasion for writing a new poem in their own language. And I feel that this is what has happened with the McMillan version. His poems are based on the poems of the anthology, but the connection to them is so remote as to be more like an inspiration for McMillan’s own poems.
Readers of this blog will guess that I prefer Galt’s translation. This is because the Galt translation is, quite simply, most faithful to the formal parameters of the original. For Galt, the form itself has meaning. And it is here where I think the literal approach to translation surpasses the meaning based approach, or the third way compromise exemplified by Porter. Porter was correct that odd-line syllable count is rare in English formal verse (especially in 1909). But that is precisely why I think mimicking the actual syllable count of the original is significant; because it opens up a new way of looking at lineation for the English reader and the English poet. If we follow Porter’s approach, we miss an opportunity to expand our poetic resources. This is something that both Porter and McMillan miss. McMillan misses it because, it seems to me, he is simply form deaf. McMillan is an extreme example of a meaning based approach where all attempt to mimic the form are rejected in favor of a strictly conceptual approach. It is a hyper-intellectual, a pure mental, approach to translation.
But clearly the form of the Tanka was significant to the 100 poets gathered in this collection. They all wrote the same form; but you would never know that by reading McMillan’s version. It isn’t only that McMillan’s version misrepresents the formal nature of the individual poems, it also misrepresents the relationship the poems have to each other. This relationship is a formal one, embedded in a tradition of formal verse.
Porter’s version is, I think, preferable in that the relationship among the poems is retained. That is to say the reader can recognize that all the poems were written in the same form; that all 100 authors wrote in the same pattern. I am sympathetic to Porter wanting to make these poems more accessible to English readers of 1909. But I also think this was an opportunity lost; in the sense that a good translation into 5-7-5-7-7 could have opened English poetry to non-iambic formal structures.
For those who are interested in composing Tanka in English, the Galt translation can serve as a guide, just as the McCullough translation of the Waka Kokinshu can. This is its great virtue. It is a genuine bridge between Japanese and English poetry. This kind of translation brings us to the deep of well of Tanka and nourishes our own efforts in that form.
In the often-overlooked, elegantly written, preface of the King James Bible, ‘From the Translators to the Reader’, it says, “Translation it is that openeth the window, to let in the light; that breaketh the shell, that we may eat the kernel; that putteth aside the curtain . . . that removeth the cover of the well, that we may come by the water.” This is an optimistic view of translation, an inspiring vision. Implicit in this vision of translation is that language is not a barrier which divides us; rather it is a characteristic which we all share, and that it is the translators’ job to build a bridge between two linguistic communities, the better to broaden our understanding of each other.