Meta: translating Japanese
Feb. 20th, 2012 11:36 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
lnhammer posting in ![[community profile]](https://www.dreamwidth.org/img/silk/identity/community.png){kind=small}
poetreeThe Kokinwakashu ("Collection of Old and New Japanese Poetry"), or Kokinshu for short, was the first imperially commissioned anthology of poetry in Japanese, compiled around 905 CE by a panel of four leading young poets of the day. This was a cultural watershed -- it firmly cemented the reputation of Japanese as a valid language for poetry, as opposed to the Chinese in vogue for the prior century and a half, and it set what was acceptable in terms of diction, subjects, and images for court poetry (and by omission, what was not acceptable) for the next thousand years minus a couple decades.
The collection has 1111 poems (in the manuscript tradition of my base text) divided into 20 thematic books. The first 6 books are seasonal, organized chronologically to follow the cycle of the year beginning and ending with New Years. Another group of 5 books are love poems following the progress of a typical aristocratic affair. Other books have travel poems, elegies, felicitations on important birthdays, miscellaneous topics, and so on. All told, the collection contains all the types of poetry a well-rounded courtier or court lady needed to be able to write -- and write they did, as part of almost every private interaction and public occasion.
Almost all the poems, including this week's, are in the short form now called a tanka, though that term wasn't invented until much later: 31 syllables grouped in lines of 5-7-5-7-7.* (No haiku, as the form wasn't invented for another six centuries.) The syllable count was not completely strict, however -- occasionally there's a line with an extra syllable, often but by no means always where two vowels can be elided together. I've chosen to imitate the original form in English -- including the looseness, by allowing myself the occasional off-count line, typically one beat short, never more than one a poem.
Aside from the usual translation problem of how words do not match one-to-one across languages, but rather overlap in meaning and tenor and connotation, the biggest difficulty with Japanese is that it's what linguists call a pro-drop language. That is, any information that a listener can understand from context can and usually will be omitted. The attitude is something like, If you have enough context to understand who a pronoun refers to, why bother with the pronoun? In everyday conversation or an extended prose passage, this generally isn't hard to deal with as there's a lot of context, but in a short, detached poem, the lacunae can be hard to fill, leaving you to ponder whether a verb describes the action of "I," "you," "us," or some other person or people.
Further complicating this is that in classical Japanese, the markers for the grammatical roles of nouns sometimes evaporate, as every syllable counts when you're working in a very tight space -- and indeed some technically optional markers (for the subject or direct object when next to the verb) routinely disappear. As a result, you can find yourself scratching your head over whether these cherry blossoms are the sentence topic, the verb's subject, a direct object, a direct address, or even an exclamation.
While it is true that in Japanese poetics, it is not necessarily expected that a poet can be pinned down to any one meaning and each reader will come away with her or his own reading, this does not change the fact that as a translator you are giving the reader your understanding. In English, things like an explicit subject are not optional and a body has to chose -- so it comes down to how you personally interpret the poem. Sometimes you can maintain an ambiguity, but not often, especially when it's grammar-based.
For what it's worth, my philosophy of translation is to render my best understanding of the original's sense while reproducing as much as possible of the poem's quality and structures -- sonic, linguistic, rhetorical, and so on. This involves not only interpretation, but trade-offs. For example, the normal sentence order of Japanese is almost, but not entirely, the complete reverse of English. This means that in a literal prose paraphrase of a poem that's one long clause, the nouns, verbs, and adjectives are in the opposite order in English from the original. If the sequence of images these substantives call up is not important to the poem's effect -- for example, if the mainspring of the poetry is a witty metaphor -- then this reversal does no real damage. But if the images form a careful progression, such as sweeping from a wide scene down to a small detail, it is probably better to maintain that sequence as best as possible, at the expense of breaking up the smooth swoop of a single clause.
Trade-offs. Just like in any poem, or any other work of art.
For this week, I've chosen three Kokinshu poems that highlight different problems of translation. I'm afraid there's far more prose than verse in these posts -- partly because the poems are so small, but mostly it's that it takes a lot to unpack these lovely puzzle-boxes that are Japanese verse. Starting on the next rock, coming tomorrow.
Until then, for discussion: those of you who translate, what is your philosophy/practice? What do you focus on the most?
* By convention, poems were and still are written in a single line, or broken as writing space dictated, just as ancient Greek poetry was; for all reasonable definitions of a poetic line, however, these have five.
---L.
The collection has 1111 poems (in the manuscript tradition of my base text) divided into 20 thematic books. The first 6 books are seasonal, organized chronologically to follow the cycle of the year beginning and ending with New Years. Another group of 5 books are love poems following the progress of a typical aristocratic affair. Other books have travel poems, elegies, felicitations on important birthdays, miscellaneous topics, and so on. All told, the collection contains all the types of poetry a well-rounded courtier or court lady needed to be able to write -- and write they did, as part of almost every private interaction and public occasion.
Almost all the poems, including this week's, are in the short form now called a tanka, though that term wasn't invented until much later: 31 syllables grouped in lines of 5-7-5-7-7.* (No haiku, as the form wasn't invented for another six centuries.) The syllable count was not completely strict, however -- occasionally there's a line with an extra syllable, often but by no means always where two vowels can be elided together. I've chosen to imitate the original form in English -- including the looseness, by allowing myself the occasional off-count line, typically one beat short, never more than one a poem.
Aside from the usual translation problem of how words do not match one-to-one across languages, but rather overlap in meaning and tenor and connotation, the biggest difficulty with Japanese is that it's what linguists call a pro-drop language. That is, any information that a listener can understand from context can and usually will be omitted. The attitude is something like, If you have enough context to understand who a pronoun refers to, why bother with the pronoun? In everyday conversation or an extended prose passage, this generally isn't hard to deal with as there's a lot of context, but in a short, detached poem, the lacunae can be hard to fill, leaving you to ponder whether a verb describes the action of "I," "you," "us," or some other person or people.
Further complicating this is that in classical Japanese, the markers for the grammatical roles of nouns sometimes evaporate, as every syllable counts when you're working in a very tight space -- and indeed some technically optional markers (for the subject or direct object when next to the verb) routinely disappear. As a result, you can find yourself scratching your head over whether these cherry blossoms are the sentence topic, the verb's subject, a direct object, a direct address, or even an exclamation.
While it is true that in Japanese poetics, it is not necessarily expected that a poet can be pinned down to any one meaning and each reader will come away with her or his own reading, this does not change the fact that as a translator you are giving the reader your understanding. In English, things like an explicit subject are not optional and a body has to chose -- so it comes down to how you personally interpret the poem. Sometimes you can maintain an ambiguity, but not often, especially when it's grammar-based.
For what it's worth, my philosophy of translation is to render my best understanding of the original's sense while reproducing as much as possible of the poem's quality and structures -- sonic, linguistic, rhetorical, and so on. This involves not only interpretation, but trade-offs. For example, the normal sentence order of Japanese is almost, but not entirely, the complete reverse of English. This means that in a literal prose paraphrase of a poem that's one long clause, the nouns, verbs, and adjectives are in the opposite order in English from the original. If the sequence of images these substantives call up is not important to the poem's effect -- for example, if the mainspring of the poetry is a witty metaphor -- then this reversal does no real damage. But if the images form a careful progression, such as sweeping from a wide scene down to a small detail, it is probably better to maintain that sequence as best as possible, at the expense of breaking up the smooth swoop of a single clause.
Trade-offs. Just like in any poem, or any other work of art.
For this week, I've chosen three Kokinshu poems that highlight different problems of translation. I'm afraid there's far more prose than verse in these posts -- partly because the poems are so small, but mostly it's that it takes a lot to unpack these lovely puzzle-boxes that are Japanese verse. Starting on the next rock, coming tomorrow.
Until then, for discussion: those of you who translate, what is your philosophy/practice? What do you focus on the most?
* By convention, poems were and still are written in a single line, or broken as writing space dictated, just as ancient Greek poetry was; for all reasonable definitions of a poetic line, however, these have five.
---L.
no subject
Date: 2012-02-20 10:15 pm (UTC)no subject
Date: 2012-02-21 12:03 am (UTC)Scholar, I am not. Fortunately, there's a lot of material aimed at Japanese non-scholars that explains the archaisms.
---L.
Yes...
Date: 2012-03-11 08:51 pm (UTC)Re: Yes...
Date: 2012-03-22 03:45 am (UTC)Re: Yes...
Date: 2012-03-22 03:52 am (UTC)Arabic calligraphy can be amazing when the letters are woven together to make pictures. The script really lends itself well to that.
Re: Yes...
Date: 2012-03-22 05:30 am (UTC)(Yeah, I look at that and go, Okay, this is a と, and that's a や, and after a couple squiggles we get the loop of a う -- and then decide that no, really, my kanji fu is weak against calligraphic attacks.)
---L.
no subject
Date: 2012-02-25 02:11 pm (UTC)One problem I met when translating Chinese poems was that verbs, pronouns and a lot of thing can be omitted to produce an ambiguity and double meaning, which when translated into English I have a make a choice between maintaining them or more clarity. The former won, but I did sometimes wonder I did too much because of my limited ability.
no subject
Date: 2012-02-25 04:16 pm (UTC)---L.
Thoughts
Date: 2012-03-11 08:49 pm (UTC)Re: Thoughts
Date: 2012-03-12 03:28 am (UTC)---L.