Or, in this case, not translating them. Matt of No-sword has a typically irresistible essay in Néojaponisme, discussing the bizarre haiku translations of Harold J. Isaacson, who rather than trying to render the kireji (meaningless words that “supply structural support to the verse”) in English simply leaves them there, little lumps of undigested material, to baffle and alienate the reader. Sure, he explains them in his introduction and provides footnotes for other undigested words (“water is poured out to/ the fukujusō*”), but as Matt puts it, “This style of translating is almost passive-aggressive in its demands on the reader. Shiki is serious business, it says. If you want to read him, there will be homework.” I’m all in favor of a little ostranenie, but this is ridiculous.


  1. kana i believe can be translated as ‘maybe’
    ya has meaning of ‘and’

  2. Not really, and besides, as I commented over there, “kana is an English word of 300 years’ standing, meaning ‘Japanese syllabic writing,’ which irrevocably conflicts with its use in this context.”

  3. I’d think that that meaningless words as in refrains of songs and filler in nursery rhymes would be closest, though most of them don’t seem to fit the tone or the rhyme of haiku. E. G., “Green grow the rashes, oh”, where the “oh” isn’t really the vocative or the common interjection, I don’t think.

  4. But I can’t see inserting “dilly-dilly-down” into a haiku.

  5. It would fit perfectly!
    Ah, the ancient pond;
    frog jumps in, water splashes:

  6. “This style of translating is almost passive-aggressive in its demands on the reader. Shiki is serious business, it says. If you want to read him, there will be homework.”
    Passive-aggressive? How about childishly pompous?It’s “I know something you don’t know!”
    Lucky for that translator he was not working with a language that had evidential markers or hard stuff like that.

  7. Oh, Carpathia,
    Run quickly to the lifeboats,
    Dilly-dally drown.

  8. A.J.P. Crown says

    The trouble with haikus is the amount of math involved.

  9. Actually, though, the translator is saying “I don’t know any more than you do about this word”.

  10. Have collations of the various kareji provided any insight? IIRC, a given kareji will normally be attached to a certain word (“mountains tiring to the feet”), or maybe several words. Do they collate with moods or themes?
    Traditional Chinese dictionaries invented definitions for both syllables of the few two-syllable words in the oldest Chinese texts. “Hudie” is butterfly, and Hu is supposedly the male, and Die the female — or something like that. Do Jpanese scholars do the same with kareji?
    It’s worse with modern Chinese, because there are a lot more two-syllable words, and often the two syllables are synonyms, but in combination they have a very specific meaning. And often the second syllable has more than one written form and more than one pronunciation and half a dozen seemingly unrelated meanings. And in some cases, neither of the two components, nor the combination of the two, has no relationship at all to the meaning of the binome.
    UPDATE: I was thinking of Makura kotoba or “pillow words”. Kireji look like Chinese final particles, which are functional but often can’t be translated by English words. For example, a question ending in “ma” is just a question, whereas a question ending in “ne” is more like a humble suggestion, or a request, or a plea, or a nag.

  11. John, they can be translated into English very well, but often the translator has spent more of his life studying Chinese than his native English, and simply assumes he is competent in it.
    So for instance ‘ne’ usually corresponds to ‘what about…’ or as for’.
    ‘Ma” as a question is different from V notV questions, more insistent, and you can use ‘Well,…..?’ or simialr expressions.
    ‘Ba’ is similar; you use ‘let’s’ or ‘why don’t you…….?’

  12. What I meant is that the sentences can be translated, but you don’t do it by finding an equivalent English word.

  13. “but you don’t do it by finding an equivalent English word. ”
    That goes for a lot else – you can find very common everyday sentences where that would be true for about every word.
    Eg. “He took the boy home.”
    “Ta1 ba3 hai2zi bao1huiqule.”
    Not one exact correspondence in the entire sentence.

  14. Thanks for the link again!
    Read, to expand on LH a bit, those are the modern Japanese words “kana” and “ya” that are only distantly related (if that) to the ones found in haiku.
    Jim + John: Right… Isaacson apparently believed that since kireji are integral to the original, any translation must have some component that corresponds to kireji 1:1, or something vital is lost. When he found that English has nothing that readily corresponds to kireji, rather than rethinking his theory (or just abandoning haiku translation altogether as impossible), he decided to bring the kireji across unchanged. This was his big point of departure from the rest of the translation world.

  15. but the intonation is the same, kana expresses wonder, as if one suggests something perhaps is this or that or it might be….
    ka also sounds as if it’s something like one is asking some, kinda rhetoric, question and ya also sounds like one describes one image and another one in contrast or in support of the first image
    and my Japanese colleague said haiku is just simple ‘asobi kokoro de yaru mono de’ and it’s nothing like that serious or deep metaphysical as it sounds in English or other translation

  16. I don’t know what to tell you, read. You’re wrong about the words — you’re confusing modern Japanese with the Japanese of centuries ago — and as for what your Japanese colleague says, I’m sure I could find a few English-speakers who’d say “Poetry? Eh, just a bunch of rhymes about flowers or whatever” too.

  17. I say Read has got it exactly right. She has filled out the detail on the wiki. Here is Hat’s haiku reworked with the meanings of the meaningful sounds added. I had to give the frog some eyes to get the math to work out right.

    The ancient pond, yes?
    Eyes blink and water splashes:
    Frog jumps in, maybe.

    “right?” “yes?” “no?” “isn’t it”(ka)…….. [wiki]: at the end of a line to indicate a ([Read]: rhetorical) question
    “and” (ya)…….. [wiki]: implies an equation [Read]: contrast two images
    “maybe” (kana)………. at the end to indicate wonder
    Maybe wonder (kana) would be better expressed as “ah, yes”, and I’ll substitute “then”=ya, and “right”=ka…so here’s the same poem again, but with different English words used for the kireji:

    The ancient pond, right?
    Eyes blink then water splashes:
    Frog jumps in. Ah, yes!

    For comparison, here it is with the kireji just in Japanese.

    The ancient pond, ka?
    Eyes blink ya water splashes:
    Frog jumps in, kana!

    I rather like Hat’s rendition as it seems so effortless, but he has also Americanized it.
    I much prefer adding some word in English that captures both the meter and the meaning (emotion?) of the missing syllables.(is that what Read means by capturing the same “intonation” and “sound”?)(Perhaps this is something like the convention of the Greek chorus, which tells the audience how it is supposed to react?)
    I always thought that, as far as traditional subjects for haiku, the haiku was not supposed to have a heavy philosophical or serious meaning, but be more of a snapshot of a simple frozen moment or a fragment of nature. The constraints of the math (17 syllables) rather force one to be succinct–any deep meaning would have to be by metaphor.

  18. Oh, dear I suppose I’d better go look at Hat’s comment on the other blog.

  19. Nijma, I think you’ve gone a bit off the rails. The original of that haiku is “furuike ya/ kawazu tobikomu/ mizu no oto”. There’s no kana in there, for one thing. I’m not going to discuss the meaning of the words any more as I think my position is pretty clear. We can agree to disagree on that.
    Personally I’m very sympathetic to translators reworking old material into something new, but only on the understanding that it IS something new rather than an example of what most people mean when they say “translation.” By adding blinking eyes, you are changing the original in a very important way, to the point where I would say you are really just riffing off Basho rather than translating him. (For example, there are many ways to interpret the original, but one is that it’s reconstructed in the speaker’s mind starting from the end: he remembers the old pond outside — because a frog has jumped into it — which he knows because he just heard the noise. Bringing a non-deducible visual into the mix destroys this.)
    I much prefer adding some word in English that captures both the meter and the meaning (emotion?) of the missing syllables.
    Some translators agree with you in re the meaning, and use punctuation or “oh!” or whatever — others disagree and think that English can do its own equivalent thing. This is mentioned in the original article.
    As for the meter, though, there really isn’t any reason to insist on 17 syllables in English, and most serious English haiku poets don’t any more. A Japanese mora isn’t equivalent to an English syllable. Again, serious English haiku poets tackle this in different ways; 2/3/2 is quite common (that is, 2 strong beats + any number of weak beats on the first line… and so on).
    As for profundity, yes, I am quite happy to concede that the profundity of haiku is often overstated. Certainly there are any number of frivolous and lighthearted haiku, even among those haiku considered exemplary of the form, and as you say it is usually considered gauche for a haiku to have a heavy philosphical or serious “meaning.” (Issa got away with some meaningful stuff, but only because he was so self-deprecating and playful.) Generally speaking, haiku are about implication and juxtaposition, which can indeed be much more than just a word game. Certainly Shiki did not believe he was playing Scrabble; he took his haiku very seriously, and had no time at all for haiku writers who didn’t.
    furu yuki ya/ meiji wa tooku/ nari ni keri
    “The falling snow/ Meiji; how distant/ it is become”
    (This one is by Nakamura Kusatao.) Some people think this is a mawkish, sentimental nothing; others praise it highly as a moving glimpse at a man’s thoughts on a snowy day, as he is reminded of how long he has been alive. There’s no right answer; there’s no attempt at “Ask not for whom the bell tolls” style profundity. It is even, arguably, playful in a way. But that doesn’t necessarily make it trivial.
    Anyway, I’ve dominated this thread enough, so I think I’ll bow it gracefully. Thanks for reading, folks.

  20. DOn’t bow out! We welcome invasions of this sort!

  21. Somewhat on topic, my favorite translator of poetry from East Asian languages is Kenneth Rexroth, who is also one of my favorite critics / cultural critics. His translations convey the compactness of those poetries better than most translators, notably Waley. I don’t read Japanese and I’ve never checked his translations against the Chinese, but that’s not really an issue for me. What he says is that some translations are nearly literal and some are imitations, interpretations, or take-offs on the original.

  22. Matt, you’re the only one here who actually knows what they’re talking about, so don’t you dare wave your hat and gallop away!
    Nijma, I assure you Matt is right and read is wrong.

  23. admittedly i’m not an expert in haiku, i told just my impression of reading and listening to it
    i don’t believe that language would become that changed in 2-300 yrs only so that modern words would have evolved into something entirely different or that in haiku people would use just meaningless syllables, for just ornament or size? or that it can’t be translated, there should be some meaning and those ‘modern’ equivalents i imagine would have the closest meaning to them, the untranslateables
    if you say it has different meaning because of its different kanjis or something i would accept that i’m wrong, not until then i guess

  24. All right, if you insist…
    Read, here’s an example for you:
    1) Haiku /kana/ is from /ka/+/na/, both exclamatory particles appearing after the conclusive form. This /kana/ arose to replace /kamo/ during the Heian period when it became uncool to end sentences with /mo/.
    2) Modern /kana/ is from /ka/ (interrogative particle) + /na/ (exclamatory). That’s why you often see it written “ka na” — it’s so recent that it hasn’t had time to become a single word, so to speak.
    The essential difference is that modern /kana/ contains the interrogative meaning while haiku /kana/ doessn’t. Also although kanji are really the last thing you should be using to judge Japanese etymology, haiku /kana/ is often written 哉 but modern /kana/ never is. They really are different words, treated differently, with no direct line of descent.
    I can respect that you don’t want to just accept what some random guy on the internet says when it goes against the information you do have. But sometimes we random guys on the internet actually do know what we’re talking about.

  25. What? You mean Hat was translating a REAL haiku called “furuike ya/ kawazu tobikomu/ mizu no oto” and not just making one up to use as an example with “dilly-dilly-down”? Hat is always full of surprises. But of course, if you are translating a known poem, you can’t add eyes where there weren’t any before–it would be better to add a few she-bops and du-wops to make up the extra syllables, if you insisted on keeping the 5-7-5 thing going in translation.
    Matt seems to be saying here that the kireji are totally meaningless. But on the other blog he says “A Japanese reader may not know the long and storied history of kana and keri, but at the very least they are known to be Japanese words often found in haiku. They are not jarring to a Japanese reader.” In other words, they are poetry conventions that Japanese-speaking aficionados of haikus are familiar with, much as English speakers would be able to derive some meaning from archaic English like “lo!” and “alas and alack!”. So now Matt has only confused me further.
    The wikipedia article also lists meanings for the various kireji, as the Japanese article linked appears to, and they dovetail quite nicely with Read’s meanings.
    I don’t think everyone is “read”ing Read correctly. I don’t know how many languages she dabbles in–Mongolian, Chinese, Russian?–but English is one of her more recent ones. In some of those languages sound (pitch?) is important in ways we in the west can not understand. (I am American, from the midwest.) Also she is a scientist, so she thinks in symbols. And she studies classical Asian poetry for kicks. You can’t look at the conventional face meanings of what she says word by word as if you were speed-reading, you have to think of the concept she is painting.

  26. Oh dear, a cross post.
    So “kana” doesn’t express wonder, only excitement? or exclamatoriness?

  27. i didn’t dispute that you don’t know what you are talking about, i just told what my impression is, and after your satisfying explanation i admit that kana and ka na are then different in haiku
    what about ya, though
    is it ‘and’? or something else
    nari keri i think is some action that is still in the state described in the haiku as if it’s the past continuous tense, as if the image described is captured and presented as it was before the haiku writer, no?
    and basically any poem/art is for the reader/ perceiver and that’s important what the reader perceives and imagines, the poem itself, haiku or orther style, its forms and canons, maybe it’s all science and art and i don’t get it
    so what i said was that haikus are perceived by me as being playful and inviting to share the emotion of the haiku writer, that’s all

  28. So Hat’s comment over there makes a lot of sense. The whole idea of translating a poem is to make it accessible to people who do not read the language. This translator has failed to translate words or word fragments that do add meaning to the poem. Leaving them transliterated without even an explanation is highly unsatisfactory, although an improvement over the convention of leaving them out altogether or substituting dashes. And then there was Hat’s point about mixing transliteration and translation. If the sound itself is that important the whole thing should have been transliterated. Homework for the reader, yes, but without notes it’s homework only those who don’t need a translator are capable of doing. But if you want to read Shiki, and don’t read Japanese, I suppose this is the only translation available. So if you’re lucky you’ll stumble across a blog or two that explains the kireji a little better.

  29. thanks, Nijma
    furuike ya/ kawazu tobikomu/ mizu no oto”
    the old pond and the frog jumping into(it)/ the sound of water
    so i imagine that sound, something blip i guess and that’s all
    i’m not sure about all the articles, if i could i would omit all of them
    so what i’m missing, tranquility and all that is included into the blip i guess

  30. -r

  31. “but English is one of her more recent ones. In some of those languages sound (pitch?) is important in ways we in the west can not understand. (I am American, from the midwest.) ”
    Nijma, not so, not so – I am in the west – Seattle to be exact, so that makes me more western than you? – and I have absolutley no problem understanding how important pitch tone is in Cantonese or contour tone is in Mandarin, in the same way a Mandarin speaker can understand how important our gnat’s-ass tiny vowel distinctions are eg. bet/bat or our inaudible distinctions eg. dog/dock or worse yet east/yeast – we both learn the language where those distinctions are important. I can assure you that when I am trying to talk in Mandarin to someone from Fuzhou who never made it past the thrid grade, that the way he mangles the tones is at least as confusing as the hash he makes out of any other phonetic feature of the language.
    And pitch and tone are very important in English. Intonation works like punctuation in English to show where clauses begina dn end and how they fit with each other. If you doubt this, go call a ccallcneter in India try to get some business done. Set aside a spare afternoon when you have nothing else to do.

  32. furuike ya/ kawazu tobikomu/ mizu no oto”–by the way, is that Shiki?
    “ya” marks the equation, the division between the two images that are suppose do be contrasted. On the one hand the pond, its ancientness, its waterness, its lack of movement in time and space. On the other hand the frog, a temporary, short-lived creature frozen in a moment of movement. The two opposite qualities of pond/frog intersect and are balanced in the moment of the poem.
    Native American tradition has stuff about water as an element and frogs as magical creatures that span the elements of water, earth, and air, but now I’m probably reading too much into it.

  33. “Ribit / splash / plop!”

  34. is that Shiki?
    No. Basho.
    Hiroaki Sato has two books titled One Hundred Frogs: the short one giving a bunch of translations of this particular poem, and the longer one with additional theoretical essays.

  35. is that Shiki?
    It’s Bashō.

  36. Curses, pipped again!

  37. Imagine that, Hat has snuck in the back door with only the most famous poem of a terribly influential poet while acting like he was just playing with some random words over his cornflakes.
    And for some reason, out of all the words on the whole thread, those were the ones that resonated with me. No wonder, as wiki says, it “became instantly famous”. As Read says,”what the reader perceives and imagines”. For Read it was the emotion of tranquility and then, blip! For Matt the sound comes first and the unseen pond visualized from it.
    And now we have 102 translations.
    Hat: Ah, the ancient pond;/frog jumps in, water splashes:/dilly-dilly-down.
    Read:the old pond and the frog jumping into(it)/ the sound of water
    What a treat to be introduced to this poem without knowing first what it was!

  38. Read, I don’t want to turn this thread into Medieval Japanese Particle Corner, so please forgive me if I leave “ya,” “keri,” etc. alone. Of course everyone’s free to think what they like of haiku and enjoy them (or not) in their own way. I just hope that I have demonstrated that in Japanese as in other languages, even if words seem the same, they’re often not. There are technical issues surrounding words (rather than interpretation) that do have a correct answer.
    This translator has failed to translate words or word fragments that do add meaning to the poem. Leaving them transliterated without even an explanation is highly unsatisfactory, although an improvement over the convention of leaving them out altogether or substituting dashes.
    Not neccessarily. Some translators believe that the function of words can be performed by other things in English, like punctuation, sentence structure, word choice elsewhere in the poem, line breaks. Like Jim says far above, translation is not about lining up words that correspond 1:1 with the words in the original. The standards you propose would render basically _all_ translation “highly unsatisfactory.” A valid position, but a rather extreme one that doesn’t allow much room for discussion afterwards.

  39. I don’t want to turn this thread into Medieval Japanese Particle Corner
    Aw, you’re no fun.

  40. The standards you propose would render basically _all_ translation “highly unsatisfactory.”
    Yup. There’s no ideal, but since I don’t read the language I would like to see an English rendition, as short as possible, and not necessarily grammatically exact, but to give the flavor, and side by side the transliteration with copious footnotes on the same page as the poem that I could ignore or not.
    I can forgive this translator anything except the use of endnotes.

  41. Nijma, I think that Rexroth’s translations of Japanese poems have the transliteration plus the translation. Waley’s “The Uta” has notes and a sketch grammar, but I don’t think his translations are as good.

  42. David Marjanović says

    Native American tradition

    Which one?

  43. “Which one?”
    Don’t remember offhand. I’ve been in so many museums. When I travel around I like to drop into the little local ones and look at artifacts. I remember the frog is special because it metamorphasizes from being a water creature to living on land. For some reason I seem to associate it with the Kankakee flood exhibit, so maybe it was in downstate Illinois. That doesn’t help much though since around 800-1200 that area was a sort of confluence of the continent’s three major native civilizations.

  44. The Paraverse translations give the Japanese in kanji and transliteration with literal translation, and then (usually) several possible translations side-by-side (example). Plus notes on translation issues (and whatever else Robin was reminded of). They also gather together thematic collections, rather than just greatest hits.
    Also see an earlier LH post on the frog poem, also inspired by one of Matt’s No-sword’s posts.

  45. Why is the “Old Pond” by Basho translated as if there are three kireji in it?

  46. A.J.P. Crown says

    ‘Bash’, in Norwegian, (it’s actually spelt bæsj, but pronounced ‘bash’) means ‘shit’, with special reference to animal droppings. So, ‘The Old Pond by Basho’ sounds pretty odd in Norway.

Speak Your Mind