My Sort of Translating

by Edward G. Seidensticker

My kind of translation is literary translation of works that I like and admire.

My subject today is “My Sort of Translating.” By this I mean a kind of translating which I think probably not very many of you are much involved in. I’m an amateur. I have always been very much an amateur. An “amateur” is someone who does something because he likes it—because he loves it, literally; but I don’t like strong words like “love”; I prefer “like.”

To give you the definition from my favorite desktop dictionary, a dictionary in the tradition of the Scottish enlightenment, Chambers: “Amateur: an enthusiast; admirer; one who cultivates a study for the love of it and not professionally, often implying that he is superficial, trifling, dilettantish, or inexpert.” Well, that’s me; every one of those is me.

On the other hand . . . the word “professional” means “undertaken as a means of subsistence, as opposed to amateur; showing the skill, artistry, demeanor, or standard of conduct appropriate in a member of a profession or of a particular profession.” If you’re professionals and I’m an amateur, you’ve got it all over me! Because I’m a dilettante, I’m superficial . . . “Professional” generally is a much nicer word. “Now, that was really a professional job”—that’s high praise. “That was really an amateur job” is an insult.

That is the sense in which I use the word “amateur.” I have for the most part done translating only because I have wanted to. For the most part—there have been exceptions—I have limited myself to works that I admire. I have long had the theory, and I think it’s rather a good theory, at least in literary translation—that a good translation will not emerge unless the translator has an affinity with the original work.

I never considered myself a professional, because back in the days when I was doing it fairly ambitiously I don’t think that there was such a thing; at least not from Japanese into English. This takes us back to the 1950s. I don’t think that there was such a thing as a professional: somebody who could make a living from translating. My kind of translation has another aspect: It’s literary translation of works that I like and admire, and authors that I like and admire.

I have often said—and I think it’s true, although I’ve never really checked up on this statement—the only long literary translation I have ever done that paid me sufficiently for the time I spent on it was The Tale of Genji. None of the others would have kept me in bread and water for the length of time it took it took to translate. The Tale of Genji didn’t immediately, of course; I did not get enough money immediately to pay me for the (awful thought!) dozen years that I was at work on it. But over the years I have. I don’t think that’s true of any other long literary translations I have done.

Japanese are always asking “What Japanese writer is the most popular in the West?” They expect me to say Mishima. They’re always disappointed when I say it’s Murasaki Shikibu. But this is true: Murasaki Shikibu is the most popular Japanese writer in the West. That’s a tribute to Miss Murasaki, but it’s also a tribute to the West: the West shows very good taste in liking her that well.

I have done professional work; that is to say, work on order. Mostly it has been short work. The only long work I ever did on order was the only Mishima novel I’ve translated: The Decay of the Angel. And it was largely my fault that it happened.

I’d had too much to drink one night with Harold Strauss, the chief editor at Knopf who was responsible, more than anyone else, for bringing Japanese literature to America and, thereby, to the West. Japanese are always giving names like mine, or Donald Keene’s, as the person who did it. Well, this is wrong. The person who did it was Harold Strauss [link to “The Knopf Translation Program, 1955–76, by Larry Walker, SWET Newsletter No. 117, June 2007].

Harold Strauss and I were out drinking one night and I seem—I have no recollection of having done it—to have said to him that I would translate the last Mishima novel, The Decay of the Angel. Then, quite a while later I had a letter from him saying “How are you coming with the Mishima?” I replied, “Mishima? What Mishima?” And he said, “Well, but you promised to do his last novel.”

Well, Harold Strauss was in many ways a rather nasty man. He was an arrogant man. But he was an honest man. And if he told me that I’d promised to do The Decay of the Angel then I had promised to do it. And so I did it.

Yukiguni: Misjudgment Perhaps, But Not Mistranslation

I think I’ll turn immediately to a nice reference. In the last issue (No. 102, July 2003) of the SWET Newsletter I am referred to, and this will lead on to many things I would like to say. “At the height of the Nihonjinron, the mystique of Japanese uniqueness, in the late 1970s and 1980s, [Charles] De Wolf found ethnocentric attitudes that were an obstacle—no foreigner can possibly understand Japanese and that’s why Edward Seidensticker ‘mistranslated’ the first sentence of Snow Country—etc., etc.”

I don’t see how anybody could call these first sentences “mistranslation.” They might be misjudgment; I’m perfectly willing to admit that. But not mistranslation. I’m sure all of you know both the original and the Japanese, and the translation of that hugely famous passage: kokkyō—or maybe it’s kunizakai—that’s just like Kawabata: nobody knows the pronunciation of the first word in the novel! That’s real Kawabata!

Kokkyō no nagai tonneru o nukeru to yukiguni de atta. Yoru no soko ga shiroku natta.

Those are the opening sentences. The second sentence is the important one. But the thing that people keep quibbling about [in my translation] is the first one, which is far less important. The first sentence, in my translation, is: “The train came out of the long tunnel and it was the snow country. The bottom of the night was white.”

People object to that first sentence on two grounds. One is that it has a subject, whereas the Japanese does not. Well, this is perfectly true. My answer is that one has to show a certain respect for the target language. The target language in this case, of course, is English; one has to show a certain respect for it. And the target language in this case likes subjects. So I put a subject in.

The Japanese language is perfectly happy to go without subjects; it can go endlessly without subjects. As those of you who have read the Genji know, the subjects are usually all tangled up in the endlessly long verbs; that’s how you figure out what they are, and it takes some figuring!

But English asks for subjects, and so I put a subject in. Why so much fuss over it? And what does it add, really? What does my putting in the train as the subject add? I can think of only two or three things that would be coming out of a long tunnel. One is a train, another is a rat, another is perhaps a bat or a cockroach. Have I really destroyed—done anything violently wrong in putting a subject in that sentence? I don’t really think I have.

The other thing that people complain about is that I’ve left the first word out, the kokkyō. The kokkyō or kunizakai means a boundary between two provinces. Well . . . All right, I did it; and I don’t feel at all apologetic for having done it. Again, what is lost?

And to explain it would be very tiresome. You would have to say “the tunnel between two provinces,” and then you would have to go on and say, “these aren’t prefectures, mind you, these are provinces.” You would have to put all of this information in the hands of the reader to be perfectly, completely faithful. I don’t see that it’s all that important. That’s the sort of thing that they complain about.

The second sentence is another matter. Now, in the second sentence, I do think I used bad judgment. The second sentence is the really famous one. “The bottom of the night was white,” is what it literally is; “Yoru no soko ga shiroku natta,” “became white” . . . “The bottom of the night turned white.” It’s a very, very striking image, and I think it was rather blurred in my translation.

I had a reason for doing it; I wanted to separate “night” and “white” as far as possible. Because I think rhyme is bad in prose, because it calls attention to itself. And one of the essentials of good prose is that it does not call attention to itself. So I went through some acrobatics to keep them apart, and I did succeed. But I think I should have translated it literally: “The bottom of the night turned white.” “White,” “night,” there they are, clashing with each other, a rhyme. But I think that’s what I should have done.

I have retranslated it, but, unfortunately, in a place that not many people can get to. The Limited Book Club brought out a retranslation of Yukiguni—Snow Country—which cost about a thousand bucks, and since the Limited Book Club has only a few hundred members and not everybody’s willing to pay a thousand bucks for a translation from Japanese, not very many people have seen it.

I think if I had known at the time what I have become painfully aware of since then—if I had known at the time how extremely famous that passage is, I probably would have thought several times before I came up with the translation that I did.

Those two first sentences of Yukiguni are without question the most famous passage in modern Japanese literature. Incidentally, they weren’t the first sentences when he originally published it. He showed very great canniness in digging them out from somewhere along the way and putting them first. That’s how they became famous, because they’re right there at the beginning. I tell aspiring translators—and now I’m telling you—that they should be particularly careful about the beginning and ending of a translation, because those are the only parts anybody’s going to notice.

Pick Something Hard

If you want to be interested in a translation, pick something hard. Easy translations are bores. Much the most interesting translation I’ve ever done is The Tale of Genji. But among modern writers the most interesting one to work with has been Kawabata. And that is because he is difficult. Difficult works keep you interested and easy works don’t. The two greatest Japanese writers—aside from M. Shikibu—that I have worked with, I would say without question, are Kawabata and Tanizaki. Kawabata was always fun to work with; Tanizaki was a bore. And the reason is exactly that: he’s easy to translate.

When people say that Japanese is an obscure language, what they mean is that if Japanese wants to be obscure, it can be devilishly, fiendishly obscure. But if it wants to be perfectly lucid, it can be. There’s nothing inherent in the language to make it obscure.

Well, Tanizaki was very strongly under the influence of English literature, and he wrote very lucid English sentences. The Tanizaki style could come straight out of the English nineteenth century, where I think lucidity was perhaps at its very, very best. That’s the Tanizaki sentence.

The Kawabata sentence is not that way at all. The Kawabata sentence is a very classical Japanese sentence. You never quite know what’s going on. Not always, not always. I knew exactly what was going on in those opening words of Yukiguni. I think I just didn’t translate them very well. But very, very frequently in Kawabata there is blur, there is fog, there is mist. You don’t quite know what it’s about. That’s very, very interesting—trying to figure out what it’s about and then setting about finding adequately obscure English.

Which is impossible, really! English isn’t an obscure language. The nature of English is in the direction of lucidity. The nature of Japanese—again, don’t understand me as saying it must necessarily be so—but the nature of Japanese is in the direction of obscurity.

People often ask me whether I consulted with my authors when I was working on them. My answer is, “Well, yes, I did try occasionally; but it never did me the slightest bit of good, so I stopped.” I think it was because of Yukiguni, the first long Kawabata translation I did, that I went around to Mr. Kawabata one day and said, “Mr. Kawabata, isn’t this passage a little bit ambiguous?” And Mr. Kawabata looked at it and said, “Yes, it is a bit ambiguous.” Period. He didn’t say a word more. That was how much help he was to me. But Kawabata is that way; Kawabata is obscure; and that makes him interesting. As also it makes Murasaki Shikibu interesting.

There was another Kawabata translation that I did something to which I think was badly judged. That was The Izu Dancer. There are deletions, and I think that they were badly done. I did them because it was to be printed in a magazine that didn’t have space for the whole story. I deleted parts I didn’t much like. They were the somewhat unseemly parts . . . You know, such as the business of old man with the shakes at the very beginning of the story, as the boy starts down over the path.

I’m sure you’re all familiar with The Izu Dancer; it’s about a Tokyo schoolboy who makes a walking journey down the Izu Peninsula, from Shuzenji all the way to Shimoda, in the company of a family of wandering performers. The dancer of the title is a member of the family.

But there are ugly things in it. It’s all very pretty insofar as it’s concerned with the dancer herself. The family isn’t all that pretty, but they’re on the whole fairly decent people. But there are nasty bits in the story, and I cut them. I think they should have been left there. If I had wanted to cut “excesses” I should have cut excesses of prettiness, not excesses of ugliness. Kawabata put them there for a reason and they should have been left there.

So another piece of advice I would give you is to be honest with your readers. I should have told the readership that this was what was happening, and I didn’t. And I have long felt guilty about this. The readers should have been informed that the excisions were made, and they should have been allowed to decide whether they were wise or not. Fortunately, that has been remedied: there has since been a complete translation of The Izu Dancer, I’m very happy to say.

I’m always being given lists of mistakes in my translations. Do the rest of you get lists of mistakes? I get rather elaborate lists. Some of them are categorized: “itsy-bitsy mistakes; considerable mistakes; howlers” . . . Somebody has gone to a very great deal of trouble to do this, and I suppose I should be grateful, but I may admit that I am not usually terribly grateful.

Sometimes, however, I get something rather good. This brings me around to one of the really essential things about Kawabata and why he’s so interesting to translate. The postman brought to me one day a letter from a professor of Japanese with a list of the ways, in my translations of Kawabata, in which I had translated the single verb omou, “to think.” He had gone through several of my translations of Kawabata and listed all of the ways in which I had translated that one word omou. Well, it was a pretty staggering list! There were about forty ways in which I had translated that one verb.

That’s not perhaps as surprising as it might seem. Complex words are words that have many meanings; every language has complex words. And the verb “to think” is a complex word in English and the corresponding omou is a very complex word in Japanese.

This tells us something very interesting about Kawabata. You know, Tanizaki is always being likened to Murasaki Shikibu. Well, there are resemblances, but the chief resemblance between these two is the interminable sentence. You look at the page in front of you and you don’t see any periods, and you say, “My goodness, this sentence is going to go over the whole page?” And sure enough, it does! And you say to yourself, “Well, now I will go overleaf, and surely I will come upon a period!” But you don’t! It just goes on some more.

Tanizaki is that way, too. Murasaki Shikibu goes in for very, very long sentences—sometimes it’s a little bit phony: she simply changes periods to commas and goes on.

But in a far more important sense, Kawabata is much closer to her. Both Kawabata and Murasaki were people of few words. Murasaki Shikibu wrote a very, very long book—it extends to some eleven hundred pages—but she was very sparse in her vocabulary. Being sparse in your vocabulary means that you put the same word to a great many uses. Murasaki does, certainly: aware—“pity” or “shame,” or something like that, she uses endlessly. And she uses it in a very great variety of meanings. So does Kawabata. And that is a point that this gentleman who sent me the list of my translations of omou brought home very forcefully: that Kawabata does exactly the same thing. He, too, is a man of few words, words which he uses in a very great many ways.

The thing that annoys me most is the implication that we who translate into foreign languages from Japanese do a worse job than they who translate from foreign languages into Japanese. I think there is a very strong implication in this: that we do a bad job, the corollary of that being that they do a good job.

Good Translation

I’m never quite sure myself what I mean by “good translation.” I’ve been talking thus far about what I think is “inadequate translation.” I think that good translation imitates the original—to the extent that it is possible to imitate the original. I think of it as counterfeiting. A counterfeiter is not a good counterfeiter if he makes George Washington handsomer than he is on the one dollar bill.

People are always saying, and it is said of me, that the translation is an improvement on the original. I don’t consider this praise, because I don’t think that that is what the translator should be doing. The translator should not be improving. The translator should be imitating, counterfeiting, to the best of his ability. Perfect counterfeiting, I would argue, is not possible. But the translator ought to be doing it to the utmost of his ability.

There are limits beyond which he is not permitted to go. One of them is that he is not permitted to be as bad as the original; editors won’t permit that. But I think that if the original has its bad spots, the translation should also have its bad spots. But editors will immediately be after you to improve on the bad spots and make them presentable. I would argue that it is not the translator’s function to improve, and therefore that the ones who are credited with improving should not take this as praise. The translator is being congratulated on doing something that he should not have set out to do.

Translation is a constant series of decisions—how to do it—you are immediately aware that a perfect translation is not possible at this point. What are you going to do? You usually have to choose among several alternatives, none of which is really satisfactory. I’m sure you've all been chided by this or that Japanese critic—hyōronka—I think “critic” is a bad translation of hyōronka, a hyōronka is really a “wise guy.” The wise guys are always telling us that we have done dreadful damage to the original by leaving out proper names. What are you going to do when a Japanese original is absolutely full of Japanese proper names that mean absolutely nothing to the reader? You have two choices: one is to cut them out and the other is to explain them all. The first is unfortunate because it unquestionably damages the original, but the second is bad, too, because it is a very awkward device. Both of these solutions are inadequate; so it is also with proper names. You have no doubt run into the absolutely maddening circumstance of being unable to tell with assurance how a name is read in Japanese—somebody asks you what the name 長田is—well, it could be Nagata or it could be Osada . . .“I don't know which,” you say; and then they say, “I thought you knew Japanese!”

Japanese are of the opinion that wabi and sabi must be dreadful problems. (I like calling them “wobby” and “sobby”—that’s a little personal quirk.) But they are not difficult [for me] at all, because writers never use these words—they appear in criticism but they don’t appear in fiction at all, and therefore there’s no problem at all. To people who say that wabi and sabi must present dreadful difficulties, I retort that the things that present the greatest difficulties are the most natural things in the world, and this is literally true of plants.

Plant Names

Plants present virtually insurmountable difficulties. The plants that are the most common in the Genji are unquestionably the cherry and the plum—although we have a little trouble with the plum—are you aware of the pedants who insist that the plum is not a plum but an apricot?—therefore, whenever you get an ume in a poem, you have to call it an apricot, which sometimes doesn't work very well. But pedants aside, the sakura and the ume do have fairly close equivalents in the West. The next two most frequently mentioned plants, I think, in the Genji are the yamabuki and the hagi. Those are the very popular garden plants. The yamabuki is sometimes called the “yellow rose,” but that sometimes brings to mind “the yellow rose of Texas,” doesn’t it?—not the connotation we want—and it isn't a rose, but it's very frequently translated as “a rose.” The hagi is even more difficult—it's sometimes translated as “bush clover,” but it isn’t a clover—and here the objection is all too obvious to describe—the other translation is Lespedeza japonica—try putting Lespedeza japonica into a poem of 31 syllables . . . These are impossible decisions.

There are several possibilities but none of them is really ideal. In the case of hagi, you can leave it in the original, you can find a rough equivalent—and I suppose bush clover will do as well as any—or you can give the scientific name. Plants present absolutely impossible problems, and they're right in there with proper names.

Tense in Japanese

We’re always being told, and there’s no question that it’s true, that the Japanese tenses are not really tenses, past, present, future and all that. I think that it is unquestionably true that they are not tenses in the European sense of the term. They don’t clearly indicate time as the Indo-European tenses do. What do they indicate? It’s sometimes said that they indicate moods or modes—I'm not quite sure what these are—but that they do indicate something other than time is quite clear. I sometimes think they indicate degrees of certainty; that what are called the past, present, and future in Japanese indicate degrees of certainty. Suppose you dropped a coin in a lawn somewhere and you’re looking for it. And you say “Aru de arō.” And that means something like “Well, it has to be there somewhere.” And then you bring it down to what is called the present, “Aru.” And that means something like “Well, I’m sure it’s right around here somewhere.”

But when you find it, what do you say? You say, “Atta!” You put it in the past, which indicates complete certainty. And I think the Japanese verbs do indicate something like that. Aru de arō, arō, aru, and atta indicate degrees of certainty, and atta is the most certain of them, putting it in the past.

I suspect that that’s what it means. But if you’re going to try to convey whatever is conveyed by the mood of the Japanese original, you’re going to end up with a very wordy and very dull translation—which I think is probably not what most of us want to end up with. If you have to explain every word—for example, “I have just said ‘She sat down,’ but you must understand that this doesn’t really mean that she sat down, it means that she’s in the process of sitting down”—if you have to do that with every verb you come to, you’re going to end up with a pretty turgid translation.

So it’s not an advisable thing to do. But what are you to do? I think, again, coming back to a point I’ve already made, that you have to pay deference to the host language or the target language. You have to subscribe to its preferences. And the preferences in English are for verbs that indicate, by their tenses, time—quite clearly indicate past, present, and future. I think that’s what you have to do. You have to give way to the tendencies and inclinations of the host language.

Translating Rhythm

I think the problem that I fret over most is the problem of rhythm. If you’re working on a technical manual, as some of you perhaps do, I don’t think rhythm really makes very much difference—unless you start putting the reader to sleep with your wordiness, in which case you’ll have a lawsuit on your hands.

I don’t think that rhythm really matters terribly much in that kind of translation. Nor does it matter in purely practical matters—the sign on the airplane, “No Smoking,” in English, French, German, Spanish, and Latin. It doesn’t matter whether they follow the same rhythm or not.

It is very interesting, however, that always, with “No Smoking,” French takes the most room. The French think of French as a very concise language: well, it isn’t, it’s a very wordy language. All you have to do is look at “No Smoking” to figure that one out. But I don’t think it matters in technical translations and it doesn’t matter in things like “No Smoking,” or “This way out,” or things like that.

In translations of literary works, however, rhythm is extremely important—the flow, the sequence of emphases, the amount of time that expires in a certain passage— these are very important. Perfect rhythm may be possible—a translation that perfectly imitates the rhythm of the original. It may be possible between German and English, which are two very closely related languages. Germans, who may be prejudiced, are always giving us as an instance of a perfect translation, the Tieck-Schlegel translation of Shakespeare. Well, they may be boasting a bit, and they may be prejudiced a bit, but for all I know it may be true.

I’m inclined to think that it isn’t. I’ve looked here and there and come upon things that seemed to me a little bit ugly in German, when they’re very beautiful in English. Take, for example, the end of Othello. Just before Desdemona is murdered she sings the “Willow Song.” It’s very beautiful. And much of that beauty is in the word “willow,” which is a much more beautiful word than the German equivalent.

It’s impossible to have a perfect equivalent in English of Japanese rhythm. I say it’s impossible . . . If you can prove the contrary, do. I’ll be happy to hear it. But I think it’s impossible. There are so many reasons why it’s impossible.

The second sentence of The Tale of Genji contains two ten-syllable words. The two words are omoiagaritamaeru and otoshimesonemitamou. Each of these has ten syllables. I’m sure that if I looked further through the Genji I could find verbs and adjectives that are even longer—they would be verbs and adjectives; the verbs and adjectives do the hard work in a Japanese sentence, both modern and classical; the verbs and the adjectives are the worker bees, so to speak; everything else is a drone.

Well, there aren’t many ten-syllable words in English. The word that is commonly said to be the longest word in the English language, antidisestablishmentarianism, is twelve syllables. But that’s a very, very exceptional word; you don’t find many words like that in English. But there are two ten-syllable words in the second sentence of the Genji. For this kind of flow of syllables I think there is no English equivalent. Therefore, any translation into standard English is bound to distort the rhythm of the original.

Fukuda Tsuneari, one of the eminent translators of Shakespeare into Japanese, once said that it is impossible to present a complete Shakespeare play in Japanese—because it’s too long. It takes too much time in Japanese. It takes longer to say something in Japanese than it does in English. And I think this is true of all European languages, including verbose French—that it takes longer to say something in Japanese . . .

Great Writers I Knew

I’d like to say just a little bit, because I think it’s something that interests people, about the great writers I have known personally. I have known two writers personally, rather well, whom I would classify as great writers; and a third, very famous writer whom I’m not so sure I would classify thus. You can guess that the third is Mishima. The first two are Kawabata and Tanizaki.

I think, although I’m not quite sure, that Kawabata was the first famous Japanese writer that I met. I think it must have been in around 1950. It was when his translation of Maihime was coming out in the Asahi Shimbun.

Kawabata was a strange man. He was a silent man. I think I learned about silence from Kawabata. I learned from Kawabata that silence is not necessarily hostile; silence does not necessarily indicate boredom; it does not necessarily indicate antipathy. It merely indicates that you have nothing to say. And that I learned from Kawabata.

I learned it very early. At first I was made very uncomfortable by Kawabata’s silence. The two things about Kawabata that I think were immediately most striking were his remarkable eyes—those piercing eyes that seemed to look straight through you, and as they were going through you, simultaneously to pass on without noting anything, but to see everything Extraordinary eyes. I remarked in my diary entry on the night he died that for the first time, when he lay dead, I saw his face without the eyes; and it was an utterly different face.

I learned about silence from Kawabata. And so I started being silent too. And we got on splendidly. He became really quite a good friend. He took me to Stockholm with him for the Nobel Prize award . . .

He was a very silent man, a very mysterious man. As we always used to say about the Japanese, and this was very true of Kawabata: you never knew what he was thinking. I was with Kawabata at a dinner party one night and he sat through the party the picture of boredom, the absolute quintessence of boredom. And we chanced to go home together afterwards. I said, “I’m sorry, sensei; you must have been terribly bored.” And he said, “No, I was fascinated!” You never knew what he was thinking. But he was a very interesting and very kind man, and he was certainly very, very good to me, taking me off to Stockholm with him.

Did he kill himself? Nobody knows, really. I went down to the Kawabata house the night he died, and Mrs. Kawabata kept saying, over and over again, “I don’t understand it. I don’t understand it.” I’m sure she was saying what she thought—which was exactly that: she didn’t know what had happened, or why it had happened.

Some of his friends were convinced that he didn’t kill himself, that it was an accident. The same thing, they said, had happened on an earlier occasion and it had been accidental. Well, I don’t know; I simply cannot say. I would like to think that it wasn’t suicide, because Kawabata was not a suicidal man. I was really very disappointed to think that he should have killed himself. He was a man of akirame—he accepted things. He accepted good things and bad things. He accepted. And he went on being that way, I think, to his death; and therefore I want to think that he did not kill himself. But I don’t know. I think that if he did kill himself, probably his reason was sheer exhaustion. He won the Nobel Prize a couple of years before he died, you know, and what happened afterwards was a rat race. He was kept constantly busy. I think he may have been exhausted and saw a way out and took it.

Tanizaki was interesting, but in a very different way. He was known as a “devilish writer,” you know—akumashugi is the word that’s associated particularly with the early Tanizaki, a devilish writer. I was very, very pleased, one day, when Tanizaki told me he hated that word and wished people wouldn’t use it. I hate the word too, and wish they wouldn’t use it.

Tanizaki was never really devilish. All you could say about him was that he was impish. And an imp and a devil aren’t the same thing at all. He was a very impish man, but he wasn’t a devilish man. There was a childish quality about him that was endearing.

And one cannot mention Tanizaki without mentioning the third Mrs. Tanizaki, who was Sachiko of The Makioka Sisters. Mrs. Tanizaki was always present when I was with Tanizaki. And she always kept things going; she and the cat. We were always a foursome: Tanizaki and Mrs. Tanizaki and a cat and I.

I think really my favorite Tanizaki work—it’s a minor work, but I think it’s a perfectly delightful work, a masterpiece—is Neko to Shōzō to Futari no OnnaThe Cat and Shōzō and the Two Women. I think that’s a perfectly marvelous work. I’ve never tried translating it, because – another instance of the “untranslatable” that I should have mentioned is dialect. Much of Neko to Shōzō is in dialect; it’s in the Osaka dialect. Much of Sasameyuki is, too; but it doesn’t really matter in Sasameyuki: the characters aren’t dependent on it. But they are in Neko to Shōzō.

It’s virtually impossible to imagine the characters in Neko to Shōzō apart from the Osaka dialect. They are the Osaka dialect—which I dearly love, by the way—I like the Osaka dialect far better than standard Japanese; I really do. You think I’m fibbing, but I’m not.

There was an element of pretending about Tanizaki. He pretended things that weren’t really true. Among his most famous writings is the essay In Praise of Shadows, part of which I translated, in which he describes an ideal Japanese house—a house of darkness, a house of toilets made fragrant by cedar boughs; that sort of thing.

But he wasn’t that way at all. His houses weren’t in the least like that. I visited several of them. I never went to his Kyoto house; I don’t know what that was like. But I did visit more than one of his Atami houses and the house in which he died in Yugawara. And they were all very bright, sunny houses. There was nothing shadowy about them.

Mrs. Tanizaki told a delightful story. She said that once when Tanizaki was building a house down in the Kansai he asked to have the architect come around, that they might consult. And the architect sent back, “Well, I know exactly what the sensei wants. I have no reason to consult,” and Tanizaki said, “Oh, my God, I hope he won’t build me that kind of house!”

I was very glad that Mrs. Tanizaki was there, because in a very different sense from Kawabata he was a man of few words. He used Japanese with extreme lucidity, which is absolute proof that it’s possible to write not only good, but beautiful Japanese, and still be very lucid. Tanizaki was living proof of that.

Rhythm and the Genji

The problem of rhythm, which strikes me as completely insoluble, brings up my relations with my predecessor, Arthur Waley. Waley’s translation is a highly decorated translation compared to mine. I’m often asked why I undertook to do a new translation. The answer is really quite simple: I undertook a new translation because I felt one was necessary.

My translation is much more literal than Waley’s. There may be a few omissions; they’re completely unintentional, if there are. But it is a complete translation. The Waley translation is not. He omits the Suzumushi chapter. I don’t know why he omitted it, but he does. And it’s a very beautiful chapter.

One of my students had a theory. It was Cliff Royston’s theory that Waley read the original Genji over breakfast and that marmalade stuck the pages together.

He omits things, but I think more to the point, he adds things. He embellishes very liberally. It has been suggested that he does this because of precisely the rhythm problem that I have been discussing. He is aware of the fact that the English translation is moving more rapidly than the Japanese, and wants to do something about it. And he does something about it by embellishing it. It may be true; I don’t know. I never met Waley and therefore never asked him. But if Waley did it intentionally, then I think it was misguided. But who knows? I don’t know why he did it.

And I think it’s a very bad solution. I don’t think anything can really be done about it. I think that the Genji is bound to move more rapidly in English than in Japanese, because English doesn’t take as much time as Japanese. And in all the Japanese renditions of the classical Japanese that I have looked at, it takes longer to render it into modern Japanese than it does to render it into modern English.

The reason for this, I think, is perfectly clear: classical Japanese was an agglutinative language and modern Japanese is an agglutinative language. And you know what an agglutinative language is: you start out with a verb that makes sense, and then you add suffixes that make it progressively “senselesser”—that’s what an agglutinative language is. Japanese is an agglutinative language, and it’s inevitable that it’s going to take a long time.

Translating Dialect

[In answer to a question from the audience] Well, I think that dialect falls in the category of an impossible problem. And you know, impossible problems are no problem at all. When you have to do something about something, it’s a problem; but if you can’t do anything about something, then it’s not a problem!

And I think that dialect is an impossible problem. I certainly enter that into the category of the untranslatable. In the case of Sasameyuki, The Makioka Sisters, I thought of an American equivalent, I think a very good American equivalent, of the Osaka dialect—a way of speech which is elegant but not standard. And that is true of the Osaka speech: it’s a very elegant way of speech, but it’s not standard.

I thought of Southern American. But I decided it wouldn’t do; I couldn’t have Sachiko talking like Scarlett O’Hara. And that really is the problem: dialect is too specific. If you put it into Southern American English, it’s Scarlett O’Hara; it’s not Sachiko.

I don’t think that anything can be done about it. Well, I say that, and yet I did try to do something. I did attempt a device. I attempted to distinguish between standard Japanese and the Osaka dialect by having standard Japanese move more rapidly, with a great many contractions, and translations of Osaka Japanese move more slowly, with no contractions at all.

Well, all that can be said of this experiment is that it was a failure. Nobody even noticed. It was an utter, utter failure. I think dialect is an instance of something that you can only ignore. It can’t be translated. I’m really convinced of that.

I think I said already that it didn’t seem to me to matter very much, in Sasameyuki, The Makioka Sisters. It seemed to matter hugely in Neko to Shōzō, and that’s why I didn’t translate it. I thought dialect was so important in that that it couldn’t be ignored. But Paul McCarthy, who translated it, didn’t agree with me and did ignore it. And it’s a good translation; I don’t wish to suggest that it isn’t.


This article is based on the transcript of the talk given by Edward Seidensticker for SWET at Tokyo-Azabudai Seminar House on September 27, 2003. Translator of numerous works by Tanizaki Jun’ichirō, Kawabata Yasunari, and other authors as well as The Tale of Genji, Seidensticker taught at Stanford University in the early 1960s, then at the University of Michigan, and later at Columbia University. He is also author of Kafu the Scribbler (1965), Low City, High City (1983), Tokyo Rising (1990), and other non-fiction works. His memoir, Tokyo Central, was published in 2003. Seidensticker was guest of honor at the SWET 2005 25th anniversary party, see “Fond Remembrance of Things Past” [link], SWET Newsletter, No. 110 (December 2005). He passed away in 2007, see “Remembering Edward G. Seidensticker (1921–2007)” [link], SWET Newsletter, No. 117 (October 2007).

Originally published in the SWET Newsletter, No. 104.

© 2004 Edward G. Seidensticker