poem index

Interview: John Hollander

Year

2001

Type

Interview
Printer-friendly version

Edwin Honig conducted the following interview with John Hollander, which originally appeared in The Poet's Other Voice (University of Massachusetts Press, 1985).


Honig: You have done translations and written poems, and have perhaps even been translated. So to begin, I'd like to have your views on how a translation is made, and if it's possible, to relate this to a theory of translation as I know you've written about it. The distinction is between thinking of translating as a prescriptive exercise and translating as something in the making, a live performance.

It might be best to talk first about your essay in Brower's book, where it appears that you're trying to establish a way of looking at translation that would facilitate thinking of it realistically as a "version" rather than a faithful rendering.

Hollander: Yes. I wrote that a long time ago and, I think, a little brashly. Certainly at too great a length. I was interested in trying to show that any particular literary translation will be a version based on the literary style of the translator. Even if he thinks he is surrendering everything to the meaning that he wants to embody, he will all the more be betraying stylistic conventions, so that the only thing to do is consciously decide upon a stylistic analogue for that of the original and carry the meaning over to that.

Honig: Interesting, but I don't remember your saying that in your essay.

Hollander: No. In the essay I brought up the difference between Latin prose composition, where there was a correct answer—where you were trying to approximate to Cicero in Latin and did your exercises—and the translating of a Ciceronian sentence into English, which had a great many possible solutions.

Honig: Yes, I thought that very valuable as a start.

Hollander: In a larger sense, all literary translations are "versions" that way. And what I just said to you was perhaps an afterthought on that essay fifteen years or so later.

Honig: What's happened to your notion of a "version"?

Hollander: I think it has implications for nontranslative writing as well. I think that a certain amount of self-awareness about style is absolutely necessary in learning how to write by learning how one is writing. What puts a lot of young poets off their true course is some sense that they're starting from scratch. And the relation of translation to original creative writing in any tradition is rather interesting. These questions have been raised in recent books on the subject. Robert Martin Adams raises that notion. Frederic Will does too.

Honig: In your essay when you bring in T. S. Eliot and the interpretive style and suggest that translation is interpretation, you evidently situate the whole drift of modernist poetry from Eliot and Pound as partly an active engagement with translation.

Hollander: Well, I won't say that it was all ideological from modernism, although I know I did pick up that idea. No. Before having any real contact with modernism I simply felt obliged to do translations. That is, before I ever did poems of my own. The first undergraduate poems I published were translations of Baudelaire. I felt that translating Baudelaire was a necessary step in an apprenticeship. I don't know why and I don't know who told me.

Honig: I've often given my writing students exercises in translation or urged them to write versions of poems from other languages.

Hollander: I had written humorous light verse in high school but never did anything I called a poem until after I'd translated Baudelaire.

Honig: I think I see a connection between what you just said about producing poetry via learning to translate and what you said before about deliberately choosing an analogue in order to make a translation, back of which is also your notion of translation as making versions of the original.

Hollander: I think so, yes.

Honig: But then you said that you are no longer interested in doing translation.

Hollander: I find myself no longer wanting to translate now.

Honig: Why is that?

Hollander: I don't know. My last experiences with it were most fortunate. The last things I did were a lot of poems from the Yiddish for an anthology by Irving Howe and Eliezer Greenberg. In the course of that work I discovered the poet Moishe-Leib Halpern, and my translations of him were lucky. More than that, they seemed to help me develop a certain tonal mode in my own poems. That is, what I had to do to translate certain poems of Halpern's, I've now retained as a vocal element. Doing Halpern provided a way of unlocking certain things. My Yiddish isn't very good. It's learned, a secondary and artificial thing, since nobody spoke Yiddish in my family. But I knew some German and I'd been taught a little Hebrew, and I learned how Germanic Yiddish is transcribed in Hebrew letters. Also, I worked very closely with Irving Howe, who is a friend, and when there were difficulties he would discuss a word or two with me.

Honig: I know that anthology; I contributed translations to it myself. How well does Howe know Yiddish?

Hollander: Very well. Perfectly, yes.

Honig: But it's not "learned" Yiddish.

Hollander: No, it's native. He could point out the resonance of a particular Yiddish word, especially one with a Slavic origin or with a special use. I knew enough language to tell immediately whether it was a Yiddish word or a Hebrew word that had entered into Yiddish. This helped in separating out tone. I'd have to use a high diction, for example, to translate a resonant Hebrew abstraction, then shift to a very vibrant low diction sometimes, for other effects. I knew enough to see that immediately, although I cannot jabber the language. But I felt I didn't trust myself to translate that.

Honig: Isn't it true that Yiddish is more of an oral, colloquial language than a literary language? My question is prompted by the fact that I picked up what Yiddish I know in my grandmother's house as a child. I didn't learn it as a literary language though I studied Hebrew in the Talmud Torah.

Hollander: Well, Yiddish has a short literary tradition, nineteenth-century mostly, and of course this foreshortens the poetic tradition in many strange ways.

Honig: So that in translating Yiddish, for example, one is aware of the vernacular more than if one were translating German. Well, let me go back to something else, and perhaps ahead at the same time. The business of your learning something about the writing of poetry from first translating, then the business latterly of your having given up translation after suddenly making a momentous discovery with Halpern indicate that you have assimilated a great deal. It makes me think again that in the work of other poets—Pound and Eliot, say—translation is a large assimilated element.

Hollander: Oh, it's essential there, but Pound and Eliot are both poets with grave problems of originality and grave problems about confronting their lack of originality. It seems inevitable that they would propound. Like Longfellow they propound a corpus of poetry largely based on translation. Corpus in both cases. They are both, I think, much more like Longfellow than we've admitted.

Honig: But doing what they did with translation, they paved the way for others to write differently.

Hollander: Well, yes, in one way. As far as I know, our greatest poet in the twentieth century, at least our greatest American poet, never did any translation: Wallace Stevens.

Honig: But I always felt Stevens had assimilated French.

Hollander: He may intone a lot of French in his poems, but he doesn't sit down and do translations. I've stopped translating because it takes so much time. Also, I think there is so much indifferent verse translation going on now by people who don't have any particular skills in writing English verse but who proceed to translate from languages that they don't know. I am a little ashamed of some translating I have done from languages I don't know well enough. I have never translated from a language I didn't know at all—I won't do that. In the case of some translations of Voznesensky, I have worked from minimal Russian. I worked with Olga Carlisle on those. I did translate from the Russian text except that my Russian text was annotated after hours of going over it with her.

Honig: But you hadn't studied Russian before?

Hollander: Yes, I'd studied some Russian.

Honig: So you knew the grammar.

Hollander: I knew the grammar. As I say, I worked from the Russian text, which I could read, and I knew the grammar, but I don't know very much Russian. I know even less now. But I still am a little ashamed of having done that. Except that the versions turned out rather well.

Honig: I was talking with Aleksis Rannit yesterday, and he illustrated rather pointedly the unexpected and unequal results of knowing and not knowing a language well. Yakobovich, a Russian poet jailed in Siberia for twenty-five years, spent his time there writing version after version of Les Fleurs du Mal. Aleksis reports that the final results were an abysmal failure. But another Russian poet, Fyodov Sologub, who knew much less French than Yakobovich, did a much quicker and vastly better job of translating Baudelaire. So . . . .

Hollander: There is another dimension to this matter. A very, very good poet can do a version of something from another language, even if he doesn't know the language. That is, he can write a poem based on somebody else's prose paraphrases of the thing. But this is purely and simply a matter of the translator's having a certain kind of poetic skill, a very rare thing to find. By and large, I disapprove of my having done translations from a language I didn't know well enough, and want now not to do that any more. I also feel I have done my bit to a degree, that is, helped out in certain projects. That Borges book [pointing to it] you have there is a unique case. I don't really know Spanish well. I can read it with a dictionary, particularly when it is clear and simple and has as few syntactic problems as Borges's poetry, which I find relatively easy. I did a number of poems because Norman Thomas di Giovanni approached me, and this all centered on one poem.

Did I tell you that anecdote? It's a little spooky. It's essentially a Borgesian anecdote. About 1968 di Giovanni said that he'd been thinking of various people to assign particular Borges poems to, and he thought that I might like to do the poem about the golem. I was startled at this because my mother's family traditionally believes that my mother's father's family is descended from the Rabbi of Prague about whom the golem stories have circulated. Without telling di Giovanni anything about this, I said, "Yes, all right, I will do the poem." I followed the original meter and rhyme scheme, and the syntax of the poem made it quite easy to do. I could preserve the rhyme of "Golem" and Gershom Scholem, who is the great commentator on Cabalism—that's a very Borgesian rhyme, rhyming a myth with its exegete—and I could hold those things over from the original, and it worked out rather well. When the work was over I did want Borges to know that there had been a kind of loop in time. In the same meter of the translation I wrote him a verse letter about having done this, and about the curious historical accident, and everything else.

Honig: When did you do this—in 1968?

Hollander: Yes. I was in England at the time. This verse letter to Borges I remember starting, "I've never been to Prague, and the last time that I was there, its stones sang in the rain . . ."

Honig: That's interesting. Here's your translation of the Borges poem on the golem [indicating it in the anthology]. I've read it in Spanish but didn't look at your translation. You said you followed the original meter. I remember it as being almost prose. In Spanish there's usually only syllabic count. Did you find accentual meter?

Hollander: Yes, in rhymed quatrains.

Honig: I mean linear meter.

Hollander: It's a kind of pentameter.

Honig: [quoting the first stanza] "Si (como el griego en el Cratilo) / El nombre es arquetipo de la cosa, / En la letras de rosa está la rosa / Y todo el Nilo en la palabra Nilo."

Hollander: Would you say that's according to a syllabic count?

Honig: Well, it seems hendecasyllabic, and also like mixed meters. At any rate, it's rare to find pentameter in the Spanish. One of the things about Borges is that I think he wants to be an English poet.

Hollander: Oh, without question. But he frequently does that in the sonnets. He moves toward a pentameter.

Honig: OK. Another question I have for you is really three questions in one. Where or when does the translator, or the translation itself, begin? How does the translation develop? And where does it end? The implication is that the poetic translation doesn't start when you put your pencil to paper, but before that. What do you think?

Hollander: It would start with a sense of what shape, what form, the finished product is going to have. One of the confusing things about this matter in the modernist tradition is that the poem format for English that Pound virtually invented looks as if it were a prose paragraph. That is, a kind of Poundian free verse in end-stopped lines he used for the Chinese poems, for example. And so one has to be aware of that as an alternative and a possibility too. There has to be some notion of how the shape is to be carried over or what it is to be carried over into. That is, when you've finished, what it will be and what it will look like. I'm not saying a verse form necessarily precedes the translation, but something like it does—an overall sense of form which may have surfaced, with clear surface manifestations. Or it may be a deeper, more abstract sense of form. You could say, "Well, I know this is written in complicated stanza structures, but I'm going to do it in one blob because there is something that I want to get out of it that is best represented by that." That's a formal idea, just as with writing a poem something happens like that. By which I suppose I mean that doing a translation is very like doing a poem.

Honig: Right. I'd imagine you'd think so. I want to know about one particular area now. You spoke of translating sonnets, and you thought of the job as that of writing an equivalent or correspondent sonnet.

Hollander: Well, that was because it was Borges, and because of what the form meant to him, I thought it important to get that relation to the English sonnet into my translation, although I could certainly conceive of translating some other poet's sonnet in another language and not trying to keep that form. On principle I don't think one should trash the poem. It's the problem of finding a viable analogue, and in so many traditions there are viable analogues. There's one of putting French into English, and that tradition involves substituting pentameters for the alexandrines. Now any translation of a contemporary French poem that doesn't have anything palpable to do with earlier French formal conventions, nevertheless still has to draw on the history of that relationship, and this is the difference between a good and a not-so-good translation. In some languages there are no traditions at all of bringing things over, in which case the problem is a very different one.

Honig: That's a good point.

Hollander: Translating from some languages into English, even though they have a long literary tradition, might well be the same as translating from a textless language. And translating from a textless language is a totally different process, I would think. It now seems very popular among a lot of people who despise textuality and despise tradition.

Honig: You mean as from American Indian languages?

Hollander: Yes.

Honig: A lot of that is being done nowadays.

Hollander: Yes, it's being done—done by people who don't know the languages at all.

Honig: Mostly, yes.

Hollander: And it's a very safe kind of hackwork. Also its ideological content is sufficiently belligerent to give the piece an edge. There's something politically ideological about translating American Indian poetry and that sort of thing. I don't mean the very careful versions done for the Department of the Interior or the Bureau of Indian Affairs.

Honig: I understand the Smithsonian Institution Reports or Transactions usually serve as a base for many of the translations of Indian poetry.

Hollander: Yes, and I'm thinking of the whole idea of versifying them. They are versified into the flagrant gestures of what are called naked forms, I believe, by some of the practitioners, which simply means the received style of the moment. Professor Harold Bloom of Yale compares W. S. Merwin to Longfellow interestingly with respect to two notions. One is that both of them based a large part of their work on translations. And, secondly, that both wrote—that is, helped create and then wrote—in what was the received style of their time. If you look at magazine verse from the 1860s and 1870s in America, in Godey's Ladies' Book, Peterson's Magazine, and that sort of thing, all of it will be imitation Longfellow. And, similarly, if you look at poetry magazines today, a lot of it is imitation Merwin. Now the relation of that to translation I think is very interesting. You see: I think Longfellow's Hiawatha is an example of just the thing we're talking about. It's taken to be "Indian" but based on the Kalevala. Yet the meter comes not from the Kalevala, which Longfellow couldn't read—he didn't know a word of Finnish; it comes from a German translation which converts the octosyllabics of the Finnish into trochaic tetrameter, a pounding meter in German. You know the one that Heine used unrhymed for so many poems. That's where the heavy beat comes from, because in the Finnish you can't really say it's trochaics . . . .

Honig: What kind of meter would you call Hiawatha?

Hollander: Trochaic tetrameter unrhymed.

Honig: You know, it now occurs to me that you're also speaking of the meter of Pound's first Canto: "and then went down to the ships . . . ."

Hollander: Well, there, in the first canto, Pound is playing. Originally a lot more of it was iambic pentameter. In the original draft of that canto it's almost pure Browning. Then Pound jumped back and developed the notion but just in the first canto as we now have it, or particularly there—the notion that there was an analogue for him in the two parts of the possible hexameter line, separated by a caesura, and the two parts of the German line, the Germanic four-stressed line, separated by the scholar's artificial line break in the text. You know he'd been interested in the relation between visual format and a structural marker very early, which is why he takes Cavalcanti's endecasillabo line and writes it on the page in three successive lines, each one three successive line thirds—each one shoved over one step to the right so as to give you the three lifts.

Honig: It's a line divided into three distinct parts.

Hollander: That's right. But descending. Written in three lines descending toward the right.

Honig: It's what Williams does.

Hollander: Well, Williams probably copied that variable-foot format from Pound's earlier use, except that what Williams says about the variable foot is sheer garbage. It doesn't make any sense. What Pound did was to see that relation. It would be a little bit like taking the first line of Dante’s Inferno and writing it as three lines: "Nel mezzo / delcammin / di nostra vita," which would show the three lifts of the Italian hendecasyllabic line. He did that with a couple of Dante things, and so got interested in the original format. What he finally came to was a meter that is the six cut in half that way, stacked this way, sometimes echoing against the four-stressed line—and then every once in a while he'll have an absolutely pure hexameter come out. "Ear, ear for the sea surge. Murmur of old men's voices"—which is an accentual Homeric line.

Honig: OK. I'd like to go back to that psychological matter of where the translation begins and how it proceeds and ends. Obviously it's different each time. Do you have anything to say about the experience—commissions aside now—of deciding to do a poem out of love, for its author or for the thing itself?

Hollander: That's interesting. In going on with you about translating; I've been talking about some rather formal, commissioned translating. In some of my recent poems I have embedded translations. But those are thefts, as it were, not formal translations. For example, I have a poem in my last book which is an expansion—I simply call it "After Callimachus." It's an expansion of an epigram of Callimachus but it's been changed—it's put into a different metrical frame. Some of the imagery is changed and expanded. It's an imitation, just the way a lot of seventeenth-century English poems are imitations of Catullus, not strict ones by any means. But that sort of thing started out naturally and differently because it wasn't done with the task of translating in mind. It was just preserving something. And I have done this with bits and passages of poetry in the past—just put them in. Well, for example, I once did a half-translation, half mistranslation-adaptation of a great little poem of Hölderlin's.

Honig: How did that start?

Hollander: That started simply by my wanting to get inside the Hölderlin poem, which I've known a good part of my literate life.

Honig: But you'd never translated it?

Hollander: No, I'd never translated it. And to begin with, I found myself playing with a mistranslation of it. In my version, the third line is not a literal translation of the German, it's a mistranslation, which produces a new image. I'm interested in that. And so I translated some of it, then in the middle wrote about five lines, completely mine but just generated by the translation, then continued by closing off the translation. I used the piece as the dedicatory poem of my book, The Night Mirror. But when it was published in the Partisan Review, without any identification, an angry letter came in from somebody claiming that I had stolen it from Hölderlin, which amused me, because it's one of the most famous poems in German. I suppose I've done this sort of thing a few times.

Honig: We're talking in some way about the old idea that all writing is a kind of collaboration. And maybe now it's time the sterile polemics and argumentation induced by the question of being faithful to the original is countered by showing that one form of faithfulness is a matter of doing a new work.

Hollander: In some cases a great new work comes from a terribly faithful translation. I can think of one in English where a great English poem in translation is made of a great Italian poem. And that is Rossetti’s translation of Dante's "Stony Sestina." Just a plain masterpiece. It's one of the greatest English poems of the nineteenth century, and very accurate as a translation.

Honig: Nineteenth-century translators of that stature—Rossetti, Longfellow, Fitzgerald—actually had a great deal more on the ball than most twentieth-century English and American translators.

Hollander: Oh, I think so. Rossetti, particularly in that very, very great poem. It's because of what the poem's about. I mean it's for all the right reasons—one of those sestinas in which the terminal words make up a poem in themselves: ombra, colli, erba, verde, pietra, donna (shade, hills, grass, green, stone, lady), and really give you a distillation of the poem, and he could keep those and work with them. The Rossetti poem is not certainly the most allegorized reading of Dante, but an unallegorized reading of the poem would be the obsessive one for Rossetti in his own imaginative, erotic mythology, and it was an absolutely perfect thing for him to do and he did it magnificently. I have used some lines from that poem and some lines from Pound—and, mind you, the Rossetti was done in the 1850s or 1860s. I gave both to students without identifying the poems and said one is by a pre-Raphaelite poet, the other is by Ezra Pound, who believed in precision.

Honig: That's a good trick!

Hollander: And naturally they all assumed that Rossetti was the real poet, and the Pound manner limping, lumpy, fussy.

Honig: I want to ask you more about the question of a unified theory of translation, a theory that would accord with the practice of translators and present an imaginative confrontation of the possibilities. What you were telling me about your own practice is very close . . . .

Hollander: Well, the theory of translation would have to be a theory of literature in general.

Honig: Yes, all right.

Hollander: And I think this is a point that Adams gets to and a point that Steiner doesn't get to in his, for me, disappointing book.

Honig: Well, Steiner in his second chapter, I think, is more imaginative. At any rate, one of those early chapters goes into the question of the mystic notion of language having originated in the first word of God. The attractive thing there is that the idea allows for the work of translation to be considered as much an original as the primary text is, where both are striving to achieve something like the lost but reconstituted word.

Hollander: Yes, but I much prefer to read Milton on that subject: the invocation to Book One or Three of Paradise Lost goes into that.

Honig: You have a point.

Hollander: I think all one can do in surveys of that kind is to look at what translations have actually been done by which people under what circumstances for what purposes, and generalize from that. I think that's very interesting. For example, I think you could give in fairly concrete linguistic terms some of the reasons why for an English speaker the Douay Bible in French sounds silly. One thing Steiner doesn't go into which is absolutely essential to literary translation, is the whole question of what the Germans call Sprachgefühl, the language sense you have. What is it about speaking English that makes you think. . . . Well, put it this way: I say to graduate students, "I want to give you some English monosyllables and I want you to tell me whether they're French or Germanic in origin," and I give them a list including the word "push." Without thinking, they might say, "Well, that's German." But of course it isn't.

Honig: "Push"?

Hollander: Yes. But they assume it's German for good reason. It's part of the Sprachgefühl of English.

Honig: So Steiner does not . . . .

Hollander: Wait, just a second. So Sprachgefühl is very important for things of this kind. Lichtenberg has a great aphorism. He says, "A donkey is a horse translated into Dutch." Now, that is funny if you're a speaker of (a) German or (b) English. Otherwise it isn't funny, because Dutch is for speakers of both English and German something like a recognizable but too highly distorted version of their language.

Honig: Yes.

Hollander: Dutch is midway between German and English in that respect, so that the relation between correct "horse" and bungled "donkey" is like the relation: correct English or German, bungled Dutch. You see. Now matters of this sort are very interesting. They would lead one in English to say, "Oh, the Douay Bible, I'll just pick a passage from the Old Testament and read it in French and it sounds funny. It sounds as if it weren't serious." These are interesting linguistic questions, but ones that Steiner doesn't go in for.

I think that certain canonical translations in the history of certain languages and literature have a great shaping force. The English Bible has had effects on the structure of English poetry that have nothing to do with doctrine. For example, if I were trying to write a book on translation (and I would not attempt to do so) one thing I would comment on would be this—a simple tiny matter, but with vast consequences for English poetry. The King James translators handled a particular Hebrew syntactic problem in one way rather than another: the so-called Hebrew construct-state, which puts two nouns in a certain relation to each other. Hebrew is uninflected, but the two nouns are put together in a combinatorial way, and it's not a specifically genitive relationship, so that, for example: literally in Hebrew you say "house of the book," for "school"; it should be translated in the German or Greek mode of English as "book-house," and it has that sense of book house. It does not have the genitive sense of "books' house," you see. Nevertheless, we have another option for combination in English from the Germanic or the Greek, which are the same, and that is the French, the romance tradition, which is to make a phrase out of it, "house of the book." Now that "of the" is very ambiguous in English. It could be a genitive construction or it could be a partitive one, and the King James translators, using that partitive construction throughout, thus generate implicit allegorizations and personifications. Take the phrase "the house of the book." It is a house in which the book dwells, it is the house that belongs to the book, it is the house infused with the book, it is the house which is itself a trope for the book. You see?

Honig: Right.

Hollander: Whenever you have those constructions in the King James Bible, then you have a part allegorization. It's what gives the Bible its poetic richness all the time and is a basic building block of English poetic vocabulary; so that when you end up with a phrase, a resonant phrase in Wallace Stevens, say, like "the malady of the quotidian," you ask, "Well, what does that mean? Does it mean the fact that there's a quotidian which is in itself a malady? Or, does it mean that the quotidian brings particular maladies of its own with it?" And of course it means both, and of course Stevens is playing on that resonant ambivalence of the construction that is traceable to the Bible.

Honig: Yes. Well, by saying that, you're also implying what you said before—that to have a unified theory of translation means nothing more or less than a theory of literature.

Hollander: It "means nothing more or less"? No. Put it this way: a theory of literature is a necessary, perhaps insufficient, condition for a theory of translation, but I think a theory of translation is part of the theory of literature.

Honig: But in the example from the King James Bible, you're also talking about the style, the literary style, of English verse and its products into our own time.

Hollander: No. I would go on to talk about the literary style of English verse by saying that a construction, "house of the book," rather than "the book-house," lends itself more to accentual syllabic verse, to regular iambic verse, with few inversions, than does the Germanic-Greek recompounding, which gives you a lot more spondees.

Honig: All right.

Hollander: And you'll notice that as new words come into English—say with the Industrial Revolution—you get a lot more words that will be spelled with hyphens and that will be spondaic, because they will be that kind of compound. Mr. Fulton invented a steamboat, which was stressed búnk-búnk (like names, Jóhn Smíth)—steámbóat. Those compounds tend to show the boundaries of the iambic alteration. One of the things that happens, of course, is that when steamboat eventually gets to be an accepted compound, the secondary stress is removed, and it becomes stressed on the first syllable. That's how you know the compound has become a thing, and say stéamboat.

Honig: Do you think that for the theorist of translation there is something to be gained from a study of linguistics? I know you have been a student of linguistics.

Hollander: No, I'm not, though I've learned a little about it.

Honig: Well, I was thinking of transformational theories, like Chomsky's. Steiner thinks that he has to answer or contend with Chomsky. How do you feel about that?

Hollander: I'd rather not talk about Steiner and Chomsky because Chomsky has made clear what he feels about Steiner's understanding of his work.

Honig: All right.

Hollander: I think linguistics is very important. I don't necessarily mean that one particular mode of analysis of one set of problems in one philosophical context is what linguistics is. Since I'm not a linguist, I'm free not to have to worry about what the boundaries of the subject are. I think historical grammar is very important: knowledge of the structures of language, knowledge of the relations between grammatical change and semantic change are very important, and the relations that those things have to trope are very important. I mean, I do think that we should know—because it's part of the life of poetry to deal with this—something about how, when the Indo-European languages began to be studied, one inevitable conclusion was that there had originally been a small stock of words, and that these had numerically expanded by processes of trope. That's certainly a very nineteenth-century theory; it looks most like biological recapitulation—that is, that in the ontogeny of a particular bit of synchronic metaphor, the phylogeny of the history of the language has been recapitulated, et cetera, et cetera. Well, I think that these matters are certainly important, yes.

Look, Milton uses a phenomenon of etymology as a very important figure throughout Paradise Lost. The relation between the primary quality of the meaning of a word that we ordinarily use and an antithetical kind of primary quality, that of its prior etymological meaning, and how these come up against each other, are for him a basic metaphor of the then and the now, of the fallen and the unfallen.

Honig: Right, exactly. So one can say with Milton, without being Milton, that there's a way of approaching the subject of translation in terms of an imaginative adaptation of theories of literature, in the general sense, and particular linguistic theories.

Hollander: Yes, I think the truest poetry is the most feigning, and probably the most satisfactory and effective translations will have the virtue of being appropriate to their literary and historical milieus. A certain kind of accuracy—one of definition, one sense of what accuracy means—has been appropriate to certain aspects of modernism, but there are great loose, free, adaptive translations. Compare Ben Jonson’s and Campion’s versions of the Catullus poem, the "vivamus mea Lesbia atque amemus," which do totally different things with it. I mean, Campion translates the first few lines, comes to the line, "the ever-during night" (the "nox est perpetua una dormienda"), "the one ever-during night," has that as his line, and loves it so much that he takes it as a refrain, and builds a new poem in two successive strophes; using that as a refrain, he leaves Catullus and writes a wonderful little poem of his own that ends up with that fine image of "When I die I want people to be screwing on my tomb," et cetera. He gets to it by starting from Catullus, and then taking off, having seen the resonance of one particular line. Ben Jonson moves right through it and does something else. However, you have these two great Catullan versions, and that's an age, of course, in which people dwelled so much with classical texts that they could do what they wanted with them. In one sense, to let somebody know what Martial is really like, I would send him not to any particular translations of Martial but to J. V. Cunningham’s Epigrams, even ones that aren't direct. Cunningham has translated some of Martial, but some of his own original ones are absolutely it. They're the best ever, the best resuscitations of that kind of thing ever done.

Honig: Well, you seem to be saying in another way something that we started with: namely, that much of the activity of translation is implicit in learning how to produce poems, and doing that is a completely self-educating process.

Hollander: Oh, yes, absolutely.

Honig: You mentioned Campion and Jonson's reworkings of Catullus. That's one example.

Hollander: Look, in English the experimental aspect of the problem starts not with Chaucer getting French into English, but with Wyatt trying to get Petrarch into English and not knowing how. It really starts there with that kind of experiment, and then is repeated again and again in the history of English poetry. Sidney doing it, getting into Petrarch successfully.

Honig: You mean by getting into Petrarch, the sonnet?

Hollander: I'm talking about the form and the diction of the sonnet.

Honig: What about the subject?

Hollander: Well, the first getting of the subject into English occurs in Troilus and Criseyde. There is an inset bit, which is actually a translation of one of the Petrarch rime there. But it didn't have consequences of that kind; it wasn't the same thing. It wasn't Petrarchan, but the first Petrarchan attempt till Sidney, and then Surrey solves the problem immediately thereafter, and gets it right, and with his good ear manages to decide that the iambic pentameter line is the one to do the hendecasyllable in, although Wyatt tried every possible kind of thing as a way to do it. I mean, those poems are truly experimental and Wyatt possibly didn't know what he was doing.

But this problem, whether it's one kind of technical problem at one level or another, is really at the heart of the matter and keeps going through. Tennyson has so much of the Greek and Latin poetry that it just keeps flowing out all the time, and so many poets as different as Dryden and Tennyson have in common, say, the Vergil in their heads. When Dryden writes that beautiful elegy to John Oldham and when he ends up with that beautiful line, "Then night and gloomy death encompass thee around," he is doing a free translation of a line in the Sixth Book of the Aeneid that he himself translates a little more accurately and tightly in another place when he actually does the Vergil. But he feels free simply to do that, whether it is—as the late Ben Brower said—whether in Pope it is a poetry of allusion, allusion as a kind of trope in itself, or whether it is simply there; it is built into the language. Now, does one call that formal translation, that kind of allusion, or not, or what?

Honig: Well, you're talking about the business of the poet, I suppose.

Hollander: I think that's always there, and I think as Adams pointed out there is such a thing as translating from earlier phases of English into our own.

Honig: Yes.

Hollander: And I don't mean just Pope's formal redoings of Donne’s Satires, and things like that, you see. I mean simply keeping the continuity of the language going.

Honig: Yes. One common device is to ask students to translate Shakespeare into modern verse without knowing whether in the beginning they know anything about Shakespeare or much about modern verse. Assuming they knew a little about both, they would then begin to see that there's a problem, or what the problems are. Then, also, you feel that certain crucial texts to illustrate changes in style, or the inauguration of a new style, would be necessary to solidify the translation, as in the Bible?

Hollander: Yes, I think the Bible is very interesting in that aspect.

Honig: Well, we've reached high noon. Thanks very much.

Hollander: Right, my pleasure.