The Art of Translating Poetry 9780271072807

Citation preview

The Art of Translating Poetry

ALSO BY BURTON RAFFEL CriticisnlHistory The Develooment of Modern Indonesian Poetrv The ~ o r k & ~ o n ~ uAeStudy : of the Translation Process Why Re-Create? American Victorians: Explorations in Emotional History Politicians, Poets, and Con Men Robert .... . Lowell . . T. S. Eliot Ezra Pound: The Prime Minister of Poetry Introduction to Poetry How to Read a Poem Biblwgmphy Guide to Paperback Translations in the Humanities: A Teacher's Handbook AnthologiesICollections Poems The Signet Classic Book of American Stories The Signet Classic Book of Contemporary American Stories Forty-One Stories of 0. Henry Possum and Ole Ez in the Public Eye: Contemporaries and Peers on T. S. Eliot and Ezra Pound Tmnslatwns Poems From the Old English Beowulf Sir Gawain and the Green Knight ChrBtien de Troyes, Yvain Chairil Anwar: Selected Poems (with Nurdin Salam) The Complete Poetry and Prose of Chairil Anwar An Anthology of Modem Indonesian Poetry Ballads and Blues: Selected Poems of W. S. Rendra (with Harry Aveling) Gems of Chinese Poetry (with Zuxin Ding) From the Vietnamese: Ten Centuries of Poetry A Thousand Years of Vietnamese Poetry (with N. N. Bich and W. S. Merwin) Horace: Selected Odes, Epodes, Satires, and Epistles Horace: Ars Poetica (with James Hynd, David Armstrong, and W. R. Johnson) The Essential Horace Lyrics From the Greek The Bull Hide (La Pel1 de Brau) Russian Poetry Under the Tsars Selected Works of Nikolai S. Gumilev (with Alla Burago) Complete Poetry of Osip E. Mandelstam (with Alla Burago) Selected Poems of Alexander Pushkin (with Alla Burago) Poetry Mia Poems Four Humours Changing the Angle of the Sun-Dial Grice Evenly Distributed Rubble Man as a Social Animal Fiction Short Story 3 (with Robert Creeley and others) ARer Such Ignorance Founder's Fury (with Elizabeth Raffel) Founder's Fortune (with Elizabeth Raffel) -

-

-

--

THE ART OF TRANSLATING POETRY

Burton Raffel

THE PENNSYLVANIA STATE UNIVERSITY PRESS University Park and London

Library of Congress Cataloging-in-Publication Data Raffel, Burton. The art of translating poetry. Includes index. 1. Poetry-Translating. I. Title. PN1059.T7R34 1988 418l.02 87-43124 ISBN 0-271-00626-9

Copyright 0 1988 The Pennsylvania State University All rights reserved

Contents

vii xi

Preface Acknowledgments P a r t O n e : T h e o r y a n d L in g u is t ic s

1 The Specific Constraints of Language; The Unavoidable Linguistic Bases of Translation 2 The Constraints of Specific Languages 3 Forms and Genres 4 Prosody and Comparative Prosody

3 23 63 80

P a r t T w o : P r a c t ic e

5 6 7 8

The Subjective Element in Translation Collaborative Translation The Translation of Oral Poetry The Translator’s Responsibility

Notes Index

97 129 138 157 187 199

This page intentionally left blank.

Preface

My first book on translation, The Forked Tongue, appeared almost twenty years ago. Its subtitle was "A Study of the Translation Process." This second book retains a practical orientation, especially in its second part, but is much more linguistic- and theory-oriented. To some degree this is the natural result of two decades of reading and reflection. But it is also a result of disquieting realizations about general knowledge in the literary field-and it is to that point that I want to address the bulk of these brief introductory remarks. Academics are by nature specialists. Their training, their life's work as scholars, and their work as teachers all tend to reinforce specialization. But translation is by its very nature an interdisciplinary subject, whether one is practicing, theorizing about, or evaluating it. As I have argued elsewhere, and as I argue again in this book, the translator is (or should be) a literary person in the oldfashioned sense. The translator of poetry must be himself a poet, and the translator of literary prose is best able to do his job properly if he is himself a writer of literary (as opposed to scholarly or critical) prose. But the translator must also be something of a scholar. He I am indebted to Dr. Richard Ziegfeld, formerly my colleague at the Computer-Baaed Training Systems Group of McDonnell Douglas, and now at Sverdrup Technology, for the careful reading and informed challengeswhich have helped shape this preface. The flowcharts at the end of the preface are entirely his idea.

viii

Preface

must know more than simply the language with which he works. He must be aware of literary and cultural history both in that language and in his own tongue. And as I argue in the first part of this book, he must also have some awareness of basic linguistic realities. He should also have thought about the act of translation, both practically and from some theoretical point of view. Anyone who comments on, and anyone who wants to make optimal use of, translations should have at least some knowledge in these same areas. The real-life situation, however, is unfortunately that, while the best translators largely meet these expectations, most of those who comment on (scholars, critics) and almost all who use (academics) translations largely fail to meet them. Specializationis of course the key. Compartmentalized approaches lead to minds that are not only deeply habituated to a single perspective but also to minds that are inherently unaware even of the existence of other perspectives. As a result: -most academics are virtually indifferent to literary quality in the translations they assign for student use; they not only see little or no difference between poetry translated into verse or into prose, but they seem usually to feel that prose translations of poetry are better (more "faithful") -most academics know nothing of linguistics and see no reason to know anything, and even those who do know linguistics can see no application of their knowledge to translation(s). These are not casual problems. Hosts of students who have had to struggle through, say, Dorothy Sayers's execrable rendering of Dante, or Elinor Marx Aveling's ghastly butchering of Flaubert's Madame Bovary, or any prose translation of Beowulf, regularly find themselves bored, irritated, and worst of all on the basis of these mis-translations are unable either to understand or to feel any sympathy with the critical and otherjudgments p r o f e d by their teachers. A superb medievalist of my acquaintance, deeply learned in many languages and literatures and also in linguistics,was attracted to my recent version of Chrbtien de Troyes's Yvain but troubled about its suitability for a college honors course (limited to freshman students with an average SAT score of over 600). "Can they read it?" I was asked. I replied that I had read portions of the translation

Preface

ix

to my then four-year-olddaughter, and she had understood it without the slightest difficulty. I was (and am) quite confident that college students, and without any question high-ranking college students, would respond at least as well. I have had more conversations like this than I care to remember. I have also seen scholars teaching courses on, say, the grammar of Paradise Lost without any reference whatever to linguistics. I have read elaborate studies of prosody which are similarly innocent of linguistics and assert, inter alia, that different languages have different sorts of prosodies because that's just the way it is. In short, I have written this book to meet what seems to me an almost desperate need for some reasonably unified presentation of both the theoretical and linguistic and the practical aspects of translation. In dealing with translation practice I have however had to recognize that much of what I myself have to say has already been said in The Forked Tongue. And in dealing with the still more difficult problem of deciding a t what level to write this book, especially its chapters on theoretical and linguistic matters, I have had to accept the fact that relatively few potential readers are likely to bring to their reading equivalent levels of knowledge in all the areas discussed. To those who are expert, accordingly, some of this book will at times seem perhaps unnecessarily elementary. To those who are not expert, some of this book will a t times seem unnecessarily complex and demanding. I do not see any clear path through this difficulty. I hope the ad hoc solutions I have devised are at least not too disturbing to readers a t either level. Some amelioration may be available, too, through the following flowcharts. They are designed to codify and perhaps simplify the book's presentation: some readers may want to refer back to these charts at intervals.

Flowchart 1: The nature of translation (1)language -, -,(3) original -,(4) new

work (2) literary history/ -+ cultural history

language/+ (5) translation literature/culture

x

Preface

Flowchart 2: Translation decision points (A) If the languages are in some way(s) related (never identical): (1)phonological resemblances: nature, extent (2) syntactical resemblances: nature, extent, reliability (3) lexical resemblances: nature, extent, history, reliability

(B) If the languages are not related: (1)importance of phonological considerations; possibility of devising partial equivalents (2) importance of syntactical considerations;difficulty of devising even partial equivalents (3) nonimportance of lexical considerations; literary translation is not verbal transposition; one does not translate words

Flowchart 3: Decision points in translating forrnslgenreslprosodies (A) If comparable forrns/genredprosodiesexist:

(1)Are there comparable formdgenredprosodies in both languages? (2) Are the histories of those formdgenredprosodies similar? (3) Are the functions of the formdgenredprosodies similar in both languagedcultures? (B) If comparable formdgenres do not exist: (1)Are there formdgenredprosodiesin the new language which are in some central way(s) similar, though not fully similar? (2) Are there formdgenredprosodiesin the new language which at least have similar functions, if not similar naturesltechniques? (3) Can a fodgenrelprosody in the new language be adapted? (4) Can a fodgenrelprosody in the new language be invented? (5)Must the fodgenrelprosody be wholly or in part abandoned in the translation?

Acknowledgments

Acknowledgment is gratefully made as follows for permission to reprint copyrighted material: Earl Miner and Hiroko Odagiri, translation of BashoIKikaku, "Poetry Is What I Sell," from Earl Miner and Hiroko Odagiri, trans. The Monkey's Straw Raincoat and Other Poetry of the Basho School. Copyright O 1981by Princeton University Press. Excerpts reprinted with permission of Princeton University Press. Joanna Richardson, translation of five lines from Baudelaire's "Le Balcon." This translation copyright O Joanna Richardson 1975. Reproduced by permission of Curtis Brown Ltd., London.

J. B. Leishman, translation of Rilke's "Herbsttag." From Rainer Maria Rilke: Selected Works, 11. Copyright O 1960 by The Hogarth Press, Ltd.,Translated by J. B. Leishman. Reprinted by permission of New Directions Publishing Corporation. Ezra Pound, translations of "The Seafarer," a poem by Heine, and six lines from Arnaut Daniel's "Lo ferm voler." Ezra Pound: Translations. Copyright O 1954,1963 by Ezra Pound. All Rights Reserved. Reprinted by Permission of New Directions Publishing Corp. and Faber & Faber Ltd.

xii

Acknowledgments

Dudley Fitts, three lines of translation. Reprinted from Dudley Fitts: Poems From the Greek Anthology. Copyright 1938, 1941, 1956 by New Directions Publishing Corp. Reprinted by permission of New Directions. Kevin Crossley-Holland, translation of Caedmon's "Hymn." Reprinted from The Battle of Maldon and Other Old English Poems O 1965. John Moyne and Coleman Barks, translation of Rumi poem. From Open Secret, Versions of Rumi, translated by Coleman Barks and John Moyne. Threshold Books, Putney, VT 05346. Richard Hamer, translation of Caedmon's "Hymn" and "The Seafarer." Reprinted from A Choice of Anglo-Saxon Verse, translated by Richard Hamer. Faber and Faber Limited. Stanley B. Greenfield, eleven lines of translation of Beowulf From A Readable Beowulf: The Old English Epic Newly Translated by Stanley B. Greenfield. Copyright O 1982 by Board of Trustees Southern Illinois University. Reprinted by permission of the publisher. Christopher Logue, thirty-six lines of translation. From Patrocleia of home^ A New Version, translated by Christopher Logue (Ann Arbor: University of Michigan Press, 1963). Willis Barnstone, two lines of translation of Sappho. Reprinted from Greek Lyric Poetry, translated by Willis Barnstone, Schocken Books. Robert Kotewall and Norman L. Smith, A moonlit night by Tu Fu from The Penguin Book of Chinese Verse translated by Robert KOtewall and Norman L. Smith (Penguin Books, 1962), translations copyright O N. L. Smith and R. H. Kotewall, 1962. Reproduced by permission of Penguin Books Ltd. Arthur Cooper, four lines from The Ballad of the Army Waggons and Spring Scene from Li Po and TuFu, translated by Arthur Cooper (Penguin Classics, 1973), copyright Arthur Cooper, 1973. Reproduced by permission of Penguin Books Ltd. Michael Alexander, five lines from The Seafarer from The Earliest English Poems translated by Michael Alexander (Penguin Classics,

Acknowledgments

xiii

1966,1977),copyright O Michael Alexander, 1966,1977. Reproduced by permission of Penguin Books Ltd. Dimitri Obolensky, plain prose translation of Hamlet by Boris Pasternak from The Penguin Book of Russian Verse translated by Dimitri Obolensky (Penguin Books, 1962,1965), copyright O Dimitri Obolensky, 1962,1965. Reproduced by permission of Penguin Books Ltd. Kate Flores, translation of Rilke's "Herbsttag." Reprinted from An Anthology of German Poetry From Holderin to Rilke, ed. Angel Flores, 1960. Richmond Lattimore, translation of one-line poem of Sappho, reprinted in Greek Lyrics, translated by Richmond Lattimore, 1960, and translation of nineteen-line passage from the Iliad, translated by Richmond Lattimore. University of Chicago Press. A. J. Arberry, translation of six-line Rurni poem. Reprinted from Mystical Poems of Rumi, translated by A. J. Arberry, 1968. University of Chicago Press. Aaron Kramer, translation of eight-line Heine poem "Im wunderschoenen Monat Mai," and Louis Untermeyer, translation of four lines from a Heine poem, from The Poetry and Prose of Heinrich Heine edited by Frederick Ewen, Copyright O 1948. Published by arrangement with Lyle Stuart. Allen Mandelbaum, first nine lines of translation of Dante's Inferno, 1980; Mary Barnard, five-line poem from Sappho, 1958; and Guy Davenport, three lines from Archilochus, Sappho, Alkman, 1980. The University of California Press. Vladimir Nabokov, first ten lines of The Song of Zgor's Campaign. O Copyright, 1960, by Vladimir Nabokov. All rights reserved. C. H. Kwock and Vincent McHugh, translation of Tu Fu's "Spring Poem." Excerpted from: Old Friend From Far Away: 150 Chinese Poems jiam the Great Dynasties, Copyright O 1980 by C. H. Kwock and Vincent McHugh. Published by North Point Press and reprinted by permission. Frederick Goldin, translation of poem by Arnaut Daniel, from Lyrics of the Troubadours, Doubleday, O 1973 by Frederick Goldin.

xiv

Acknowledgments

Lydia Pasternak-Slater,translation of "Hamlet" by Boris Pasternak. Reprinted from Fifty Poems, 1963, translated by Lydia PasternakSlater. Unwin Hyman, London. Witter Bynner, translation of Tu Fu's "Moonlit Night" and "Ballad of the Army Carts" reprinted from The Jade Mountain: A Chinese Anthology by Witter Bynner, translated by Witter Bynner from the texts of Kiang Kang-Hu. Alfred A. Knopf, Inc., 1929. Robert Lowell, excerpt from "Hamlet" from Imitations by Robert Lowell. Copyright O 1958, 1959, 1960, 1961 by Robert Lowell. Reprinted by permission of Farrar, Straus and Giroux, Inc. Andrew Sinclair, a two-line and a six-linepoem from Selections From the Greek Anthology, 1967, translated by Andrew Sinclair. Serge A. Zenkovsky, seven-line excerpt from the Slovo (The Tale of Igor), from Medieval Russia's Epics, Chronicles, and Tales edited and translated by Serge A. Zenkovsky. Copyright O 1963 by Serge A. Zenkovsky. Reprinted by permission of the publisher, E. P. Dutton, a division of NAL Penguin Inc. M. Musa, first nine lines of translation of Dante's Inferno O 1971 Indiana University Press.

Part One Theory and Linguistics

I am very much in doubt whether Molbech in his translation of the Divina Commedia has done wisely to use only feminine rhymes. If the Italian language had contained masculine rhymes, Dante would naturally have made use of them; and I do not see any sense when translating the work into Danish to burden the translation with a defect from which the Danish language does not suffer.

-Henrik Ibsen, letter to Frederik Gjersten, 21 March 1872

This page intentionally left blank.

The Specific Constraints of Language; the Unavoidable Linguistic Bases of Ranslation

The Specific Constraints of Language There is no human society without some form of culture. "Culture," in this sense, is a basic aspect of human existence, a set of ways of living specific to a group of human beings and usually passed by them from generation to generation. Language is one such pattern. Just as human culture can change, so too can human language. Neither culture as a whole nor the subaspect of culture which we call language is or ever has been a closed system. Neither cultures nor languages are immortal. The social grouping (tribe, nation) which has created cultures and languages can cease to exist, or the forces of change can operate so powerfully, and so fundamentally, that one culture gives way to another, one language is superseded by another. In the long, largely unrecorded history of human language, we know of many such deaths and transformations. The Frankish tribes of Gaul, living under Roman rule, gave up their Germanic tongue

4

THE ART OF TRANSLATING POETRY

in favor of provincial Latin, and their form of Latin eventually became French.' The Egyptians gave up their original language, known as Egyptian, and entirely replaced it with Arabic. Under the rule of the Normans, beginning in 1066, the mass of Anglo-Saxon speakers in England did not completely abandon what we today call English, though many individuals among them did, but English was drastically and permanently changed by three hundred years of linguistic as well as military colonization. The number of English words of French origin is incredibly high, though we no longer think of such words as having been borrowed (or imposed). Once people stop speaking a language we classify it as "dead," though the nature and existence and some of the records of the language may continue and, to some extent, be studied and even used by scholars and a few select others. Sanskrit is a "dead" language which is still employed by a small number of Hindu religious devotees; Latin is a "dead" language studied in schools and colleges, and still fairly widely read, though rarely spoken and no longer anyone's native tongue; Hebrew is a "dead" language which has become revivified. Though they are not closed systems, languages are cultural systems, not in many ways much different from other patterns of social behavior. We correctly attach enormous importance to language. In its primary, spoken form it enables us to communicate both with others and with ourselves, for in addition to its vast role in our daily lives language also operates in our dreams and in our unconscious minds. Actually, thinking is most mysterious, and by far the greatest light upon it that we have is thrown by the study of language. This study shows that the forms of a person's thoughts are controlled by inexorable laws of pattern of which he is unconscious. These patterns are the unperceived intricate systemizations of his own language. . . . His thinking itself is in a language. . . . And every language is a vast pattern-system, different from others, in which are culturally ordained the forms and categories by which the personality not only communicates, but also analyzes nature, notices or neglects types

The Specific Constraints of Language 5 of relationship and phenomena, channels his reasoning, and builds the house of his consciousness. Brain research has made enormous progress since 1941, when Benjamin Lee Whorf wrote these words. But the central facts remain the same, at least for our purposes here. As Whorf goes on: In linguistic and mental phenomena, significant behavior. . . . [is] ruled by a specific system or organization, a "geometry" of form principles characteristic of each language. This organization is imposed from outside the narrow circle of the personal consciousness, making of that consciousness a mere puppet whose linguistic maneuverings are held in unsensed and unbreakable bonds of att tern.^ Plainly, the same is true as well of language in its secondary, written form (knowledge, information). Written language is not entirely identical to spoken language, but differs from the primary form only in less basic matters than those with which Whorf is dealing. The evidence for this view of language as a cultural system, one which both reflects and in its turn shapes and even determines verbal expression (and the thought and feeling which underlie it), is not hard to come by. Since this book is being written in a Western language, let me make the point still more emphatic by choosing my most prominent examples from three non-Western tongues: Japanese, Chinese, and Indonesian. Donald Keene writes: The sounds of Japanese are very simple. Each syllable generally consists of one consonant followed by one vowel. The restricted number ofpossible sounds has inevitably meant that there are many homonyms in the language, and countless words contain within themselves other words or parts of words of quite unrelated meanings. . . .From such a multiplicity of word associations evolved the kake-kotoba, or "pivot-word," one of the most distinctive features of Japanese verse. The function of the "pivot-word is to link two different images by shifting

6

THE ART OF TRANSLATING POETRY

in its own meaning. . . . [Thus] it was even possible for poets to keep two different sets of images going at the same time through an entire poem without any awkwardness [example omitted]. . . .One may give two almost entirely different translations of these lines. . . . [But] neither of these translations is a full rendering, because in the poet's mind and words there is a constant shifting of the two sets of images. . . . The author meant both to be understood at the same time, to draw as it were two concentric circles of meaning, each complete but indissolubly linked to the other. The effect achieved in this poem was naturally possible only because of the variety of word-play that Japanese afford^.^ (emphasis added) There are other striking dissimilarities between Japanese and the Western languages. Keene summarizes a few: Japanese sentences are apt to trail off into thin smoke, their whole meaning tinged with doubt by the use of little particles at the end, such as "perhaps," "may it not be so?" . . . The prosody of Japanese has been determined by the nature of the language. Stress accent, or quantity, the two most common features of European poetry, are ruled out by their absence in Japanese. This is true, of course, of French poetry as well, but the excessive facility of rhyme in Japanese, where every syllable ends in a simple vowel and there are no consonant clusters, deprives the language of this mainstay of French p~etry.~ Syntax (grammar) inevitably plays a major role in defining what is and is not possible, as well as what is and is not likely, for anyone using Japanese: Japanese grammarians classify the adjective along with the verb as a "working word." In most Indo-European languages the adjective is closely related to and dominated by the noun it qualifies, whether attributively or predicatively. The great part of our linguistic experience leads us to make the link

The Specific Constraints of Language

7

adjective-noun, in the matter of number, or gender flexions. But it is not relevant for the Japanese adjective that there is no nominal flexion that will specify any of these three factors, for the adjective acts in a manner much more akin to that of the verb and frequently stands duty for the verb in that it is really a fusion of adjective and copula. In fact, the adjective conj~gates!~ Bownas and Thwaite, whose comment I have just quoted, go on to observe that "the whole of the Japanese's linguistic experience moulds him to express himself in terms such as 'I probably shall not think As Arthur Waley put it, with a laconic calm that should not mask the high significance of the distinctions drawn, "There is no article in Japanese. . . .The Japanese verb lacks person and number. . . . Tense in the Japanese verb is usually implied rather than expressed. The verb-endings tend to define the nature of the action of the verb (e.g. whether it was complete, continuous, etc.) rather than the time when it occ~rred."~ I have heard (and seen) Donald Keene nicely illustrate the power of these distinctionsby displayinga scroll on which were written both a Japanese poem and an illustration of that poem drawn shortly after its composition. The poem deals with a bird or birds sitting on a branch or branches: the text neither was nor-since it was written in Japanese-could be clear about either the number of birds or the number of branches. And the illustration showed something like twenty-six birds sitting in &en branches! Though there are many differences between Japanese and Chinese, for our purposes the differences are less striking than the similarities. James J. Y. Liu explains: 9%

For instance, in the lines [example omitted] it is of no consequence whether "mountain," "bird," and "valley" are singular or plural. . . .As Chinese does not require any indication of "number," the poet need not bother about such irrelevant details and can concentrate on his main task.8 Liu draws other distinctions of equal importance: Th[e] sense of timelessness and universality is further enhanced by the frequent omission of the subject of a verb. . . .

8 THE ART OF TRANSLATING POETRY

Such omissions of the subject allows [sic] the poet not to intrude his own personality upon the scene, for the missing subject can be readily identified with anyone, whether the reader or some imaginary person. . . . Sometimes even verbs are omitted in Chinese poetry, and lines can consist of a series of nouns shed of all the connecting links such as conjunctions, verbs, and particles. . . . Another grammatical feature of the language of poetry, one that is closely connected with its syntax, is the fluidity of "parts of speech." In Chinese prose, words already enjoy a high degree of freedom in this respect, and the same word can be used as noun, verb, adjective, etc., according to the context. This freedom is increased in poetry. . . . Chinese grammar is fluid, not architectural. Whereas in a highly inflected language such as Latin, words are solid bricks with which to build complicated edifices of periods and paragraphs, in Chinese they are chemical elements which form new compounds with great ease.9 In Indonesian, a Malayo-Polynesian tongue which has no direct links to either Japanese or Chinese, one cannot ask if the person being addressed has a brother or a sister. It is not that Indonesian is an androgynous tongue, or that Indonesian culture does not make important distinctions as between males and females. But in family matters both the culture and the language make far more of age than of sex distinctions. Ada adik? one can ask, meaning "Do you have a younger sibling?" One can also ask Ada kakak? meaning "Do you have an older sibling?" But a straightforward query like "Do you have a brother?" or "Do you have a sister?" is linguistically impossible. So too one cannot indicate, in English, just who is included and who excluded in a word like "we." "We are going to the storem-but who is going? Do we include or do we exclude the person being addressed? In Indonesian there is never any such uncertainty. Kami pergi means that "we, excluding the person addressed, are going." Kitapergi means that "we, includingthe person addressed, are going." Unlike Chinese, Indonesian has tenses. But like Japanese it pays very little attention to them. Kita pergi, "We are going," can be turned into a past-tense statement by adding the word-particle su-

The Specific Constraints of Language 9 dah. Kita sudah pergi thus means "We went9'-but so too does kita pergi. Kita akan pergi means "We will go"-but, once again, kita pergi can also constitute a future statement. And an exactly parallel ambiguity exists with the use of Indonesian's reduplicative plural. Pulau means "island," so pulau-pulau (often writtenpulau2) means "islands." But pulau can also mean "islands." Indonesian forges linkages between words having to do with a similar subject. In English we say, for example, that a student studies because he wants an education, a student learns his lesson, a teacher teaches. Many root meanings are involved; most of the words I have italicized have very different etymologies and linguistic histories. But in Indonesian one starts with a single root, ajar, and extends it to all the various aspects of studying and learning: ajaran = teachings, doctrine belajar = to study mengajar = to teach mengajari = to train mempelajari = to study carefully mengajarkan = to teach a specific subject pelajar = a student pengajar = a teacher pengajaran = education pelajaran = a lesson terpelajar = educated, learned Although the languages of Europe are almost all of the same IndoEuropean family, these languages too differ in clear ways. Russian, for example, uses verbs as English never does. ''There are certain verbs of motion, the so-called 'double-imperfective' verbs which have two imperfective forms, the 'indeterminate' form which expresses repeated, habitual motion without definite aim or direction and the 'determinate' form indicating motion which proceeds in a specific direction andlor at a specific time."'O It would not be difficult to imagine uses for such a distinction, in English, but the imagining affects nothing. These are specifically Russian distinctions. For better or worse-and it really does not matter which-English does not have them.

10

THE ART OF TRANSLATING POETRY

German has prepositions that take only the dative case (aus, bei, mit, nach, von, zu) and prepositions which take only the accusative (durch, fir, gegen, ohne, um, wider). There are some prepositions which take only the genitive case-but the point is I trust sufEciently clear without listing them. English once had dative and accusative cases; it now has only their bare wreckage when indeed it employs cases at all. Links which are syntax-forged in German thus have no existence, and consequently no meaning, in English. There is no need to prolong these examples indefinitely. Let me close the list with Erich Auerbach's fine explanation of how classical Latin sentence structure became Vulgar (provincial) Latin sentence structure: Classical Latin possessed a very intricate system of means of subordination which made it possible to classify a great many facts, in their reciprocal relationships, in one syntactical unit: a sentence sometimes very long, but nonetheless very clear and limpid, which was called a period. . . . But it seems . . . that Vulgar Latin did not feel a need to classifyand systematize facts. Consequently, the art of the periodic sentence . . . fell into decadence. The participial constructions and those employing the subordinated infinitive were less used; the large number of richly shaded conjunctions was considerably reduced; the sense of those which survived lost its clarity; and relationships among facts, especially relationships of cause and effect, were no longer expressed with classical precision. . . . It was only much later, when the Romance languages had themselves gradually become literary instruments, that this state of affairs was modified." (emphasis added) There is nothing normative in saying that every human language is different from every other human language, sometimes exceedingly different, or that languages evolve and change. It is simply factual. We can choose to regret facts, as Auerbach does in the passage just quoted. We can choose to hail them, as Whorf did, exulting at "beautiful, effective, and scientific devices of expression unknown to western Indo-European tongues or mentalities."12 But our attitude, whether negative or positive, does not affect the facts

The Specific Constraints of Language

11

in any way. We must live with them, and with their consequences, whether we like it-or them-or we do not.

The Unavoidable Linguistic Bases of 'lkanslation Yet who would wish to discourage the peoples of the world from translating, merely because it is fundamentally impossible? -Thomas Mann13 The impossibility of translation is in a sense not debatable. If every human language is distinct (as it is) in structure, sound, and vocabulary, and if every language contains unique features, then clearly it is literally impossible to fully render anything written in one language into another. This is not a judgment about the translatability of poetry: it is a judgment about translatability in general. The catch, of course, is the word "fully." If it is not possible to fully render anything written in one language into another tongue, it is certainly possible to satisfactorily translate-that is, to translate most things and to translate them well. So-called literal translation, again, is on the face of it literally impossible. Exact linguistic equivalents are by definition nonexistent. But good translation can and frequently has been achieved. It is good translation with which this book is concerned, and specifically good translation of poetry. The primary linguistic facts, for good translation of poetry, are the linguistic facts of the language into which one is translating, not those of the language from which one is translating. The source language is naturally of high importance,for it is within its linguistic system that the literary work has been created. But by definition, too, that work already exists; the translation comes into being only when the translator re-creates the original work in the new (sometimes referred to as the "target") language. The translator must work with what the original work gives him, and those givens include the use made of the original language.14 But the inevitable, unavoidable differences between languages ensure that there are al-

12 THE ART OF TRANSLATING POETRY

ways significant aspects of the original literary work which cannot be reproduced in the new language. A few of these impossibilities, and their linguistic bases, should be highlighted. Let me number them for ease of discussion:

1. No two languages having the same phonology, it is impossible to re-create the sounds of a work composed in one language in another language. 2. No two languages having the same syntactic structures, it is impossible to re-create the syntax of a work composed in one language in another language. 3. No two languages having the same vocabulary, it is impossible to re-create the vocabulary of a work composed in one language in another language. There are also exceedingly important corollaries to these primary statements, corollaries which are founded upon the interaction of linguistic and literary fact:

4. No two languages having the same literary history, it is impossible to re-create the literary forms of one culture in the language and literary culture of another. 5. No two languages having the same prosody, it is impossible to re-create the prosody of a literary work composed in one language in another language. This is not a complete list; it is however more than sufficient to indicate the scope of the problem and some of the unavoidable perils with which all translators of poetry (and indeed of other literary work) must deal. Before discussing these five statements, it is worth emphasizing that each of them, like the assertion of the basic impossibility of full and complete translation, is deliberately framed in absolute terms. No one would argue, for example, that the phonology of any IndoEuropean ('Western") language is even remotely similar to the phonology of any Semitic, Sino-Japanese, or Malayo-Polynesian language. Nevertheless, there are individual resemblances,and so there also are approximations. And as between languages within, say, the

The Specific Constraints of Language

13

Indo-European family, those individual points of resemblance are naturally both more numerous and more marked. The phonological distance between French, for example, and Spanish or Italian is very much less than the phonological distance between French and Hebrew, Chinese, Japanese, or Indonesian. The same can be said of syntactic and vocabulary differences; it can even be said of differences in literary form and in prosody. The impossibility of exact recreation does not preclude the very real possibility of approximation-and it is precisely on approximation that good translation of poetry must be built.

1 . No two languages having the same phonology, it is impossible to re-create the sounds o f a work composed in one language in another language. Leonard Bloomfield's great pioneering study, Language, published in 1933 (and a revised version of a book first printed in 1914), is on these matters still both accurate and relevant: English (in contrast, say, with French or German) retracts the jaw; the Central and Western type of American English adds a tendency to raise the tip of the tongue. German and French (in contrast with English) advance the jaw and use the muscles more vigorously-German in large, sweeping movements, French in smaller and more precise ones, especially in the front of the mouth. Danish draws the muscles in toward the median line.ls In short, not only are there phonological differences between languages, but such differences have the usual physical bases which underlie all human sound production. One can measure, and comparative phonologists have indeed measured and charted, just what the phonological differences between languages are and how the vocal apparatus of people who speak each of the languages at issue makes each of the sounds a t issue.16 "The knowledge of the organs of speech, their relation to each other, and the way in which they are used in speaking, affords a universal framework in which the

14

THE ART OF TRANSLATING POETRY

sounds of all languages can be classified. This is possible because all human beings, no matter what race or other [ethnic]sub-division they belong to, have the same organs of speech, and can make whatever sounds are customary in the language they are brought up speaking."" More detailed examination of these differences, and some of their consequencesfor the translation of poetry, will be found in the second chapter. What needs to be emphasized here is that differing language sound systems, like different languages, are a given-a preexisting fact with which the translator must deal and which he cannot avoid. To pretend that English, say, is capable of the sounds of classical Greek (assuming that we indeed knew what the sounds of classical Greek were) is simply foolish. To base an attempt at translation on so foolish a pretense is doubly foolish.

2. No two languages having the same syntactic structures, it is impossible to recreate the syntax o f a word composed in one language in another language. The study of syntax, Noam Chomsky declared in the first sentence of his first book (significantly entitled Syntactic Structures), "is the study of the principles and processes by which sentences are constructed in particular languages."l8 If phonology is the weft of the particular social fabric we call language, then syntax is its warp. One governs the sounds a language chooses to work with; the other governs the structure into which those sounds must fit. Anne Cluysenaar neatly analyzes the reciprocal fitting together of a language's basic components: Modern linguistics has shown that natural languages are not ragbag aggregates of elements, but systems in which all elements are ultimately interrelated. At any one period of time, a speaker has a t his disposal a fixed number of sounds, words, and ways in which these may be combined. . . . These the linguist describes in the phonology, vocabulary and grammar [syntax] of each language.lS

The Specific Constraints of Language

15

It is hardly news that no two languages organize themselves in the same way. W. Nelson Francis puts it very plainly: "Each language . . . has its own unique system of behavior patterns. Parts of this system may show similarities to parts of the systems of other languages, particularly if those languages are genetically related. But different languages solve the problems of expression and communication in different ways."MPaul A. Olson takes a more playful approach: If we imagine language as a kind of game we play with words, sounds, and strings of words and sounds, we shall perhaps gain a certain insight into parts of its working. That is, we can think of words, sounds, and strings of words and sounds as pieces in, say, such a game as chess. The pieces in such a game can be moved about to form a variety of patterns, but the patterns which can be formed are determined by something we call "the rules of the game". . . .We know that every move we make must obey a rule. We cannot "remake the board" to our tastes and still be playing chess. . . . There are certain kinds of moves that we can make with certain words in relation to their sentence in the same way that there are certain moves that we can make with a chess piece in relation to its board. There are other moves which are not allowed.21 Winfred P. Lehrnann emphasizes the basic contrast between syntactical organization and nonhuman communication systems: "Syntax," he explains, is "the central component of human language. Of all the known communication systems-those of bees, of birds, of other animals-only human language has a highly developed syntax." 'What do we mean by grammar?" ask the linguists Angus McIntosh and M. A. K. Halliday. And they promptly answer: "When we are dealing with a closed system we are concerned with grammar. . . . When the choice is open, we are dealing with a lexical [vocabulary] selection, not a grammatical one."23 A very few examples, largely drawn from Indo-European languages with a high degree of familial closeness, will serve to support these statements. iQue hay? asks the Spaniard. (More or less literally, 'What's there?," though the meaning is closer to "How are

16

THE ART OF TRANSLATING POETRY

you?") Or he says: iQue pasa? ("What's happening?") But "How are you?" or "What's going on?" asks the native speaker of English, or sometimes "How's it going?" Qu'est qu'il y a ? says the Frenchman, meaning something very like the Spanish iQue hay?, but using a very different language form to express it. Or he asks: Comment ca va?, which is in meaning very close to the English "How's it going?" The structure may not seem terribly different, but in fact English uses the copula "is" and divides the verb into two parts, with the pronoun between those parts. The French, on the other hand, uses no copula and employs a one-part verb. For an utterance of only three words, these are in truth large differences. The German asks: Wie geht es Ihnen?, which again means something very like "How's it going?," though the structure is again different. ("How goes it with you?") If we step across the Pacific Ocean for just a moment, we find things utterly changed. Apa kabar? says the Indonesian, and then usually adds, Kabar baik? ("What news? Good news?") Even without a working knowledge of these languages, it is easy to see that the same fundamental query can be structured in a host of different ways; it can also employ many different vocabularies. "Language channels those behavior patterns which underlie, or perhaps constitute, social structure," observes C. F. Hockett. "Specific linguistic patterns vary from community to community along with almost everything else."24 If it is necessary to approximate in order to achieve a sort of partial similarity of language sounds, it is far more necessary, and far more difficult, to approximate syntax. This too will be discussed in more detail in chapter 2.

3. No two languages having the same vocabulary, it is impossible to re-create the vocabulary of a work composed in one language in another language.

Even words of close genetic kinship can have enormously different impact in different languages. English "sensitive," for example, is totally different from French sensitif, though the former was at one point a borrowing from the latter. To be "sensitive" is to be aware of others' feelings or of what is going on: it is usually a highly positive

The Specific Constraints of Language

17

description, a compliment. When English wants to indicate an unfavorable meaning for "sensitive," it usually employs a qualifying adjective: "He's hyper-sensitive," "He's over-sensitive," or "He's too sensitive." We also say "He's sensitive to criticism," which comes a bit closer to the French sensitifi ordinarily a neutral description at best but quite often a negative comment. Elvire D. Bar's Dictionnuire des Synonymes lists impressionable ("impressionable") as the synonym for sensitif; again, the connotation is at best neutral and quite commonly negative. The Indonesian phrase terima h i h (more or less literally, "acknowledgment of received giving") is usually considered a satisfactory equivalent of the English phrase "thank you." In many contexts it is indeed a functional equivalent-but not in all contexts. Offered something we do not want to accept, we say in English "No thank you." But an Indonesian in the same situation will say terima kasih. A phrase that means, in different contexts, both "thank you" and "no thank you" is clearly not a fully satisfactory equivalent of either of those two meanings, and must be used and understood with a keen awareness of its unlikeness to the English phrase. "Passion" in German is Leidenschaft. We think of "passion" as usually good, and always strong. But leid in German is "painful, disagreeable," and in its nominal form is "harm, hurt, injury, wrong, suffering, pain, mourning." Can one doubt that the two cultures, and their languages, thus express significantly different attitudes toward high emotion? (Leidenschaftlichcan mean "enthusiastic" as well as "passionate.") Again, vocabulary differences in the context of translation will be discussed in more detail in chapter 2.

4. No two languages having the same literary history, it is impossible to re-create the literary forms of one culture in the language and literary culture of another.

Literary forms (structures), like other linguistic structures, have deep roots in a particular culture and a particular language. "Every artist's work is conditioned by the limitations of the medium within which he works, by the cultural background in which he has grown

18 THE ART OF TRANSLATING POETRY

up, and by the demands which his culture makes on him. Hence the literature written in any given language is of course channeled . . . by the structure of the language."= "Cultural conditioning," adds C. F. Hockett, provides patterns; the poet fills them in. "The poet works from the outset with a more or less precise pattern into which words and phonemes must fit in order to qualify as a poem of the particular variety he wishes to write. . . . Poems are thus specific to language in a way in which prose literature is not."26 Some poetic forms travel reasonably well (the ode, the epic, the sonnet), some forms do not (the Indonesianpantun, the Malay shuir, the Japanese tanka). The sonnet is a distinctively European form, originating in Italy and migrating first to other Western countries and ultimately, though much less pervasively, to the rest of the The form's name comes from the Italian sonetto, meaning "a little sound, a little song." The history and development of the sonnet are long, complex, and rich. In the European cultures to which it spread, it has developed new and often exceedingly different variants and styles. The English sonnet has become distinctively unlike its Italian parent; so too, each in its own way, have the French and the German sonnet. No careful reader could mistake the Italian, the English, the French, or the German sonnet for the form as it is employed in any of the other languages. Gerard Manley Hopkins's sonnets, for example, are as conclusively English as Rilke's are German or Baudelaire's are French. Without anticipating the more detailed discussion of chapter 3, it may be helpful to consider briefly what happens when this same form suddenly h d s itself transported to a culture and into a language without most of the cultural and language-family ties existing between and among Italian, English, French, and German. Perasaan siapa tidakkan nyalu, Melihut anuk berlagu dendung, Seomng saja ditengah padang, Tiada berbaju buka kepala. Beginilah nasib anuk gembala, Berteduh dibawah kayu nun rindang, Semenjak pagi meninggalkan kandang, Pulang kerumah disenja kalu.

The Specific Constraints of Language 19 Jauh sedikit sesayup sampai, Terdengar olehku buni serunai, Melagukan alum nun molek permai. Wahai gembala disegara hGau, Mendengar puputmu menurutkan kerbau, Maulah aku menurutkan dikau. This is Gembala ("A Shepherd"), by the early modern Indonesian poet Mohammad Yamin (1903-1962). Yamin had learned of the sonnet through the work of Dutch poets, and used it extensively: nineteen of the first twenty-one poems he ever published were sonnets. But without for the moment turning the poem into English or considering its literary merits, let us examine its formal aspects. A good many things can readily be seen. There are fourteen lines, divided into two quatrains and two triplets. Like the Italian sonnet there are only two rhymes in the first eight lines; unlike any recognized Western version of the sonnet, the two triplets each contain only one rhyme (no great trick in a language as rhyme-rich as Indonesian). Without knowing Indonesian prosody (or phonology), one cannot I suspect tell a great deal about the metric employed. But it is plainly not the traditional metric used for the sonnet in the West. It is in fact the traditional Indonesian metric of four words per line. Every single line is end-stopped:this is highly unusual in the sonnet as the West knows it. It must be added, I think, that strictly speaking there is no such thing as a shepherd in Indonesian culture, since there are no sheep anywhere in the archipelago. The derivative nature of the form, that is, is paralleled by the derivative nature of the subject-and its treatment: I feel like someone on fire, Watching this boy sing, Alone in the middle of the meadow, Wearing no shirt, his head bare.

So it has to be for a shepherd's son, Resting under a shady tree, With the herds from morning on, Going home in the twilight.

20

THE ART OF TRANSLATING POETRY In the distance, vaguely, There's the sound of a flute, Singing of beautiful things. Oh herdsman on a green ocean, Listening to your high flute lead your flock, I want to follow along behind.

(N.B.:In an attempt to convey all the poetry the original has to offer, this version does not rhyme.) The description of the shepherd is plainly not a description of an Indonesian but of a European. The setting, from animals to "shady tree," is European. The typical Indonesian herd-boy does not play a flute or any other instrument: he either sits on the back of the kerbau (water buffalo) he is guarding or else is busy with his small friends. There are no "flocks." In short, the material content of the poem is as foreign to Yamin and to Yamin's country as is the sonnet form itself. And his handling of that form is hardly inspired. He breathes very little life into this alien construct. Having commented elsewhere on the literary qualities of this poem, let me simply quote from those earlier remarks: Yamin's work is not without a certain craft, it is not incompetent; I am sure, too, that he had persuaded himself of the emotional validity of the shopworn scene he portrays. He wanted, he perhaps needed to feel as Europeans feel-or felt. Looking back, [six] decades later, it is not hard to feel the thinness of Yamin's poeticisms, their tepidity and even unreality. I do not mean that in those days Indonesians could not honestly feel the kind of romanticism that . . . [European poets] express in their lines; not at all. . . . I mean only that Yamin's decision . . . to employ the sonnet form, previously unknown to Malay[-Indonesian] literature, imposed more of a burden on his poetry than that poetry could bear.28 Later Indonesian poets, notably Chairil Anwar, were able truly to domesticize the sonnet. Yamin's failure is in good part the failure

The Specific Constraints of Language 21 of the pioneer. But the significance of that failure here is that it underlines the peculiar difficulty of trying to transport lock, stock, and barrel the literary forms of one culture and one language into another. Literary forms can be adapted across cultural and linguistic boundaries; they can be transformed, shaped to fit a new context. But like all other expressions of culture and language they cannot simply be lifted, unchanged, into a new cultural and linguistic environment. "It seems to me," exclaimed Robert Payne, "that the world's languages all resemble infinitely complicated grids, and the basic patterns of these grids scarcely ever coincide, and even the units that make up the grid . . . never or very rarely coincide."29

5. No two languages having the same prosody, it is impossible to re-create the prosody of a literary work composed in one language in another language.

The discussion in chapter 4 will much more fully illustrate this statement: it is a subject which peculiarly requires illustration, both because there is little consistency in the many theories broached and also because many of the theorists, as well as a good many of the more practically oriented commentators, are unfortunately either totally or largely innocent of linguistic knowledge. The linguistic facts have been stated by Robert A. Hall, Jr., more clearly than by anyone else I have encountered: The way in which the utterance of syllables is timed, i.e., the rhythm of their utterance, differs from one language to another. . . . In one type (as found in Italian, French, Spanish, etc.) the rate of utterance of a succession of syllables remains approximately the same no matter how many stresses there are or where they fall in the stream of speech. . . . This type of rhythm is known as syllable-timed, since it is determined by the number of syllables (stressed or unstressed). In another type of rhythm (such as we have in English), the stressed syllables in an utterance come a t evenly spaced intervals, and any unstressed syllables falling in between the stressed ones

22

THE ART OF TRANSLATING POETRY are simply fitted in with greater or lesser speeding up as may be necessary. . . . [Tlhis type of rhythm . . .is known as stresstimed. . . . Since English and the other Germanic languages are stress-timed [citation omittedl, the meter of our verses is determined by the number of recurrent stresses in the line. . . . In the Romance languages, which are not stress-timed, but syllable-timed [citation omitted], a system of versification depending on recurrent stresses is not possible. . . . Romance versification depends on the number of syllables in the line.g0

So too the classical Western languages, Greek and Latin, being differently organized, have yet another prosodic system. "Unlike English poetry whose rhythm is largely the result of the arrangement of stressed and unstressed syllables, Greek verse is measured by predominantly quantitative standards. Down to the end of the fifth century B.C. and beyond, most poetry exhibits certain identifiable metrical patterns; long elements and short elements are distributed according to traditional rules. . . .With the exception of the Saturnian meter [citation omitted], all meters of classical Latin poetry are based on Greek prototype^."^^ As Maurice Bowra correctly says, because of its linguistic structure and its very different prosody "Greek is capable of effects quite beyond the capacity of English. Nothing can be done about this, and nobody has succeeded in the smallest degree in turning Greek choral meters into E n g l i ~ h . "The ~~ prosody of nondndo-European languages is different in yet other ways-but that discussion is best left for chapter 4.

The Constraints of Specific Languages

This chapter is a more detailed examination of three of the propositions, numbered one through three, in chapter 1, dealing with linguistic differencesin (1)phonology, (2)syntax, and (3)vocabulary, and with some of their significances for the translator of poetry.

Phonology Music, of course, is concerned not with beauty of sound but with organization of sound, and beauty has to do with the form of the organization. . . .The sound-patterns of poetry are . . .very complex. . . .The study of the complex sound-patterns of poetry has greatly lagged behind the study of the complex patterns of meaning.' The structure of any thing is simply the totality of all the relations that obtain among the elements of which it is com-

24

THE ART OF TRANSLATING POETRY

posed. The elements of which literary structures are composed are of course the linguistic elements of sound and meaning.2 The complex sound-patterns of poetry are necessarily built upon the sound systems of the language in which the poetry is composed. This includes, of course, everything embraced by the term "phonology," which covers the full range of sounds produced by any given language. The Javanese language, for example, makes a phonemic ("linguistically meaningful") distinction between two types of t and two types of d, just as Chinese (along with other tonal languages) makes a phonemic distinction between and among various tonal patterns. The famous sentence, known to all beginning students of Chinese, which in English runs "Mother scolds the horse" is in Chinese simply a repetition of the monosyllable ma in different tonal patterns. But poetic sound-systems also include the aesthetic traditions of what might be called the poetic culture. The linguistic and the aesthetic of course meet, blend, and mutually influence one another. "Modulation in pitch," explains James J. Y. Liu, "plays a role in Chinese verse comparable to that of variation of stress in English ver~e."~ We can also say, more startlingly, that poetic composition is a manifestation of culturally determined elements-simply another set of social acts. Craig La Drihre puts the case elegantly: When we speak therefore of the intrinsic properties and the corresponding relations of the sounds and meanings of language as the basis for an aesthetic or poetic structure in them, it is to properties and relations established in them by their social character as much as to any others that we are in fact referring. Not all their properties and relations, of course, are socially established, and in the character of these elements there remains a great deal that is not socially determined. But there is in the intrinsic nature of the sounds and the meanings of language a social aspect or element. Aesthetic structure in language is free of the social aim of other uses of language as instrumental to social action; but it is by no means free of social attachment, since that is already a constituent of the

The Constraints of Specific Languages

25

very nature of the elements of which the structure is composed.4 (emphasis added) Phonological peculiarities are inevitably nontransportable. It is as impossible to reproduce the phonology of one language in another as it is to translate literally any language's words or syntax. "The attempt to render the source text 'sound for sound,' " as Andre Lefevere explains, is on its face something of a fraud. Such a translator "does not render the sound of the source text in the target language," that being on its face impossible. At best, "his target text is merely an approximation to the sounds of the source text as filtered through the 'phonemic grid' of the target language."6 If that approximation is constructed intelligently and carefully, as in Ezra Pound's rendering of the Old English poem "The Seafarer," we can get a very good idea ("approximation") of how the original's phonology works-and without doing too much violence to the phonology of the new language. Notice how Pound pushes and stretches modern English's phonology (and its vocabulary and syntax as well) as hard as he can, without ever breaking out of the language's basic patterns: May I for my own self song's truth reckon, Journey's jargon, how I in harsh days Hardship endured oft, Bitter breast cares have I abided, Known on my keel many a care's hold, And dire sea-surge, and there I oft spent Narrow nightwatch nigh the ship's head While she tossed close to cliff^.^ Somethingof the sound and rhythm of the original from which Pound worked, the verse movement which he labored mightily (and successfully) to capture, can be understood just from seeing that Old English original set out in normalized spelling. Notice in particular how ingeniously Pound has balanced a very deliberate aping of the Old English with just as deliberate approximations. His "journey's jargon," for example, stems from sithas secgan ("to telVspeak of journeya/voyages"). His phrase alliterates, but not according to the

26

THE ART OF TRANSLATING POETRY

pattern used in the original. His "bitter breast-cares," on the other hand, stems just as directly and much more echoingly from bitre breostceare. His "reckon," at the end of line 1, translates only sound, not sense: wrecan means "to tell":

Maeg ic be me sylfum sothgied wrecan, sithus secgan, hu ic geswincdagum earfothhwile oft throwade, bitre breostceare gebiden huebbe, gecunnud in ceole cearselda felu, at01 ytha gewealc, thuer mec oft bigeat nearo nihtwaco aet nacan stefnan, thonne he be cliffumcnossath. I will return to Pound's "The Seafarer" version in part 2. There is much more to be said, and on both sides of the argument. It is easy enough, however, to see that Pound has taken modern English about as far as it will go, in an attempt to accommodate as much as possible of the phonology of Old English. To go even a little further is to court disaster. To go a great deal further, as does Louis Zukofsky (in many ways a Pound disciple) in his translations of Catullus, is to fall off the edge. Catullus's #loo, eight lines long, will be more than sufficient: Caelius Aufilenus' mate, Quintius Aufilena's am flow'rs Verona's 'n sum they pair and hunt (you venom) "he" 1 ' 1 frat him, he'll lace her or hum. Who quest,' quoth thick with ore, 'a lewd frater new way re th'whole cuss't sodality mm.' Who we favor-whom-put to us? Caeli, tibby: now to your (know bees?) perspective egregious "best" unique ah me kitty ah, come wasting on my eyes t'her reared at flame my medulla's. Seize fay licks, Caeli, see as in and more potence.' Again, a quick look at the Latin original shows all too clearly what Zukofsky is up to:

The Constraints of Specific Languages

27

Caelius Aufilenum et Quintus Aufilenum flos Veronensum depereunt iuvenum, hic fratrem, ille sororem. hoc est, quod dicitur, illus fraternum vere dulcem sodalitium. cuit faveam potius? Caeli, tibi: nam tua nobis perspecta egriegiest unica amicitia, cum vesanu meas torreret flammu medullas. sis felix, Caeli, sis in amore potens. When a so-called translator can produce "you venom" for iuvenum ("a youth") and "know bees?" for nobis ("me, us"), we have left the world of translation as this book is concerned with it. Here, for as unprejudicial a comparison as possible, is Horace Gregory's very plain translation: Caelius is desperately in love with Aufilenius, Quintius would surrender his life to Aufilena (these two men the flower of their families in Verona). Here's the golden age of brotherhood, now risen again and bloomingone or the other claims my favor. Then, you, my Caelius, take all I have; you've shown me that your friendship served mine. And at a time when Love, fire in my blood, consumed me. Here's good fortune, Caelius! May your true love pro~per.~ Lefevere's sober judgment, based on a very detailed examination of another of Zukofsky's Catullus renderings, is admirably restrained: On the whole, then, phonemic translation only very rarely achieves an acceptable rendering of the source-language sound in the target text while at the same time producing an acceptable paraphrase of its sense. All too often the much-sought equilibrium between dominance of sound and undercurrent of meaning is shattered. What remains are a few blessed oases

28

THE ART OF TRANSLATING POETRY of plain sense, devoid of successful sound-imitation, between vast bewildering stretches of moderately successful soundimitation either altogether devoid of immediate sense or running contrary to the sense of the source text. The result, in other words, is a hybrid creation of little use to the reader, testifying at best to the translator's linguistic virtuosity and inventivenes~.~

Translators more balanced than Zukofsky, and less single-minded than Pound, must still contend with the absolute nonequivalence of linguistic phonology-even between dialects of the same language. In the late-fowteenth-century poem Sir Gawain and the Green Knight, written in an obscure north-country dialect of Middle English, there is a stunning passage which no one can possibly fully re-create in modern English. The brave-but not entirely brave-Arthurian knight Gawain has come to honor a grisly pledge. The huge, oafish, obviously magical Green Knight has permitted (has in fact encouraged) Gawain to hack off his head. The Green Knight readily and macabrely survives the process, riding off, head in hand, to await Gawain's coming exactly a year later, so the game can be played in reverse. Gawain has of course serious doubts about his own ability to survive minus his head. But he comes, finds the appropriate site for the rendezvous, and as he approaches hears the Green Knight readying his axe, sharpening it for Gawain's neck. Quat! hit clatered in the clyffjas hit cleuve schulde, As one vpon a gryndelston h u d ~grounden a sythe. What! hit wharred and whette, as water at a mulne; What! hit rusched and ronge, rawthe to here.1° Middle English in general is far more ruggedly consonantal than our modern speech. This north-country dialect is still more so. The hissing harshness of the repeated exclamations, What!, fades into bland innocuousness in our clearly cognate equivalent, "what." The heavy alliteration, based on Old English patterns, works hand in glove with this consonantal emphasis. Words like w h d and whett~, again, have a sibilant harshness that no word in modern English possesses. The r all through the passage is both higher and more

The Constraints of Specific Languages

29

palatalized than our modern English r; it is "rolled" even more markedly than the r in some contemporary dialects of British English. Vowel placement is of course very different: the forwarding of English vowels had not yet taken place. Vowels placed further back in the mouth are by their physical nature heavier. The grim humor is thus significantly abetted by the poem's phonology; the anonymous author was well aware of what his language would let him do. The modern-day translator, similarly, must deal with the language he has available: What! It clattered on the cliff, as if To split it, like a grindstone grinding a scythe. What! It whirred like water at a mill. What! It rushed and it rang, and it sang Miserably.ll "It whirred like water at a mill" seems, and in fact is, considerably more effete than hit wharred and whette, as water at a mulne. On the other hand, the translation is decked out in a veritable flurry of internal and slant rhyming. The aural dimension added by this technique heightens the sound-density of the passage. It does not re-create the density of the original, because it cannot. But it creates a kind of approximation, in the language being translated into, which more or less re-creates (suggests, hints at, echoes) the effect in the original language. The almost pounding alliteration of the original cannot be re-created: neither the phonology nor the technique of modern verse in English will allow it. Instead of the full substance, we have the flavor of alliteration; once again we also have an added feature, the very different verse movement engendered by the use of enjambement in lines 2 and 4. There is no point pretending that the translation is the original: no translation ever is or can be the original from which it takes its life. The only valid standard remains: how successful is the translation as an approximation of that original? Indonesian phonology is vastly more unlike English than is the phonology of older forms and dialects of English. We speak of English as essentially a "lenis" language, meaning thereby that when a person speaks English the muscles on each side of the mouth are

30 THE ART OF TRANSLATING POETRY

relatively relaxed. To some extent, of course, this varies from speaker to speaker-but among native speakers of English there is nothing like the variation that exists as between English and any of the "fortis" languages (Indonesian, Japanese), in speaking which the muscles on either side of the mouth are extraordinarily tense. The effects of fortis as opposed to lenis vocal production pervade a language's entire phonology. In English, for example, the sounds p, t, k are aspirated in initial position. This, plus the generally lenis nature of English vocal production, also affects the vowels which follow these aspirated initial consonants. In these three examples, the vowels are diphthongal and therefore long. But in a fortis language like Indonesian, there are no long vowels: all vowels are short. Nor are there aspirated initial consonants. Initialp, t, k in Indonesian are therefore exceedingly unlike their English "counterparts." To a native speaker of English, a word like "pie," pronounced as in Indonesian, would sound more like a half-strangled attempt to say "buy." The shortness of Indonesian's vowels, too, adds to the tense, almost explosive phonic quality. Again, everyone who has studied Indonesian a t Cornell University recalls the dreaded early utterance Pukul bempa kapal terbang berangkat ke Surabaya? ("What time does the airplane leave for Surabaya?"). Spoken properly, this sentence splays out of the mouth rather like the explosive sounds made by a submachine gun. Native English tongues have a good deal of accommodating to do before they are comfortable (or accurate) a t such vocal production. The consequences for poetry are inevitably immense. Since I have already mentioned Indonesia's greatest poet, Chairil Anwar, in connection with the use of the sonnet in that country, let me set out one of his most brilliant works in the sonnet form, Tuti Art& Antara bahagia sekamng dan nunti jumng ternganga, Adikku yang lagi keenukan menjilat es artic; Sore ini kau cintaku, kuhiasi dengan sus + coca cola. Isteriku dalam latihan: kita hentikan jam berdetik. Kau pintar bemr bercium, ada goresan tinggal terasa -ketika kita bersepeda kuantar kau pulangP a m damhmu, sungguh lekas kau jadi dam, Mimpi tua bangka kelangit lagi menjulang.

The Constraints of Specific Languages

31

Pilihanmu saban huri menjemput, saban kali bertukar; Besok kitu berselihjalan, tidak kenal tuhu: Sorga hunya permainan sebentar. Aku juga seperti kau, semua lekas berlalu Aku dan Tuti + Greet + Arnoi . . . . . . huti terlantar, Cinta adalah bahuya yang lekas jadi pudar. (Since 1907, for the record, Indonesian has been printed in Roman characters.)Without knowing a word of the language,one can readily see both that this poem is a great deal closer to the sonnet as the West knows it than the Yamin poem cited earlier-longer, richer lines; a more appropriate rhyme scheme, especially in the "sestet"and also that the phonology is extremely different from that of any Western language. It is impossible fully to explain Indonesian phonology in this book: even a recording, though helpful, would not explain all the language's sound-patterns.But add to what has already been said about Indonesian phonology the fact that there is frequent use of the glottal stop (almost unknown in standard modern English, though it exists, marginally, in certain dialects); that the sound represented by the letter j is something like the first consonant in the English word "George," but far tauter, tenser; and that the sound represented by the letters ci is something like the first consonant in the English word "cheek," but again much tauter and tenser. Try to imagine what the Indonesian would sound like, as pronounced by a native speaker. Plainly, there is virtually no phonic resemblance to English. How then translate Anwar's poem? After many, many versions, here is what I came up with: Between present and future happiness The abyss gapes. My girl is licking happily at her ice cream: This afternoon you're my love, I adorn you with cake and coca-cola. Oh wife-in-training, we have stopped the clocks' ticking. You kissed skilfully, indelibly -When we cycled I took you home -Your blood was hot, oh you were a woman soon, And the stiff old man dreamed dreams that leaped at the moon.

32

THE ART OF TRANSLATING POETRY Every day's beau invited you on, every day's beau was different. Tomorrow we'll meet in the street and not know each other: Heaven is this minute's game. I am like you, everything ran by, Me and Tuti and Hreyt and Amoy . . . . . . dilapidated hearts. Love's a danger that quickly fades.12

Not much of Indonesian's phonology has been transferred into English. Some of the basic rhythm survives, but not consistently. For example, lines 1and 2 of the translation (which render line 1of the original: Indonesian is a vastly more compact language) echo the verse movement of the Indonesian. But line 3 of the translation (which renders line 2 of the original) is not in any way echoic of the Indonesian. Adikku yang lagi keenakan menjilat es artic ("My sweetheart [little sister] still [continuously] to her heart's content licks arctic ice [a form of ice cream]") is a bit like the sentence I quoted earlier from the Indonesian language training program at Cornell University: it moves too much unlike English for basic sound transfer to occur. There is simply no English equivalent for either the phonology of the sentence as a whole or for any of the individual words. Some small sound adjustments of course can be made. The Dutch name Greet is almost certain to be pronounced like the English word "greet." So I have invented a kind of English equivalent, to prevent at least some linguistic malformation. Lines 9 and 10 of the translation rhyme; there is part-rhyme as between lines 13 and 16 (plus internal rhyme in line 12 and part-rhyme, internally, in line 10). Alliteration being more natural to English than to Indonesian, I have rendered sus + coca cola as "cake and coca cola." It may help to suggest at least some of the clipped, terse Indonesian poetic effects. But what path could there possibly be for Cinta adalah bahaya, which in Indonesian positively soars, in part because of the aspiration that hangs after the last sound in the second word, in part because Chairil Anwar is quite simply a magnificent poet and uses Indonesian magnificently. (His fragmentary translations from English poetry, es-

The Constraints of Specific Languages 33

pecially perhaps those of T. S. Eliot, show a poetic and phonological mastery that is nothing short of wizardry. He somehow actually makes Indonesian sound like Eliot's lines, I do not know how.) The tonal qualities of the Chinese language naturally have had a major effect on Chinese poetry. By about A.D. 1000 the basic principles were (a) repetition of tones and (2) contrast of tones. "The resulting tone patterns," writes Julia C. Lin, "invariably decide the verse movements, which are predominantly musical. This innate musical quality of the language is the basis for the remarkable sense of tone color and musical nuance developed by traditional poets, and may explain why there is so much exquisite lyricism in traditional Chinese verse."ls "The effect of this orchestration cannot, of course, be reproduced in English a t all, belonging as it does to the phonetics of the Chinese language."14A fine summary is given by James J. Y. Liu: On the whole, Chinese verse has a stronger but perhaps less subtle music than English verse. The variation of tones creates a sing-song effect characteristic of Chinese, and in fact most Chinese readers chant rather than merely read verse aloud. At the same time, the relative paucity of vowels in Chinese and the lack of marked contrast between stressed and unstressed syllables render it more liable to monotony than English. Finally, the clear-cut quality of Chinese syllables, the absence of elision and liaison, and the fact that there are usually few syllables in each line, all tend to produce a staccato effect, unlike the more flowing, legato rhythms of English or French verse.ls There is little point in reproducing Chinese characters; here, instead, is a romanization, in the Pin-yin system sanctioned by today's Chinese government, of an eight-line poem by Du F'u (712-770; his name is also anglicized as Tu F'u), "Moonlit Night":

34

THE ART OF TRANSLATING POETRY

XirEng wh yan-hu6n shi Qing hui yh-bi h4n. ~ d - s hyl t xn hu&ng, ShurEng zh& &i-hdn gcin? It is clear even from this romanization (the accent marks indicating tones) that lines 2, 4, 6, and 8 rhyme. The poem's precise sounds could of course be communicated only by a sound recording or a live performance, but with the assistance of the preliminary comments, at least the general nature of the phonology should be more or less decipherable. A dog trot translation, arranged line by line to correspond with the Chinese original, may help to make still clearer how difficult the translator's task is (I will discuss some aspects of Chinese syntax in the next section of this chapter): Tonight Fuchou moon, My wife can only alone watch Distant sorrow for little sons daughters Not yet understand remember Ch'ang-an. Fragrant mist cloud-hair wet Clear night jade arms cold. What time lean empty curtain, Double shine tear-marks dry. Here is a translation by Ding Zuxin and myselE This moon above Fuzhou Must be watched from my bedroom, alone. My poor children, there in Zhangan, Cannot understand what their mother feels. She sits staring, stares long, stares hard. Her silken hair will be wet From the dew, from the mist. Her arms will be wet, will be chilled. When will we be able to lean at the curtain, together, Seeing tears dried on shining faces?16

The Constraints of Specific Languages 35

Stephen Owen has succinctly summed up the problem. "The distance that separates a modern English reader from an eighth-century Chinese poem can be crossed in only two ways-moving the reader or moving the poem. . . .We may learn and assimilate a new poetics; or we may remake the Chinese poem to answer the established values of English readers."17 Since translation assumes that we have chosen to move the poem rather than the reader, it will be additionally helpful to look at three more translations of this poem, two in verse and the final one in prose. Far off in Fu-chou she is watching the moonlight, Watching it alone from the window of her chamberFor our boy and girl, poor little babes, Are too young to know where the Capital is. Her cloudy hair is sweet with mist, Her jade-white shoulder is cold in the moon. . . .When shall we lie again, with no more tears, Watching this bright light on our screen? (Witter Bynner)ls This night at Fu-chou in moonlight, In her chamber she alone looks out; Afar I pity my little children That they know not yet to think of Ch'ang-an. In the sweet mist her cloud-like tresses are damp; In the clear moonlight her jade-like arms are cold. When shall we two nestle against those unfilled curtains, With the moon displaying the dried tear-stains of us both? (Robert Kotewall and Norman L. Smith)19 Tonight in Fu-chou my wife will be watching this moon alone. I think with tenderness of my far-away little ones, too young to understand about their father in Ch'ang-an. My wife's soft hair must be wet from the scented night-mist, and her white arms chilled by the cold moonlight. When shall we lean on the open casement together and gaze at the moon until the tears on our cheeks are dry? (David H a w k e ~ ) ~ ~

36

THE ART OF TRANSLATING POETRY

Several clear facts emerge from the comparison of these translations. First, though David Hawkes is one of the best of Western Sinologists, a brilliant critic, and a stunningly fine translator of prose, his clear, straightforward prose rendering is inferior to every single one of the verse translations, no matter what their quality. It is impossible to tell that Du Fu's poem is in fact a poem, reading only Hawkes's version. "The translations are intended as cribs," Hawkes explains. "They are not meant to be beautiful or plea~ing."~~ But is this prose rendering in any functional sense a translation at all? None of the translations, whether verse or prose, can do much with the phonology of the Chinese original. But each of the verse translations has created, to one degree or another, a sense of poetic approximation. Hawkes has created only a clear-eyed, romantic, cultivated gentleman's nocturnal musings. Second, though I do not want to make qualitative comparisons as between the Bynner translation and the collaborative one by Ding Zuxin and myself, it should be very clear that both these translations are superior in fundamental, and fundamentally important, ways to the translation by Kotewall and Smith. The difference, in a word, is that both Bynner and I are practicing poets, and Kotewall and Smith are not. Kotewall and Smith have an advantage over David Hawkes: they have at least tried to create a poem, and to that extent they are fairer both to the poem and to their readers. But they are painfully unskilled at poetic composition. Third, none of the translators, no matter their differences, makes the slightest attempt to capture the phonology of the Chinese original. As I will indicate in the next section of this chapter, something-not a great deal, perhaps, but something-can be done with the syntax of other languages, though they are necessarily very different from the syntax of English. But anyone who thinks that even the phonology of moderately closely related languages is transferable should remember the poor German translator of Shakespeare who turned "Hark, hark, the lark!" into Horch, hiirch, die Lerche! As Stephen Owen wisely says: Two languages contend for dominance in translation. It is a struggle that occurs all along the disputed text, and each language must yield points to the more forceful configurations

The Constraints of Specific Languages

37

of the other. . . . If we wish to escape this strange romance involving "fidelity" and "liberties taken," a field of warring identities and the erotics of power, we must conceivethe translator's art differently. . . . If we choose the way of translation, to move the poem rather than the reader, we must sacrifice Chinese poetics. . . . But we will choose the best analogues from our own literature.22

Syntax Le public doit comprendre l'art pictural; le lecteur doit comprendre l'art de la traduction. . . . On peut seulement cr&r une nouvelle muvre podtique qui ressemble a l'original comme un fr2re d son fr2re, comme un enfant 13 ses parents. . . .La liben% est dcessaire au traducteur non pas en tant que satisfaction de sa vanitd, mais pour qu'il ne soit pas l'esclave des &tails, pour qu'il puisse &re W l e au muximum a l'esprit de l'original et atteindre une ressemblance de portrait pour ce qui est de l'es~entiel.~~ ("People need to understand visual art; readers need to understand the art of translation. . . . One can only create a new poem which resembles the original as one brother resembles another, as a child resembles his parents. . . . The translator doesn't need freedom because it satisfies his vanity, but to keep him from becoming a slave to details, and to allow him to be muximally faithful to the spirit of the original, so he can achieve a fill and fair likeness, because this is what is truly important.") [Tlhe grammatical pattern of a language determines those aspects of each experience that must be expressed in the given language.24 Human languages are as incomplete, as faulty, as humans themselves. It would be hard enough to achieve a successful approxi-

38

THE ART OF TRANSLATING POETRY

mation of some perfectly regular but alien syntactical system-and, again, all syntactical systems are by definition alien to all other syntactical systems. But "system" does not mean perfection. All it means is a set of communal linguistic tools in good enough working order so that speakers of any given language can readily communicate with other speakers of that language. "Unfortunately-or luckily," Edward Sapir remarkedjustly, "no language is tyrannically consistent. All grammars leak."26 The fallibility of syntax immensely complicates the translator's task, especially if he holds to the literally impossible (because literally untrue) notion that an utterance in one language is fully transferable into another language. But the translator is also oddly freed by the "leakiness" of syntax: that which is inherently imperfect in the original language cannot require any sort of ideal perfection in the new language. To understand truly requires more than merely mechanical application of "rules." The translator must first be able to decipher, through the shifting, sometimes miasmic web which language weaves out of phonology, syntax, and lexicon (vocabulary), the "true" meaning of what he is translating. That is, the translator must first understand, as fully as possible, his text. I have deliberately omitted from consideration, for the moment, all the literary dimensions added by history, all the additional difficulties created by poetic form and genre, all the complications caused by prosody. Language of course precedes and controls poetry, which cannot be written in a linguistic vacuum. To focus better on syntax as an aspect of language in general, not limited to poetry, let me turn back to the quotation from the Russian poet, translator, and critic Vilgelm Venyaminovich Levik, reproduced at the beginning of this section in French and then translated (by me) into English. A sentence-by-sentence examination of that translation may help to illustrate more concretely this discussion of translating syntax. We can then move more comfortably and knowledgeably to a discussion of translating poetic syntax. A dogtrot rendering of the first sentence might run: "The public must understand pidorial art,the reader must understand the art of translation." But there are important reasons why this sort of rendering is unacceptable. Public is the lineal ancestor of "public"; one naturally tends to translate in the light of such plain linkages.

The Constraints of Specific Languages 39 But the true sense of public is a rather different English word, "people," and on further consideration the latter word is obviously a better rendering. French syntax involves formal considerations of both number and gender which no longer exist in English. Although in French public is masculine and singular, its sense, again, is the sort of mass plural which, in English, we express either through such generalized nouns as "people" or by the use of a generalized plural. Le public and le lecteur parallel one another; so too must the English, with "people" and "readers." To translate le public as "the public" is falsely to assume syntactical identity; to translate le lecteur as "the reader," once public has been rendered as "people," would be to break a significant rhetorical parallelism in the name of the fanciful whimsy of so-called literal translation. The verb devoir carries the sense of obligation, of duty. Doit could therefore be translated in a number of ways. "Must" seems too strong: Levik would have written something like I1 faut que le public cornprend (''It is necessary that people understand") had he intended to make his statement quite that strong. "Need to" is therefore a better rendering than "must." L'art pictural ("pictorial art"), similarly, is what the French-speaking mind will think. But "visual art" more accurately expresses how the English-speaking mind is likely to frame the same thought. A dogtrot version of the second sentence might go: "One can only create a new poetic work which resembles the original as a brother his brother, as a child his parents." The French-speaker thinks of an oeuvre H t i q u e . The English-speaker might think of a "poetic work," especially if he is an academic, but "a poem" better and more plainly expresses the meaning. Un @re ct son fi&e ("one brother to his brother") involves a perfectly natural repetition in French. English more idiomatically expresses the same idea as "one brother resembles another." The French phrase d son provides a syntactical parallelism lacking in English, so the final translation deliberately repeats the verb "resembles," though it does not need to be repeated in the original language. The third and final sentence is syntactically the most complex. It uses a typically French periodic construction, built around the balancing off of non pas en tant que ("not so much in order to") and mais pour . . .pour (''but in order to . . . in order to . . ."). This sort

40

THE ART OF TRANSLATING POETRY

of construction is usually, and this one is specifically, nontransportable into another tongue. Accordingly, I have had to substitute a similar but nonidentical English construction ("approximation"). "Liberty is necessary to the translator not so much in order to . . . ," a dogtrot rendering of the opening words in French might run.After shifting from the linked word "liberty" to the more accurate word "freedom," I have altered the syntactical construction to "The translator doesn't need freedom because . . ." Satisfaction de sa vanit4 could be rendered "satisfaction of his vanityn-but this works, in French, because of the non pas en tant que construction, which has not survived the linguistic move into English. So too,therefore, this construction must be shifted to "it satisfies his vanity." The second, positive-assertion part of the French sentence, built around two iterations of pour ("for, in order to, so that"), both conditioned by the key balancing word mais ("but"), might be rendered: ". . . but in order that he might not be the slave of details, in order that he might be able to be faithful to the maximum to the spirit of the original, and attain a portrait-like resemblance, because this is of the essence." The differently launched syntactical construction of the English translation mandates certain changes, as does the fact that we no longer have much functioning subjunctive left in our language. The French writer can make great play with soit ("might be") and puisse ("might be able"); the English writer lacks this resource and must accept that ineluctable fact. So the final translation says "to keep him from becoming a slave to details," because "becoming" carries for the English-speaking mind a reasonable and functionalapproximation of the French subjunctive.Pour qu'ilpuisse &re W l e au muximum ("in order that he might be able to be faithful to the maximum") similarly becomes "to allow him to be maximally faithful." "Attain" is not so truly accurate for atteindre, though it is historically/etymologicallycloser than "achieve"; "a portrait-like resemblance" is not so idiomatic a rendering of une ressemblance de portrait as "a full and fair likeness." The last clause, finally, though it too is introduced by pour, does not function as part of the pour + subjunctive verb construction. One might almost translate pour ce qui here as "which" rather than "because" ("a full and fair likeness, which is what . . ."). But "because this is what is truly important" seems to carry the thrust of the French better: there is a causative impulse in pour ce qui which would be muted by "which." Essentiel

The Constraints of Specific Languages 41 could be rendered by "essential"; de l'essential would then emerge as "of the essence," which seems distinctly too legalistic. "Truly important" avoids that rhetorical flaw. The late-sixbenth-century Spanish poet San Juan de la Cruz ("Saint John of the Cross") masterfully exploits syntactical possibilities. His Coplas del mismo hechas sobre un extasis de altu contemplacion ("Stanzas composed about an ecstasy of deep contemplation") contains eight seven-line stanzas, but begins with this three-line preface: Entreme don& no supe y quedeme, no sabiendo toah cieryia tm~endiendo. San Juan here makes high use of morphological features that English either lacks or does not employ in anything like the same fashion. He twice incorporates the pronoun into the past-tense verb (entre + me = "I entered"; quede + me = "I remainedlstayed"); he omits the subject pronoun entirely (no s u p = "I did not know"; yo for "I" omitted); and he uses the present participle (sabiendo, "knowing") much as English would use a verb form with full person and tense indicators (e.g., "I didn't know"). One might translate, dogtrot style: "I entered where I did not know (or: know how to) and I remained there, not knowing (how) all knowledge was achieved" (or, if one ignores the absence of a comma after sabiendo, one might also render this last portion of the three-line preface as "reaching [or: coming to] all knowledge"). The syntactical ambiguity is clearly deliberate. Much of the mystical power of this poet in fad depends on his ability to harness the compressions of Spanish syntax to his churning, only part-rational message. The second stanza, of full seven-line length, makes San Juan's syntactical patternings still clearer. Y o no supe donde entrava, p r o quando alli me vi, sin saber donde me estuva grandes cosas entendi. no d i d lo que senti que me quede no sabiendo toah cieryia tras~endiendo.

42

THE ART OF TRANSLATZNG POETRY

Dogtrot style: "I did not know where I had come Wentered, but when I saw I was there, without knowing where I had gotten to I understood great things. I can't say what I felttperceived which left me not knowing how all knowledge was achieved." Phonology plays an unusually large role in San Juan's poetry-but it constantly works hand in glove with his manipulation of syntax. He repeats words and phrases like incantations; this is a technique he uses in all his poetry. He weaves these repetitive phrases and words into a tonal fabric deepened and enriched by intricate, elaborate rhyme structures, but also into a structure of simple-seeming but in truth deeply ambiguous syntactical usages. The compressed form no supe, in the three-line prefatory stanza, becomes the fully elaborated yo no supe in the first line of this full-length stanza-but yo does not again occur anywhere in the poem. Such a pattern of what might be called deliberate syntactical omission (perfectlyproper in Spanish)places enormous stress on this one iteration of yo. The refrain, as it does in most of the poem, continues to use the participial sabiendo, but in this first full-length stanza San Juan also uses the infinitive form of saber ("to know"): sin saber ("without knowing"). He also introduces yet another verb with much the same meaning, entender ("to know"): grandes cosas entendi ("I understood great things"). Y quedeme ("And I remained1 stayed"), he writes in the three-line prefatory stanza, picking up that same verb (as in later stanzas he also picks up the verbs entender and saber) in this first full-length stanza, but in a different morphological form: me quede no sabiendo ("I was left not knowing"). The que at the start of the stanza's sixth line is again pivotally ambiguous: "which left me not knowing" or "because I was left not knowing." There is no need to trace out every syntactical ambiguity, every play on form and meaning. San Juan intends them all-and how does the translator who must work with the crisper syntax of English supply San Juan's poem to the reader who cannot know it in the original Spanish? I think two parallel columns of translation, the first by John Frederick Nims, the second by myself, may graphically show what a desperate wrestling match the translator is engaged in:

I entered-yes but where? knew nothing being there, burst the mind's barrier.

I reached that unknown place And, staying there, knew Nothing true.

The Constraints of Specific Languages

I entered-where, who knows?but being where I would (where, who dare suppose?) great things understood no telling if I could. Knew nothing being there, burst the mind's barrier.

43

I came to that unknown place, Knowing I had come But not how, or where from: All laws and boundaries were erased. Who could say How, staying there, I knew Nothing true?

It is unusual to find two responsible translators, working from one Western language to another, coming out with such different versions. Both translations attempt to weave a phonological web something like San Juan's; both attempt to reproduce as much as possible of the questing, ambiguous syntax. Nims has used three parenthetical interjections (not all marked with parentheses) and ended each with a question mark, though there are neither parenthetical constructions nor question marks in San Juan's Spanish. (The single question mark in my translation occurs here in this first full-length stanza.) Nims also relies on two fundamentally ambiguous English verbs, "would" and "could," to capture the elusiveness of San Juan's poem. Indeed, he relies heavily on verbs of deliberate vagueness: quedeme in the prefatory stanza becomes "being," and so too does quede in the penultimate line of the full-length stanza. My version attempts to find more precise syntactical equivalents for San Juan's verb-play, without resorting to verbs of elusive content. "I reached," "I came to," runs my version, in place of Nims's repeated "I entered," "I entered." Quedeme and quede are both translated "staying," but "knew nothing" is paralleled by "knowing" in the second line of the full-length stanza. Both translations are plainly aware of San Juan's fondness both for repetitions and for participial constructions. But there is, I think, a much greater simplicity in the syntax of my version, and I suspect this better echoes the deceptive simplicity of San Juan's Spanish. "But being where I would" and "no telling if I could," in Nims's version, echo nothing in San Juan, for the obscurity in the Spanish is subtle and integral, not strained. San Juan never strains: that is one of his greatnesses as a poet. It seems to me that the translator must, if he can, attempt to represent that supreme quality. And to

44

THE ART OF TRANSLATING POETRY

represent it here means in good part keeping the syntax at least superficially clear and limpid, as the Spanish is. I think that phrases like "burst the mind's barrier," though ingenious, tend to complicate this simple-seeming, flowing movement in San Juan's Spanishwhich is why I devised a perhaps equally simple-seemingequivalent, "nothing true." I had in mind, too, that as a rhyme word "true" gave me access, in the refrain, to a more basic, concrete rhyme of the sort San Juan favors. "Barrier" is I think far more limiting. Nims rhymes "barrierlthere" in his translation of the refrain; I rhyme "knewltrue," which seems to me more representative both of San Juan's technique and also of what he is talking about. Without the full knowledge of God, says the poem, he knew nothing true. It is not "barriers" of which San Juan writes, but the knowledge which is not knowledgewhich recognizes that toda cieqia ("all [usual]knowledge") is simply not good enough. Without this time reproducing the Spanish, let me add the next full-length stanza, once again in both translations: Nims

Of goodness and of peace many a thing I knew; deserts wide as space and one road leading through, clear, yet hidden too. I stood stammering thereburst the mind's barrier.

Raffel

I knew all knowing Of peace, of worship, Locked in silence stripped Of all but truth, yet growing Darker, not glowing, As I stayed and stammered what I knew: Nothing true.

The Spanish runs, in dogtrot style: "Peace and devoutness were perfect knowledge, in a profound solitude of comprehension (or: understood in profound solitude) (the right road), were a thing so secret that I remained stammering (how) I'd reached all knowledge1 knowing." Again, Nims has tried to elaborate and I have sought to approximate. There is no desert in San Juan, and nothing about space: via recta ("the right road") is the single parenthetical expression in the entire Spanish poem, and Nims takes it as something to be expanded, rather than, as I think San Juan intended, as a sup-

The Constraints of Specific Languages 45

porting, essentially sideways-moving reference. Its separateness is heavily underscored by the parentheses-which are a serious syntactical signal and should be so taken by the translator. Here, finally, is the Spanish:

De pax, y de piedad era la ciencia perfecta en profinda soledad entendida, (via recta) era cosa tan secreta que me quede balbuciendo toda cieqia tras~endiendo. The appearance at least of direct simplicity is fundamental to San Juan's entire poetic-and to the glowing religious message which it embodies. To complicate that simplicity, which finds expression in phonology, syntax, and lexicon (and also in literary form and in prosody), is, I believe, to translate in error. Different but parallel problems occur in Chinese. The word order is something like that in English, but Chinese employs particles unknown to English and does not strictly parallel English's analytical syntax, though word order influences meaning in both languages. As I noted earlier, Chinese is simply not concerned with verb conjugations, or plurals, or any of the niceties of tense so basic to English. A. C. Graham nicely says that "late T'ang poetry, which explores the Chinese language to the limit of its resources, can be damaged severely by the irrelevant precisions imposed by Indo-European person, number, and tense."28 He adds: A strictly word-for-word translation . . .disrupts English syntax without teaching the reader the syntax of the Chinese; it also requires, to be intelligible at all, a near-perfect equivalent of every word, since it allows no opportunity to compensate for deviationsby the phrasing of the sentence. [Such] a version . . . capable of standing on its own might therefore have to take more liberties with the original than would be necessary in standard E n g l i ~ h . ~

46

THE ART OF TRANSLATING POETRY

The first five lines of Du Fu's famous Bing-ch2 xtng, variously rendered as "Ballad of the Army Carts" or, as Ding Zuxin and I have translated it, "Chariot Song," illustrate Graham's point: Ch2 Itn -1tn,

Md x&-xi&, Xtng-rdn gling-jan g2 zdi y&, Y&-niiingqi-zl z6u xZng-sdng, CMn-di br2 j a n XZn-yting-qirEo. ("Carts rattle-rattle, Horses whinny-whinny, Soldier-men bows-arrows each one at waist, Fathers-mothers wives-children run see-off, Dust not see Hsien-yang Bridge.") Reduplicative verbs are all very well for children, in English. They do not work in adult poetry any better than they work in adult speech or prose. Note that the word order of line 5 shows why Chinese syntax is not analytical in exactly the way that English syntax is: it is not the dust that cannot see the bridge, but dust which makes it impossible to see the bridge. Witter Bynner's translation shows both how much of Chinese syntax can be readily turned into English and how much cannot: The war-chariots rattle, The war-horses whinny. Each man of you has a bow and a quiver at his belt, Father, mother, son, wife, stare at you going, Till dust shall have buried the bridge beyond Ch'ang-an.30 So too Arthur Cooper's version: The din of waggons! Whinnying horses! Each marcher at his waist has bow and quiver; Old people, children, wives, running alongside, Who cannot see, for dust, bridge over river.g1

The Constraints of Specific Languages 47

Both translators necessarily make changes, and supply syntactical bridgework, though Cooper has made an attempt to reflect the Chinese fairly closely by leaving syntactical holes unpapered over. Nee Wenyei's translation neither embroiders so much as Bynner's nor attempts, like Cooper's, to echo some of the precise movement of the Chinese: Chariots rumble and roll; horses whinny and neigh; Men are marching with bow and arrows at their hips. Their parents and wives hurry to bid farewell, Raising clouds of dust over Hsien-yang Bridge. (Nee W e n - ~ e i ) ~ ~ Ding Zuxin's and my version, unlike Cooper's, does not ape or echo the precise syntax of the Chinese, but tries, instead, to reproduce something of its terseness.= Like Nee Wen-yei and Bynner, we have tried to write in standard English: Chariots creaking, Horses neighing, Men marching, bows at their sides. Parents and wives see them ofE Clouds of dust smother the great Bridge.a What emerges from a comparison of all these versions, however, is that none of them is able simply to reproduce either the Chinese word order or indeed the words (in no matter how literal a fashion). Bynner and Cooper supply a possessory verb, to tie bows and arrows to the men carrying them; Nee Wen-yei accomplishes this by using "with"; Ding and I stick to the Chinese "at." But Ding and I "explain" the dust of line 5 with "smother," while Nee Wen-yei "raises clouds of dust," Bynner has the bridge "buried," and only Cooper manages to rely on the Chinese "see9'-though he does find himself having to tie that verb to the specific people enumerated in the previous line. None of the translators is able to pack into that fourth line all the named persomgw. The Chinese specifies "fathers, mothers, wives, children." Bynner gives us the somewhat altered "father, mother,

48

THE ART OF TRANSLATING POETRY

son, wife"; Cooper offers "old people, children, wives"; Nee Wen-yei and Ding and I limit ourselves to "parents and wives." To see perhaps the outer limits of syntactical translatability, working from Chinese to English, consider Du Fu's eight-line Chan wdng, translated as "Spring Scene," "Looking at Spring," "Advent of Spring," and so on. Here is a dogtrot rendering of the Chinese: "The state may be ruined, but mountains and rivers survive. In the city it is springtime and grass and trees grow thick. Flowers weep for the bad times, and separation startles the birds. Signal fires have been burning for three months. A letter from home would be worth ten thousand twls (unit of currency). Scratching at my white hair has made it even shorter: soon it won't hold a hatpin." Nee Wen-yei translates, in standard English: The city has fallen; only the hills and rivers remain. In spring the streets were green with grass and trees. Sorrowing over the times, the flowers are weeping. The birds startled my heart for fear of departing. The beacon fires were burning for three months. A letter from home was worth ten thousand pieces of gold. I scratch the scant hairs on my white head, And vainly attempt to secure them with a hairpin.= The translator's obviously half-desperate elaboration has in good part ruined the poetry. The tense structure, totally his own idea, jumps about wildly; the role of the birds has become obscure at best; and even the poignant wry humor of the last two lines has been smudged into triviality. Nor is Nee Wen-yei's desperation unique. Arthur Cooper has been similarly buffaloed by the untranslatable syntax (as the poetry has been largely destroyed, too, by his exceedingly lame attempt to employ English meter and rhyme and even English quatrains): In fallen States hills and streams are found, Cities have Spring, grass and leaves abound;

The Constraints of Specific Languages 49 Though at such times flowers might drop tears, Parting from mates birds have hidden fears: The beacon fires have now linked three moons, Making home news worth ten thousand coins; An old grey head scratched at each mishap Has dwindling hair, does not fit its

Six and a half closely printed pages of notes to this little poem do not save Cooper's inept rendering. But there is yet another approach, as best represented by C. K. Kwock and Vincent McHugh: My country in ruins Hills remain rivers Spring coming to the city The grass grows tall These sad days even the flowers wet with dewy tears When I grieve at our separation even a bird can startle me Fighting goes on and on these fist three months

50

THE ART OF TRANSLATING POETRY A letter from home worth ten thousand pieces of gold The more I scratch my white hair the shorter it gets -almost too short to hold a hairpin!37

There seems to be a good deal more elaborationhere than in fact there is. Spatial arrangement has been extensively used to suggest a good bit of the syntax and verse movement of the Chinese; the translation seems, but in truth is not, much longer than Nee Wen-yei's. The word count is virtually identical, with seventy-four words being used by Nee Wen-yei and seventy-five being used by Kwock and McHugh. And despite slightly unfortunate interpolations like "dewy tears," Kwock and McHugh have echoed not only the structure but also the bite and passion of the Chinese original. Dealing with what he calls "pidgin" translation, rather than the sort of spatial echoingused here, A. C. Graham wonders if "more can be done with this technique than has yet been a~hieved."~" (I do not think he was aware of the Kwock and McHugh rendering.) I suspect Kwock and McHugh take matters as far as they can reasonably be expected to go.

Vocabulary (lexicon) We use the term speech-sound . . . to eliminate grunts,clearings of the throat, screams, and all other possible sounds which do not function as part of the system of human speech. The term significantmeans "making a difference in meaning": . . . such a significant difference is phonemic, and a unit of speechsound which has phonemic significance is a ph0nerne.5~

The Constraints of Specific Languages 51 The term parts of speech is traditionally applied to the most inclusive and fundamental word-classes of a language, and . . . the syntactic form-classes are described in terms of the parts of speech that appear in them. However, it is impossible to set up a fully consistent scheme of parts of speech, because the word-classes overlap and cross each other.* Those who have not reflected on the translation process, and especially those who have not themselves attempted literary translation (and in particular the translation of poetry), frequently think of it as essentially the moving of words from one language to another. The notion of literal translation, plainly, is founded on some such conception. If all language were simply a matter of words arranged in strings, then of course it would be relatively easy-given acknowledged differences in available vocabulary-to transpose the language strings of one tongue into the language strings of another. Alas, translation is nothing like that easy. All the same, the building blocks of language include words, and words must therefore to some extent be translated. Just how much one can actually translate words might seem to vary enormously, depending on the phonological and syntactical similarities of the two languages between which one is working; literary and other formal considerations of course also play a role (to be discussed in the next chapter). But the realities of poetic translation are not what one might expect them to be. Let me start with the lexical aspects of translation from Indonesian, since in both phonology and syntax it is about as unlike English as can be, and there has been virtually no borrowing back and forth between the two languages. Most of Indonesian's linguistic loan words come either from Asian or Middle Eastern tongues; what Western loan words the language has acquired tend to come from Dutch-Indonesia having been a Dutch colony for some three hundred years-and to a lesser extent from English. (Indonesian's sister tongue, Malay, of course borrowed far more extensively from English, since Malaya was for a very long time a British colony.) Here is Chairil Anwar's most famous poem, Aku ("Me"), in an earlier version also entitled Semangat ("Spirit"):

52

THE ART OF TRANSLATING POETRY Kalau sampai waktuku 'Ku m u tait seorang 'kan merayu Tiduk juga kau Tak perlu sedu sedan itu Aku ini binatang jalang Dari kumpulannya terbuang Biar peluru menembus kulitku Aku tetap meradang menerjang Luka dun bisa kubawa berlari Berlari Hingga hilang pedih peri Dan aku akan lebih tiduk perduli Aku m u hidup seribu tahun lagi

("When my time arrives I wish that no one will mourn Not you either There's no need for (all) this sobbing I'm here a wild beast Expelled from the herd Let bullets pierce my skin I'll go on being excited, attacking I'll carry forward my wounds and my pain Running (forward), running (forward) Until the way of pain has vanished And I won't care any more

I wish to live a thousand years more")

The Constraints of Specific Languages

53

It is hard to do justice to the vocabulary dissimilarities as between a poem like this and anything ever written in English. Take, for example, the simple word Aku ("Me"), which constitutes both the title and the subject matter. It occurs in a variety of forms, not immediately obvious to the non-Indonesian speaker. Waktu, the last word in line 1,means "time." It has appended to it, as a sort of suEix, ku, which is a form of aku and indicates what we would call a genitive (though no such category of course exists in Indonesian). Waktuku therefore means "my time." 'Ku, the first word in line 2, is again a colloquial, even slang form of aku, and means "I." The normal form of aku occurs in line 5 and means "I"; the possessive usage occurs again in line 7, where it is appended to kulit ("skin"), giving the meaning "my skin." Aku as "I" occurs in line 8. The ku form occurs once again in line 9, where, however, it is attached to the word bawa ("carry") as something like a prefix, the combination meaning "I carry." Aku occurs, finally, in each of the last two lines. But not only is aku unlike any word in English in terms of sound; it is also unlike any English word in terms of syntactical/morphological function. In this thirteen-line poem, forms of the word are used as in English we would use "me," "I," and "mine," and the forms meaning "I" are used both separately (aku) and as a prefix-like ku. There are additional possibilities, all of them part of the word's range in Indonesian. Beraku means to use familiar speech with someone. Keakuan means "egoism." Mengaku means "to confess." Mengakui means "to acknowledge." Akuan or pengakuan means "acknowledgment, confession, creed, recognition."Beraku-akuan means "to promise." In parts of Malaya the word has close linkages with communal prayer; it can also refer to friendly familiar spirits. We can say that we have satisfactorily rendered the word when we use "I,me, mineY'but plainly we have crammed a vastly different set of Indonesian meanings into an English package, leaving out enormous associations of great importance. And aku is one of the simpler words in Anwar's poem. How do we deal with the range of lexical associations in rayu, the root form of merayu in line 2? It has to do with sadness and melancholy; it has to do with seduction and charm; it has to do with coaxing, in both amatory and political contexts. Rayuan means "a sentimental song" or "sentimental singing." Merayu means "to mourn, grieve over,"

54

THE ART OF TRANSLATING POETRY

but it can also mean "to flatter, to deceive, to seduce, to persuade." Or take the key word kumpul, the root of kumpulannya in line 6 (kumpul, "gather," + -an = "collection, gathering," + -nya, a suffixlike particle indicating either a possessive or, more loosely, something like a possessive "the"). Indonesian society is based on adat, meaning "customary" as opposed to Moslem law. Ia tahu adat, which to the uninitiated might seem to mean "he knows adat," in truth means "he has good manners." That is, it means that a person understands and knows how to act according to the rules of customary behavior. Indonesian society being a communaYgroup culture, the force of custom and the importance of fitting in with the group's standards, the group's requirements, cannot be underestimated. Indeed, when Anwar wrote this poem, in March of 1943, the single most striking fact about it, for most Indonesians, was precisely that the poet declared the relative unimportance to him of being driven out of the "herd" (i.e., the collective, the group). The vast number of iterations of aku-based words supports and abets this-to the Indonesian mind of 1943-terribly shocking declaration. Indonesian social doctrine held that one would be crushed by being expelled from the collective; no one could even carry on, much less thrive, in such a solitary state. Anwar forcefully asserts exactly the opposite. Here is my own translation of Aku:

When my time comes I want to hear no one's cries Nor yours either Away with all who cry! Here I am, a wild beast Driven out of the herd Bullets may pierce my skin But I'll keep on, Carrying forward my wounds and my pain, Attacking, Attacking Until suffering disappears

The Constraints of Specific Languages 55

And I won't care anymore

I want to live another thousand years4' One must know what the words truly mean in Indonesian; one must also know what the words mean in English. Since berluri means ''to run," I have seen a translation of this poem which reads: "Running, I carry with me the wound. / Running / Until the pain is 10st."~The problem is, however, that while one may run at something in Indonesian, in English "run" without the preposition "at" signifies to run away from, rather than toward, something or someone. The Chinese-speaking scholar who translated berlari as "running" did not know the effective lexical contrast-as also he did not know that luka dun bisa (more or less literally, "wound and wound-poison") cannot be translated as the singular noun "wound" when it is the plural "bullets" which pierce the persona's skin. Nor was he aware that for somethingto be "lost" is, in English, for it to suffer a negative experience: a lost dog, lost money, and so on. Hilung can mean "lost," among other things. But one cannot simply translate words. In Aku what is hilang is "suffering"-and to be rid of "suffering" is not a negative but a positive development. Rather than "lost," I have translated hilung here as "disappears," which is I have no doubt what Anwar meant. Aku is a militant, denunciatory, and also celebratory poem, not a negative or even a passive poem. So too this same Chinese-speaking translator renders Dan aku akan lebih tidak perduli as "And I'll be more indifferent," which is sheer nonsense in so rousing a poem. Lebih means either "more" or "any more," and though it is clearly the latter which Anwar intends, the translator has automatically chosen the somewhat more common usage. But, again, this is to treat words as if they controlled the poem, rather than the other way around. One cannot so translate and hope to translate well. Working h m languages lexically closer to English of course makes for some differences. But it remains true that one never translates simply by translating words: it is concepts and structures with which one must work, and words are only one of the many building blocks of which concepts and structures are composed. Vocabulary similarities, as well as phonological and syntactical similarities, can to

56

THE ART OF TRANSLATING POETRY

some degree ease the translator's burden. Just how much easing is problematical, as it is also highly variable (and therefore not to be relied upon). In general, a safe rule is to assume that the translator receives far less aid and comfort from lexical similarities than either he or anyone else might assume. Take, for example, the early lyric by Heinrich Heine, beautifully set to music by (among others) Franz Schubert, Zm wunderschiinen Monut Mai: Zm wunderschonen Monut Mai, Als alle Knospen sprangen, Da ist in meinem Henen Die Liebe aufgegangen. Zm wunderschiinen Monut Mai, Als alle Vogel sangen, Da hub' ich ihr gestanden Mein Sehnen und Verlangen. ("In the very beautiful/marvellous month of May, When all the buds/sprouts were springing open, Then in my heart Love sprouted. In the very beautifuVmarvellous month of May, When all the birds sang, I told/confessed to her My passionate longing and desire.") German is a cousin language to English. Apart from the vast importation of French expressions in the centuries after the Norman Conquest, the basic vocabulary of English is much more like German than it is like French. We can easily see those lexical similarities in Heine's eight-line poem: im wunder Mai alle

- in - wonder - May - all

The Constraints of Specific Languages 57 sprangen ist in meinem sangen und

- springing - is - in - my, mine - sang - and

The phonology is quite different, but less so than is that of Indonesian. The syntax is also different, but less so than is that of Indonesian. Literary history brings German and English fairly closely together: both have ballads (the form of Heine's poem), the technical aspects of which are quite similar. One can analyze Heine's prosody here in familiar terms: two replicative quatrains, each with a first line of iambic tetrameter and three succeeding lines of iambic trimeter. But what do these similarities mean, once we come to actually translating the poem? In May, the magic month of May, When all the buds were breaking, Oh then within my heart The fires of love awakened. In May, the magic month of May, When birds were merry-making, Oh then I told my darling Of how my heart was aching. (Aaron Kramer)'s Although Kramer is both a poet and an experienced translator, this version plainly does not succeed-and neither does it profit from the lexical similarities of the two languages. Kramer in fact discards such lexical similarities as wunderlwonder,sprangenlspringing,sungenlsang in favor of totally different word choices: "magic," "breaking," and "merry-making." He even mufiles such obvious lexical similarities as idin by using "within," and completely eliminates the undland linkage in the last line.

58

THE ART OF TRANSLATING POETRY

Kramer's focus has been on linguistic units larger than the word; his focus has also been far more on literary matters. But since his rendering is not a successful one, perhaps it is his focus which is to blame. Here then is the recent version by Hal Draper: In May, the magic month of May, When all the buds were springing, Into my heart the burning Bright arrow of love came winging. In May, the magic month of May, When all the birds were singing, I told her of my yearning, My longing and heart-wringing.44 Draper's translation uses more of the lexical similarities, but to no particular purpose. Sprangen is "springing" here, and sangen is "singing," and und is "and." But the poetry is as limp and fustian as Kramer's. Neither version comes close to capturing the lyrical sweep, the freshness, the charm, of Heine's German. Regardless of their treatment of lexical similarities, both versions feature large quantities of poetic fustian. "The fires of love" and the persona's aching heart are neither better nor worse than "the burning bright arrow of love" or "my . . . heart-wringing." These are stiff, stale, fundamentally ungraceful and unlovable renderings. Neither Schubert nor any other composer would be moved to music by this sort of artifically sweetened, saccharine verse. Concededly, lyrical poetry is one of the hardest genres to translate; no one in the last hundred and fifty years has done much to bring Heine into English. To date, only those who can read him in German have any true understanding of his deservedly large reputation. But one might still argue either that (a) these particular translators are simply not equipped to do the job correctly, or (b)German is too close to English: the very similarities may get in the way of effective translation. Let us therefore move further away, linguistically speaking, and choose a poet who has had more than six hundred years to make his way into English, Dante Alighieri. Here are the opening lines of the Inferno, the first book of the Divina Cornmedia:

The Constraints of Specific Languages

59

Nel mezzo del cammin di nostm vita mi retrovai per una selva oscum che la diritta via era smurrita. Ahi quanto a dir qua1 era 6 cosa dum esta selva selvaggia e aspm e forte che nel pensier rinova la paum! Tanf 6 amum che poco 6 piu morte; ma per tmttar del ben ch'i vi trovai, d i d de I'altre cose ch'i v'ho scorte. The lexical similarities are not only vastly lessened, they are almost reduced to the level of Chairil Anwar's Indonesian. One might find something of "me" in mi; there could be a distant echo of "fort" or even "forceful" in forte. For the rest, to those who cannot read Italian these lines might be written in almost any human language. Translators galore have tackled Dante: Midway the journey of this life I was 'ware That I had strayed into a dark forest, And the right path appeared not anywhere. Ah, tongue cannot describe how it oppressed, This wood, so harsh, dismal and wild, that fear At thought of it strikes now into my breast. So bitter it is, death is scarce bitterer. But, for the good it was my hap to find, I speak of the other things that I saw there. (Laurence Binyon)& Midway in our life's journey, I went astray from the straight road and woke to find myself alone in a dark wood. How shall I say what wood that was! I never saw so drear, so rank, so arduous a wilderness! Its very memory gives a shape to fear. Death could scarce be more bitter than that place! But since it came to good, I will recount all that I found revealed there by God's grace. (John Ciardi)*

60

THE ART OF TRANSLATING POETRY

Midway along the journey of our life I woke to find myself in some dark woods, for I had wandered off from the straight path. How hard it is to tell what it was like, this wood of wilderness, savage and stubborn (the thought of it brings back all my old fears), a bitter place! Death could scarce be bitterer. But if I would show the good that came of it I must talk about things other than the good. (Mark M ~ s a ) ~ ' When I had journeyed half of our life's way, I found myself within a shadowed forest, for I had lost the path that does not stray. Ah, it is hard to speak of what it was, that savage forest, dense and difficult, which even in recall renews my fear: so bitter-death is hardly more severe! But to retell the good discovered there, I'll also tell the other things I saw. (Allen Mandelbaum)" There are not many "perfect" translations; the adjective is not applicable to any of these versions, though each has its strong points and, on balance, Mandelbaum's seems to me distinctly the best. From the perspective of lexical translation, however, there is literally nothing to choose from, between and among these versions. It becomes very clear, comparing them, that none of these translations in fact translates words. Each takes a tonal and a technical perspective, fitting the words into those larger structures. Binyon employs a somewhat archaic diction; he also works much harder at preserving the full linked rhyme scheme-even transforming Dante's syllabic metric into a rugged but recognizable iambic pentameter. Not much of the flow and grace of Dante is left, though the basic thought is clear, as it is in all of the versions quoted from.

The Constraints of Specific Languages 61 Ciardi intensifies Dante, turning him into a poet who could have sat at table with Milton and exchanged poems with Donne. A modified form of the tena r i m is employed, as is an iambic pentameter metric. But Ciardi's Dante is exclamatory,turbulent, emotional. The rhetorical level is high; the diction strains and heaves. Mark Musa's Dante is distinctly less overwrought than Ciardi'sperhaps too laid-back and conversational. The t e r n r i m has disappeared, becoming simply a three-line stanza in blank verse. The rhetorical level is low; the diction is on the colloquial side. Mandelbaum's Dante is thoughtful, precise. The diction does not strain; the rhetorical level is neither elevated nor low. The tena r i m is presented much as in the Ciardi version; again, the metric is iambic pentameter. This is not a passionate so much as a meditative Dante. But what have any of these considerationsto do with lexical translation? Dante's first line, nel mezzo del cammin di nostm vita, is translated in four slightly different ways: Midway the journey of this life I was 'ware (Binyon) Midway in our life's journey, I went astray (Ciardi) Midway along the journey of our life (Musa) When I had journeyed half of our life's way (Mandelbaum) Three of the translators turn mezzo ("half, middle") into "midway"; one uses "half." Three turn nostra vita ("our life") into "our life"; one uses "this life." Three translate the noun cammin ("road, journey") into the noun "journey"; one uses the verb "journeyed" but also adds "way" to the phrase "our life." But these are differences without essential significance. No one judges or ever will judge the merits of translations in such terms. When I came to know the Italian and found myself, as I remain, dissatisfied with all the existing translations, I considered translating the poem myself. My rendering of the first nine lines, however, was based, like all the other translations, on a differing critical and tonal and structural perception of the original. There were no lexical issues other than those connected with larger matters such as diction, rhetorical tone, and the like:

62

THE ART OF TRANSLATING POETRY Half done with this road of life I woke in a murky wood Where the straight road was lost. Oh, a wood so wild, so bitter wild, So hard to describe, That even remembering brings back my fear! As painful, almost, as deathBut to tell the good things I found Let me also tell the rest.4s

The tenarima is echoed by intermittent rhyming; there is no regular metric. But what I was after was the clear limpidity of Dante's Italian. What I wanted and strove to achieve was something akin to the deceptively simple gracefulness of Dante's verse movement: I sought to translate the clarity without sacrificing the inner complexity. None of these considerations-none-has anything to do with the translation of words. And neither does good poetic translation of any kind, or from any language.

Forms and Genres

This chapter is a more detailed examination of the fourth of the five propositions, set out in chapter 1, dealing with the impossibility of re-creating the literary forms of one culture in the language and literary culture of another. Preliminary discussion of this proposition, in chapter 1, was largely concerned with transportability of the sonnet form into Indonesian; the discussion centered on a sonnet by the Indonesian poet Mohamrnad Yamin. Here I propose to examine the translatability of a variety of literary forms and genres, from a variety of languages and cultures. Again, no one would argue that literary forms and genres are linguistically landlocked, sealed forever into one language and one culture. All literary forms develop in one language and, over time, may be used in others-but always in adapted fashion. The history of the sonnet makes the point admirably. Of Italian origin, the sonnet spread all across Europe; more recently, it has come to Asia. But in each of its new homes the form has become something differentand the longer its tenure in a new language and culture, the more different it grows. The sonnet, with its Classical, Italian, and Catholic syrnbolism, evolved into a sort of national art form that was used for

64

THE ART OF TRANSLATING POETRY

many themes, addresses, and purposes, including political ones: the morning after a political happening in Florence, dozens of sonnets would appear on the walls. . . . We find the sonnet spreading more slowly into the Gerrnanies,not actually reaching them until the seventeenth century, and practiced in a relatively intermittent and uneven way in subsequent years. . . . The sonnet has been a major form in Greece, forming a sort of companion to the Italian sonnet, and it has been important in Yugoslavia for similar reasons. But in general the sonnet reached the Slavs very late-four to six centuries after Dante-and, for cultural reasons, has played a far lesser role in their literatures. . . . The sonnet is often controlled by formal and cultural factors that are relatively subtle and remain to be explored. For one thing, the greater formal variation among the Italians (and persons imbued with Italian such as Pushkin and D. G. Rossetti) suggests that the "native" imagines il sonneto more in the generative terms I have outlined, whereas, except for master experimentalists such as Frost and Robinson, the English-speakingpoet thinks of the Italian sonnet in terms of a memorized rhyme scheme.' Plainly, to become truly rooted, a transplanted literary form needs to grow indigenous roots. Paul Fussell explains:

In The New Criticism, John Crowe Ransom set forth a usehl distinction between the texture and the structure of a poem. The texture, he found, is "local," unique to a given poem, while the structure involves the larger elements of form which ally the poem with a tradition or with history or with a wider world of recurring shapes. . . .The fact that, like the limerick, [the sonnet] has meaning as a form alone is demonstrated by such a historical phenomenon as its lack of popularity among poets in the eighteenth century. These poets were acute enough to sense not merely that the sonnet as a form tends to imply a particular, highly personal, usually somewhat puzzled or worshiphl attitude toward experience, but also-and more importantly-that they did not possess that attitude nor could affect it convincingly. It was for these reasons rather than

Forms and Genres

65

because of any technical incompetence that they eschewed the sonnet and chose to work in other forms more expressive of their view of things and of themselves. Poetic forms are like that: they tend to say things even if words are not at the moment fitted to their patterns. As Louis MacNeice has said, "In any poet's poem the shape is half the meaning."2 Let us examine more closely certain central facts about the forms and genres of poetry in Japanese. Donald Keene sets out the basic goals and aims of the Japanese writer, and their consequences as well: In Japanese literature the unexpressed is as carefully considered as the expressed. . . . Seldom has it been desired to present the whole of any sight or experience. . . .The attempt to represent larger entities by small details resulted in a realism and concretenessin the images which contrast strangely with the misty ambiguity of the general effect. . . . In this we may detect the influence of the philosophy of Zen Buddhism which taught, among other things, that enlightenment could come from any sudden perception. . . . It may be seen that the effect of suggesting a whole world by means of one sharp image is of necessity restricted to shorter verse forms, and it is in fact in such forms of expression that the Japanese have in general excelled. . . . Certain genres of literature have developed to a greater extent in Japan than in other countries, perhaps as a result of the difficulty experienced by Japanese writers in organizing their lyrical impressions and perception~.~ Neither the epic form, nor true narrative poetry, is possible given this set of goals and aims. Bownas and Thwaite refer to "the lack of sustaining power in the Japanese man of letter^";^ Keene has noted that "in the late Tokugawa Period many patriots who found that they could not adequately voice their burning thoughts within the tiny compass of a Japanese poem turned to poetry in Chine~e";~ Arthur Waley declares flatly that "the Japanese poets quickly realized that they had no genius for extended cornpositi~n."~

66

THE ART OF TRANSLATING POETRY

What the Japanese developed, instead of the narrative or epic long forms created in the West, is a "linked" form, a sequence of short poems tied together not by any overarching plot, or theme, but at most by some common experience, like a festival or a shared night of poem-making. The connection that is established between the elements of these linked-verse sequences is poem by poem. That is, poem A is linked to poem B, but poem C is linked only to poem B, poem D to poem C, and so on. "Plot is impossible to sustain, and the lyric narrative is grounded solely in a sequence integrated by means of progression and association."' Subject matter, too, shifts, often according to more or less prescribed patterns: poems on seasons, poems on marriage, and so on. But as Keene emphasizes, "it was . . . considered undesirable to pursue the same subject beyond a few verses. . . . It is a poem unlike any ever written in the West, as far as I know, in that its only unity is from one verse to the next. . . . No matter which segment. . .we see at one time, it makes a beautiful composition, although when we examine it as a whole it possesses no more unity than a river landscape seen from a moving b~at.''~ Earl Miner, who has published the fullest account of the linkedverse poem sequence in English, explains that though "the method of composition seems casual, as close to randomness as intelligibility might allow[,] . . .linked poetry is also governed by elaborate canons, rules, and principles. It has been said that two decades are necessary for a poet to learn the rules well enough to compose naturally and show a talent." He adds:

In fact a sequence of linked poetry has many means of joining into wholeness those links which do not touch. There are parts defined by a "rhythm derived from music, parts defined by the method of setting down on sheets, and parts defined by topics that may well run well beyond two stanzas or that must not be repeated for seven or a hundred stanzas. The sheets on which sequences are set down must have flower stanzas, and the sides of the sheets must have moon stanzas, usually in particular places. Some topics are in place and some out of place on the front side of the first sheet.1°

Forms and Genres 67

What all of this means for translation is on its face simple enough: no one can possibly klly translate Japanese linked verse into any other language.A translator may give up in despair and not translate everything he sees on the page in front of him, leaving in Japanese what seems to have no effective equivalent. "If the translators have often not accepted Western approximations for particular Japanese andlor Chinese terms," declares the Translator's Preface to one rendering of Basho linked-sequence poems, "it is not to create undue diaculties for readers, but rather to admit an exactitude otherwise impossible." Of course, that being a Japanese rather than an English exactitude, these translators add at once that "as a result, notes may be needed in greater profusion than before."" Earl Miner takes a very different approach: "The reader of these translations will encounter a kind of English almost stripped of its capitals and punctuation." He adds that this is "a procedure new to my translations of Japanese poetry."12 Nobuyuki Yuasa, in the introduction to his translation of Basho linked-sequencepoems, elaboratelydefends the substitutionof a four-line haiku-stylestanza for the three-line stanza used in the original. "In my opinion, a three-line stanza does not carry adequate dignity and weight [in English?]. . . . I [also] had before me the task of translating a great number of poems mixed with prose, and I found it impossible to use the three-line form consi~tently."~~ The translator of a medieval linked-sequencepoem, framed around the ritual of rice planting, notes that "the original text is a difficult one even for Japanese scholars. It would of course be doubly so for the English reader who wishes to understand it in all its detail."14 In short, "like all other literature, linked poetry derives from contingencies of a nation, a language, a time, and individual poets."16 Here, finally, are the first six stanzas of Basho and Kikaku's Shi Akindo No Maki ("Poetry Is What I Sell"), as translated by Miner and Odagiri. These first six stanzas constitute the prescribed introduction to the linked-sequence form, the sho-ornote. (The second portion is known as the Development, the third as the Fast Close.) Stanzas 2 and 6 are by Basho; stanzas 1 and 3 through 5 are by Kikaku. I set the Japanese and the translations in parallel columns, as Miner and Odagiri do:

68

THE ART OF TRANSLATING POETRY

Shi akindo toshi o musaboru sakate kana

1. Poetry is what I sell short life not my debts concerns me so I drink the year out 2.

Shi akindo toshi o musaboru sakate kana toko hi kurete u r n ni nosuru koi

Poetry is what we sell short life not our debts concern us so we drink the year out as the sun sets on the wintry lake and horses bear away the carp

Toko hi kurete u r n ni nosuru koi hoko nibuki ebisu ni seki o yurusu ran

3. The sun sets on the wintry camp and horses bear away the carp they let the barbarian dull both in spear and head destroy the barrier-guard post

Hoko nibuki ebisu ni seki o yurusu ran Samisen. Hito no oni o nakashimu

They let the barbarian dull both in spear and head destroy the barrier-guard post. The sharnisen playing. It would calm the hearts of fiercest warriors

Samisen. Hito no oni o nakashimu tsuki wa so& korogi nemuru hiza no ue ni

5. The shamisen playing. It would calm the hearts of fiercest warriors on my sleeve the moon glistens as a cricket sleeps upon my lap

Tsuki wa sode korogi nemuru hiza no ue ni shigi no ha shibaru yoru fukaki nari

On my sleeve the moon glistens as a cricket sleeps upon my lap the longbill's wings are fettered to silence it in late night hours16

4.

6.

Forms and Genres

69

It is probably sufficiently apparent, even in the face of an excellent translation, that this is not native English-language poetry. No one can imagine a group of professional English-language poets sitting in a circle and composing verses such as these-verses which move laterally rather than in linear style, sliding from considerations of poetry to those of fishermen and border guards and finally (though this is only one-sixth of the actual linked sequence) to moonlight and crickets and tethered birds. A few of Miner and Odagiri's notes to these stanzas should, however, make the point obvious beyond dispute. The note to stanza 1reads: "This relates to the couplet by Tu Fu [Du Ful given [earlier in their book]. Kikaku was a famous tippler, but here we have mixed pride and shame over a drinkingpoetry life. Economic associations hover in akindo, musaboru (also meaning excessive enjoyment), and sakate, suggesting sai (debt)." The note to stanza 3 reads: "More Chinese atmosphere. Nibuki (dull) applies both to hoko (spear, halberd) and ebisu (barbarian). Since taking great fish out of water by the hooked spear is also implied with [stanza] 2, it may be, as some interpret, that ebisu also designates the fisher (although for marine fish) among the seven gods of happiness. Yurusu chiefly means 'destroy' but also 'allow.' "I7 Still, to nail the point down even more firmly, consider the following remarks about aspects of the phonology behind these Japanese originals: "Japanese is without stress accent, and as sung or recited as poetry, although not when spoken, has very little quantitative difference of syllables. All syllables are open vowels (in classical Japanese poetry all final nasals are vocalized), there are no true diphthongs, and in the classical language, all consonants are very simple. . . . Japanese poets, and probably before them the singers of Japanese folksong, developed a complex and subtle pattern depending mostly on the pitch of the vowels, certain echoes and repetitions which are not the same as rhyme, and a number of peculiar devices of meaning. . . . Classical Japanese poetry is read in a slow drone, usually a low falsetto; that is, the voice is kept lower and more resonant than its normal pitch, with equal time and stress on each syllable."1s We have in Miner and Odagiri's translation (and with the help of their detailed commentary) about what we are able to receive, in English, of this linked-sequence poem. What we do not have, and

70

THE ART OF TRANSLATING POETRY

cannot have, is the form itself. As I said in chapter 1,there are poetic forms which travel well and poetic forms which do not. Poetic exchanges, usually epistolary, have frequently taken place between English-language writers-but they are not a t all the same thing as the Japanese linked sequence. The very act of poetic "improvisation"-oral composition, without prior written preparation-is distinctly foreign to Western poetry as a whole, though there have been isolated pockets of improvised poetry-making. (Alexander Pushkin was apparently a master of the art; I know of no Englishlanguage poet of whom the same can be said.) We need to be grateful for what devoted craftsmen like Miner and Odagiri can give us. But we also need to be sharply aware of how much we cannot have. The Malay-Indonesianpantun is, as I have said, yet another poetic form that has not traveled well. Jikalau ti&k karena bintang Masakan bulan terbit tinggi? Jikalau tiduk karena abang Masakan saya &tang kemari? ("If not because of stads) How could it be the moon risesfappears high? If not because of older brotherlolder male/husband/loved one How could it be I cometarrive here?") A quatrain form, the pantun is divided into two couplets, frequently (though not necessarily) interrogative, and ordinarily rhyming like this example, a b a b. The inevitable linkage of the two couplets is plainly demonstrated in part, and nailed down, by this interlinked rhyming: not many pantun abandon interlinked rhyming for a straightforward and more Western-like a a b b rhyme pattern. The two linked couplets are in a sense "opposed" rather than, in the usual sense, consecutive. That is, the first (known as sampimn, a word which does double duty as a "clothes rack/peg/hook and as the foundation lines of a pantun) poses an evocative, challenging situation. The second couplet starts from ("hangs on") the sampimn and expands, sometimes very imaginatively, even fancifully, into the

Forms and Genres

71

pantun's true content. It can be a very clear expansion, as in the example quoted, or it can be almost abstruse. I will give an example of the latter in a moment. Here is a more polished but perhaps overliterary translation:

If not for the stars on their string Would the moon be pulled so high? Could anything But you keep me from walking on by?19 This translation attempts to preserve the Indonesian rhyme pattern; it also attempts to reproduce some of the tension, the excitement, of the second couplet's expansion by turning the sampiran into a fairly regular trimeter, but then substituting monometer for the third line and tetrameter for the fourth. The interrogative mode of both couplets is also reproduced. I made this translation more than twenty years ago and have come to feel uncomfortable with it. There would be a much less cramped feeling, I suspect, and far more accuracy, if the translation ran:

If not because of the stars Would the moon climb so high? If not because of you Would I keep coming by? The patterned repetition is preserved here; the metrical pattern remains unchanged across all four lines; and though the a rhyme is abandoned, the b rhyme-far more important in English practiceis retained. And of course lexically the second translation is distinctly a closer approximation. But neither version can possibly capture the true thrill of a basically folk form's carefbl, subtle variations. The Indonesian poem features exactly four words in each line. Line 3 repeats the first three words of line 1,then introduces its only departure in the fourth word. But substituting abang ("older brotherlolder male/husband/loved qne") for bintang ("star[sl") also sets up an implicit identification between the glowing beauty, the celestial power and romance, of the heavenly body and the earthly, all-too-mortal one. Several linguistic

72

THE ART OF TRANSLATING POETRY

barriers prevent this identification from being realized in English. For one thing, Indonesian's very different approach to singularlplural distinctions makes it necessary to insert a definite article in the translation, turning bintung into "the stars." This immediately breaks the pattern of the Indonesian poem. Neither "stringlanything," in the first translation, nor "the starstyou," in the second, can match the various levels of fit as between bintang and abang. It's simply not possible. And there are yet more impedimenta. Malay-Indonesians remain, for all the industrializing, modernizing effects of the twentieth century, infinitely closer to the physical realities of the stars and the moon, which have for the most part become mere literary symbols for us. There is nothing to be done about that, either. The obliquity of the pantun, its natural, normal way of aiming to the side of the target in order, metaphysically as it were, the better to hit it straight on, is not a Western approach. When we want to hit a target, we aim at its center. If we aim to the side, it is either because we do not want to hit the bull's-eye or because we are bad shots. The Indonesian does not think as we do. Once, when I was first in Indonesia, I said to a shopkeeper who was offering me somethingwildly unlike what I had asked for, Itu saya tidak suka ("I don't likelcare for that"). The friend who had brought me to the shop turned pale and, taking me to one side, demanded whether I was really determined to get us killed. I did not understand. "You can say Itu saya kurang suka, 'I like that less,' " he explained. "That's as strong a negative as you're permitted. If you say Itu saya tiduk suka, 'I don't like that,' it's as if you were trying to insult him-as if you were saying 'Your grandmother's a whore' or something like that." How does one carry over such built-in cultural biases, deeply embedded in their native language and culture, into another language and culture which would not recognize them even if it saw them? And how does one carry over into the new language the effect of having grown up with the pantun in one's ears, the effect of having heard, over and over, just this sort of subtle, delicate variation, always built into this same immensely popular form? That kind of heritage is as nontranslatableas the most eccentricidiomaticexpression or the most alien bit of syntactical construction-and perhaps even less translatable, since mere words cannot hope to summarize

Forms and Genres

73

years of intensely personal experience. There is a vast background, in rural Malaya and Indonesia, of pantun used in courting contests; one does not need to draw on that particular sort of experience, though it persists to this day, in order to make clear how embedded in cultural existence the pantun remains. When I said goodbye to one of the classes I taught there, the students sang, among other farewell songs, this cleverly varied pantun: Laju, laju, perahu laju, Lajunya sampai ke Surabaya. Lupa kain dan lupa baju, Tapi jangan lupa kepada kami. ("Quick, quick, boat, quick, Go quickly to Surabaya. Forget your shirvsarong, forget your jackethlouse, But don't forget us.") Kami, again, means "we excluding the person being addressed"and the cleverness, as well as much of the emotional force, came from the fact, known to everyone in the country and to everyone who has even an elementary knowledge of the pantun, that the expected rhyme word at the end of the last line was saya, "me." That is, the students were delicately transforming a personal love poem into a sort of collective love poem, indirectly expressing for their guru ("teacher") something like the degree of affection one would ordinarily express for one's sweetheart. It was deeply affectingand how in God's name would one translate it? How would one convey the marvellous indirection, the alternation of quick and slow, used in this pantun, the speed of the swiftly moving boat in the sampiran beautifully balanced against the hoped-for eternality (slowness, unchangeability) of remembrance expressed in the second couplet? But the pantun, as I've said, can be still more indirect: Telur itik dari Singgora Pandan terletuk dilangkahi; Darahnya titik di Singapura Badannya terhantur di Langkawai.

74

THE ART OF TRANSLATING POETRY

("A duck egg from Singgora," A silver pine stands too high; His blood drips in Singapore, His body lies neglected in Langkawai.**") The distance between sampimn and second couplet here stretches so tenuously that many Malay-Indonesian commentators have argued that there is in fact no connection and these are simply two entirely separate couplets-furnished with lovely, exotic imagery but meaning nothing in particular. The majority of scholars, however, have explained the connection as an implicit contrast between something very far off (the far-off north) and something very close (the "silver pine" of line 2). Using that implicit contrast as the poem's foundation, which is of course exactly what the sampiran is supposed to enable the pantun to do, the poem then comments on still another and more grisly contrast. A man has been killed (the language suggests murder: there is a perfectly good verb, berdamh, meaning "to bleed," but the poet chooses to say darahnya titik, "his blood drips"). The death (or murder) occurred in Singapore, but the dead man's body lies on the island of Langkawai, far from the scene of his death. It is perhaps a hypersubtle connection, but it is both palpable and unsurprising: subtlety is the hallmark of the pantun form-and is plainly one major reason why the pantun has taken no particular hold on Western poets. There is too much linguistic, literary, and cultural baggage attached to the pantun. And other cultures have more or less comparable forms of their own, far better designed to express what their languages, cultures, and literary histories have over the years chosen to express. How could a poet like Alexander Pope possibly prefer the pantun, had he known of it, to his beloved iambic pentameter end-stopped heroic couplet? And for the same reason, how can a translator readily find or invent a truly viable equivalent for the pantun in English? The approximation here is necessarily very approximate indeed. Japanese and Indonesian are of course languages, literatures, and cultures distinctly distant from English. What of a poem like that *Singgom = northern Malay peninsula, part of Thailand. **Langkuwai = island off northwest Malay coast.

Forms and Genres

75

by San Juan de la Cruz, discussed (from another perspective) in chapter 2? Let us consider the first two stanzas, as before, but this time from the perspective of literary form and tradition.

Entreme donde no supe y quedeme, no sabiendo toda cieqia tras~endiendo. Yo no s u p donde entrava pero quando alli me vi, sin saber don& me estava grandes cosas entendi. no d i d lo que senti que me quede no sabiendo toda cieqia trascendiendo. The regular stanza length in this poem, as I have noted, is the sevenline form; the three-line introductory stanza is followed by eight seven-line stanzas. But not only are the syntactical and verbal repetitions much the same in all the stanzas, no matter of what length, but the rhyme scheme is much the same as well. Stanza 1rhymes a b b. Stanza 2 rhymes a b a b b c c. The double rhyme of stanza 1 is not only repeated in the final three lines of stanza 2, but is additionally echoed in the a b b rhyming of lines 3 through 5 of that second stanza. The prosodic pattern is very similar, though not identical. Stanza 1has lines of 8, 8, and 10 syllables; stanza 2 has lines of 9, 8, 9, 7, 7, 8, and 10 syllables. Taken together, all the literary and cultural markers point to a form of literary ballad, which is exactly what the poem is. Writes Willis Barnstone, a Hispanic scholar (and a very fine translator), addressing a different but essentially similar poem: The parallelistic structure . . .-the repetition of the estribillo [refrain; pet word or phrase] and certain key phrases . . .-is characteristic of the Portuguese and Galician cosante. It is also a central prosodic device in Hebrew poetry of the Old Testament, as in Job and the Song of Songs. In fact we see the same device of parallelism used in the jarchas, poems

76

THE ART OF TRANSLATING POETRY

written in Spanish by Andalusians in Moslem Spain as early as the tenth century, which later appears in the Portuguese and Galician popular t r a d i t i ~ n . ~ ~ From a formal, literary perspective, what use have translators made of this plainly apparent background? Let me begin with Barnstone:

I came into the unknown and stayed there unknowing, rising beyond all science.

I did not know the door but when I found the way, unknowing where I was, I learned enormous things, but what I felt I cannot say, for I remained unknowing, rising beyond all science.21 Lines 2 and 5 of the seven-line second stanza rhyme; there is otherwise no rhyming in the translation. The diction and tone are emphatically neither popular nor ballad-like: Barnstone obviously knows the literary facts but finds no application for them in the actual translation. There is even a footnote explaining that "science" is "used in primary sense, in Spanish and English, of systematized knowledge"-hardly the frame of mind in which one reads or composes poetry with emphatically folklike roots. Prosodically, Barnstone uses a loose webbing of what might be called dimeter, trimeter, and tetrameter lines, but is probably more accurately termed vers libre and without any conventional patterning. These are not dispositive matters: if the translation sang, formal discrepancies would not matter very much. But it does not sing, and I suggest that the formal, literary discrepanciesbetween San Juan's Spanish and Barnstone's translation are a great part of the explanation. Consider the version by Roy Campbell:

I entered in, I know not where, And I remained, though knowing naught, Transcending knowledge with my thought.

Forms and Genres

77

Of when I entered I know naught, But when I saw that I was there (Though where it was I did not care) Strange things I learned, with greatness fraught, Yet what I heard I'll not declare. But there I stayed, though knowing naught, Transcending knowledge with my Campbell's version does sing, though somewhat archaically. "Naught" and "fraught" are not words of popular usage-at least not in the century for which Campbell was writing. And Campbell's rhyme pattern is echoic, though not an exact replication, of San Juan's. His first stanza rhymes a b b; the b rhyme is then picked up as the first rhyme of the second stanza, which rhymes b c c b c b b. That is very close to San Juan's pattern, and in an approximation very close is quite good enough. Far more of the ballad-like movement and feeling comes through here. Campbell also employs a regular iambic tetrameter, not only a metric associated with ballads in English, but also much the sort of regular prosodic beat which, though not strictly required (San Juan himself is, as I have noted, not quite so regular), can help capture the true feeling of the original. Nims's version, cited earlier, should be repeated here: I entered-yes but where? knew nothing being there, burst the mind's barrier. I entered-where, who knows?but being where I would (Where, who dare suppose?) great things understood no telling if I could. Knew nothing being there, burst the mind's barrier.z3 Nims too rhymes, and like Campbell not according to San Juan's pattern. Indeed, in the short first stanza Nims employs only one rhyme, accordingto the pattern a a a. In the second stanza he rhymes

78

THE ART OF TRANSLATING POETRY

b c b c c a a-again echoic of rather than replicating San Juan's pattern. The diction and tone are not quite so consistent as Campbell's, veering between plain and straightforward ('entered-yes 'I but where?") and rather elaborately formal ("where, who dare suppose?"). But the colloquial touches are nicely reflective of what San Juan himself does in Spanish. Rather than a regular tetrameter, Nims chooses a regular trimeter, thus speeding up the verse movement and removing the slightly stodgy feel of Campbell's version, but at the same time sacrificing some of the ballad-like associations. This version too sings, though its inconsistencies make it a less effective song, and less truly like San Juan's than is Campbell's. My own version, finally, should also be repeated:

I reached that unknown place And, staying there, knew Nothing true. I came to that unknown place, Knowing I had come But not how, or where from; All laws and boundaries were erased. Who could say How, staying there, I knew Nothing true?25 Like Campbell's and Nims's, but unlike Barnstone's, this version too rhymes, and according to a regular pattern which is yet not exactly San Juan's. The rhyme here is a b b a c c a a b b. Like Nims's, this version uses a great deal of internal rhyme; Campbell uses some, though not so much as Nims and I employ; Barnstone of course does not rhyme; and San Juan too uses a good deal of internal rhyme. The prosody, though patterned, does not follow either the vers libre approach of Barnstone, the tetrameter of Campbell, or the trimeter of Nims. This version scans as trimeter, trimeter, and dimeter for the first brief stanza, and then trimeter, trimeter, tetrameter (irregular), tetrameter, dimeter, trimeter, dimeter. The diction is I think both more consistent than Nims's as well as more emphatically informal (not quite colloquial, but not so formal as Campbell's). I

Forma and Genres

79

think the translation sings, and with something more or less like the voice of San Juan. But the important thing to observe, after examining all these renderings, is that none of them attempts literal transposition into English of either the literary form or the literary tradition of the Spanish. With the exception of Barnstone, whose translation appears pretty conclusively to do the least justice to the original, in its own way each of these versions chooses certain formal aspects of the original to approximate (not replicate) in English, necessarily placing less emphasis on other aspects. And because they are more successful, more consistent, in their approximations, I think either Campbell's or my version is to be preferred, depending on one's own taste. Again, none of the four translations, and especially not the version by the best Hispanist among the four translators (Barnstone), makes any serious attempt to transpose what is nontranspoeable: the exact literary form, the exact literary tradition, or the exact literary genre of the original. There is no point to multiplying examples, for what we have found even in a language and tradition reasonably close to that of English, and found too in languages and cultures far removed from English, is what we will find elsewhere. No two languages having the same literary history, it is impossible to re-create the literary forms of one culture in the language and literary culture of another.

Prosody and Comparative Prosody

To realize that in nearly all activities of language, rhythm carries much more than its traditional and formal definition limits it to, brings one to alter the definitions both of rhythm and of meaning, and their relationship. . . . Translating . . . with [the] rhythm, and in [the] rhythm, is not more d i f i u l t than translating within the frame of the sign. It is just a different program, for a differentpurpose, another way of being in the language. -Henri Meschonnicl Knowing . . . that [Beowulfsl prosody was traditional . . . ,a translator could then respond to those aspects by using what modern English speakers would consider a traditional verse form. . . . Translators wishing to produce an equivalent to the original would clearly not choose to imitate the Anglo-Saxon prosody, since this form is just the opposite of traditional to our ears: it is startlingly new and different.

Perhaps the most important, and least understood, fact about prosody is that it is neither fortuitous nor serendipitous. Verse rhythms

Prosody and Comparative Prosody

81

are determined by language rhythms. Even prosodies skewed by historical or sociopolitical pressures-like the prosody of our language, which was sharply affected by the Norman Conquest-eventually return like homing pigeons to their linguistic roots. When Tartar invasions destroyed native Russian culture, including of course Russian poetic culture, it was centuries before any sort of restoration was possible. And when in the seventeenth century Russian poetic culture finally began to rise from the ashes, poets and literary theoreticians, who had by then lost conscious knowledge of what Russian prosody ought to be like, tried to borrow the prosody of a sister language which had not experienced Tartar devastation and had preserved a continuous poetic culture, namely Polish. But for all its similaritiesto Russian, Polish is a language in which linguistic stress is not phonemic (i.e., shifts in stress do not change lexical meaning). The prosody of Polish is therefore necessarily syllabic. Accordingly, these early pioneers wrote in a syllabic measure for the most part of thirteen syllables, usually broken by a caesura (pause) after the seventh syllable. But Russian is a language in which stress is phonemic (i.e., shifis in stress do change lexical meaning), and so its prosody must inevitably feature linguistic stress. "The effect on Russian poetry was predictably disastro~s."~ Russian poets struggled for roughly a century, writing awkwardly and with great difficulty, and not understanding why what they were writing simply did not work. Then suddenly the poets Vasily K. Trediakovsky (1703-1769) and Mikhail V. Lomonosov (1711-1765) realized both the problem and the solution, working out the prosodic rules which have governed Russian poetry ever since-and which made possible, not too many years later, the magnificent work of Alexander Sergeyevitch Pushkin (1799-1837). They suddenly came to understand that, while syllables were of interest and importance in the construction of Russian verse, stress was far more important. The precise nomenclature is not here important (the relevant term that was developed is "syllabo-tonic" verse). What matters is that, now employing a prosodic stress, poets could begin writing real poetry. (Lomonosov even rewrote his earlier verse accordingto the new prosodic rules.) By the time of Gavrila Romanovitch Derzhavin (17431816) the clumsy syllabic metric had died a natural death; it has never been resurrected, nor can it be.

82

THE ART OF TRANSLATING POETRY

No one can say why one language is stress-phonemicand another is not. Indonesian, for example, which is emphatically not stressphonemic, naturally employs a syllabic prosody. One can say saya pergi, "I go," and stress either syllable in saya or pergi without affecting meaning. But in English, which is just as emphatically stress-phonemic and so naturally employs a stress-based prosody, shifts in stress can transform meaning. The "CONtent" of a box refers to what is in that box. But a person who is "conTENTWis satisfied, happy. A "black BOARD" is a board which is colored black. But a "BLACK board" is something attached to a wall, on which one writes, usually with chalk. And even when in English meaning is not totally transformed by stress shifts, it is nevertheless affected. A "HOUSE paint" is a paint used in or for a house, rather than in some other context. A "house PAINT" is a household object which is used for painting purposes. These are givens, not subject either to our control or, most of the time, to any variation. Very long periods of time, and immense political and cultural shifts, transformed Latin from a quantitative language, in its original classical form, to a syllabic language (with some stress aspects), in its late medieval form. But such transformations-and the extraordinary continuity of linguistic history which might make them possible-are rare. Even under the conquering onslaught of French, English never took on a syllabic prosody. Instead, English poets learned to combine syllabic counting with stress principles-which is of course what our traditional iambic pentarneters and tetrameters are all about. Indeed, it can be argued that the five-hundred-year span,roughly 1400 (when what might be called the "Chaucerian Compromise" had more or less settled into place) to 1900, shows a gradual but persistent movement away from syllable-counting in English pr~sody.~ In the same half millennium, one sees a profound though still only partial turn in the direction of the exclusively stress-governed prosody which had prevailed before the arrival of the Normans. That five-hundred-year evolution is not the subject of this book; this is not the place either to present or to argue the theory. But even the possibility (let alone the probability) that the theory is accurate shows, I think, the overriding strength of strictly linguistic factors in determining a language's working prosody. Since every language develops its own unique prosody, the trans-

Prosody and Comparative Prosody

83

lator cannot be expected to attempt the impossible: the reproduction of one language's prosody in another language. For the same reason that Language A employs a syllabic prosody, but Language B uses a stress prosody, it is linguistically impossible to reproduce either language's prosody in the other. Even two sets of syllabic of two sets of stress prosodies are not transferable. For example, the standard twelve-syllable French poetic line, the alexandrine, is as clearly syllabic as the traditional eight-syllable Indonesian poetic line, but there are additional, limiting factors which prevent us from turning either into the other. German has as distinct an iambic metric as does English, but it would be foolish, as well as inaccurate, to equate a five-stress German iambic line with English iambic pentameter. Approximation is thus once again the necessary rule, in prosody as it is in other aspects of translation. This chapter is thus an expansion of the fifth of the five statements of principle set out in the first chapter: No two languages having the same prosody, it is impossible to re-create the prosody of a literary work composed in one language in another language. The bulk of the discussion will test and provide demonstrations for a more detailed exposition of that basic principle. Let me start with two syllabic prosodies, the same two mentioned a moment ago, French and Indonesian. Here, first, is a French version, and a very good French version, of Chairil Anwar's Aku, a poem discussed in chapter 2 in connection with its lexicon (pp. 3033). Lorsque mon heure sera venue, Je veux que personne ne me regrette, Pas &me toi. Bien inutiles seraient de tels sanglots! Me voici, animal traqut!, De son troupeau rejt!t4. Qu'une balle me transperce,je n'en ai cure, Sans repit, exa@rt!, je me debattrai, Blessure et poison duns ma course emportant, Duns ma course, Jusqu'ct ce qu'aient disparu peine et tourment.

84

THE ART OF TRANSLATING POETRY

Et tout me sera encore plus indifferent, Car je v e w vive mille andes e n ~ o r e . ~ The crudest sort of prosodic comparison is perhaps the best place to start. Here, in parallel columns, are the number of syllables in each line of the thirteen-line, free-verse (vers libre) Indonesian original and in Louis-Charles Damais's French translation:

Number of syllables per line: Indonesian French 7 8 10 10 5 3 9 11 9 8 9 7 11 12 10 12 11 11 4 3 8 11 12 13 11 11 Should we conclude, from the bare fact that only three of Darnais's lines match the syllabic length of Anwar's, that the translator has failed? Hardly: a much more sensible question to ask would be whether Damais appeared to have given any thought to reproducing an exact syllable count, line for line. Fairly clearly, the answer is no. The variations, which occur in ten of the poem's thirteen lines, range from one to three syllables-that is, from relatively minor to moderately severe. And there is no pattern to Damais's matching or not matching the syllable count of the two languages. More interestingly, and more significantly, a line-by-line comparison of the Indonesian and the French demonstrates very convincingly that the two syllabic prosodies are simply not doing the

Prosody and Comparative Prosody

85

same thing. There is no need to march doggedly through the entire poem: the three-line initial strophe should be more than sufficient:

Kalau sampai waktuku 'Ku m u tak seorang 'kan merayu Tidak juga kau

Lorsque mon heure sera venue, Je veux que personne ne me regrette, Pas m&metoi.

The visual is here an accurate clue to the phonological. The Indonesian is so much more compressed, taut, even at times explosive than the French, that it necessarily occupies much less space. Put differently: the phonology of the two languages is so different that what we call a syllable in the one language is not necessarily equivalent to a syllable in the other. That is, even leaving aside all other considerations (and some of them are of powerful importance), a bare syllable-for-syllable equivalency is manifestly impossible. But some of those other considerations play a very important role in poetry. 'Ku m u tuk seorang 'kan merayu ("I wish that no one will mourn," in dogtrot translation; "I want to hear no one's cries," in my own version) is deliberately harsh, challenging. As I've noted, Anwar was consciously setting his face against the rule of the customary (Indonesian adat). But Je veux que personne ne me regrette ("I wish no one to regret me") is inevitably far more urbane, civilized, calm. (It is hardly accidental that the greatest playwright and poet in English is Shakespeare, an often moody, often dark, always passionate writer, while the greatest playwright and one of the greatest poets in French is Racine, a writer who is exceedingly controlled, working with an exceedingly plain and very carefully limited vocabulary.) That is, while there are clear linguistic limitations which affect Damais's translation, so too there are strong cultural and historical factors at work. Kalau sampai waktuku ("When my time arrives," in dogtrot rendering; "When my time comes," in my own translation) is similarly belligerent, provocative. Lorsque mon heure sera venue ("When my time will have come") is infinitely more philosophical, even a bit jaded, weary. French, and the people who speak that language, are in plain truth apt to be like that. Indonesians, and Chairil Anwar even more than most of his countrymen, are not

86

THE ART OF TRANSLATING POETRY

at all like that. It makes for large differences, in poetry as in other matters. I referred earlier to that standard French line, the twelve-syllable alexandrine. Consider what happens when one tries to turn an alexandrine into English. Here is the first stanza of Charles Baudelaire's Le Balcon (''The Balcony"):

M2re des souvenirs, mattresse des mattresses, 0 toi, tous mes plakirs! 6 toi, tous mes devoirs! Tu te rappeleras la beaut2 des caresses, La douceur du foyer et le c h a r m des soirs, M2re des souvenirs, mattresse des mattresses! ("Mother of memories, mistress of mistresses, Oh you, all my pleasures! oh you, all my responsibilities! You will remember (recall) the beauty of caresses, The sweetness of the hearth (fire-place) and the charm of evenings, Mother of memories, mistress of mistresses!") Baudelaire's poetry is experimental in many ways; in prosody, however, it is entirely regular. Every line in the stanza is an impeccable alexandrine. But since prosodies are not transferable, how does one render this rigorously traditional French line into English? Marthiel and Jackson Mathews's many-hands Fleurs du Ma1 (''The Flowers of Evil," Baudelaire's chief collection and the book in which Le Balcon appears) gives this version by F. P. Sturm: Mother of memories, mistress of mistresses, 0 thou, my pleasure, thou, all my desire, Thou shalt recall the beauty of caresses, The charm of evenings by the gentle fire, Mother of memories, mistress of rni~tresses!~

I have discussed elsewhere the larger issues involved in translating this poem.7 Here we are concerned only with prosody, and Sturm's solution is on the surface an informed and intelligent one. He has

Prosody and Comparative Prosody

87

taken the standard French alexandrine and equated it not with a twelve-syllable English line (and especially not with iambic hexameter, a disfavored meter in English: it is heavy to the point of boredom) but with the standard English line, the iambic pentameter. But the question remains: Is this a true or only an acceptable solution? Does it represent actual facts, or is it no more than an accommodationto surface identifications? Is there indeed any way that we can expect a particular prosodic pattern in one language always to equate with a specific prosodic pattern in another language? Is there sound linguistic reason behind such an automatic identification, or is this pedantry rather than poetry? If it is possible to discuss prosody without also discussing other factors, Sturm's iambic pentameter is clearly awkward at best. The first line in fact has twelve syllables, making it almost intrinsically awkward; the fact that Sturm closely apes the lexicon of the French original is only one of the things that helps tie him into poetic knots. It is not easy to find a conventional iambic pattern in this first line: as Sturm has constructed it, the line falls naturally into four beats, which fall (again quite naturally in a language like English, where alliteration has from the beginning played a much larger role than in the prosody of most other languages) on the four words beginning with the letter m. The second line is only arbitrarily iambic, since there is no linguistic reason to stress either "all" or "my," and one may end by stressing "my" simply in order to accommodate the pattern. This is of course exactly the opposite of how good prosody ought to work: first come the linguistic imperatives, and then, making use of them, working with rather than against them, come the metrical imperatives. Line 3 is made a bit limp by the necessity of stressing "of"; line 4 is standard and acceptable iambic pentameter; and line 5 of course repeats line 1. Frank Sturm was a minor English poet, author between 1905 and 1921 of three volumes of his own verse. Standard reference works make no mention of him, nor have I attempted to evaluate his poetry. We must assume, however, that he was basically competent in his own writing. Nor is Baudelaire an easy poet to translate. Martin Turnell speaks justly of "his dominating position in the European literature of the past hundred years. He stands at once for tradition and experiment, for discipline and revolt.'" That sort of constant

88

THE ART OF TRANSLATING POETRY

tension between the old and the new, between the traditional and the experimental, is hard to capture in another language, and also in another time. Additionally, for a great many reasons, prosody not at all the largest among them, Le Balcon is one of the most difficult of his poems to translate into English. It has had relatively few translators in the past hundred years; most selections from Baudelaire's work seem to omit it. The fact that Sturm was something less than competent in handling the prosody of Le Balcon helps emphasize both the difficulty of translating Baudelaire generally and the particular difficultiesof translating this poem, which Turnell nicely describes as "sorcellerie evocatoire ['evocative witchery']. . . . We exist for a moment in a luminous, trance-like atmosphere. We are separated from the everyday world. Time seems to have stopped. Everything is larger, more real than in 'ordinary life' "9 Here is Joanna Richardson's version of the first stanza: Mother of memories, mistress of mistresses, All I delight in, all I venerate! Thou wilt recall the joy of each caress, The gentle fire, the evenings exquisite, Mother of memories, mistress of mistresses!1° Whatever its literary merit, clearly not great, Richardson's translation opts for exactly the same safe and apparently unexceptionable prosodic solution, iambic pentameter. Her versification is frankly wretched. this is dogtrot prosody. But then, neither iambic pentameter version seems to me to capture the basic movement, and especially the pulsing vitality of the movement, of Baudelaire's versewhich is, again, syllabic verse movement and totally unlike the stress-based verse movement of English. Four of Baudelaire's five lines here employ a formal caesura after the sixth syllable. Line 3 has a faint caesura, again in this same medial position-which is obviously made possible by the structure of the twelve-syllable alexandrine line. Iambic pentameter by its very nature cannot ever have that sort of balanced cadence, one half the line set against the other half: half of five is simply not a whole number, and in addition English verse employs feet, in which there are two sorts of components, one less and one more stressed. I do not want to launch into

Prosody and Comparative Prosody

89

a disquisition on Baudelaire's handling of the alexandrine: these bare indications should, however, be suflicient to indicate that he wields the tools available to him like the master craftsman he is. But how can one capture in another language that utterly different verse movement? My own solution was to vary the prosody-preserving the basic notion of form, preserving rhyme, but employing a metrical mixture and adjusting it, line by line, to the Englished necessities of the translation: Womb of Memories, queen of desires, Oh you, my shrine, my delight, You summon the beauty of all brushing hands, The soft peace of the fireside, the grace of night, Womb of memories, queen of desires!ll These five lines are tetrameter, trimeter (if scanned traditionally) or tetrameter (if scanned less rigorously), pentameter (if scanned traditionally) or tetrameter (if scanned less rigorously), pentameter, and tetrameter. Twenty years ago, in explaining my reasons for this metrical solution, I wrote: "I have preferred the irregular line pattern for several reasons: it allows me to emphasize what seems to me to need emphasis, in a translation, without being bound to the futile attempt to duplicate French rhythm in English-and I am myself more comfortable with it, since it comports with my own practice as a poet. (Pope did Homer in heroic couplets for the same reason.) Nor do I think the reader of translations should delude himself that a translation in some way is the original. Only the original is the original: it becomes, again, a matter of decision-making. If poetry is what is lost in translation, what is it that one chooses to preserve?"12If syllabic prosody allows a writer to develop effects which cannot be literally transported into a language requiring a very different, stress-based prosody, is it not the obligation of the translator to evolve, as best he can, an equivalent of those effects in the stress prosody? Is it not his duty to avoid the obvious but ultimately self-defeating choice of what is in theory an equivalent-iambic pentameter in English for the alexandrine in French-and instead to construct a true, particularized equivalent which can carry into the host language as much as possible of the actual movement of

90

THE ART OF TRANSLATING POETRY

the original? My translation employs, exactly as does Baudelaire's poem, an emphatic caesura in lines 1, 2, 4, and 5 and a faint but palpable caesura in line 3. Further, in lines 1 and 5 the caesura occurs at the exact midpoint of a tetrameter line; in line 2 it may occur at the midpoint of a tetrameter line, if we scan the line as tetrameter; the same is true of line 3. Only line 4 follows the native English tradition-never worked so beautifully, or so delicately, as in the poems of Alexander Pope-of a caesura unequally placed. Sturm's translation, in its awkward pentameter, of course cannot match verse movement with the original-though it would seem to me that such matching is one of the major requirements of any translation. Additionally, his second line employs not one, not two, but three caesuras, which drastically alters verse movement, and his third and fourth lines cannot truly be said to have any caesura a t all. Richardson's version, too, employs pentameter, and so scanned her lines represent a consistent inequality of caesura placement. Her third line has no caesura. Once again, in short, the apparently obvious turns out to be false rather than true, and the literally nonrepresentative turns out to be a more truthful representation than the apparently " f a i t h f ~ l . " ~ ~ Moving from a stress-based prosody to either another stress-based prosody or to a syllabic prosody presents just as many problems. Here is one of Rilke's most transcendently glowing poems, Herbsttag ("Autumn Day"):

Herr: es k t Zeit. Der Sommer war sehr gross. Leg deinen Schutten auf die Sonnenuhren, und auf den Fluren lass die Winde 10s. Befiehl den letzten Fruchten voll zu sein; gieb ihen noch zwei siidlichere Tage. Drange sie zur Vollendung hin und jage die letzte Susse in den schweren Wein. Werjetzt kein Haus hut, baut sich keines mehr. Werjetzt allein k t , wird es lunge bleiben, wird wachen, lesen, lunge Briefe schreiben und wird in den Alleen hin und her unruhig wandern, wenn die Blatter treiben.

Prosody and Comparative Prosody

91

("Lord: It is time. Summer was very grand (wonderful, great). Put (set) your shadow on the sun-dials, and let the wind loose in the fields (meadows). Let it happen that the last fruits are full; give them two more southern days, push (urge, hurry) them to completion (perfection) and drive (pursue, chase) a final (ultimate) sweetness into heavy wine. He who now has no house, won't build one any more. He who is now alone, will stay so for a long time, will watch, read, write long letters and will wander up and down the avenues restless, as (when) leaves drift (float).") Rilke's language is stress-phonemic; the prosody of his language is stress-based; and this poem is in a clear iambic pentameter. (He does not always use traditional metrical forms.) But German prosody partakes of a very different literary tradition, in addition to being in a very different linguistic setting, from its English cousin. Iambic pentameter in German, accordingly, is not the same thing, and does not have either the same effect or the same significance, as iambic pentameter in English. Still, most translators assume a complete identity and use English iambic pentameter for their renderings: Lord, it is time. The summer was so great. Impose upon the sundials now your shadows and round the meads let the winds rotate. Command the last fruits to incarnadine; vouchsafe, to urge them on into completeness, yet two more south-like days; and that last sweetness, inveigle it into the heavy wine. He'll not build, now, who has no house awaiting. Who's now alone, for long will so remain:

92

THE ART OF TRANSLATING POETRY sit late, read, write long letters, and again return to restlessly perambulating the avenues of parks when leaves downrain.

(J.B. Leishman)14 Leishman is scrupulous about both rhyme and meter, though he is utterly unscrupulous about everything else. The translation is almost a parody, an example of how not to traduce a great poet. And much of that gross inadequacy stems, in fact, from his decision to (a)rhyme exactly as Rilke rhymed, and (b) to employ as an equivalent to the German iambic pentameter an English iambic pentameter. Forcing his rendering into these straitjackets obliges a translator of limited poetic ability to lose whatever chance he may have had of turning out a satisfactory re-creation of his original. Not that a degree of prosodic flexibility is any guarantee of success. Still, even for a translator of limited ability, prosodic (and other) flexibilities can help. Witness the not-quite-so-awful version of Kate Flores: Lord, it is time. Most great the summer was. Lay your shadow down upon the sundials now, And across the meadows set the wind loose. Make the last fruits mellow on the vine; Spare them but two further southern days, Speed them to fullness, and press The last sweetness through the heavy wine. Whoever does not have a house will build one now no more. Whoever dwells alone now will long remain alone, Will wakefully write long letters, read, And restless to and fro Will wander in the alleys, when the leaves are blown.16 Five of the first seven lines here are iambic pentameter: line 2 is a clear hexameter, and line 6 is trimeter. But none of the lines in the

Prosody and Comparative Prosody

93

last stanza is iambic pentameter. Flores is similarly more flexible about rhyming. Hers is hardly an ideal translation, if such a thing is possible for Rilke, probably as language-locked a poet as exists. But recognizing what she cannot do has clearly aided Flores to do a good deal, and a far better job than Leishman has done. The prosodic and other freedoms of my translation speak for themselves: Lord: it's time. Summer was wonderfully full. Throw your shadow over the sun-dials, Turn winds loose in the meadows. Let the last fruit be round and soft; Give it two more southern days, Push it to perfection, and pour A final sweetness into heavy wine. A wanderer, now, will stay a wanderer. Whoever's alone will stay alone, Will watch and read and write long letters And drift, restless, up and down the lanes As leaves blow back and forth.16 We are fortunate to have a translation of Rilke's poem into one of the syllabic languages to which a good deal of attention has been given, in these pages. The great Indonesian poet Chairil Anwar translated Herbsttag:

MUSIM GUGUR Tuhan: sampai waktu. Musim panas begitu megah Lindungkan bayanganmu pada jarum hari dun atas padang anginmu lepaslah. Titahkan buuhan penghabisan biar matang beri padunya duu hari selatan lagi Desakkan mereka kekemurnian dan buru jadi gula penghubisan &lam anngur yang garang.

94

THE ART OF TRANSLATING POETRY

Yang kini tiduk berumah, tidukkan menegak tiang Yang kini sendiri, 'kan lama tinggal sendiri, 'kan berjaga, membaca, menyumt panjang sekali, dun akan pulang balik melalu gang berjalan gelisah, jika duunun mengalun pergi.17 Since it is not literary quality which is here at issue, you will have to take my word for it that Anwar has done a magnificent job. He is indeed Rilke's equal as a poet: Indonesia is lucky that, although Anwar died before he was twenty-seven, he left as much finished work behind him as he did. Our interest is for the moment prosodic. And it is not difficult to see that Anwar is far too knowing a craftsman even to attempt any sort of prosodic equivalence. Here is a list of the number of syllables in each of the translation's twelve lines, grouped according to the stanzaic pattern of the poem:

Not only is there no observable pattern, and large variations in number of syllables per line, but Anwar's translation bears not the slightest resemblance to any standard Indonesian poetic line. The basic number of syllables per line in traditional Indonesian poetry is, as I have said, eight: not a line here has fewer than eleven syllables. The translation, again, is magnificent-and a good part of its magnificence comes from the translator's unmistakable awareness that it is impossible to re-create the prosody of a literary work composed in one language in another language.

Part Two Practice

On the grounds that nothing is less interesting or valuable than theorizing about principles without example, we asked each speaker to let his generalizations flower among his particulars or not at all. -William Arrowsmith and Roger Shattuck, preface to The Craft and Context of Translation

This page intentionally left blank.

The Subjective Element in Tkanslation

Literary criticism o h n pretends-and it is only that: a pretensethat it proceeds from entirely objective motives, that it operates entirely objectively, and that it arrives a t scrupulously objective conclusions. Writers too perpetrate the same usually wishful fraudon themselves as well as on their readers. Theodore Dreiser, notoriously self-deceptive, began his autobiographical book, A History of Myself, in order to prove "how little an individual had to do with what he became." He hoped that his autobiography "would be read as compelling fact and his family would be viewed with . . . scientific detachment." And he asserted flatly that "life and the individual should be judged on their chemical and physical merits."' Translators are no different. Translation too is a literary process, if in many ways a relatively minor one. And the decision to work in one literary genre rather than (or in addition to) another is rarely made with the sort of conscious forethought and careful, rational pre-planning that we sometimes attribute (again, inaccurately) to "scientific" decisions. As Norbert Wiener, a distinguished scientist, Portions of this chapter appeared, under the same title but in very different form, in Tmnslatwn Review 7 (1981):22-31.

98

THE ART OF TRANSLATING POETRY

has written, the "recognition [that there is] an element of incomplete determinism, almost an irrationality in the world, is in a certain way parallel to Freud's admission of a deep irrational component in human conduct and th~ught."~ Nobel Prizewinning scientist James D. Watson is equally clear: "science seldom proceeds in the straightforward logical manner imagined by outsiders." He also speaks of his search for the structure of DNA as "an adventure characterized both by youthful arrogance and by the belief that the truth, once found, would be simple as well as pretty." And he refers to "the incomplete and hurried way" in which all sorts of human decisions, scientific ones included, are frequently made, and to "a scientific world complicated by the contradictory pulls of ambition and the sense of fair play."3 Literary translation, like the primary literary arts it refleds, can be divided into genres: fiction, poetry, drama, essays and belles lettres, and more. But first we need to make one further distinction, unconnected either with genre or indeed with any of the considerations here discussed. This concerns the intensely practical difference between translators who are assigned work by publishers and translators who are free to choose the books or other material they will work with. Readers of translations often wonder why writers in other languages are so wordy. The chances are that they are reading the handiwork of a translator who was paid according to a word count and who received two cents a word for 53,000words as opposed to the 41,000 which the original contained. Probably publishers ought to pay flat fees based on the word count of the original, to discourage ad ding.^ Among the translators who are assigned their work, and who often draw all or a significant part of their income from that work, "far too few . . . are interested in translation for its own sake, much as they may protest to the contrary. Although a number of American writers have proved that it is possible to get rich writing, any translator who claims that it is possible to translate for more than a subsistence living is either dreaming or highly s~spect."~ It remains true, as one editor wearily puts it, that "difficulties notwithstanding,

The Subjective Element in Translation

99

translations are produced even in this country. Overcoming the occupational hazards of insomnia, hypochondria, and alcohol, translators do from time to time deliver a completed manuscript.'" But this side of the tracks is not the side I wish to discuss. I propose to limit myself to the work of literary translators who do not attempt to make their living by translation (though to be sure they have no objection to being paid for their labors), and who are therefore freeto about the same degree as writers of original material-to decide for themselves what, when, and even whether they will translate. Genre decisions tend to be made in similar, and deeply subjective, fashion both by writers of original material and by translators. It will perhaps be clearer to begin with writers of original material. And the clearest formulation of the importance of genre decisions may be this: in all of English and American literature, and in all the various genres of literature, there is I think only one major writer who has achieved approximately equal stature in more than one primary genre: Thomas Hardy. (T.S. Eliot is as great a critic as he is a poet, but criticism is not a primary genre.) There are many writers who have worked in more than one genre, but they have not achieved anything like equivalent stature in both. (Shakespeare's standing in both drama and poetry is misleading, since those were not in his time the separate genres they have since become.) This is both a deeply significant and a seldom discussed fact. It is unarguably true that writers cannot excel in both verse and prose: if they could, they would, and plainly they have not. But why? The career of Henrik Ibsen is instructive. We know that Ibsen began as a poet and wrote virtually all his early work in verse, but also that "after Peer Gynt Ibsen never wrote another play in verse.''' We also know that Ibsen's career as a dramatist took a completely different and far more powerful and influential turn once he began to write in prose. In a letter to Edmund Gosse, dated 15 January 1874, Ibsen offers a partial explanation: You say that the drama [Emperor and Galilean]ought to have been written in verse and that it would have gained by this. Here I must differ from you. As you must have observed, the play is conceived in the most realistic style. . . . If I had employed verse, I would have counteracted my own intention

100 THE ART OF TRANSLATING POETRY

and defeated my purpose. The many ordinary, insignificant characters whom I have intentionally introduced into the play would have become indistinct and indistinguishable from one another if I had allowed all of them to speak in the same meter. We are no longer living in the days of Shakespeare. . . . What I sought to depict were human beings, and therefore I would not let them talk the "language of the Gods" [i.e., p0etry1.~ Nine years later, writing to an actress, Lucie Wolff, Ibsen goes still further: The prologue would of course have to be in verse, since that is the established custom. But I will take no part in perpetuating this custom. Verse has been most injurious to the art of the drama. . . .It is improbable that verse will be employed to any extent worth mentioning in the drama of the immediate future since the aims of the dramatists of the future are almost certain to be incompatible with it. . . . During the last seven or eight years I have hardly written a single verse, devoting myself exclusively to the very much more difiult art of writing the straightforward, plain language spok.en in real life.9 (emphasis added) The coin glitters in exactly the same way when turned over. Sinclair Lewis, famous for his intensely realistic fiction, began his literary career as a passionate writer of what his first wife aptly called poetry of the "Tennyson and water" school. PRINCESS, Princess, silver maiden, Throw your casement open; seeOn the terrace I am singing; Come and take the road with me: All your gentlefolk are silken, And your knights are stately proudBut they do not know the hillside Where the daffodilies crowd,

The Subjective Element in Translation

101

Where the lark is mad with morning And your heart goes mad with May, Jesting with the wastrel breezes, Racing with the laughing day. "From the beginning," Grace Hegger Lewis has recorded, "he chose for himself the roles of Jacques the Jester and Francois the Troubadour who sang to the Lady Grace of the holidays and holydays we spent together. . . . In later years Lewis himself referred contemptuously to the medieval poems he had written and published while a t Yale as appalling nonsense." She concludes her discussion of this early poetic phase by noting that Lewis continued to write romantic, rather sloshy verse until "he met romance in person, and ceased writing about it except to that person. . . . Except for me he had closed the gate on his young man's garden of verse."1° Nor is Lewis a solitary example of the writer who starts in one genre, works exceedingly hard at it, then finds himself leaving it forever and rededicating himself to another genre. Aldous Huxley's first book was poetry, and in his early years as a writer he considered himself exclusively a poet. William Faulkner, too, began as a poet, and published a volume of exceedingly bad verse, before finding where his true talent lay. The relevance of these matters for translation seems clear, for in translating too one chooses both what one does and what one does not translate. It is every bit as basic a choice as the decision to translate in the first place. To the extent that the translator is a literary artist, his inner promptings are very much the same as those of the creator of original work. It is still somewhat astonishing to me, as a longtime writer of original prose fiction, to realize that in more than thirty years of translating I have confined myself virtually exclusively to poetry-twenty published volumes to date, just one of which incorporates a few pieces of prose fiction and a play. One of my critical books has a fair-sized appendix of translated literary criticism. I have no explanation, though I can say that I somehow do not feel entirely comfortable translating prose-it does not feel "natural," whatever exactly that means. I am and always have been completely comfortable translating poetry. Again, though I am myself a practicing poet, with six volumes of original poetry to date, I

102 THE ART OF TRANSLATING POETRY

am also a practicing writer of original prose fiction. About all I find myself able to say, consideringthe written record, is that apparently I translate poetry because I like to and do not translate prose because I do not like to. I do know that in order to translate poetry one has to be a poet. It surely helps to be a writer of prose fiction if one translates that genre. But it is not necessary. D. H. Lawrence clearly brought superb stylistic and technical resources to his collaborative translations of Russian writers and to his other prose translations, notably the fiction of Giovanni Verga, the Sicilian novelist. Francis Steegmuller, whose version of Flaubert's Madame Bovary seems to me one of the greatest translations done in our time, has written and published novels, though he is not the novelist Lawrence was. Steegmuller has actually-miraculously-evolved an English style for Madame Bovary which closely and beautifully matches Flaubert's style. But there have been myriad fine, even magnificent, translations of fiction produced by people with no particular record of achievement as writers of original fiction. I do not believe that the late John Butt, a distinguishedscholar of eighteenth-centuryBritish literature, ever wrote fiction. But his translation of Voltaire's Candide seems to me, after close and carefid comparison of the English version and the French original, far and away the best rendering ever produced. Translators' affinities and dis-afiities of course do not end with a choice of genre. Three additional decisions occur at once when one has decided to translate either poetry or prose: 1.what language? 2. what specific author (or authors)? 3. what literary period?

Let me begin with the choice of language, for that is the easiest of these three issues to discuss. Apart from collaborative translations from languages which one knows inadequately, or not at all (a subject discussed in the next chapter),a translator starts by knowing, or occasionally by learning, a language other than his own. One need not be bi- or tri- or multilingual in order to translate and translate well. One need not command a speaking knowledge of the language being translated

The Subjective Element in Translation

103

from (and indeed there are some languages, like Latin and classical Greek, where a true speaking knowledge is impossible). By definition, however, one cannot be a monolingual translator. A translator's choice of a language to work from is-or in my view should be-every bit as subjective as the choice of literary genre. It is a very rare bird indeed, even in this aviary of rare birds, who can display a universal sprachgefihl, who can equally well sense and then re-express the subtleties of all languages. That great tiller of oriental fields, Arthur Waley, made the mistake of translating one of Franz Kafka's short pieces, Der Kubelreiter ("The Bucket Rider7'; Waley's aberrant translation of the title is "The Coal-Scuttle"). The tale begins, in German:

Verbraucht alle Kohle; leer der Kubel; sinnlos die Schaufel; Kalte atmend der Orfen; das Zimmer vollgeblasen von Frost; vor dem FensterBaume starr in Reif; der Himmel, ein silberner Schild gegen den, der von ihm Hilfe will. Willa and Edwin Muir beautifully (and accurately) capture the staccato, almost gasping tone of the German: Coal all spent; the bucket empty; the shovel useless; the stove breathing out cold; the room freezing; the leaves outside the window rigid, covered with rime; the sky a silver shield against any one who looks for help from it." But here is what a great translator from Chinese and Japanese, but not, alas, from German, has done to Kafka's fiercely aggressive prose: Coal? None. Empty the scuttle, senseless the shovel, icy the breath of the stove, the room glazed white; stiff with frost the trees a t the window. Heaven a silver shield blocking the path of prayer.12 It would be easy to mock this inept rendering-one would, for example, have thought that most shovels were by nature senselessbut not productive. Waley plainly had no more feeling for Kafka7s German than did that senseless shovel. Even a great translator, one

104

THE ART OF TRANSLATING POETRY

of the very greatest, is unable to handle every language in the world with the same sureness, the same authority. Waley's failure should be a lesson to us all. Choosing what author one will translate is not anything like so easy as choosing a language to work from. One either knows and has a feeling for a language, or one does not. But a specific author, or a specific work in those cases where we do not know who the author was, can present more serious problems. The basic rule must always be: it is impossible for anyone to translate everything. Affection and respect for an author or for a great anonymous work are a good place to start, but they are not enough. There must be some special affinity, some special relationship, that subjectively, powerfully links translator and original author. If, for example, we read a version of a poem by the indisputably greatest mystical lyric poet in all of Persian literature, Maulana Jalal-uddin Rumi (1207-1273), usually known simply as Rumi, we expect to find what standard reference sources tell us is "the most fertile and, at the same time, the most fascinating mystic poet writing in Persian. . . . His lyric poetry is, for the largest part, extremely rich and beautiful, full of images and music."13 If, however, what we read is the following translation, by the distinguished scholar (and non-poet) A. J. Arberry, we may wonder why either he or we ought to bother: Time bringeth swift to end The hour men keep; Death's wolf is nigh to rend These silly sheep. See, how in pride they go With lifted head, Till Fate with a sudden blow Smiteth them dead.14 This is as clearly poetaster blather as anything ever penned. There is no affinity whatever between Arberry and Rumi. But if we read instead the translation by John Moyne and Coleman Barks, we feel (especially in the second half of the poem) that we are to some extent in the presence of the Persian poet; we feel somethingof his grandeur,

The Subjective Element in Translation

105

something of his spirit; and we know that the translators have felt these things too and tried to transmit them to us: This bleating eventually stops. The wolf appears. We run off in different directions, with always some thought of how lucky we are. But nothing floats for long. Death floods in the mouth and the ear. Every head goes under and away.16 Similarly, if we read Arberry's version of a poem also translated by Robert Bly, one of the most accomplished translators alive, we again see the enormous difference between poetic a f h i t y and mere scholarly knowledge: At the night prayer, when the sun declines to sinking, this way of the senses is closed and the way to the Unseen is opened. The angel of sleep then drives forward the spirits, even as the shepherd who watches over his flock. To the placeless, towards the spiritual meadows, what cities and what gardens he there displays to them! The spirit beholds a thousand marvellous forms and shapes, when sleep excises from it the image of the world. You might say that the spirit was always a dweller there, it remembers not this world, and its weariness does not increase. Its heart so escapes from the loan and burden for which it trembled here, that no care for it gnaws at it any more. (A. J. Arberry)16 Night and Sleep At the time of night-prayer, as the sun slides down, the route the senses walk on closes, the route to the invisible opens. The angel of sleep then gathers and drives along the spirits; just as the mountain keeper gathers his sheep on a slope. And what amazing sights he offers to the descending sheep! Cities with sparkling streets, hyacinth gardens, emerald pastures!

106 THE ART OF TRANSLATING POETRY

The spirit sees astounding beings, turtles turned to men, men turned to angels, when sleep erases the banal.

I think one would say the spirit goes back to its old home; it no longer remembers where it lives, and loses its fatigue. It carries around in life so many griefs and loads and trembles under their weight; they are gone, it is all well. (Robert Bly)17 The contrast becomes all the more remarkable when we read that "Robert Bly's versions [of Rumil come from his work with the translations of the late British professor of Persian, A. J. Arberry "18 That is, since Bly does not know Persian, he has reworked Arberry's translation. Everything he has found has been unearthed through the medium of Arberry's inert verbiage. There is no clearer or stronger proof of the necessity for affinity. Choice of a literary period, of a particular time in the life of a literature, is sometimes hard to distinguish from either the choice of an author or of a particular book or even from the choice of a particular language. But one might well argue that what made Kafka's story an inappropriate choice for Arthur Waley to translate was neither the fact that it was German, rather than Chinese or Japanese, nor that it was Franz Kafka rather than Li Bau (Li Po) or Du Fu (Tu Fu), but simply that Kafka represents in one of its fiercest and most urgent forms the hard, clangorous work of the twentieth century. Waley, after all, longed to be a writer, not a translator-nor is that an unusual situation. Many of the best translators are indeed writers manque. But as a creator of original work Waley failed almost totally. And one might argue that one reason for his failure-only one, but an important one-was that he had remarkably little sympathy for the century in which we live. He led a notoriously tormented and convoluted life: that sort of existence frequently spawns writers like Kafka, which may perhaps account for Waley's attempt at translation. But Waley could not find open channels for expression in the modes and manners of his own timeor in his own language. He turned for solace to the writers of another culture, and we are all grateful that he did, for he has given us countless wonderful versions of material most people would have no

The Subjective Element in Translation

107

knowledge of without his mediating services. But, once again, no one can possibly translate everything-and Waley was sufficiently out of tune with our time so that working with material like Kafka's was completely inappropriate for him. But the clearest example known to me of a translator choosing to translate from the wrong period is the late John Ciardi's fluent and sharply skewed Dante. Ciardi's grasp of Italian seems to me unarguable. He was a h e poet in his own right, and his respect, admiration, and love for Dante is I think obvious. He also has the poetic skill to handle, and handle well, much of the Divinu Commediu's swift-paced forward movement: Ciardi is a first-rate narrative poet, and therefore a first-rate translator of narrative verse. But for all these admirable qualities, Ciardi remains, unfortunately and to my mind fatally, unable fully either to understand or to identify or even to sympathize with the medieval aspects of Dante's verse. Ciardi's subjective affinity is not in truth with Dante as Dante actually exists, but with Dante as a kind of combination of John Donne and John Milton, Dante as a passionate, dramatic voyager into unknown realms. I have spoken, in chapter 2, of some of the problems created by Dante's limpid, pellucid diction. Ciardi's difficulties go still further. Here, once again, are the opening lines of the Inferno:

Nel mezzo del cammin di nostra vita mi retrovai per unu selva oscura, che la diritta via e m smurrita. Ahi quanto a dir qua1 e m 2 cosa dum esta selva selvaggia e aspm e forte che nel pensier rinova la paura! Tanf L a m r a che poco 2 piti morte; mu per trattur del ben ch'i vi trovai, dird de l'altre cose ch'i v'ho scorte. ("In the middle of the journey of our life I found (came to) myself in a dark wood for (where) the straight road was lost. Ah, how hard a thing it is to tell (speak of) that wood, wild (savage), dense (rugged), harsh, that thinking of it renews my fear!

108 THE ART OF TRANSLATING POETRY

So bitter is it that death isn't much more bitter; but to deal with (tell) the good things I found there, I will speak of (tell) the other things that I saw (noted) there.") It is worth emphasizing that all beginnings are crucial, whether for writers of original work or for translators. Capturing the exact tone, the precise mood, in language and form ofjust the right weight, with just the right pace, is critical. Who can forget the image of that masterful and vastly experienced novelist, Leo Tolstoy, struggling and sweating to get the novel that eventually became Anna Karenina off the ground, flailing and half-desperately fighting to get himself started properly? It was a long time before the novel opened to himthough he was already the creator of that enormous, complex masterpiece War and Peace. I have recorded elsewhere how long and wearily I myself struggled with the opening lines of B e o w ~ l fThe .~~ scop (Old English "bard") seems to have had the right tone easily at his fingertips-unless of course he too struggled and sweated, though we have no record of anything of the sort and no way of knowing. But I unquestionably found the opening lines immensely difficult and crucially important, for how could I go on unless I knew how I was to go at all? The beginning of Dante's great poem is similarly crucial and equally difficult for the translator. Dante's verse has that translucent medieval capability: he can seem to be saying very little and saying it with great quietness, when in fad he is sasng a very great deal and saying it with crystalline force and power. Critics have often noted Dante's sparseness of metaphor. Mark Musa, himself a translator of the Divina Cornmedia,has neatly pinpointed the danger for the translator in this metaphorical sparseness: "Most of all, the diction should be simple when Dante's is. And this is where the translators have sinned the most . . . [, creating] original striking rhetorical or imagistic effects where Dante has intended none.'=O Ciardi's sins are striking and obvious: Midway in our life's journey, I went astray from the straight road and woke to find myself alone in a dark wood. How shall I say

The Subjective Element in Translation

109

what wood that was! I never saw so drear, so rank, so arduous a wilderness! Its very memory gives a shape to fear. Death could scarcely be more bitter than that place! But since it came to good, I will recount all that I found revealed there by God's grace.21

I have put in italics those phrases which simply are not Dante, but are Ciardi's inventions. Dante does not say "I went astray," but mi retrovai per una selva oscum, 1 che la diritta via era smarrita ("I found myself in a dark wood, for the straight way was lost"). Dante does not say that he was "alone," but only that he was there. He tells us that the wood was dark, and wild, and rugged, and harsh, and he stresses these unpleasant qualities by exclaimingAhi quanto a dir qua1 era b cosa dura ("Ah how hard it is to tell"). But what Ciardi has him exclaim is far more like late Renaissance than like medieval poetry: "I never saw so drear, / so rank, so arduous a wilderness!" Ciardi's adjectives can justly be termed Shakespearian, even Miltonic. The very form of the exclamation, as he has phrased it, is both extravagant and false to the original, for Dante, again, does not say that he "never saw" anything so this and so that, but simply that it was wild, rugged, and harsh, as well as dark. Ciardi is even more extravagant and false to the original in translating Dante's che nel pensier rinova la paura! ("The thought of it renews my fear!"). "Its very memory gives a shape to fear!" writes Ciardi, and the contained (and restrained) power is lost, scattered to the winds. Dante wants the power of what he is describing to assert itself because it is horrible, and because he and his readers are in perfect agreement on the subject. But Ciardi, who has no patience with either medieval clarity or with his readers, cannot allow things simply to assert themselves. As a result, in Ciardi's wonderfully readable but strongly unreliable Dante almost everything is larger, bolder, brighter, and also louder than anything in the fourteenthcentury original. Medieval man did not think he had to invoke God at every step, as Ciardi makes Dante do in the ninth line of his translation: "all that I found revealed there by God's grace." Dante does not speak of revelation; he does not speak of God's grace. He

110 THE ART OF TRANSLATING POETRY

says, plainly, dirt3 de l'altre cose ch'i v'ho scorte ("I will speak of the other things that I saw there"). In short, medieval man felt God, saw God everywhere, and had no doubt that He was always and inevitably and justly in the world, seeing and caring for His creatures. For all his energy, for all his skill, for all his love of Dante, Ciardi was wrong to translate the Divina Commedia. Hundreds of thousands of unsuspecting students have read in his version a Dante who never existed. Ciardi's act of love is unfortunately much more like an act of betrayal. It might be argued that throughout this discussion I have been assuming and yet never stating a basic premise, namely, the existence of a particular kind of audience and therefore of a particular kind of translation suited to the needs of that audience. The argument would be valid: the third and last of the basic choices I want to discuss here does indeed deal with the question of audience and the appropriateness of a translation for that audience. I have to this point been quietly assuming a general, nonscholarly audience and translations which can be read by that audience. But it is time to admit that there are other audiences and other sorts of translation. It is time to admit, too, that these other sorts of translation are widely practiced and always have been, and that in some people's judgment they are more appropriate than the literary translation which I advocate (and try to create). I classify translations into four broad types and identify each type with a different audience: 1. formal translation, aimed primarily at scholars and those taught by scholars, largely for scholarly rather than literary purposes;

2. interpretive translation, aimed primarily at a general audience which reads for literary reasons; 3. expansive (or "free") translation, aimed not simply at those who read for literary reasons but at those who usually prefer to read something, anything, new rather than anything old; 4. imitative translation, which in plain truth I think just barely translation at all; it is aimed at an audience which wants the

The Subjective Element in Translation

111

work of the particular translator rather than the work of the original poet; Robert Lowell's accurately titled volume Imitations is the model of this type of translation in our time. Before I discuss these four classifications, in the order just given (though it will be impossible to discuss each category in entire isolation from others: some overlapping is inevitable), it needs to be stated quite clearly that most translators, as well as most of those who write about translations or translation itself, are unlikely to approve of my classifications. Very few of those who write about translation, in particular, are prepared to admit that no single type of translation is "best" for each and every audience. In part, I believe, this is because these commentators also fail to understand, or at times even to be aware of the existence of, the linguistic and cultural matters discussed in the first four chapters of this book. "And yet," as Sidney Alexander says, "we must translate. Else we are locked in our separate boxes. We must communicate across cultures; we must devise a simulacrum that seems-at one and the same timeto reproduce the original work and yet exist as a work of art in itself."22

Formal Panslation Formal translation, as I understand it, is not primarily concerned with literature or literary values. Rather, it deals with scholarshipexamining, classifying, categorizing, even comparing. The formal translator is apt to believe that what he is fond of calling "literal" translation (the word "faithful" is also often used) is not only possible but highly desirable. What the formal translator is after is what he calls "exactness." He is apt to be much concerned with notions like "fidelity" and with the "exact" reproduction of literary form, prosody, and so on. Though he often thinks that his goal is a kind of mirror image of the original, a precise reproduction achieved by a process of bodily transference, in fact he is most likely to give his readers

112 THE ART OF TRANSLATING POETRY

the ideas, the social and philosophical orientations, the information, and the historical data contained in his original. He struggles, often honorably and sometimes very successfully, to avoid any intrusion of an alien presence, any sort of interference between the original and his rendering thereof. Frequently (though not always or even necessarily) he makes no objections to making available, through translation, something of the literary experience of his original. But that is almost never his goal; he will regularly and even combatively sacrifice literary to scholarly values. So much for a generalized description. Let us consider some actual cases. Alan R. Press, editor and translator of a bilingual edition, Anthology of Troubadour Lyric Poetry, is without question a formal translator. "Elegance and grace have been-not always without remorse-sacrificed to literal accuracy," he explains in his introductory essay. He is sufficiently aware of linguistic reality-not always true of the formal translator, though most would be horrified to be told it-to concede also that "on occasion, of course, the demands of clarity and of simple English grammar have imposed some modifications of an absolutely literal rendering . . . and even the interpolation of words corresponding to no form in the original text, though such interpolationscan at least be-and of course have beenadmitted by the use of par en these^."^^ Press's appreciations of the poets he has collected and translated are equally formal. His concerns are plainly technical, and often distinctly abstract. Of Arnaut Daniel, for example, much beloved of Ezra Pound, Press writes that "such fragmentary indications . . . ,coupled with the almost complete absence from his work of any specifically professional themes, motifs, or attitudes, suggest that he may not have been the 'minstrel' which his mediaeval biographer reports him to be." There is a certain sort of literary analysis: "One is aware above all of an exceptionallylively imagination which endows his work with a wealth and variety of imagery drawn not only from traditional literary sources but also from the everyday, practical realities of feudal existence." The "literary" side of this account is generalized and vague; the material about sources and social matters is clearly what Press truly relishes. When he mentions Daniel's "constant striving after richness and variety of rime and other acoustic effects," Press is pretty clearly saying what he thinks needs to be said. When, however, he quickly

The Subjective Element in Translation

113

goes on to emphasize that "brilliance of imagery and versification . . . [are] integrated into a coherent unity of structure in which art and inspiration are inseparable,'% Press is obviously on home ground. Questions of "art and inspiration" are far more meaningful to him than is any mere literary necessity for there to be a subjective element in good translation. He is not concerned with good translation, but only with "accurate" translation. In a word, Press is a scholar, translating for others of his kind-or who are to be taught by his kind. Here is how, operating from that perspective, Press translates the first stanza of Arnaut Daniel's poem Lo ferm voler qu'el cor m'intra ("The firm desire which enters my heart"): The firm desire which in my heart enters cannot be torn from me by tale-teller's beak or nail-albeit he arms himself to speak evilly; and since I dare not beat him with branch or rod, then at least by stealth, there where 1'11 have no (guardian) uncle, will I enjoy (love's) joy, in bower or bedroom.26 Press has, true to his lights, translated Daniel's glowing love poem into dogged prose-though not quite into English prose, for no native speaker of English would ever use, in either verse or prose, such syntactically anomalous constructions as "the firm desire which in my heart enters.'' But neither English nor poetry is what he is concerned with, as he tells us himself, in discussing this poem. "'I'he real problem is to appreciate how far the ironic effect. . . was intended to destroy, to dominate, or simply to contend with the mood of emotional intensity created, in particular, by the arma and cambra motifs."26That truly is the real problem for Press, and for those who think as he does. And his translation, relatively unreadable though it may be, manifestly stems from those beliefs and concerns (though it is odd even to try to discuss "ironic effect" on the basis of a translation so patently defiant of all literary values, irony included). But even the formal translator does not have to so drastically flatten out a great poet. The plain, unmistakably formal version of this same poem by Frederick Goldin proves that literature does not have to be ignored by the formal translator. It is significant both that Goldin makes no generalized statement about his goals in trans-

114 THE ART OF TRANSLATING POETRY

lating and also that his analyses of the poets are a great deal more literary than are those by Press. Of Arnaut Daniel, for example, he writes that "with a few exceptions he composed in a style characterized by brilliant ornamentation, particularly by elaborate rhyming and by the use of rare words chosen for their sound effects, the sense falling where it may: the vida ["life"] says his songs were hard to understand and learn." And of the poem Lo ferm voler he notes, with an admirable dry humor, that "this is a sestina, a form that Arnaut seems to have invented. It consists of six strophes of six lines, the same rhymes appearing in a different order in each strophe. Lines that do not make too much sense in the translation are not clear in the original either.'"' It is apparent from the translations that Goldin is after as close a formal transference as he can achieve. It is also apparent that he is a good deal more sensitive to the literary values of those poems, and at the same time to the literary needs of his audience. Here is the same opening stanza that Press translated: The firm desire that enters my heart no beak can tear out, no nail of the slanderer, who speaks his dirt and loses his soul. And since I dare not beat him with branch or rod, then in some secret place, at least, where I'll have no uncle*, I'll have my joy of joy, in a garden or a chamber.28 It would be hard to deny that, without striving to translate fully, at least in the sense I have argued for, Goldin has done infinitely better than has Press. What has he done differently? First of all, he has cheerfully given in to the inevitable and conceded that modern English requires a different word order (syntax)from that of ancient Provengal. Recall how Press translated Lo f e n voler q'el cor m'intra: "The firm desire which in my heart enters." Press tries to preserve P r o v e n ~syntax l in modem English-more or less. If he were truly consistent, and truly "literal," he would have translated "The firm desire which it heart to me enters." But even the literalist balks at *That is, to keep a watch on him [Goldin's note].

The Subjective Element in Translation

115

such monstrosities. Has Goldin given up his formal goals by writing "The firm desire that enters my heart'? I think not. And from my perspective, just as plainly, Goldin has acquired so much higher a level of sheer readability that he has made Daniel's poem vastly more accessible. Indeed, not only is Goldin's formal translation better and more readable than that by Press, it is in fact more accurate as well. Daniel's third line reads, in part, qui pert per ma1 dir s'arma ("who loses, by evil speaking, his soul"). Goldin accurately translates this as "who speaks his dirt and loses his soul." Press mangles the phrase, inaccurately, into the unreadable and unfaithful "albeit he arms himself to speak evilly." Formal translation does not require bad translation, even though bad translation is what it usually produces. There are, alas, many, many more like Press than there are like Goldin.

Interpretive 'lkanslation I do not think we can do better, in moving to this distinctly more literary style of translation, than to set out the version of the same Arnaut Daniel poem made by the father of modern poetic translation, Ezra Pound: Firm desire that doth enter My heart will not be hid by bolts nor nailing Nor slanderers who loose their arms by lying And dare not fight with even twigs and switches. Yes, by some jest, there where no uncle enters I'll have my joy in garden or chamber.29 Pound translated only twelve of the poem's thirty-nine lines; he also mistranslated a fair number of the lines he did translate. Hugh Kenner, fiercely protective of Pound, argues that "since he doesn't translate the words, he may deviate from the words, if the words blur or slide, or if his own language fails him. . . . [But] if he doesn't

116 THE ART OF TRANSLATING POETRY

translate the words, the translator remains faithful to the original poet's sequence of images, to his rhythms or the effect produced by his rhythms, and to his tone."g0I have put the matter rather differently: Ezra Pound the translator of poetry into poetry is by no means a consistent, unitary figure. I have said that he is essentially an appropriative translator: that is, he goes to an original because it has something he wants. There can be no question that he admires, and frequently passionately admires, the beauties of the texts he chooses for translation; there has seldom been anyone, poet or critic or scholar, whose relish of poetic beauty remained all his life so keen. . . . [But] appropriation is basically as indifferent to consistency as it is, in the final analysis, to the integrity of the text being translated. . . . Things in the original which may get in the translator's way, "obscurities [in Pound's words] not inherent in the matter, obscurities not due to the thing but to the wording, are a botch, and are not worth preserving in a translation. In plainer terms, anything that troubles you can be omitted, as long as you do not think it central; the mere words of the original are in no way binding."s1 But it would be too facile, and deeply imperceptive, to say that in translating Lo ferm voler Pound's interpretation is simply, or even largely, linguistic error. To put it differently, and without attempting to excuse the errors, which are as unmistakable as they are horrendous: not all linguistic errors are equal to all other linguistic errors. Some of Pound's errors are obviously due to lack of knowledge. However, they are not simply due to lack of knowledge. Nor are his errors mere accidents, scattered mistakes of the sort made by amateurs like the hapless translator (male) who came on the word soutien gorge and guessed, since gorge by itself means "throat" and soutien means "supporting," that put together in a compound noun the word referred to a scarf of some sort, worn around the throat. Alas, soutien gorge means '%rassiereW:the example is one put forward, long ago, by Roger Shattuck, I do not now recall where. That

The Subjective Element in Translation

117

is an amateur error, let me explain, because the translator has been wildly, purposelessly in error. Not only does he not truly know what his original means, but he does not care what effect his errors, if he makes any, may produce on his readers. Pound, on the other hand, though he is obviously not in full control of the original, is very firmly in control of the translation as an independent poem. And he knows more than enough about the poetry of the period, as well as the poetry of Arnaut Daniel, and cares enough about that poetry (that is, has precisely the sort of subjective linkage with that poetry that I have been stressing so determinedly), that when he makes a linguistic error he partially redeems it by creating somethingthat might very well have existed in the original. Knowing what other poets of that time and place might and often did put into their work, Pound is able-even in error-to interpret, and to interpret from knowledge as well as from affection. What the reader receives, accordingly, may be skewed in detail but remains true in general import. Kenner's defense, in short, is perhaps too all-forgiving. But it is not silly, and it is not wrong. The best interpretive translators try for exactly Pound's goal, though to be sure they also try, and try harder than Pound usually did, not to commit errors. One of the most knowledgeable orientalists I know, Professor Donald Keene, once assured me in a private conversation that Pound's interpretive rendering of the Chinese poem "The River Merchant's Wife: A Letter" was both inaccurate in detail and at the same time the very best translation of that poem he had ever seen. For the interpretive translator necessarily recognizes, more clearly than translators of any of the four types I have enumerated, the basic linguistic and cultural problems faced, willy-nilly, by any and every translator.

When we talk about literature and translation of literature, we're not talking about just anybody's experience: we're talking about a very high-level synthesis by extremely intelligent people. This synthesis involves a central fact of the art world; that is, a man who is a great painter is as intelligent as, but differently intelligent from, a great physicist, and a great physicist is as intelligent as, but differently intelligent from,

118 THE ART OF TRANSLATING POETRY

a great writer. To say that Einstein is more intelligent than Shakespeare is nonsense. To say that Shakespeare is more intelligent than Einstein is nonsense.s2 The interpretive translator frequently says, forthrightly and in my view correctly, that he is translating for a literary audience and for people who do not have access to the original. That frank admission simply emphasizes what should not need emphasis or even restatement, namely, that no translation is, was, or ever will be the original which it translates. The interpretive translator faces this uncrossable gulf and tries to give his reader as much of the original as he can. He does not delude himself into believing that he possibly can give true access to the original: only knowledge of the original's language, and culture, and literary tradition can in truth accomplish that. The formal translator, on the other hand, frequently does so delude himself (and his readers), assuring them that what he offers is in fact genuine access to the original (or at least access to what the formal translator considers of importance in the original). Rowland L. Collins, introducing the translation of Beowulfby Lucien Dean Pearson, emphatically a formal rendering, claims that "the Pearson version not only faithfully represents the Old English text but it recalls the vigor of the original through a vigor of its own. The antique vocabulary is suggested, as are the older poetic techniques. Modern English, however, is the vehicle, and it remains unviolated. The translation is an excellent one because it is at once accurate and natural, at once ancient and m ~ d e r n . "The ~ words are eloquent; the translation, alas, does not in any way justify them. Two prose paragraphs should be a sufficient demonstration: Hear! We have been told the glory, in days now gone, of Spear-Danes, people-kings, and how the nobles did their feats of arms. Scyld Scefing [footnote omitted] o h n took the mead-benches [footnote omitted] from his banded foes and many a clan; he cowed the Eruli [footnote omittedl. First found in want, he lived to see amends., he grew beneath the clouds, and throve in honor, till each neighbor on the whale-road must obey and yield him tribute; that was a good king!g"

The Subjective Element in Translation

119

The 'banded foes" sound like birds with naturalists' tracers tied to their legs. "He lived to see amends" is neither natural nor English. The presence of a multiplicity of footnotes is itself proof that the translation is not considered capable of standing on its own feetand if it is not so capable, why read it in the first place? The original is more than capable of standing on its own feet; it has been doing so for over a thousand years, and shows no signs of faltering. Stanley B. Greenfield, a superb scholar, is another self-deluded formal translator (though he turns Beowulf into a kind of syllabic verse!). "I wanted my translation to be not only faithful to the original but . . . I wanted it to 'flow,' to be easy to read, with the narrative movement of a modern prose story; yet to suggest the rhythmic cadences, including variation, of the Old English poem."gs Indeed, we have heard of the Spear-Danes' [footnote omittedl glory, and their kings', in days gone by, how princes displayed their courage then. Often Scyld Scefing shattered the hosts [footnote omittedl, unsettled many a nation's mead-hall, terrorized tribes, since first he was found abandoned; comfort and abundance later came his way, and worldly fame, until neighboring nations, near or far over whale-big seas, obeyed him, gave tribute: a good king in deed!%

If the footnotes are disturbing, as before, the "unsettled . . . mead hall" is all too redolent of barroom scenes that have no place in Beowulf. The pregnant seas, "whale-big," are, I'm afraid, ludicrous, just as the academicpun at the end of this passage-"in deed" instead of "indeed": one can almost hear the scholarly chortling that must have accompanied this-is totally out of place. Parts of this formal version are in truth readable, but parts are distinctly not, arid the claims made by the translator are patently untrue. Not that all formal translators believe so naively in their own virtues and achievements. Here are the first three lines of Beowulf

120 THE ART OF TRANSLATING POETRY

in a once moderately popular formal translation by Gordon Hall Gerould, a Princeton professor: We have heard of the Spear-Danes in the days of yore, what glory came to the kings of the people, the deeds of might of the doughty leader^.^' After a brief discussion of the meter and language of the original, Gerould confesses that "in the translation that follows, an effort has been made to reproduce these characteristic features of Old English poetry, though it is difficult to do this successfullybecause of changes in our language-notably the loss of inflectional endings. The translator can only hope to give the reader some inkling of the imaginative " ~ ~in his introductory essay to the prose power . . . [oflB e o w ~ l f .And translation by John R. Clark Hall, J. R. R. Tolkien flatly disclaims any magical superiorities for the rendering: "Beowulf is not merely in verse, it is a great poem; and the plain fact that no attempt can be made to represent its metre, while little of its other specially poetic qualities can be caught in such a medium [as prose], should be enough to show that [this translation] . . . is not offered as a means of judging the original, or as a substitute for reading the [Old English] poem itself. The proper purpose of a prose translation is to provide an aid to study."a9That is, this sort of formal translation is no more than a crib. Tolkien is clearly right; Collins and Greenfield are just as clearly wrong. One formal translator, E. Talbot Donaldson, not only turns Beowulf into prose but argues, remarkably, that this is the reader's only true pathway to the original. "I have eschewed verse . . . , for I am persuaded that only a prose translation, made with no other end in mind than fidelity to the original, can bring out the distinctive qualities of the work. To make it a modern poem is, inevitably, to make it a different poem. . . . Rather than try to create a new and lesser poem for the reader, it seemsbetter to offer him in prose the literal materials h m which he can re-create the poem" (emphasis addedhqO Professor Donaldson utterly ignores the operative facts: (1)any translation, whether in verse or prose, is not the original which it translates, and (2) the only method by which a reader can truly "re-create the poem" for himself is to learn to read the original. Professor Tolkien has the clarity of mind to identify the formal prose version he writes about

The Subjective Element in Translation 121 as a crib. Donaldson's is a modest, plain rendering; it is I believe the best of the prose translations now available. But in this version Beowulfappears neither as a poem nor as any sort of material from which anything resembling a poem might be constructed-and constructed, it should be emphasized, by readers neither learned in the language nor in the culture which produced Beowulf. Again, interpretive translators almost never assert that their versions manage, in some magical fashion, to reduplicate the original. In the introductory essay to my own Beowulf, I explain that "the translator's only hope is to re-create something roughly equivalent in the new language, something that is itself good poetry and that at the same time carries a reasonable measure of the force and flavor of the ~riginal."~~ The translation has these and only these goals: Hear me! We've heard of Danish heroes, Ancient kings and the glory they Cut for themselves, swinging mighty swords! How Shild made slaves of soldiers from every Land, crowds of captives he'd beaten Into terror; he'd traveled to Denmark alone, An abandoned child, but changed his own fate, Lived to be rich and much honored. He ruled Lands on all sides: wherever the sea Would take them his soldiers sailed, returned With tribute and obedience. There was a brave King!42

Expansive (Free) Tkanslation Not many translators practice this approach. Perhaps the clearest example, and one that can nicely stand for the positive side of this approach, is Christopher Logue's version of the "Patrocleia" section of Homer's Iliad. Exactly how expansive this version is can be seen by setting out parallel passages, first in the more or less standard translation of Richmond Lattimore, and then in Logue's translation. (Note that the italics and the ellipses are as given in Logue.) The speaker is Patroklos; he is addressing Achilleus:

122 THE ART OF TRANSLATING POETRY

But you, Achilleus; who can do anything with you? May no such anger take me as that you cherish! Cursed courage. What other man born hereafter shall be advantaged unless you beat aside from the Argives this shameful destruction? Pitiless: the rider Peleus was never your father nor Thetis your mother, but it was the grey sea that bore YOU

and the towering rocks, so sheer the heart in you is turned from us. But if you are drawing back from some prophecy known in your own heart and by Zeus' will your honoured mother has told you of something, then send me out at least, let the rest of the Myrmidon people follow me, and I may be a light given to the Danaans. Give me your armour to wear on my shoulders into the fighting; so perhaps the Trojans might think I am you, and give way from their attack, and the fighting sons of the Achaians get wind again after hard work. There is little breathing space in the fighting. We unwearied might with a mere cry pile men wearied back upon their city, and away from the ships and the shelters. So he spoke supplicating in his great innocence; this was his own death and evil destruction he was entreating. (Richmond L a t t i m ~ r e ) ~ ~ Is there no end to your obdurate grudge, Achilles? No. Don't shrug me off. Remember who is asking. Not Agamemnon. Not the smart Ithacan. No one save me-

The Subjective Element in Translation

123

And God forbid I share the niceness of a man who, When his friends go down, sits tight And claims their punishment as just amends For the wrongs they did him. They are dying, Achilles. Dying like flies. If you can't think of them, think for a moment Of those who will come after them, what they will say: "Achilles The Grudgebearer"-can't you hear it?-or, "Achilles The Strong"-and just as well for that, Because his sense of wrong was very heavy-or, "Somepeople say Lord Peleus and Thetis, Lovely among the loveliest of women, Were his folk. But if you want the truth, His father was a long bleak rock Thut after centuries was moved By the even bleaker, always disconsolate Pestering of the Sea, until she had a son, Achilles." Why make me talk to you this way? If it is true that God or your sainted Mother Whisper persuasive justifications for desertion in your heart, It's also true they do not mention me. Let me go out and help the Greeks, Achilles? Let me take our troops? Half of them, then!And let me wear your weapons . . . Man, it will be enough! Me, dressed a3 you, leading the Myrmidons . . . The sight of us will make Troy hesitate and say "It's him," a second look will check them, turn them, Give the Greeks a rest (although war has no rest) And, once they have turned, nothing will stop us till They squat inside their walls.

In this way, in words Something like those written above, Patroclus begged for death. (Christopher LogueY

124

THE ART OF TRANSLATING POETRY

Logue compresses at times. But mostly he expands, developing what Donald Carne-Ross (who supplied the crib from which Logue worked: Logue does not know Greek) calls, citing E. M. W. Tillyard, "the main undulations of his original." Carne-Ross goes on: ''While Logue's version is in no real sense 'the Iliud in modern dress,' it is written in the belief that no sort of fancy translationese should be allowed to m d e the impact of the original."46 Carne-Ross's argument is considerably less persuasive, as it is patently less consistent, than Logue's swaggering, vital re-creation of the Homeric moment. This is, as I have indicated, expansive translation at its best. Frederick R. Rebsamen's rendering (I am tempted to say "rending") of Beowulf, published under the title Beowulf Is My Name, represents expansive translation a t its silliest and worst. Logue is a poet; Rebsamen is a scholar-and like most scholars who tackle such tasks, an addled scholar. We have looked at other versions of the opening lines of Beowulf Here is Rebsamen's translation: I am Hygelac's hearth-companion; Beowulf is my name. Sometimes I merely rest and avoid thought, feel as if I do, yet do not, exist. SometimesI find myself marchingthrough the fields of Zealand, yet speaking an alien tongue. My gods are the gods of Northern men, and I can see them in the green shoots of spring, hear them in the snarl and slash of storms, and feel them in my arm as I lift my spear and draw it back for hurlingyet at times I speak of only one. Sometimes I myself feel something like a god-and sometimes I am like the morning fog as it slowly vanishes before the rising sun. At such moments I reach for my sword but grasp nothing; call my father's name but get no answer; search for my bountiful king but cannot think where he is. The h e s t moments come when I am awakened from a long sleep by the sweet plucking of harp strings and the strong clear voice of the poet, calling my name: "Beowulf, son of Ecgtheow. He was a good king!"&

If there is surely a degree of arrogance in attempting to translate a first-rate poem, the poet-translator is justified in making the claim.

The Subjective Element in Translation

125

The scholar is usually not thus justified-and Rebsamen's astonishing expansion bears remarkably little relationship to Beowulf, as also it makes very little sense. Rebsamen explains, with staggering arrogance, that "I have presented the whole in the first person, from Beowulf's point of view, simply because it seemed to me that it worked better that way.'"' He has indeed done even more, though the sea changes already noted would be enough to render his version worthless, not to say irresponsible. As for the rest, I have done a number of things to make this presentation of the poem as clear and self-sufficient as I could without destroying its general effect. In order to help the unspecialized reader to a better understanding of what the poem is all about, I have freely incorporated into my translation much information gathered through books and a year's study and travel in Europe: descriptions of ships and armor, geographical precisions, historical references, details of topography and architecture, and numerous small embellishments throughout. I have even, in the best tradition of the medieval poet, stolen lines from other Old English poems when they seemed to add enrichment. I have kept all the "digressions" because they were essential to the poem, but I have clarified most of them and rearranged the sequence of many.* This entire line of argument, that the translator can do essentially whatever he likes with his original so long as he feels, as Rebsamen explicitly does, "that it worked better [my] way" rather than the original poet's way, seems to me to destroy what I take to be the basic purpose of translation. Neither Logue nor Rebsamen is practicing what I call "imitative" translation, because neither is consciously trying to create his own poem quite without regard for or deference to the original. Rebsamen is very clear on this score: "For all that I have done to the poem here, I offer no apology. It has been a labor of love and respect, conceived in honor of this nameless poet and delivered as a tribute to his memory."4s

126 THE ART OF TRANSLATING POETRY

Imitative 'lkanslation There is no need to linger long over this fourth and last category. The outstanding practitioner of our time, Robert Lowell, calls his volume Imitations and makes no bones about what he is doing:

This book is partly self-sufficient and separate h m its sources, and should be first read as a sequence, one voice running through many personalities, contrasts and repetitions. . . I have been reckless with literal meaning, and labored hard to get the tone. . . My licenses have been many. My f i s t two Sappho poems are really new poems based on hers. Villon has been somewhat stripped; Hebel is taken out of dialect; Hugo's "Gautier" is cut in half. . . .I have dropped lines, moved lines, moved stanzas, changed images and altered meter and intent. . . . All of my originals are important poems. . . . I have been almost as free as the authors themselves in finding ways to make them ring right for me.60

.

.

Here is what Lowell does with Boris Pasternak's poem about Hamlet: The clapping stops. I walk into the lights as Hamlet, lounge like a student against the door-frame. and try to catch the far-off dissonance of lifeall that has happened, and must! From the dark the audience leans its one hammering brow against meten thousand opera glasses, each set on the tripod! Abba, Father, all things are possible with theetake away this cup!

I love the mulishness of Providence, I am content to play the one part I was born for . . . quite another play is running now . . . take me off the hooks tonight!

The sequence of scenes was well thought out; the last bow is in the cards, or the stars-

The Subjective Element in Translation

127

but I am alone, and there is none . . . All's drowned in the sperm and spittle of the PhariseeTo live a life is not to cross a field.61 (Lowell has combined this with another poem, but for purposes of comparison I have taken only this portion of Lowell's construct.) Dimitri Obolensky's prose translation showshow far Lowell has gone from the original: The hum has died down, I have come out on to the stage. Leaning against the door-frame, I seek to grasp in the distant echo what will happen during my life. The penumbra of night is focused upon me through a thousand opera-glasses. If only it be possible, Abba, Father, take away this cup from me. I love your stubborn design, and am content to play this part. But now another drama is being acted, so this time let me be. But the order of the scenes has been thought out, and the end of the road is inevitable. I a m alone, everything is sinking in Pharisaism. To go through life is not the same as to walk across a field.*62 Pasternak's sister, unfortunately not a terribly talented translator, nevertheless provides yet another convenient comparison: The murmurs ebb; onto the stage I enter. I am trying, standing in the door, To discover in the distant echoes What the coming years may hold in store. The nocturnal darkness with a thousand Binoculars is focused onto me. Take away this cup, 0 Abba, Father, Everything is possible to thee. *The last sentence is a Russian proverb.

128 THE ART OF TRANSLATING POETRY

I am fond of this thy stubborn project, And to play my part I am content. But another drama is in progress, And, this once, 0 let me be exempt. But the plan of action is determined, And the end irrevocably sealed. I am alone; all round me drowns in falsehood: Life is not a walk across a field.6s Lowell is not simply an expansive translator. The change from a thousand opera glasses to ten thousand is typical: Lowell's mode is extravagance; his purpose is to intensify Pasternak until he is as frantic, as neurasthenic, as Lowell himself in his original poetry of this period. "Everything is sinking in Pharisaism" becomes, for Lowell, "All's drowned in the sperm and spittle of the Pharisee." The addition of (1)drowning, (2) sperm, and (3) spittle makes it overwhelmingly clear that Lowell is simply not interested in the "tone" of Pasternak, but only in turning Pasternak's tone into his own. ''This book," Lowell says in the introductory essay to Imitations, "was written from time to time when I was unable to do anything of my own."54 Nor is this a bad thing, if we take Lowell's work for what it is, a kind of superior poetic cannibalism, and do not consider it what it is plainly not, a volume of translations. I have elsewhere advocated treating the poems in Imitations as "a kind of hybrid form, neither original poetry nor translation," and I have emphasized that "freed from the torment of his own subJects, liberated from the responsibility of full creation, Lowell [in Imitations] showed a vein of abandon, of gaiety, which does not emerge to anything like this degree anywhere else in his poetry."65Any genre which can accomplish that much for a poet of major gifts is a respectable genre. Even if most of its practitioners do not approach Lowell's level, imitative translation remains a significant and worthwhile process.

Collaborative %anslation

Collaboration is both an important and a regular feature of artistic work in almost all fields. Dancers and choreographerswork together, often so intimately evolving their work that the final result can only be attributed to two (or even more) persons. Musicians, like painters and sculptors, tend to create on their own, though they frequently rework (orchestrate, reorchestrate, rearrange, recompose) other people's productions. But novelists have participated in some notable collaborative work (Joseph Conrad and Ford Madox Ford, for example, or Wilkie Collins and Charles Dickens), as have poets, too. In all these primary creative collaborations the two parties are essentially equal. One has an idea; another shares it; and then they proceed to jointly bring their idea to fruition. There may be differences in seniority; there may be emotional reasons why one partner rather than the other dominates. But there is rarely any inherent, substantive inequality. The central fact about collaborative translation is that only rarely is it between equals. The basic dynamic tends to be entirely different. Usually, one party is literary and the other is linguistic. That is, one knows how to write (and usually how to translate) and the other knows the language and culture. The literary partner may be proficient in other languages; he may even have some knowledge of the

130 THE ART OF TRANSLATING POETRY

particular language being translated from; but he will not be able to manage a translation on his own. The linguistic partner may have literary abilities, though most often he is a scholar and does not, but his literary abilities are likely to be neither so well developed nor so well known as those of his partner. In short, while in most other artistic collaborative ventures either partner could proceed on his own, in collaborative translation neither could operate without the other. Each possesses knowledge and abilities that the other does not have. They are thus fundamentally, substantively unequal. Their working relationship is in the biological sense as symbiotic as the fungus and alga which, jointly, constitute the compound plant we know as lichen. Geoffrey Bownas and Anthony Thwaite's The Penguin Book of Japanese Verse, cited earlier, is a typical collaborative translation. At the end of the long introductory essay, which is signed by Bownas alone, it is indicated that Bownas is of Oxford University. He also extends thanks to named parties "for reading and advising on my translations before I passed them on to Anthony Thwaite."' Thwaite then appends a separate note, signed by himself alone, which more fully explains the collaborative dynamic employed: I ought to make it clear what my own part in the making of this anthology has been. Although I lectured in Japan for two years, I cannot read Japanese and, indeed, my command of the spoken language is of the kind generally called "kitchen". Geoffrey Bownas supplied me with translations of the poems, which-with the help of his copious marginal notes, consultation of other translations when they existed, occasional glances at the originals in romanji transcription-I then attempted to turn into forms of words and lines which might suggest, to the English reader, what I took the poems to be in the first place. With very few exceptions, I attempted only those pieces which seemed capable of such suggestion. Geoffrey Bownas and I then worked over my versions so that the final product is a composite effort2 Though their comments are a bit less clear, C. H. Kwock and Vincent McHugh seem to have done their splendid Old Friend From

Collaborative Translation

131

Far Away in exactly the same way. Kwock, patently the linguistic partner, is described as having been born in Honolulu, "where he founded a Chinese-language school and developed a new system of notation for the Chinese alphabet." McHugh, on the other hand, is plainly the literary partner. He "has written and published extensively, including two books of poetry and several novels. He has worked at The New Yorker and with the Federal Writers Proje~t."~ In a conversation about translation, included at the end of the volume, McHugh refers to a "literal version" given him by Kwock, and Kwock refers to having sent "the first finished version" to assorted linguistic "cons~ltants."~ Laughingly, Kwock says that his partner "must have a lot of faith in me . . . I mean, not knowing Chinese," and McHugh answers: "Perfect trust. Hardly a character do I know. And what I know about the structure I got from you. Or from [Achilles] Fang and [James J. Y.1 L ~ U .The " ~ two partners jointly critiqued and worked over the evolving translations, sometimes with the help of another speaker of Chinese, Yao, "a Fukien man from southeast China." McHugh refers to one poem where "the Mandarin lost me early," and Kwock agrees: "Yes. Yes. I remember. The trouble was in the first two lines of the second quatrain." And McHugh continues: "You two got it worked out finally and explained over and over what it could mean. I kept trying this and that and when I got it we all yelled and shouted.'" Clarence Brown explains (in an "Afterthought" which was, he notes, urged on him by both his publisher and his co-translator) how he and W. S. Merwin came to prepare a selection of the poems of Osip Mandelstam. A Slavic scholar who teaches at Princeton, Brown was at work on a book on Mandelstam; in the course of his critical and biographical labors, he explains, "I developed a habit of preparing worksheets on each poem. These included, along with notes on every aspect of the poem that struck me, notations of variant readings, semantic nuances of the diction, peculiarities of the prosody, and so on, a plain English translation, often with numerous alternative translations. Thus, when W. S. Merwin proposed to me in the spring of 1971that we might collaborate on an English translation of Mandelstam (Russian, by some inadvertance, having been omitted from his impressive array of languages), it struck me that much of my own share of the work had already been done. This

132 THE ART OF TRANSLATING POETRY

proved, though only very partially, to be true, for the first stage in our work was my simply turning over to Merwin my worksheets. From these . . . he produced the first versions. In the intervening couple of years we have, with a pleasure that I trust was mutual, debated the early results, sometimes syllable by syllable, offen by painstaking correspondence and more often still by personal meetings in Princeton and London. Bargains were struck, but no compromises were undertaken, I hope, with the English poem that was trying to be born out of Mandelstam's Russian."' Other Russian scholars, notably Olga Carlisle (granddaughter of the Russian writer Leonid N. Andreyev) and Suzanne Massie, have worked with any number of literary people in preparing quite extensive selections of Russian poetry in translation. Carlisle, indeed, has encouraged others with the requisite linguistic knowledge to work with her in the same cause. With two exceptions, she notes, "none of the adaptors whose work is included in this book knows any Russian, each had to work with collaborators who provided them with literal Russian texts. . . . In some instances I consider myself as co-translator of the poem." She adds that the final literary procedure was, as is usual in collaborative translation, distinctly a twoway street. "I wish to thank the many Russian and American poets who helped. . . .Some of them also acted as patient and indispensable editor^."^ Suzanne Massie's working procedures were somewhat similar. "With the exception of the poems of Joseph Brodsky," she notes, "all literal translations from the Russian were done by Suzanne Massie and Max Hayward [another Russian scholar, and a distinguished translator in his own right]. Paul Roche and John Stathos collaborated on some of the final poetic translation^."^ Not that all collaborative translations of this basic sort work out well. The British Anglo-Saxonist A. J. Wyatt served as linguistic collaborator for William Morris in an attempt to translate Beowulf. In public, at least, Wyatt was not unhappy about the results. The final prefatory words to his edition of the poem, in 1894, were "that Mr. William Morris has taken the text of this edition as the basis of his modern metrical rendering of the lay."lo The private reality was distinctly different. "The only English poet to turn his hand to [Beowulfl has been William Morris, and this translation is disastrously bad, being uncouth to the point of weirdness, unfairly in-

Collaborative Translation

133

accurate, and often more obscure than the original (hardly, in fact, a translation at all, since Morris 'worked up' a prose paraphrase passed to him with increasing misgiving by the scholarly A. J. Wyatt)."" Morris's best biographer, Philip Henderson, observes that "by the summer [of 1893,Morris] was regularly reading his Beowulf translation to Burne-Jones on Sunday mornings. A. J. Wyatt of Christ's College, Cambridge, supplied him with a prose translation which he proceeded to 'rhyme up' with his usual speed. Mackail admits the Beowulf to be one of Morris's 'few failures . . .in his desire to reproduce the early English manner he allowed himself a harshness of construction and a strangeness of vocabulary that in many passages go near to making his version unintelligible.' After all, Anglo-Saxon was a closed book to him. Writing to Wyatt in February 1893, Morris said: 'if we read over the original I shall soon I think begin to appreciate the language.' "I2 One of the rare instances of collaborative translation by two partners of apparently equal standing, both in language and in literary abilities, is the long-standing and fruitful work of Edmund Keeley and Philip Sherrard. They have over the years done many first-rate translations of modern Greek poets, including George Seferis and C. P. Cavafy.13 The only "inequality" I am aware of in this distinguished partnership is that Keeley is American and Sherrard is English. The identificatory pronoun in all their books that I have seen is "we"; nor is any separation of function indicated on any level. This is extremely rare, perhaps even the sort of exception that tends to prove the rule. Equally rare, and perhaps even more fortunate, is the situation of a collaborative translator like Norman Thomas Di Giovanni, who worked on the poems of Jorge Luis Borges with the poet himself as his partner. Di Giovanni, however, was part of a sort of multiple triumvirate, since he also worked with other translators: To begin with, on my own, I studied each poem and for most of them wrote out literal versions which I took to Borges and, because of the degree of his blindness, read to him. I would read a line or two of Spanish at a time, followed by an English equivalent for literal sense. [Borges was of partly English descent and fluent both in modern and in Old English.] Oc-

134 THE ART OF TRANSLATING POETRY

casionally, my preparations exhausted, we wrote out transliterations on the spot. . . . On the sheets of these rough versions I also jotted down whatever deliberate or spontaneous comments on the poem Borges would make as I read it to him, and noted any additional biographical or historical background that came up in our talks . . .-anything, in short, that I might pass on to the prospective translator in hopes of lightening his task or improving the quality of the translation or of doing both. . . .The degree of collaboration between Borges, editor [Di Giovanni], and each poet [i.e., translator] varied widely from poem to poem.14 In the remainder of this chapter I want to summarize and briefly comment on my own experience with collaborative translation. This discussion, which will be more or less chronological, is not intended as any sort of qualitative or normative statement. I want simply to set out how one experienced translator found himself thus engaged, what my thoughts about the process were, and how I came to feel about the process after a number of collaborative ventures, with different partners, in very different literatures, and with varying degrees of linguistic capability on my own part. Before 1966, when I was approached by the then Literary Director of the Asia Society, Bonnie R. Crown, and asked to help with a project for translating Vietnamese poetry, I had had several brief experiences of collaborative translations from languages and literatures I did not myself know, again on behalf of and in response to requests from Mrs. Crown, and had also done many Indonesian translations both with my former student Nurdin Salam and with others who had submitted poems for an anthology I was compiling but whose work could not be published without significant revision. Working with Indonesian, of course, I was frequently operating both as a literary and also as a linguistic partner; even with Nurdin Salam, who is a native speaker of the language, I after a time became an equal partner in both respects. In helping Mrs. Crown and the Asia Society, however, and in working with languages and literatures I did not myself command, I had largely considered myself a kind of consultant, or perhaps a sort of editorial helper. In each case-Thai poems, a t which I worked

Collaborative Translation

135

with James N. Mosel, a psychiatrist married to a Thai woman, and Urdu poems, a t which I worked with C. M. Naim of the University of Chicago-I felt considerably more like a transient traveler than any sort of legitimate resident in these cultures. Only a few poems were ever finished, in good part because I was exceedingly nervous about the entire procedure and demanded a quite extraordinary effort from my collaborators. I called the plain prose cribs I requested "literal-literal" versions, but what they truly were was a combination of starkly plain translation, in the course of which I asked for all the multiple meanings of every key word, plus fairly extended essays on the particular poet and his particular literary and technical attributes, plus a general essay on both the larger literary culture and on the poet's specific place therein. Not surprisingly, this distinctly onerous requirement did not encourage my partners to continue our joint work. But I could not feel happy with less, knowing as I did that I in fact knew nothing of Thai or Urdu literature or the languages in which those literatures are written. I felt acutely uncomfortableperhaps "trespasser" would be a more accurate word than "transient travelerw--and was continually worried that what I was doing was neither appropriate nor legitimate. Except for these brief forays, which had very little actual result and did not continue more than a few months, I had neither attempted nor even considered collaborative translation. I had by 1966 published volumes translated from both Old English and Indonesian; I had for many years been translating, though not publishing my translations, from the French, notably the French of a poet very dear to me, Charles Baudelaire; and I had from time to time, usually at someone's request, done versions of poems from Spanish and Italian. For a book on modern Indonesian poetry, drafted in 1964 but not published until 1967, I had translated a pair of Dutch poems for which I could not find existing versions. I had puttered about with several German poems, and been tempted to try my hand a t Horace's Latin, though I had not yet done so. But for all of these poems, and all of these languages, I had a basic competence in both language and literature. To venture beyond those competences struck me as neither practical nor desirable. Neither did I know how it could be done: my early experiments with Thai and Urdu had proved deeply,

136 THE ART OF TRANSLATING POETRY

and I suspected permanently, discouraging, not because the final results were poor but because of the amount of work involved, particularly for my linguistic partner. It took a great deal of straining to produce each single gnat. The turning point was the Vietnamese translation project, my collaborator in which was Nguyen Ngoc Bich. Whether what I acquired was simply a sense of ease with and confidence in this until now very uncomfortable procedure, or whether I became hardened to the discomfort, I cannot now say. But the pleasure I found in working with the strange brilliance of Vietnamese poetry, as well as the pleasure both Mr. Bich and Mrs. Crown took in the translations on which I worked, prepared me for a good many subsequent repetitions of the process. Indeed, by the time Mr. Bich, Mrs. Crown, and I were no longer on quite such good terms, and it was agreed to split the extant translations between a volume to be edited by Mr. Bich and another which would appear under my own name, I was sufficiently comfortable with Vietnamese poetry to write, at my publisher's request, a basically competent introduction to my own volume.16 That brief eight-page essay is in no sense a contribution to scholarship, nor is it (or could it be) original. But it is accurate, and to my knowledge no one has ever found fault with it. In a sense, the dam had been broken. I lived in Israel for two years, in 1968-69, and acquired something more than Anthony Thwaite's "kitchen" knowledge of Hebrew. It was (and is) hard for me to read the language on my own, but in those years I spoke it with some fluency, and with the expert help of Professor Noam Flinker, then my colleague at the University of Haifa, I was able to read the originals and, together with Flinker, turn out some publishable versions of the Israeli poet Yehuda Arnichai.16 I am of Russian Jewish background; both my parents were born in that country and speak its language. My own Russian is primitive, though I am told I have a good accent; I cannot read the language. But when I returned to the United States, and to a post at the University of Texas at Austin, I found myself collaboratingboth with Sidney Monas and, much more extensively, with one of Professor Monas's students, Alla Burago, a native speaker and currently a teacher of Russian." In fact, collaborative translation from Russian had begun even before I left Israel. I worked with secret bmigrbs from the Soviet Union,

Collaborative Translation

137

whose help I paid for but who for political reasons refused to allow me to use their names (which, twenty years later, I have now forgotten). The book I eventually published was tidied up in collaboration with Dr. Burago, who was also initially a paid assistant.18 It may be helpful, finally, to try to summarize what was learned in this second (and still continuing: I have recently done collaborative translation from Chinese, a literary culture about which I know a good deal but a language of which I know next to nothing)19phase of my collaborative translation work. Let me try to rank these comments in their approximate order of importance: 1. What the literary partner wants and needs is a linguistic partner, not another writer. In still blunter terms: the inherent inequality of this sort of translation is also an essential ingredient for its success. The language expert should control in linguistic matters; the literary expert should control the wording of the final product. Neither party should be inflexible, but fundamental control should be divided as I have indicated. 2. The more the literary partner knows, or can learn, about the language and the literary culture with which he is working, the better he will be able to understand his collaborator, and the better he will be able to translate. 3. An experienced translator does not need more than a verbal skeleton of the poem to be translated. He must of course count on his partner to correct his mistakes, both linguistic and cultural. But the sort of elaborate preparation I once insisted upon is, if not exactly a waste of time, at least largely superfluous.

4. Neither partner can afford to be possessory, despite the more or less clearly demarcated areas which each controls. The important thing-virtually the only thing-which matters is the final product, the translation. To that end, both partners must be prepared to sacrifice ego and all other merely personal emotion. It is on this issue, more than any other, that collaborations are likely to fail and fall apart. One must learn a special kind of patience, perhaps even a special kind of humility, to sustain a fruitful collaboration.

The Tkanslation of Oral Poetry

The force of oral transmission-its accuracy and integrity-is perhaps best demonstratedby comparingwritten and orally transmitted texts, both sorts of transmission covering some fairly extensive period of time. One might expect that oral transmission would be far less effective and that texts transmitted orally would contain many more errors, changes, deletions, accretions, and all manner of other divergences from the original form. Judah Goldin, however, describes the "baskets full of books," the "living texts" represented by the living men who both orally transmitted and constituted, in their own persons, effective "oral publication" of Hebrew sacred material. He adds that "to us it no doubt seems that an oral text would be less trustworthy than a written one. This was not necessarily the case with the ancients." Goldin cites the very plain passage in Plato's Phaedrus which argues that writing, as opposed to oral transmission, tends to decrease rather than to increase understanding.' It must be understood, of course, not only that the ancients were accustomed both to transmitting texts orally and to acquiring texts from others via oral transmission, but also that such transmission This chapter appeared, in somewhat different form and under the title "The Manner of Boyan," in Oral Tradition 1:l (January 1986): 11-29.

The Translation of Oral Poetry

139

is a very different thing from what we today think of as memorization. Memorization, that is, is understood by us as an essentially word-for-word affair. Oral transmission, on the other hand, works with larger blocks of material, using thematic and a variety of traditionally derived patterns to aid retention. Goldin notes that in Jewish tradition "no written text, particularly if it is meant as a guide for conduct, can in and of itself be complete; it must have some form of oral commentary associated with it."2 So too with literary rather than religioustexts: scribes and copyists plainly corrupt what they record quite as readily as they preserve it, sometimes from carelessnessor stupidity or inability to read what they are reproducing, but sometimes also from such motives as shaping a text to more modern standards or eliminating or altering something no longer either appropriate or apposite. It is for such reasons, to be sure, that we have, and that we need, textual scholars, even in dealing with material as recent, relatively speaking, as the works of Geoffrey Chaucer, dead only so short a time ago as A.D. 1400. There is no need to argue that oral transmission is always and inevitably superior to written transmission. Where a single piece of written material survives intact over some lengthy period, written transmission is in fact almost invariably superior. But in f a d single pieces (or single collections or groups) of written material do not usually survive intad. They are usually recopied, and recopied again, and that recopying process is what produces a true comparability between the two methods of transmitting texts. In this process of re-transmission, which is arguably a more accurate term for what most of the time actually takes place, oral transmission is apt to be as good as or even better than its written competitor.As Marc Slonim noted in 1950, Russian byliny ("tales-of-things-that-have-been," a form of folk epic poetry conclusively oral both in origin and in transmission) have been "collected quite recently in certain remote villages of northern and eastern Russia, where they were still being narrated in an amazingly well-preservedform by old men or women.'" In the best of all possible worlds, where written texts and oral texts might be neatly and conclusively separable, translation too would be a simpler and infinitely more straightforward process. Interpenetration is, however, a fact of life: oral texts influence written

140

THE ART OF TRANSLATING POETRY

ones, and vice versa, and at various stages of literary development it is essentially impossible to know which (if either) is ~ r i m a r yThe .~ Slovo o pulku Igoreve ("Word of the Campaign of Igor," or as Sidney Monas and I have translated it, "The Tale of Igor's Men"), for example, may well combine both aspectsof textual indeterminacy. That is, it was found in a sixteenth-century manuscript, later burned, but had probably been composed in the fourteenth century. Its subsequent textual history is totally uncertain. Transmission from the fourteenth to the sixteenth century may have been via written texts; it may also have been by oral means, since as Dmitri Mirsky says, "There existed in Kievan times [tenth through thirteenth centuries] a secular oral poetry, preserved by singers belonging to the upper military class. . . . This poetry flourished in the eleventh century; some of the poems were still remembered in the end of the twelfth. . .But it is not clear that at the time of the composition of the Slovo this oral poetry was still alive." Mirsky quite properly insists that "the Campaign of Igor itself is a purely literary work, written, and not sung." Scholars are on the whole well agreed on this. But it remains perfectly clear, too, that the author of the Slovo, to quote Mirsky again, "was steeped [both] in books and in oral tradition. The great originality of his work was that he used the methods of oral poetry in a work of written l i t e r a t ~ e . " ~ This sort of inchoate interpenetration is surely a more problematic matter, especially for the translator of such a work, than Mirsky seems willing to recognize. It seemed plain to Vladirnir Nabokov that the Slovo "is a harmonious, many leveled, many hued, uniquely poetic structure created in a sustained and controlled surge of inspiration," a work of such polished, balanced art that its very existence "attests to deliberate artistic endeavor and excludes the possibility of that gradual accretion of lumpy parts which is so typical of folklore. It is the lucid work of one man, not the random thrum of a pe~ple."~ But Thais Lindstrom, operating with fewer preconceptions, points out that "the Slovo is written in rhythmic prose and its title (slovo meaning 'word,' 'discourse') tells us that it was intended to be declaimed rather than read. It is almost certain that the minstrels, as they recited it, emphasized the rolling alliteration of its phrases with accompanying chords on the gusli, an ancient Russian harp."' And Dimitri Obolensky speaks, similarly, of "the

.

The Translation of Oral Poetry

141

highly musical texture of the poem," noting too that "the use of the repetitions and refrains, the numerous fixed epithets so characteristic of heroic poetry, and the visible signs of strophic composition leave no doubt that the [Sbvo] was intended for oral recitation; and the author himself describes his work as a 'song.' " Like other scholars, Obolensky is clear that the Slovo was a written performance: "the terseness of its style, the richness and complexityof the imagery, the subtlety of its euphonic devices, are quite incompatible with the view that it was ever improvised orally." The point can be argued, to be sure: if oral transmission is in fact based on a patterning of structures larger than single words, then the sort of acoustic patterning Obolensky describes may well be a sign that the Slovo is distinctly related to orally transmitted texts about which today we know nothing. But Obolensky also understands that these neither are nor can be black-and-white matters. Like most scholars, he recognizes without hesitation that "the author [of the Slovo] seems to have known and sought inspiration in an earlier, oral tradition.'" Serge A. Zenkovsky, finally, after pointing out that, although poetic, the Sbvo is not in the usual sense a poem, being "neither rhymed nor organized in verses, nor does it follow any metrical pattern," goes on to observe that "the rhythm and the length of the sentences to some extent replace verse organization. . . .Among other devices, the author of the [Slovo] employs the repetition of characteristic images, stylized descriptions of military action, assonance and alliteration." And he concludes, accurately, that these devices are "impossible to reproduce in translati~n.'~ A translator cannot deal with hypotheses, nor can he safely balance himself between uncertainties. How then can he approach this mare's nest of uncertainties? Here are the relevant fads. The Slovo is not an oral text, but it is heavily oral-influenced. It may or may not have been transmitted orally at some point in its history. It features a rhythmic prose but also many of the devices characteristic of oral heroic poetry. And its artistic density, above all else, is remarkable, making it "a national classic, familiar to every educated Russian and oRen known by heart by lovers of poetry."1° Nabokov's translation chooses a poetic form, and the lineation of verse, but employs a diction so remote and strained that we seem to be reading an impossibly ancient artifact, effectively neither literary nor oral:

142

THE ART OF TRANSLATING POETRY

Might it not become us, brothers, to begin in the diction of yore the stern tale of the campaign of Igor, Igor son of Svyatoslav? Let us, however, begin this song in keeping with the happenings of these times and not with the contriving of Boyan." Zenkovsky too seems to believe that an ancient poem is necessarily an archaic poem and that, as J. R. R. Tolkien passionately insisted, you must not "eschew the traditional literary and poetic diction which we now possess in favour of the current and trivial"12: Might it not behoove us brethren to commence in ancient strains the stern lay of Igor's campaign, Igor, son of Sviatoslav? Then let this begin according to the events of our time, and not according to the cunning of Boyan.13 It seemed to Sidney Monas and myself, on the other hand, that something of the sweep, the rolling prose rhythms, of the Slovo could in fact be brought into modern English, which is surely as dignified a tongue as old Russian. That which is "current," despite Tolkien's (and Nabokov's) prejudices, is not necessarily "trivial." One can adjust, one can fine-tune, any language a t any time to reflect such matters as heroism and ambition, suffering and celebration: And how would it be, brothers, to begin telling the hard tales of the men of Igor, of Igor Svyatoslavich,telling the tales as they used to be told? But let us rather be true to our own time, not to the manner of Boyan.14

The Translation of Oral Poetry

143

No one knows just who Boyan was, but everyone assumes, probably correctly, that he was a once-famous poet, a singer perhaps of orally composed songs, certainly dead, though not yet forgotten, at the time of the Slovo's composition. Like the author of the Slovo, I want the translator of oral poetry, and of oral-connected or oralderived poetry, to be "true to our time, not to the manner of Boyan." And I insist quite as fervently as Nabokov that it can be done, and done well, given a proper respect for both original and translation.

Topi saya bundar, bundur topi saya; kalau tidak bundur, bukan topi saya. This is a child's verse, from a game familiar, in different linguistic guise, to many children all over the world. In dogtrot translation, this Indonesian exemplar of that children's game would run: "My hat is round, / round is my hat; / if it's not round / it's not my hat." No one would argue, and I certainly do not propose to, that this is significant poetry. But it is without any question oral-and in this particular linguistic guise it presents both linguistic features and translation problems that make it worth some attention. Most notably, since Indonesian is a syllabic rather than a stress-phonemic language, this little oral quatrain demonstrates a quite remarkable pattern of stress. As sung, which it always is (to a melody which is similarly employed in a good many other cultures, including our own), it sounds like this (with stressed syllables marked in capital letters): TOpi SAya BUNdur bunDAR toPI saYA KAlau TIduk BUNdur buKAN toPI saYA Neither spoken nor written Indonesian ever organizes a linguistic presentation in this fashion: lines 1and 3follow a completelytrochaic mode; lines 2 and 4 counterbalance with a completely iambic mode.

144

THE ART OF TRANSLATING POETRY

But neither iambic nor trochaic means or possibly can mean anything in Indonesian, which has no prosodic pattern of a stress-based nature. Stress not being phonemic, one can as readily and meaningfully say BUNdar or bunDAR, TOpi or toPI, SAya or saYA. The prosodic organization of traditional Indonesian verse is entirely syllabic; stress has absolutely nothing to do with it. One neither would nor could hear a stress patterning of this sort either in ordinary spoken Indonesian or in Indonesian poetry. Here, for example, is a classic Indonesian pantun, the traditional four-line poem discussed, earlier, on pages 70-74: Dari manu hendak kemana? Tinggi rumput dari padi. Tahun rnana bulan yang mana, Hendak kita berjumpa lagi? The lines each contain eight or nine syllables, but the rules of Indonesian prosody, which disqualify particles and the like, reduce the prosodically recognized syllable count to eight. (All particles in Indonesian are enclitic. Ke-, in line 1, is a direction particle; the first word in line 3, tahun, is invariably pronounced with one syllable; and ber-, in line 4, is again a particle, though here indicating certain verb-like meanings rather than direction.) There is rhyme, there is a steady, stable syllable count (again, by Indonesian prosodic standards), but there is no detectable pattern of stress whatever. One possible way of emphasizing the poem in speaking or reciting (and I deliberately use the term "emphasize" rather than the more technical, linguistically oriented term "stress") would be: dari MAna hendak keMAna TINGgi rumput duri PAdi tahun MAna bulan yang MAna HENdak kita berjumpa LAgi The poem could be read in a different manner; the reading above is simply one among many possibilities. And what this reading indicates is a phrasal sort of emphasis, allied in lines 1through 3 with meaning clusters, and switching in line 4 to a greater emphasis on

The Translation of Oral Poetry

145

meaning. The pantun's meaning, once more in dogtrot translation, is: "Where are (were) you from? where are (were) you going? I Grass is taller than [wet field] rice. 1 When will it be the year? when will it be the month? I that we'll want to meet again?" (I have rendered it, in more literary fashion: "Where have you gone to, where were you from? I Weeds grow taller than grain. I What year, what month, will time have spun 1 Around to when we meet again?") The important thing about the sung stress pattern in the Indonesian child's quatrain, bluntly, is that it is not in the usual sense prosodic a t all, but melodic. That is, only in sung form, and only in conjunction with the particular, familiar melody used around the world for versions of this quatrain, do we get a stress pattern of this sort in a syllabic language like Indonesian. Material which is not sung does not and cannot have any such pattern in Indonesian. And what this means for the conscientioustranslator is that usual translation practices for dealing with poetry must be altered. That is, verse which is not only oral but also and always melodic falls into a distinctly separate category; one needs to try to reconstruct on the page a t least somethingof what the tightly joined combination of words and melody produces in performance. One might, for example, translate the words of the child's quatrain like this (that is, without trying to match the melody): I've got a hat That isn't flat, If that one's flat It's not my hat. What this sort of translation does, plainly, is reconstitute the original's overall verbal effect, but via lexically very variant usages. The formal, structural patterns are different, though related. But the lexical differences are most significant: in material of greater breadth and range than this little poem, those differences would appear enormous, not to say disabling. And then, to attempt to match not only the words but also the music would present complications of enormous difficulty, involving such issues as singability of consonants, the difficulty of certain vowels at higher pitch levels, and so on. This is why translations of texts set to music (songs, opera)

146

THE ART OF TRANSLATING POETRY

are ordinarily so awful.16 Their translators seem not to understand that lexical fidelity is not only not expected of them, not only impossible of attainment, but is in fact counterproductive. Lexically accurate translations of a text tied to a melody cannot be properly sung, cannot be properly heard, cannot be properly understood or appreciated if one does try to sing them. But this is a very special order of oral poetry. The requirement of nonlexical translation applies only to texts that in a sense do not exist apart from some particular melody. This is not the case with most fully oral poetry, and is certainly not the case with oral-connected or oral-derived poetry. The pantun, for example, is a traditional poem very often recited aloud, frequently in "battles" of two reciters (who may in some areas of Indonesia be a man hunting a wife and the young woman he is trying to win), but it is not tied to a specific melody. That there is an oral and folk background to the pantun is plain. But exactly which pantun are derived from this nonlettered background, and which have been composed by lettered authors, is often impossible to say. Nor does it matter: "I have excluded a good many which seemed to me to smell of the lamp," explains C. C. Brown in his admirable Malay Sayings, "but some had to be admitted, by reason of their being heard so often . . . that they could not well be left out." Combining oral and lettered traditions does not leave the resultant form diminished in vitality, or in endurance. Let me set out, once again, the pantun quoted earlier: Dari mana hencI.uk kemana? Tinggi rumput dari padi. Tahun mana bulan yang mana, Henduk kita berjumpa lagi? And let me, this time, give a word-for-word, syntactically absurd rendering, so that something of the flavor of the poem may be appreciated: From where wish to-where? High[erl grass than [wet field] rice. Year where month where, Wish we meet again?

The Translation of Oral Poetry

147

Note that yang, in line 3, is not translated. It is a function word, a connective, sometimes with lexical meaning, sometimes without it. Note too that "we," in line 4, is the form which includes the person spoken to rather than (as in Indonesian's other form of "we") excluding that person. In a sense, then, "we" is here understood by an Indonesian to mean something like "we two." Once more, here is my more polished, literary translation of this pantun: Where have you gone to, where were you from? Weeds grow taller than grain. What year, what month, will time have spun Around to when we meet again? The original rhymes a b a b; so too (more or less) does this translation. The assumption of a past-tense query in line 1 is only that, an assumption, since as I have said in part one of this book the Indonesian language usually does not specify tense. However, the assumption permits use of the rhyme word "from," and it is lexically quite as justifiable as the assumption of a present tense would be. More important, and also more basic, this translation followsa metrical pattern familiar to all readers of English balladry (a folk form of roughly comparable nature to the Indonesian pantun). Its first and third lines have four metrical feet; its second and fourth lines have three metrical feet (the enjambement from line 3 to line 4 causing the streases in line 4 to fall on the second syllable of "around," on "meet," and on "againn-though one certainly could also stress "when" and make this a line with four iambic feet). The translation is basically iambic, as of course most poetry in a language like English must be. It contains three trochaic substitutions: at the start of lines 1and 2, and internally in the third foot of line 3 (where the line-like the Indonesian original-repeats itself syntactically). But trochaic substitutions, most especially in the first foot of an English poetic line, are so frequent in our prosodic tradition as to be almost as regular as the iambic feet they replace. Again, the single most important fact about the translation is its use of the ballad form and the ballad metric. Such formal equivalency explicitly recognizes that comparability of poetic forms from one

148

THE ART OF TRANSLATING POETRY

language to another requires when possible the use of forms that more or less evoke the same genre-feeling in the host language as they evoke in the original. I would argue, further, that in folk and oral-connected poetry this requirement is significantly greater and more important for successful translation. It might well be possible to produce a better (higher) level of poetry in the translation of this pantun if we chose to scant the formal requirement. But comparability would be sharply reduced. For example, concentrating more on sheer poetic effect than on formal comparability, we might translate the pantun like this: Where have you come from? Where will you go? Grass grows taller than grain. What year will it be, What month will it be, When we come together again? This version (which I have just concocted) preserves the lexical shape of the original; it also keeps some of the rhyme. It may well be somewhat better, strictly speaking, as poetry. But whatever its generalized literary merit it is obviously, indeed flagrantly, less like the Indonesian original in the bedrock structural sense. True, the repetitions in lines 1 and 2, and also in lines 4 and 5, do preserve something a t least of the original's folk character. Still, the version as a whole preserves much less of that folk character than does the first, structurally parallel version, and so gives a much less satisfactory idea of the true nature of the original. And that, I would argue, is both inevitable and unavoidable: poetry is structure and genre quite as much as it is individual words or even syntactical patterns. And oral and oral-connected poetry, as I shall now proceed to argue, is even more structure- and genre-dependentthan is strictly lettered poetry. Longer poems, and especially more sophisticated longer poems, with more and more admixture of written literature's approaches and literary devices, require the translator to scramble a good deal more flexibly. Nor do we need to move to a full-length epic to exhibit this

The Translation of Oral Poetry 149 difficulty. Indeed, we can choose one of the few Old English poems we know pretty reliably to have been composed both orally and by a nonlettered (i.e., an illiterate) poet, namely, the short poem we call "Caedmon's Hymn."16 It must of course be emphasized that, though illiterate, Caedmon surely learned some significant part of his poetic technique from learned monks (he was employed at a monastery). The ultimate literary ancestorsof these nameless monks were oral bards (scops),but monkish poetry plainly incorporated as well a whole host of lettered influences, many of them non-English in origin. Caedmon, in short, is an unlettered, oral poet shaped both by oral tradition (though we do not know how o; to what extent) and by lettered poets (who were in their turn in good part shaped by both oral and lettered influences). Here then is the Old English text of "Caedmon's Hymn," typographically normalized from the surviving West Saxon version:

Nu sculon herigean heofonrices weard, meotodes meahte and his modgethanc, weorc wuldforfaeder, swa he wundra gehwaes, ece drihten, or onstealde. He aerest sceop eorthan bearnum heofon to hrofe, halig scyppend; tha middangeard moncynnes weard, ece drihten, aefter teode, firum foldun, frea aelmihtig. In plain dogtrot prose, arranged to follow the original's lineation, this might read: Now should (must) we praise the lord (keeper, guard) of heaven, the power (strength) of God (the creator; fate) and his thought, the work (action, labor) of the glorious father, as he all (each one of the) wonders (marvels, miracles), eternal (everlasting) lord, in the beginning created (established). First he shaped (created, formed) for the sons of earth

150

THE ART OF TRANSLATING POETRY heaven as a roof (summit), holy creator; then the earth (world) mankind's guardian, eternal (everlasting) lord, afterwards created (intended, appointed) the land for men (mankind), lord (king, ruler) almighty.

Large cosmological matters, and large theological ones, are here handled with great sureness and, equally, with immense conviction. The stuff of the poet's belief, that is, is of no greater importance for the poem than the quality of his belief, its intensity and persuasive power. It is a noble and a memorable poem-as is witnessed by the fad that no fewer than seventeen manuscript versions have survived, thirteen of them in the West Saxon version here reproduced.17 People obviously listened to, and also read, this nine-line hymn with engaged devoted attention. It seems to me indisputable as well as completely sensible to say, with John M. Foley, that "the phraseology is most productively understood not as a collection of prefabricated units ready to hand, but as a living tissue of language with genetic ass~ciations."~~ The signs of oral connection are necessarily different here than in either the Sbvo or in the pantun. Caedmon's poem employs formulaic or formula-like expressions, as well as a prosodic patterning which joins stressed syllables within a line by means of alliteration. As I have indicated, it is next to impossible to determine just how truly oral a poet Caedmon was. Just as we lack much understanding of the background of the Slovo, so too we do not know a great deal about Caedmon, his training, just what he had heard and had not heard, to whom he had talked, by whom he might have been instructed, and so o n . l T e lack almost entirely the poems that preceded and were contemporaneouswith the Sbvo; we have some but by no means all the poems that were roughly contemporary with "Caedmon's Hymn," a deficiency which makes it impossible to be authoritative about what is and is not oral-formulaic in the poem. In translating "Caedmon's Hymn," accordingly, we have to reckon not only with assorted uncertainties, but also with the imprecise certainty that it is in part an oral poem, that it is connected to all sorts of other poems in the same tradition, some of which we know, but most of which we probably do not and never will know. And we

The Translation of Oral Poetry

151

also have to deal with the plain fact that, no matter what its composer's nominal background (or lack thereof), "Caedmon's Hymn" is doctrinally and cosmologically distinctly sophisticated. The poem's diction fairly rings with echoes both poetic and theologicaYphilosophical. And yet, a t the same time, its diction also resonates, fervently and with a peculiarly medieval credibility, with the strength and joy of what we can only assume to have been Caedmon's personal faith. Here is the first of the translations I wish to consider; it is by Kevin Crossley-Holland: Now we must praise the Ruler of Heaven, The might of the Lord and His purpose of mind, The work of the Glorious Father; for He, God Eternal, established each wonder, He, Holy Creator, first fashioned the heavens As a roof for the children of earth. And then our guardian, the Everlasting Lord, Adorned this middle-earth for men. Praise the Almighty King of Heaven.20 This rendering, which dates from 1965, is obviously competent. I want neither to praise nor to damn it, but only to try to understand, from the perspectives so far here employed, what its rationale is and is not. The Old English scop employs three large phrase units (what contemporary poets are apt to call "breath" units). The first is of four lines, the second is of two, and the third is of three. CrossleyHolland uses four phrase units, the first of almost three lines, the second of just over three, the third of two, and the last of exactly one line. The structural patterning of his translation is thus similar, though not identical, to that of the original. Since, however, there is no reason to think the scop's phrase units inherently fixed, such minor variations are in one way not terribly important. More significant by far are the particular rhythmic effects aimed at, and created by, the translation's use of phrase units. That is, whatever Caedmon's models (again, we do not know what they might have been, if indeed there were any), he obviously was aiming a t a series

152 THE ART OF TRANSLATING POETRY

of sweeping, piled-up phrasal units, with all the usual repetitions and sidewayspoetic movement (as opposed, that is, to straight-ahead, linear poetic movement) that we inevitably and correctly associate with Old English poetry in general. The translator has plainly sensed this movement and done his best to reproduce it. His one-line final phrase unit, however, seems just as plainly a totally different sort of verse movement from anything to be found in Caedmon's poem. This may or may not be a crucial difference, but the short, crisp summary of the final line is distinctly unlike the original's more modular effect: "Caedmon's Hymn" in Old English is built like a series of waves, each sweeping in to the shore, rather than on the sort of neat, single, separate assertions that Crossley-Holland's translation ends with. The scop's architectonic, as opposed to his strictly architectural, patternings are, however, partially ignored in Crossley-Holland's translation. There is replication of the formulaic phrasing; there is no replication of the stress-alliteration prosody. One can argue that Old English prosody in fact no longer exists, and is therefore not available in the language that modern English has become. The answer to this objection, however, must I think be that there is no need fully to re-create Old English prosody. What is necessary is to suggest it, to re-create it to whatever extent may be feasible. And the loss of stress-alliteration in the translation seems to me distinctly critical, and a serious deficiency. Lexical considerations are, as I have argued earlier in this book, of a second order of importance: structure and genre must emphatically take precedence. But lexical considerations are obviously significant, and the variations as between the Old English and the Crossley-Holland translation are somewhat more serious than the variations in phrase unit and prosodic patternings. And these variations seem to me to fit a doctrinal pattern of their own. CrossleyHolland gives us weard as "ruler" rather than as 'lord, guard, keeper." Drihten (commonly used for both secular and sacred monarchs) becomes "god" rather than "lord." Teode is translated as a distinctly KingJames-Bible-sounding"adorned" rather than as "created, intended." And frea, "lord, king," which is all the scop thinks he needs to say, is turned into "King of Heaven" (emphasis added). These variations taken as a whole create an oddly nineteenth-century ver-

The Translation of Oral Poetry

153

bal atmosphere-at the same time, too, as the rendering of middangeard, "middle-earth" rather than "earth, world," sounds a note of antiquarian preciosity sharply out of key both with the original and with much of the general tone of the translation itself. There is some sense as well of lexical inconsistency. Words like "established," "fashioned," and "adorned" are considerably more formal than the rest of the translation's vocabulary. They are also distinctly stiffer, more constrained, less "popular," even less colloquial, than their Old English equivalents, onstealde, sceop, and teode. But the essential competency of Crossley-Holland's translation becomes clear if we compare it to what has sometimes been done by others. Here for example is the rendering of Richard Hamer, who sees very fully the structural principles involved but who totally fails to embody them lexically: Now we must praise the Guardian of heaven, The power and conception of the Lord, And all His works, as He, eternal Lord, Father of glory, started every wonder. First He created heaven as a roof, The Holy Maker, for the sons of men. Then the eternal Keeper of mankind Furnished the earth below, the land for men, Almighty God and everlasting Lord.21 All the same, notice how the verse movement of the last three lines here, no matter how dull and flat-footed the translation as a whole, far exceeds Crossley-Holland's both in inherent sweep and in accurate reflection of the original. Structural matters, again, almost automatically take precedence over merely lexical ones, especially in this sort of poetry, despite the equally obvious fact that inadequate handling a t the lexical level can ruin a sound structural perception. Had Hamer made some attempt to echo the stress-alliteration pattern, which in fact he ignores quite as steadfastly (and erroneously) as does Crossley-Holland, he could have vastly improved the translation. Its general sensitivity, however, seems clear. One can acquire some notion of the original from Crossley-Holland. One can acquire very little notion of the original from Hamer.

154

THE ART OF TRANSLATING POETRY Now sing the glory of God, the King Of Heaven, our Father's power and His perfect Labor, the world's conception, worked In miracles as eternity's Lord made The beginning. First the heavens were formed as a roof For men, and then the holy Creator, Eternal Lord and protector of souls, Shaped our earth, prepared our home, The almighty Master, our Prince, our God.22

This translation, made in 1954 (though it was not published until 1960),is one for which I am myself responsible. It makes use of only two phrase unit structures, one just over four lines long, the second just under five lines long. In strictly numerical terms, plainly, this is neither closer to nor further from the phrase unit arrangement of the original than is Crossley-Holland's rendering. Crossley-Holland, again, has four units, one of just under three lines, one of just over three lines, one of two lines, and a final phrase unit of exactly one line. The original, once more, has three phrase units, of four, two, and then of three lines. But numerical terms hardly settle the poetic issue, for as I have emphasized, Crossley-Holland's h a 1 phrase unit, consisting of but a single line, does not accurately reflect the verse movement of the original. I would argue-quite apart from my own authorship of the last translation quoted-that my version does in fact capture more of the basic sweep of the Old English. Furthermore, the second of the original's primary attributes, namely, its stress-alliterationprosody, is distinctly echoed, if not entirely accurately replicated. (The four-stress pattern per line, too, is more carefully adhered to than it is in Crossley-Holland's translation, where lines 5 through 8 are of doubtful four-stress authenticity.) Only lines 5 through 8 in my translation do not preserve some clear stress-alliterativepatterning, and even these lines at least hint at what they do not quite effect. Line 5, for example, has f as the initial sound of the third stressed word and the final sound of the fourth stressed word. Line 6 substitutes rhyme for stress alliteration in the first two stressed words. Line 7 blends 1 and r to at least suggest stress alliteration. Line 8

The Translation of Oral Poetry

155

uses r to the same impressionistic but nevertheless palpable effect. Lexically, though it may seem somewhat less close to the original, my translation follows a deliberate course that may not at first sight be apparent. It attempts to replicate key Old English words with a small cluster of alternate meanings, rather than merely rendering them word for word. It makes a parallel attempt to use modest syntactic rearrangement to replicate a fair amount of the original's lexical variety. Modern English is of course a very great deal freer in its word arrangements, having a decidedly analytical syntax and employing many fewer morphological markings to indicate a word's function and meaning. And Old English poetry is notorious for its insistent delight in multiple iteration of essentially the same thing. The translation thus doubly renders Heofonrices weard, in line 1, as "God, the King / Of Heaven." Frea aelmihtig, in line 9, becomes "The almighty Master, our Prince, our God." "Prince," in turn, also evokes some of the dual sense of drihten, in line 4, just as h a , which Crossley-Holland correctly translates "King," in my translation also evokes "Master." There are thus nine epithets for God in my translation, as against a total of seven in Crossley-Holland and eight in the original. Similarly, wuldorfaeder, in line 3, is divided between "The glory of God," in line 1 of my translation, and "our Father's power," in line 2. Modgethnc, in line 2, and weorc, in line 3, become in my translation both "the world's conception" and "His perfect labor." In short, my translation attempts to incorporate structural and lexical features of the Old English original, adapting those features to the very different linguistic nature of modern English. The verse movement of the partially oral original is not precisely re-created, but it is unmistakably echoed-and, just as important, at no point is it ever contravened, as it so clearly is in the last line of CrossleyHolland's version. Literary tone and rhetoric, too,are adaptations rather than precise replications. However, there is I think nothing internally inconsistent in the presentation of rhetoric and tone, as there is both in line 8 of Crossley-Holland's version and also in some of his lexical choices. Just as the original is internally consistent, so too must the translation be, if it is to convey in a new linguistic garb any of the authority of the original. Translation is inevitably approximation.

156 THE ART OF TRANSLATING POETRY

But like Janus the translator must constantly be looking in both directions, carrying out of the original as much as he is able to bring with him, but also creating in his translation a replica which can have some chance of standing for itself as well as for the original on which it is based. This is in some ways doubly difficult when-as in Caedmon's poem-the original incorporates indeterminate elements of two traditions, one oral, one lettered. Whether the scale of priorities the translator develops, and employs, be analytical or impressionistic or even inchoate and unstated-it does not matter, to my mind, so long as that scale is basically accurate, and so long as the resultant translation works-there must be some such scale, for it can do a great deal to help point the translator's difficult way. In the translation of oral or oral-connected poetry, as I have argued, the two highest priority items on that scale must be genre and structure, with the latter incorporating both formal external structure and such internal structuring devices as, in Old English poetry, stressalliteration prosody, and, in Indonesian forms like the pantun, balanced eight-syllable lines. Indeed, the translator of oral or oral-connected poetry who keeps his priorities straight can succeed, I firmly believe, far better than a perhaps more talented competitor who simply follows his nose (and his sometimes too contemporary inclinations). Bluntly, those priorities will reflect-and if properly framed will accurately reflectthe true meaning of the word "tradition." And that is inevitably the single key word of words both in reading and in translating all oral and oral-connected literature, both in verse and in prose.23 What it comes down to is that we cannot, either in reading or in translating, validly substitute our own basic priorities for those of the original poet. In the case of oral and oral-connected literature especially, his priorities are functionally embedded in the operating forms of his culture, and they are in fact what "tradition" means to him. In seeking to understand another tradition, what more fundamental error could we make than replacing one set of basic priorities with another-thus effectively replacing one tradition with another? All cross-cultural, cross-traditional understanding ought to involve as little of such substitution, and as much painstaking replication, as can possibly be achieved.

The Tkanslator's Responsibility

The title of this chapter should perhaps be in the plural rather than the singular: the art of literary translation, and above all the translation of poetry, might almost be defined as the art of balancing different claims. The strictly linguistic claim on the translator, though of course it is a substantial one, is not-though to most people it would seem to be-the most pressing. What I consider the most important, and also the most difficult, claim to deal with is what might loosely be termed the aesthetic: How is the translator to reproduce in the new language the peculiar force and strength, the inner meanings as well as the merely outer ones, of what the original writer created solely and exclusively for and in a different language and a different culture? Of course, there are also time claims: what was written a thousand or two thousand years ago necessarily requires different treatment by the translator than what was created only yesterday. And there are cultural claims: no matter how simplistically some people may regard these matters, the differences between and among cultures are not simple and mechanistic, are not mere differences in the words by which the identical phenomena are described. Those phenomena are not fixed and unchanging: the world is in fact perceived differently by different peoples, and their languages and literatures ex-

158 THE ART OF TRANSLATING POETRY

press those differences. The literary translator can neither ignore nor ever fully capture those differences. Again, his job lies in the balancing of these and all the other claims laid against him. Literary translation is not an impossibility, if we view it from a reasonably large and friendly perspective. It does become impossible, admittedly, if we view it narrowly and from the narrow, univalent perspective of any one among the claims presenting themselves for attention. I shall never forget the pre-publication reviewer for a university press (which did ultimately publish the book) who objected to one of my and Alla Burago's translations from the poetry of Nikolai Stepanovitch Gumilev on the ground that the Russian original depended for much of its subtle effect on the distinction between verbs which favored the infinite form and verbs which did not. There being of course no such distinction in English, from this sort of perspective the poem can be considered untranslatable-and anyone who attempted to translate it might be considered both foolhardy and incompetent. On the phonological rather than the syntactical level, consider a brief phrase from Baudelaire's "Hymne B la Beaut&": Ange ou Sirane. Not one syllable of this phrase has an aural counterpart in English. No translator into our language can possibly replicate the phonology of the French. But then, a translation of Baudelaire into English must necessarily be neither French nor, strictly speaking, Baudelaire. Its author is the translator; its phonology and syntax are the phonology and syntax of English. Had Baudelaire been thoughtful enough to write in English in the first place, there would be no need to bother translating him-though we might then have to deal with the problem of translating him into tongues foreign to him, one of which would be, M h ,French. Different literary translators approach, and always have approached, their responsibilities differently. What I should like to discuss is some of the shifting and sliding that goes on, both as between and among different translators of the same work and also in the career of a single translator as he confrontsdifferent originals and as he himself grows and, one hopes, matures and improves. My basic method will be the comparison of specific examples rather than the laying down of authoritative principles. I tend to have more faith in achieved realities than in rules. And since I have myself been translating for something over forty years, from a variety of lan-

The Translator's Responsibility

159

guages and cultures and times, and have at least evolved if not necessarily matured or improved as a translator, I shall feel free to examine my own work, and my own approaches to it, as well as the work of others. I can think of no better, nor any more logical, place to start than the work of Ezra Pound. From the very beginning of his career, and up to the very end of that career more than half a century later, Pound's translations seem to me peculiarly authoritative. Whether or not one likes a particular translation, or even approves of it, Pound's assumptions are hard to deny-and his example is so extraordinarily persuasive that, if one had to specify a single person as the father of twentieth-century literary translation, no other single person could possibly fill the bill. I spoke briefly of Pound's "Seafarer" translation in the first part of this book and promised to return to it. When Pound makes the Old English scop say, Bitter breast-cares have I abided, Known on my keel many a care's hold,' one senses that the poetic voice is assured and deliberate, that the unusual phrasing is the furthest thing from accidental and lies tightly in the translator's control, and that whatever strange and exotic effects the translation is producing are effects the translator intended. More: one feels, even without a knowledge of Old English (and I can affirm that I so felt before I came to know the language), that Pound is giving us something which without any doubt is to be found, and found uniquely, in Old English. We become aware, that is, that what Pound is offering us is some reasonable facsimile in our own tongue of what the ancient scop had done in his. That perception is quite simply accurate. There is no translation of any Old English poem which gives us as much of the music, the harsh rhetoric, and the piled-up effects of Old English poetry. Let me especially emphasize that statement: no translation known to me captures anything like as much of the poetics of Old English, and in particular of its prosody. None. That is a flat assertion, and I am prepared to defend it. But it must also be said that Pound has made, as I will argue that

160 THE ART OF TRANSLATING POETRY

every translator must make, a very particular, and therefore a limited, balancing among the multiple claims exerted by the original. In order to accomplish what without the slightest doubt he has accomplished, he has had to forgo, and even to betray, other aspects of the original than those he wants to offer us. (Recall that basic axiom of axioms, applicable to all manner of translations from all manner of tongues and periods. the translation is not, never will be, and cannot be the original.) There are different varieties of resemblance for which a translator may aim; there are different emphases with which he may work. But no rendering can ever capture everything. What Pounds is after, principally, is the prosody of Old English, and to a somewhat lesser extent the phonology generally. He captures what he is after-and captures it, as I have said, magnificently. The Old English of the lines from Pound's translation is: bitre breostcare gebiden haebbe, gecunnad in ceole cearselda fela. The dogtrot prose rendering might go: "I have endured (experienced) painful anxiety, / experienced on shipboard many sorrowful places." Such a rendering makes extremely clear what a huge difference there is between such criblike translations and the masterful translation of an Ezra Pound. Yet the dogtrot rendering suggests an important question. What has happened, in Pound's version, to other aspects of the original than the prosodic or the broader phonological ones? Lexically, it is possible to see something like "I have endured painful anxiety" in Pound's "Bitter breast-cares have I abided." Is it, however, possible to see anything of "I have experienced these things on shipboard in many sorrowful places" in Pound's "known on my keel many a care's hold'? Well, perhaps a lexical bit can be seen. We can surely manage to locate a "keel" on a boat; that may not put us on "shipboard," but it does at least get us out to sea. But Pound's "many a care's hold"? Bluntly, it is not what the Old English says, not by a long shot. Does that matter? Especially since I have asserted Pound's authority as a translator, and also asserted the plain successes and the undeniable iduence of his translations,2are these lexical divergences not to be

The Translator's Responsibility

161

viewed as necessary M o m s , variations required in order that Pound may in fact achieve what Pound surely has achieved? Let me step back a minute and lengthen our perspective. We have been looking only at lines 4 and 5 of a longish poem. Let us now consider the poem as a whole. It comes to 101 lines in Pound's translation. But that too presents us with a problem, for the Old English original comes to 124 lines. In short, there are twenty-three lines more to be translated than Pound has in fact translated. Nor does he give us any indication that he has in any way truncated his original. Pound is the furthest thing from an Old English scholar. But there has been some controversy over the authenticity of the Old English text that has come down to us. Mrs. I. L. Gordon, the poem's most scrupulous editor, observes that "some would end [the poem] at line 102, some would have us believe that the original consisted of the seafaring part only, to line 64, the rest being the work of a Christian reviser, and some have even distinguished originally independent poems in the seafaring part itself.'" Mrs. Gordon's firm response to such would-be editorial tinkerers is unequivocal: "The poem has suffered . . . at the hands of editors and commentators. The unusual difficulties its theme and structure present have proved a temptation to theorists who, all too often, have shown scant respect for the poem as it stands. . . . There are, in fact, no reasons for denying [textual] integrity to the poem except the difficulty of tracing in it a connected theme."4Mrs. Gordon points out, in a very full discussion, that much of the would-be tinkering was suggested early on, and that later, more thoughtful scholars have quite rejected all such approaches. But let us concede that Mrs. Gordon's work, and the more recent scholarship on which it rests, were necessarily unknown to Pound. Let us concede the possibility that Pound was in truth aware of those older, now disproven theories, and might have based his cut version of the poem on them. But a problem remains, for what Pound has omitted from the end of the poem is also what he has consistently suppressed throughout, namely all reference to anything religious. The last twenty-two lines are clearly a prayer-but the religious theme is basic all through the poem. Pound will have none of it. His intent is to present "The Seafarer" as an exemplar of secular harshness. His view of the poem makes it strictly, and exclusively, a

162 THE ART OF TRANSLATING POETRY

description and account of a gray, drizzling, p a i a landscape, a journey of the body's groans and aches. He sees the elevation of the soul as having no part in that miserable music, and he extirpates all soul-elevatingreferences with admirable but b t e d consistency. The scop says, for example, at lines 39-44 (a portion of the poem that the old theories did not question, except in their dottiest moments):

for thon his thuese modwlonc mon ofer eorthun, ne his gifena thues god, ne in geoguthe to thues hwaet, ne in his daedum to thues &or, ne him his dryhten to thues hold, thuet he a his saefore sorge naebbe, to hwon hine Dryhten gedon wilk. ("For there is no man on earth so proud-spirited (brave), of such excellent endowment, nor so bold a youth, nor so courageous in his actions, nor to whom God (or: his lord) has been so gracious, that he feels no anxiety when he puts out to sea, no fear of what God will do to him.") Pound turns this into a ruthlessly, utterly secular passage: For this there's no mood-lofty man over earth's midst, Not though he be given his good, but will have in his youth greed, Nor his deed to the daring, nor his king to the faithful But shall have his sorrow for sea-fare Whatever his lord This does not, taken as a whole, make a great deal of sense: Pound, as we have noted, is much less concerned with lexical transmission than with bringing other aspects of the Old English to us. But let us concede that the meaning of dryhten, which can in fact be either secular ("lord, king, ruler") or religious ("God, Fate"), is at least debatable here. Modern editors see dryhten in the last line of the quoted passage as a religious reference, and the reference to dryhten

The Translator's Responsibility 163

in the third line of that passage as either a religious or a secular one. But Pound was not privy to modern scholarship, so there is at least an appearance of rationality in his lexical choices. It is not a convincing rationality; the fierce anti-clericalism is pretty clearly out of place, even here. But Pound's choices come to have vastly less rationality when we get to lines 78-80:

ond his lof sithan lifge mid englum awa to ealdre, ecan lifes blaed, dream mid dugethum. ("and his glory will live then among the angels for ever, in the blessedness of eternal life, bliss among the noblest.")

To turn this into a secular passage takes ingenuity, as well as absolute conviction (not to mention irrational stubbornness, of which Pound notoriously possessed more than his share). Because of discrepancies between the original and Pound's translation, I include one preceding line and then the three-line passage principallly at issue:

So that all men shall honour him after And his laud beyond them remain 'mid the English Aye, for ever, a lasting life's-blast, Delight 'mid the doughty! The scop may have written of englum ("angels"), but Pound wants no angels in his "Seafarer." Englum thus suffers a severe (and lexically totally unwarranted) sea change and emerges as "the English." So too d m mid dugethum, '%lissamong the noblest," gets a smartly military rendition and becomes, Kipling-like, "delight 'mid the doughty." Again, there is no lexical warrant for this: none. But whatever it may take to turn the Old English poem into a secular affair-anything-Pound is prepared to do. And on one important level we ought not to criticize him for this lexical mangling. Instead, we ought to try to understand both what he has and what he has not done. Plainly, the notion of the Old

164

THE ART OF TRANSLATING POETRY

English people as anything but swaggering pagans is repugnant to Pound. To him, their poetry is a brave old blast out of the bogs and fogs and moors and fens. Fine and well, I would argue, though I do not in the least agree. There is at least some warrant for that view of Old English life and culture. It was also a much more common view in Pound's day. Old English poetry as a whole, and "The Seafarer" in particular, exhibit a far wider and deeper range, in truth. Without much hard looking we can find in that poetry, in unmistakable language, great delicacy and passionate religious feelingand much, much more. Pound is entitled to his view, however archaic, even however unbalanced we know it to be, for his view does give us something-not all, but something-of what the original does indeed contain. No translation, again, can ever give us all; Pound is to be endlessly praised for what he has given us. But still, other mixes, other balances, other sorts of accommodations as between and among the conflicting claims asserted by the original poem, are possible. In my own translation I have been much more cautious in trying to reproduce the prosody, and have done very little about trying to reproduce larger aspects of Old English phonology, because I wanted to convey more of the complexity of Old English religious feeling, more of what I perceived as the poetry's striking fusion of secular and religious elements. In the first of these two passages, where Pound gives us his resolutely secular "moodlofty man," and a bracing passel of kings and lords, I translate instead: But there isn't a man on earth so proud, So born to greatness, so bold with his youth, Grown so brave, or so graced by God, That he feels no fear as the sails unfurl, Wondering what Fate has willed and will do.' And in the second and more emphatically religious passage, I translate (again, syntactical differences make it necessary to quote a bit more than the three Old English lines at issue): and death Can only bring you earthly praise

The Translator's Responsibility 165 And a song to celebrate a place With the angels, life eternally blessed In the hosts of H e a ~ e n . ~ Do not come to my translation of "The Seafarer" for the bravura display that Old English poetry puts on. But do not go to Ezra Pound's translation for an accurate portrayal of the Old English soul or of the Old English mind. The heart probably beats harder in his translation. I think it beats more complexly, as well as rather more accurately, in mine. But both translations seem to me, in their exceedingly different ways, responsible. I have not been discussing error; I should like to make that very clear. I have been talking solely about translation choices, and about the balancing between and among competing claims which underlies and dictates those choices. Error is another matter entirely, and I neither need nor want to say very much about it. It happens. When Pound tells us, for example, in yet another truncated version from Rimbaud, that "A woman's head with brown over-oiled hair / Rises out of a theatre box,"9 he has been seriously overconfident of his French. He has mistaken one sense of baignoire ("theatre box, bathtub") for another. As he so often does, he has assumed-and thereafter been utterly unable ever to question his own assumption-that it is the "theatre box" meaning which Rimbaud intended. Alas, it is in f a d the 'bathtub" meaning which Rimbaud has employed, a fact which becomes starkly clear as the French goes on to speak, grossly and savagely, of the bathing woman's spine and buttocks and anus-anatomical features not on display in theater boxes, but regularly visible in bathtubs. The French poem is a fourteen-linestandard sonnet; Pound translates only eight lines (and again as usual makes no mention of omitting anything), probably because the impossibility of his "theatre box" setting became obvious even to him. But Pound does not admit error, nor does he change his mind. He simply walks away, having done what he has done. This is regrettable, and unfortunate, but what of it? Translators too are human, and they have always made and will always make errors, just as other humans do. In dealing with errors in translation I take my place alongside that h e translator, and even finer critic, John Dryden:

166 THE ART OF TRANSLATING POETRY

There are a sort of blundering, half-witted people who make a great deal of noise about a verbal slip. . . . True judgment in poetry, like that in painting, takes a view of the whole together, whether it be good or not; and where the beauties are more than the faults, concludes for the poet against the little judge; 'tis a sign that malice is hard driven, when 'tis forced to lay hold on a word or syllable; to arraign a man is one thing, and to cavil at him is another.1° Ezra Pound is far too great and important a translator, and we are all of us, translators and readers alike, far too deeply in his debt, for us to "make a great deal of noise about a verbal slip," or even about a series of verbal slips. All the same, I can't help recalling my own translation of an essay by the Indonesian poet Chairil Anwar, in which I make him speak (I give only the English, which for this purpose is sufficient) of "the service given to his Homeland, more and more fervently, by J. M. M. Tenno Heika." Confronted with so strange and un-Indonesian a name, I did not know what to do. I hunted and hunted for some reference to J. M. M. Tenno Heika but never found the faintest whiff of him. After the book in which the translation appeared was published, the disthguished scholar of all thingsJapanese, Donald Keene, informed me that "J.M. M. Tenno Heika" was indeed not Indonesian but Japanese, and meant "His Majesty the Emperor of Japan." Perhaps I should have realized, baffled as I had been by all the usual pathways to linguistic knowledge, that though I knew no Japanese, Chairil Anwar had lived through the Japanese Occupation of Indonesia and had been forcibly exposed to the Japanese language. Still, I do not think the translation of the essay is ruined by this silly mistake. It is marred, to be sure-but then, so is everything we humans do. Of course, in dealing with a translator of the magnitude of Ezra Pound, there is no reason to doubt the validity of what I have referred to as the aesthetic claim his translations make. In dealing with lesser translators, however, their aesthetic claim must be much more carefully scrutinized. I have stated this fundamental position many times in my study of the translation process, The Forked Tongue:"Poetry in translation is either poetry born anew or it is nothing at all";ll "The greatest sin a translator can commit . . . is to fail to breathe

The Translator's Responsibility

167

.

life into his re-creation";12 "No greater sin exists, to my mind, . . [than translation which] keeps the scholarship clean, but muddies and destroys the poetry."lSOr as I later phrased it, in the introduction to my translation-edition of Horace's Ars Poetica ("The Art of Poetry"), if a translator "is not producing a viable poem, why is he bothering to translate? Would anyone have bothered to read Horace, all these years, if he had written clumsy, confused Latin, devoid of music, hard to penetrate, flat of tone? Why should anyone read a translation of Horace, written flatly, clumsily, dully? Is that sort of translation really Horace?"14 To put it still differently: if the translator of a literary work, and above all of a poem, has not done justice to the aesthetic claim, almost nothing else that he has done can possibly be worth very much. I do not believe this can be called prejudice; it seems to me simple, clear fact. For example, translation which proceeds from the simpleminded notion-discussed in part one of this book-that the translator's chief responsibility is to the words of his original is frequently amateur (or scholarly) translation. And it is almost bound by its nature to slight, usually drastically, fatally, the aesthetic dimension of the original and the aesthetic claim of the translation. Any linguist can testify (if literary scholars would only ask them), as does J. C. Catford in A Linguistic Theory of Translation, that "a restricted kind of 'transference of meaning' from one language to another is possible; but it is equally clear that this is not what is normally meant by 'translation.' . . . Since every language is ultimately sui generisits categories being defined in terms of relations holding within the language itself-it is clear that formal correspondence is nearly always appro~imate."'~ Richard Hamer, translating the same Old English poem that Ezra Pound and I and many others have tackled, "The Seafarer," produces a numb, almost catatonic travesty of poetry: Yet no man in the world's so proud of heart, So generous of gifts, so bold in youth, In deeds so brave, or with so loyal lord, That he can ever venture on the sea Without great fears of what the Lord may bring.16 Hamer is, however, very clear about what he is up to: "I have tried to give the sense of the texts as closely as possible . . . [and] I have

168 THE ART OF TRANSLATING POETRY

tried to keep as close to the original as I could while avoiding absurdity.'''' In short, Hamer is working as much on the purely lexical level as he decently can. As the result effectively demonstrates, it is much too modest a goal. I do not think "translation" is in truth an accurate description for a rendering where the mix is so determinedly, so consciously unfavorable to the poetry. I do not think we can call Hamer's sort of rendering anything more than a trot, a crib. It is also possible to err in the opposite direction, to be so overly concerned with the aesthetic claim that it swallows up and thus distorts other and equally valid, if not necessarily equally pressing, claims. Michael Alexander, yet another of the multitude which has at one time or another translated "The Seafarer," nails his colors directly to the flag, dedicating his The Earliest English Poems to Ezra Pound. He also declares, with great passion, "that it is the business of the translator, once he has done his homework and made himself aware of everything there is in his original, to decide which of those elements can be reproduced in the living language and strive to recreate them, ruthlessly ignoring those which can not."18 Alexander's translation, unfortunately, is similarly excessive, bounding about with all the passionate uncontrol of a colt that wants to run all the races in the world at the same time. Strictly speaking, despite his obvious devotion to and love for the poetry, no style at all emerges in his translation, but rather a host of styles and levels of diction, contradicting and struggling with one another, right and left, and with might and main: For no man above mould is so mood-proud, so thoroughly equipped, so quick to do, so strong in his youth, or with so staunch a lord that before seafaring he does not fear a little whither the Lord shall lead him in the end.19 "!hose unsatisfied by my translation," Alexander notes, "should look at Mr Pound's Seafarer to see what le grand translateur of our age has done with it."20I do not wish to be unfair, but Alexander's prose, too, is a pastiche, a jumble: there is no more reason here to break into French than there is, in the translation, to set stiff upperclass expressions like "so thoroughly equipped" against false archaisms

The Translator's Responsibility

169

and would-be neologisms like "no man above mould" or "so quick to do." The whirling first three lines of this translation suddenly peter down, in the final two lines, into Wardour Street prose: the effect is of overwhelming confusion and utter literary uncertainty. It is one thing to be "aesthetic"; it is somethingquite different to waltz,juggle, tumble, and do a muscular ballet leap while at the same time singing a Cockney patter song and playing, with one's nose, upon a harmonica. But the aesthetic focus of a translation can, believe it or not, be still more excessive. Professor Rebsamen's rendering, discussed earlier on pages 124-25, is I think understandable as the aesthetic claim gone mad. No one would argue more strongly than I would for the primacy of the aesthetic claim. But no one has the right to make so much of it that, under its banner, he wantonly rewrites the original. There must be balance in the translator before there can be balance in the translation. It is, to be sure, a difficult and a delicate balance; even an excellent craftsman, successful at one moment, can fail the next. In his Greek Lyrics, something of a pioneering book in its time, the author of the best version of Homer's Iliad,Richmond Lattimore, translates one of the Sapphic fragments like this: But I claim there will be some who remember us when we are goneF1 That is all there is to it, and it is not enough. One does not need to read classical Greek to know that Sappho could never have enjoyed the reputation she has had for almost three millenia had she written such flat, tepid verse. Even the scholarly Henry Wharton, preparing a deliberately literal translation as long ago as 1895, came up with a version distinctly more attractive: Men I think will remember us even hereafterF2 This is not glowing poetry, but it a t least suggests the possibility that the original was in truth glowing poetry. Lattimore's version offers us no such possibility.

170

THE ART OF TRANSLATING POETRY

Willis Barnstone published a version of Sappho which contains the following nicely cadenced but still rather too prosaic version: Someone, I tell you, will remember This too, pleasantly, gently poetic as it is, is simply not good enough. As Dudley Fitts has said, "a reputation like Sappho's can not be wholly the result of accident. The ancient writers who knew her work [that is, all her workl agreed that she was a poet of the first importance." Fitts goes on, in praise of Mary Barnard's versions, to speak of Barnard's "direct purity and versification." He adds that "like the Greek, [her workl is stripped and hard, awkward with the fine awkwardness of You may forget but Let me tell you this: someone in some future time will think of

As I have said of Mary Barnard's version: "Sappho now enjoys as nearly perfect an English translation as one can find, a great translation, an immensely moving translation, complete, beautiful, deserving of endless praise."26Even Guy Davenport, a translator who has done gorgeously well by Archilochus, handles this Sappho fragment in too masculine a fashion, too forthright, even too blatant. The subtlety as well as the delicacy seems to me lost: Someone, I'm bold to say, Will remember us In time hereafter.27 And yet, since all translation is by definition partial, and no translation is or even can be definitive-all are inevitably different mixes; all strive for the glowing balance but, even finding it, are doomed to being equalled or surpassed or, in the end, superseded-not even

The Translator's Responsibility 171

Mary Barnard's as-close-as-one-can-get-to-perfect translations can expect wholly to occupy the field. Nor can they ever satisfy as the original alone can do. "I could hardly expect," Miss Barnard writes, "to reproduce all the virtues of a poem by Sappho in an English translation. The flexibility of Greek allows complicated tense structure and swift movement at the same moment. The ambiguities which enrich her simplest lines, the overtones and undertones, the occasional puns, which are not quite puns and seem right instead of ridiculous, are almost impossible to convey in another language.'% I have myself translated this poem, not in one but in two equally despairing versions, neither of which is of course in any sense definitive: Somebody Listen Somebody Yes Remembered us.

or: Somebody Listen Somebody Yes Will remember

us

Then.29 There is some uncertainty about the Greek text, and that uncertainty underlies the basic difference between these two versions. The search for a balance which, while giving precedence to the aesthetic claim, does not deny other claims, is a perpetual search. There are times when the original seems to yield a translation balance almost effortlessly, as in a little poem by Plato. Here is, first, the plain prose rendering of the Loeb Library, and then the version in Dudley Fitts's too-little-known Poems From the Greek Anthology:

172 THE ART OF TRANSLATING POETRY

I, Lais, whose haughty beauty made mock of Greece, I who once had a swarm of young lovers a t my doors, dedicate my mirror to Aphrodite, since I wish not to look on myself as I am, and cannot look on myself as I once was.30 I Lais whose laughter was scornful in Hellas, Whose doorways were thronged daily with young lovers, I dedicate my mirror to Aphrodite: For I will not see myself as I am now, And can not see myself as once I was.31 And sometimes, as in this two-line epigram by Lucilius, it seems to be necessary to employ a high order of craft, and even some guile, to bring the old poem into a proper balance in its new existence. Here, again, is first the Loeb version and then that by Dudley Fitts: You bought hair, rouge, honey, wax, and teeth. For the same outlay you might have bought a face.32 Darling, a t the Beautician's you buy Your [a] hair [bl complexion [cl lips [dl dimples, & [el teeth. For a like amount you could just as well buy a face.s3

Can one say that Fitts is being untrue to the original in these translations? Not, it seems to me, unless one chooses to ignore the primacy of the aesthetic claim and assume, in my view terribly wrongly, that the merely verbal (lexical) claim is primary. Indeed, in order to emphasize how rare a balance Fitts has achieved, let me juxtapose his translations against those of Andrew Sinclair:

The Translator's Responsibility 173 I am Lais. My pride of face Once laughed at all the Graecian race. At my door, lovers stood ten deepGoddess of Love, my mirror keep. As I am now, I shun my glass, And I cannot look on who I was. You bought hair, rouge, cream, teeth and paste. It'd cost the same to buy a face.g" I need hardly say that forcing delicate verse into artifical, constrained molds (i.e., meter and rhyme) has seriously damaged Sinclair's versions. His own commentary shows that in fad he has not given the aesthetic claim primacy, preferring to torment himself, uselessly, over that will-o'-the-wisp, Greek metrics in English. "The most satisfadory verse translation [of the Greek Anthology] is by Dudley Fitts; but although his style is admirably lucid and brief, his refusal of rhyme and his typographical tricks lose the studied metres of the Anth~logy."~~ But as we have seen, if we want to savor "the studied metres of the Anthology," we have no choice but to go to the Greek. Those studied metres cannot be reproduced in English. And because they cannot, it is to Fitts we must go, rather than to Sinclair, to find what Sinclair aptly calls the "savage and tender, biting and despairing" tone of the original poems.= Sometimes in his Greek Anthology translations Fitts is obligedand again, as I see it the imperative is both accurate and fully justified-to be even less lexically "faithful" than in the two examples just given, in order to keep the balance effective, in order to keep the aesthetic claim functionally primary. Here as before is first the Loeb plain prose rendering and then Fitts's translation: I was thunderstruck when I saw the rhetor Maurus, with a snout like an elephant, emitting a voice that murders on from lips weighing a pound each.37

174

THE ART OF TRANSLATING POETRY Lo, I beheld Mauros, Professor of Public Speaking, Raise high his elephant snout And from between his lips (12 oz. apiece) give vent To a voice whose very sound is accomplished murder.

I was impres~ed.~ Time claims and cultural claims both assume an incredibly diverse set of faces. Details of all sorts-names, places, dates, customshave to be pried like pearls out of their time- and cultureflanguagebound oysters, to be served up in a new shell. But what may seem eternally pearl-like in one context can seem distinctly odd, even grotesque, in another. Professor Donald Keene informs me that some Japanese terms of endearment, as spoken by a man to a woman, involve images of sea-slime, slugs, and the like, all repellent to us but wholly positive, even aphrodisiac, in Japan.39"The bluish tint of his shaved head made him look very attractive," observes Sei Shonagon in her Pillow Booka A blue-tinted head would not much attract a woman in our part of the world or of our time. In Beowulfthe dyinghero wishes to be memorialized by, and buried in, a stun-beorh, a "stone barrow." He asks his relative and heir, Wiglaf, to be sure that his stun-beorh is placed near the seacoast, so that sailors in ships will be sure to see it and remember him. The appropriate archeological translation for stun-beorh is "tumulus": David Wright's prose rendering of Beowulf in fact uses that archeologically correct term.41I do not approve of prose translations of poetry; they seem to me virtually a total abdication of responsibility. But I approve still less of the use in any literary translation of jawcracking, technical terminology like "tumulus." No matter how narrowly accurate in a scholarly sense, such word choices are off-putting. They place obstacles in the reader's path, instead of leading, as they should, ever closer to the true center of the literary experience to be re-created. To thus translate seems to me a distinct over-response to the culturaVhistorical claim, and an insuilicient response to the aesthetic obligation. Edwin Morgan, whose craggy version of Beowulf is almost always

The Translator's Responsibility

175

intelligent, tries to dodge around stun-beorh, translating the term as "burial-mound."42The attempt cannot work, since the object in question has to be significantly larger than a mere '%burial-mound if it is to be seen and recognized by sailors off at sea. Morgan tries to get around this difficulty, too, once again intelligently, and once again I think unsuccessfully, by adding that the "burial-mound" must be "a memorial. . .towering high." But to paraphrase Gertrude Stein, a burial-mound is a burial-mound is a burial-mound. E. Talbot Donaldson turns stun-beorh into "mound," which creates much the same difficulty for the modern reader." A "mound" does not rise high enough from the ground; the modern reader cannot either know or be expected to know that a stun-beorh was in fad a dual earth-and-stone construction, frequently of immense size. There would be a significant tumulus below, and then a great heap of rock above. But the modern reader is a modern reader and not an archeologist. And for him, again, a mound is a mound is a mound. His aesthetic claim is the one to which the translator should pay heed, not the claim of the scholar, or of the archeologist. So too Lucien Dean Pearson translates stun-beorh as "mound," though he also has the dying Beowulf refer to "Beowulf s b a r r ~ w . " ~ "Barrow" is indeed the consensus choice of most translations. My own translation of Beowulf predates the popularity of J. R. R. Tolkien's fiction, and in particular The Lord of the Rings,a trilogy so steeped in lore and history that it has made "barrow" almost a generally acceptable term. I did consider "barrow" as a translation of stun-beorh; in fad, I circulated a small questionnaire to some dozen people of literary taste and judgment, most of them with at least some graduate study of English literature. But roughly threequarters of my respondents conclusively associated "barrow" with the wheeled tool used in the garden and on construction sites. "Barrow," that is, had for the overwhelming majority of them a fatal association (fatal, that is, from the perspective of the literary translator conscious of the aesthetic claim) with "wheel-barrow." Never mind that the lexical claim and the scholarlykistorical claim would still argue for "mound," "barrow," or "tumulus," or perhaps for some combination of these. The aesthetic claim as I understood it argued even more strongly against any and all of these words-and it was, predictably, to the aesthetic claim that I listened. Not that I chose

176 THE ART OF TRANSLATING POETRY

to ignore competing claims: translation, as I have said, is a process of balancing out all such claims. And there had been, earlier in the poem, a similar sort of structure inhabited by a dragon who (in my translation) "slept in a huge stone tower, with a hidden 1 Path beneath."& And that tower offered me what seemed a logical solution, lexically not entirely accurate, nor historically entirely accurate, but having I would argue both the right aesthetic associations and at the same time also meeting the other claims quite sufficiently well: Have The brave Geats build me a tomb, When the funeral flames have burned me, and build it Here, at the water's edge, high On this spit of land, so sailors can see This tower, and remember my name, and call it Beowulf s tower . . .M Two prominent Old English scholars went through the translation with me line by line (very nearly word by word at times) before publication; other scholars have had their say in print. Objections there have of course been to aspects of the translation, but to this day no one has questioned "tower" as a rendering of stun-beorh. There are similar problems with proper names, which in each culture carry large burdens of time and cultural association. I once taught a course entitled "Literature and Culture," a prime text for which was a translation of the Chinese novel now known as The Story of the Stone but then in print as TheDream of the Red Chamber. My undergraduate students were so desperately tied in knots by the unfamiliar Chinese names that, after more than one attempt, I was forced to drop the book from my reading list. Anyone in the Englishspeaking world who has ever taught a Russian novel (and many who have simply read Russian novels) knows how unhappy students (and others) can become dealing with a patronymic device that simply makes no sense to native speakers of English. Pyotr Davidovich makes excellent sense to a Russian, as does Galina Davidovna. (The principle, at least, might seem less forbidding to native speakers of Arabic or Hebrew, where similarly the father's name becomes ine-

The Translator's Responsibility

177

ludably part of his child's name as well: Mohammad bin [son 04 Abdullah, Moshe ben [son ofl Yosef.) Ivan Morris, translator of The Pillow Book of Sei Shonagon, explains that except when it was essential for clarifyingpuns, I have usually not translated proper names. . . . I wished to avoid the false exoticism that can result from identifying the Emperor's residence, for example, as "The Pure and Fresh Palace." Names should not be made to sound more colorful in translation than they do to the reader of the original J a ~ a n e s e . ~ ~

I would add that, obedient to the claims of the aesthetic obligation, names should also not be rendered in any way that distorts their true function, making them either an obstacle or a diversion rather than a simple identificatory label. The great linguist Otto Jespersen, who regularly taught a number of languages, always recognized this problem at once and renamed any student the phonology of whose name would prove a stumbling block in the language being taught. A Norwegian "Olaf," studying French with Jespersen, would be turned into a French-sounding "Pierre." But "Olaf" might become "Jack" if the class was one in English. And so on. But to anyone who places time and cultural claims first, or closer to first than I believe advisable, a failure to reproduce Wealhtheow exactly as one finds it in the Old English of Beowulf, or a failure to let the reader seeDaeghrefn plain and clear, or Ecgtheow, is (to quote one of the two scholars, already mentioned, who did an exhaustive pre-publication review of my Beowulf translation) "a tragic loss, one that I feel as bitterly as I would the irreparable loss of a page of Mozart. We have no right to throw away a name or a page, not one."'8 My scholar-advisor makes his stance truly clear only when he notes that "for the Modern English reader who is familiar with Old English, [name changes] would . . .be a puzzling distraction" (emphasis added).49And there, precisely there, is the rub. The modern reader who is familiar with Old English does not need the translationand, it is hoped, will not for his own purposes make any use of it. If that modern reader, who in sober reality is either an Old English scholar or one of his students (and no one else), were in truth the

178 THE ART OF TRANSLATING POETRY

audience for my or for any other translation of Beowulf, the situation would be puzzling indeed. Observe, too, how effortlessly the objecting scholar identifies the physical destruction of a manuscript page with the alteration, for aesthetic purposes, of an unfamiliar and largely unpronounceable name. But these are phenomena of a very different order. Wealhtheow, after all, is still discernible in "Weltho," which latter name can (a) be recognized as foreign but (b) can readily be managed by a modern English tongue. So too Daeghrefh is discernible in "Dagref," and Ecgtheow in "Edgetho." Indeed, one can see but not hear a spelling change in this latter name-form, one purpose of which is in fad to guide a modern reader, unaware of Old English spelling conventions, into some reasonable facsimile of the actual sound of the original Old English name. A scholar who reviewed the translation, I'm afraid I no longer recall who or where, was just as upset about the name alterations as was the pre-publication reviewer. "Would a reader of the Iliad," he demanded, "like to be confronted with characters named Prime, Akill and Patrokle? I doubt it, nor do I like the change of names in Raffel's version." There are two obvious errors of logic here. First, the characters in Homer are considerably better known than those in Beowulf, even in some cases quasi-proverbial,so that they are likely to have an existence for the modern reader even before he comes to the poem. (One might recall the young student, encountering Shakespeare for the &st time, who was puzzled to find the plays terribly full of quotations. One might also think of the newer romanizations of Chinese, which have hesitated at older, less accurate, but very well-known and long-extant romanizations like Peking.) Of course one does not lightly trifle with long-established and well-known conventions, whether in names or anything else. And second, in spite of what I have just said, there has in fact been just such a shift in established name conventions in the Homeric epics. I grew up, in the period before World War II, with "Ajax" and "Ulysses" and the rest; later, when I had to switch to "Aias" and "Odysseus" I found it somewhat strange indeed. But the switch has largely been made. Indeed, what difference is there to the English reader in either "Ajax" or "Aias," either "Ulysses" or "Odysseus," neither of which is an English name with English associations and significance? The English reader would have been

The Translator's Responsibility

179

perfectly happy to see the characters named Blunderbuss and Cannonball, as long as all of us are reasonably well agreed that that is what they ought to be called. The characters' names do not and cannot change in the original Greek; they will always be whatever they have always been. The translator's incarnation is after all only a transient copy, not in any true sense a replacement. For those who need and use no copy, the translator's incarnation is really not of much importance. A massive and influential poem like Dante's Divinu Cornmedia, discussed above on pages 58-62 and 107-10, raises a great many time and cultural issues, intertwined, inevitably, with a host of complex aesthetic issues as well. T. S. Eliot's 1950 "What Dante Means to Me," less well known than his deservedly famous and enormously influential 1929 essay, in effect sums up the argument I have been making here: The whole study and practice of Dante seems to me to teach that the poet should be the servant of his language, rather than the master of it. This sense of responsibility is one of the marks of the classical poet. . . . Of some great poets, and of some great English poets especially, one can say that they were privileged by their genius to abuse the English lan-

It remains true, nevertheless, as I hinted earlier in this discussion, that in the course of his career the translator is quite likely to change the meaning he attributes to terms like "responsibility" and "abuse." There has not been a great deal of public comment on this subject, I cannot be sure that any degree of generality truly applies, particularly when the experience I know best is of course only my own. John Frederick Nims does say that "perhaps the translator should not be too aware of his motives; this may be one of the pleasures frustrated by self-consciousness." But Nims also goes on to say, revealingly: "I am not sure that the enjoyment I had in doing these translations is very different from that of writing 'one's own' poetry."61 The relationship of work translated and the sometime translator's own original work is a subtle and difficult one. Nims, for example,

180 THE ART OF TRANSLATING POETRY

may be making more of a statement than he intends by using quotation marks around "one's own." Vasily Andreyevitch Zhukovsky, Pushkin's immediate predecessor as Russia's leading poet, was in fact more a translator than a poet. The flow of his inspiration was slim and irregular; a t his best he was distinctly more of a craftsman than, like Pushkin, a veritable fountainhead of ideas and emotions. In 1820, after the publication of the twenty-one-year-old Pushkin's Ruslan and Ludmilla, Zhukovsky sent him a portrait, inscribed 'To a victorious pupil from a defeated master." And Zhukovsky declared, flatly, that "the translators of the poets are the rivals of the poets themselve~."~~ A poet who has trouble creating his own work, and who is sharply aware of competing both with other poets and with the very poets he translates, is apt to have a different approach to his responsibilities as a translator than someone whose own writing flows more freely. A more typical stance for Russian poets, in particular, is taken by Nikolai Stepanovitch Gumilev-a poet who had no trouble creating his own work: Clearly . . . the translator of poetry must himself be a poetand also a careful investigator and an honest, sensitive critic, who can choose an author's basic characteristics and, when he needs to, can sacrifice to these characteristics others of lesser importance. And he must forget his own personality, must think only of the author's personality. Ideally, translations should be presented unsigned.6s Nims takes a distinctly less inflexible approach, far more typical of American translators of poetry: One cannot translate a poem, but one can try to reconstitute it by taking the thought, the imagery, the rhythm, the sound, the qualities of diction-these and whatever else made up the original-and then attempt to rework as many as possible into a poem in English. Since no translator can manage equally all such data a t the same time, with so many conflicting claims to be reconciled, what he has to do is set up a constantly shifting system of priorities: now the thought has to be flexed into a rhythm, now modulated into another key of sound, now

The Translator's Responsibility 181 an image has to be refocused, now some clue given to a lost allusi~n.~ My own experience-and let me repeat that I simply do not know if I am typical, atypical, or neither-has been that the tension between my own primary work and my translations has been largely responsible for whatever changes I am aware of in my sense of responsibility as a translator. That is, as I have progressed in my own primary work, both poetry and fiction, and as I have grown more confident and clear about what I was doing, I have grown more careful to observe the distinction between original and translated work. I have, bluntly, felt less need to "appropriate" what I was translating. Ezra Pound, forever unsure about his primary work, was an appropriative translator to the day he died. As I have said elsewhere, Pound "goes to an original because it has something he wants. . . .The translation process is, for Pound, appropriativerather In Zhukovthan in the more usual scholarly sense transmissi~e."~~ sky's terms, I have felt less need to "rival" the poets I was translating. This is not to claim that I have become a better translator, but only a different one. I do not think that greater deference to a text necessarily makes one a better translator; I most emphatically do not think, for all the reasons discussed in part one of this book, that greater concern with what scholars call the "letter" of a text makes one a better translator. Again, I know of no better translation from Homer than that by Christopher Logue, discussed earlier at pages 121-24. It is also of interest that the standard Russian translation of Homer's Odyssey, in the f i s t half of the nineteenth century, was prepared by Vasily Andreyevitch Zhukovsky, who was as innocent of Greek as I am told Logue is.S6 In any case, though Gurnilev pushes his view of translation much too hard, he is without a doubt correct when he asserts that "clearly . . . , the translator of poetry must be himself a poet." If this is a requirement, as I have no doubt it is, it is also a danger. A good case could be made, for example, for Ezra Pound as a distinguished and highly unusual form of a literary phenomenon not a t all rare or unusual, namely, the poet manqu6. Pound wrote many fine poems of his own; just as plainly, he never achieved what he wanted to achieve as an original poet. Much of his long, late work, the Cantos,

182 THE ART OF TRANSLATING POETRY

is either translation, adaptation, or outright borrowing-not the sort of practice, despite the involved, tedious arguments put out by Pound's flock of adoring critics, that a secure poet well past middle age indulges in. These same critics think Pound was incompetent, at least in this regard, when toward the end of his life he declared that "everything I touch, I spoil. I have blundered always . . . my poems don't make sense."67But still another way of looking at Pound is that all his life he was strictly a translator, a man who brought pearls back to us out of the past, pearls from distant shores-but always and inevitably other people's pearls, not his own. The connection between the translator's habit of mind and a failure as an original poet is a real one, no matter how difficult it may be to trace out all the links. For despite Nims's inability to see a difference between translating and creating, a difference there is. The translator must be a poet: there should be no argument about that. And the translation he writes is his, not the original poet's: even the laws of copyright recognize that plain fact, though the critics have not fully acknowledged it. But the poet creates, in a sense, ex nihilo: he starts only with himself and a blank page. The approach of what he writes, its tone, its organization, its craft, and its ultimate thrust and meaning and significanceare the poet's-all of them. The translator of a poem is only a transient mediator-for as I have suggested, in time all translations fall away, fade, become literary curiosities, time-bound and largely of scholarly rather than literary interest. The great Elizabethan translations, Pope's splendid versions of Homer, and virtually all the nineteenth-century's many translations of many, many literary masterpieces are no longer read. But the originals of all those translations, in newer versions more appropriate to our own time, continue to be read, exactly as they always have been. And they will continue to continue, while in their turn those newer translations-my own of course among them-will go on dating and aging and will in the end become what all translations inevitably turn into, museum pieces. It is a process that can make a dedicated translator enormously sad, if he happens to live to see it happen. Arthur Waley, one of the greatest of translators, did so live, and was so saddened. In an essay collected in his little-known volume The Secret History of the Mon-

The Translator's Responsibility

183

gols, he first quite critically examines a passage from his classic translation of Lady Murasaki's great Japanese novel Genji, and then goes on: There is not any other translation of the passage with which I can compare mine [a statement no longer true]. If there were, I might suddenly feel that I did, after all, make rather a mess of it. In saying this I have in mind a passage in the No play Sotoba Komachi: Oh how fell she from splendour, How came the white of winter To crown her head? Where are gone the lovely locks, double-twined, The coils of jet? Lank whisps, scant curls wither now On wilted flesh, And twin-arches, moth-brows, tinge no more With the hue of far hills. That was the way I translated it in 1921, and it is not bad verse. But I must confess that when recently I read Sam Houston Brock's translation of Sotoba Komuchi in Donald Keene's anthology, I was rather shaken. His translation of this passage is: How was ever such loveliness lost? When did she change? Her hair a tangle of frosted glass Where the black curls lay in her neck And the colour lost from the twin arched peaks Of her brow. I felt at once that my translation was hopelessly overladen and wordy and that it tried in a quite unwarrantable way to improve upon the original.68 This is both candid and, alas, accurate. I myself grew up on Waley's translations, but I have had to give up on his books for use with today's students. They can no more empathize with his Edwardian approach and assumptions than they can with the equally dated,

184 THE ART OF TRANSLATING POETRY

though never quite so glowing, verse of Laurence Binyon's Dante translation. Pound's sadness at the end of his long life seems to me distinctly similar to Waley's. It is something like the inevitable sadness which afflicts the pianists (and there have been many of them) who would infinitely rather have been composers, the art teacher who would rather have been a painter or a sculptor. Waley too had ambitions, as a writer of both prose and, in particular, of poetry, but they never came to a fraction of Pound's accomplishment. That sort of sadness, that sort of abiding frustration, cannot help but affect one's performance in any secondary role to which one finds oneself consigned. Consider Pound's version of the first stanza of a neat, nasty Heine lyric: The mutilated choir boys When I begin to sing Complain about the awful noise And call my voice too thick a thing.59 This is wonderfully deft, tactful, funny. But it turns out, when we look at the German, that the two most notable touches, "mutilated" in line 1and the immensely skilful phallic pun, "too thick a thing," in line 4, are entirely Pound's work, not Heine's at all:

Doch die Kastraten klagten, Als ich meine Stimm erhob; Sie klagten und sie sagten: Ich sange vie1 zu grob. ("But the castrati complained When I raised my voice; They complained and they said: I sang much too coarsely.") Louis Untermeyer's version is in fact much closer to Heine's German: And still the eunuchs grumbled, Whene'er my voice arose; They grumbled as they mumbled My songs were far too gross.@

The Translator's Responsibility

185

This is very much closer to the German-and very much worse as both poetry and as translation. And the reason is obvious. It is not that Untermeyer has worked improperly, conceived the nature of translation wrongly. It is not that he has failed to assess the aesthetic claim properly. Quite simply, the reason for the comparative poorness of his translation is that Untermeyer is not the poet-or the poet manque, for Untermeyer too had creative ambitions-that Ezra Pound is. Pound's version plainly does not match the German anything like as well on the lexical level. But as poetry, and as a recreation in modern English of something very near to what we can imagine Heine writing if he had lived in 1920and been an American, Pound's version is incomparably superior. With no access to the German, and only Untermeyer's translation in front of us, we might imagine Heine to be a mildly risque bore. Bluntly, there is far more of Heine in Pound's "liberties" than in Untermeyer's "faithfulness." For as Nims has said, "The greatest infidelity is to pass off a bad poem in English as representing a good one in another language. . . . Modifications of imagery may be necessary in view of differing linguistic habits: what we want in the translation of poetry is equivalent effect, which we hardly get by writing English as if it were a foreign language."61 Or as William Arrowsmith puts it, "There are times-far more frequent than most scholars suppose-when the worst possible treachery is the simple-minded faith in 'accuracy' and literal loyalty to the original."'j2 Not many people know that D. H. Lawrence was a fairly active translator. However, he admitted to only a very little of his translation work, on the ostensible grounds that it would damage his reputation as a creative writer. That is, he was worried, since he had to live by his pen, that if it became known that he did translations, editors and publishers would be less likely to print his own writing, and would probably pay less for it when they did print it. These are I think real fears; the ostensible reason is thus a logical and even a sufiicient one. But I also wonder if there wasn't present in Lawrence's mind, or somewhere in his unconscious, some awareness of the danger of being swallowed, of being engulfed, perhaps even consumed by the powerful writers a translator must necessarily confront? Not only that: What happens to the writer's own growth if he lives too long in

186 THE ART OF TRANSLATING POETRY

another writer's universe? It can be, let me assure you, a fearsome experience. When after working for fifteen years with the great Indonesian poet Chairil Anwar, I decided to translate the complete surviving corpus of his work, and even edit the Indonesian text of his poetry (which had to that date never been printed in one place, not even in Indonesia), the work had a distinct effect on me, not always in pleasant ways. Anwar was a very great poet; he was also a madman, reckless, unreliable, fiercely egocentric, irresponsible, endlessly aggressive. When he died in 1949,just short of his twentyseventh birthday, he was suffering from syphillis, typhus, tuberculosis, cirrhosis of the liver-and he died, in fact, of a blocked intestine. When a t long last I finished the manuscript and mailed it off to the publisher, I immediately sat down and wrote the following poem. Not a translation, please note, but a poem of my own, expressing my own very clear feelings about what I had just been through: Having Anwar out of the House Louse: Like a drunken house-guest you nailed My ass to my chair, chewed My ear as my eyes faded away; You haunted the bathroom, hung Under my bed, slopped In my coffee, pissed In my beer. Bastard: this morning I packed you off, mailed you out. Let the printer sweat your guts, Let the world worry: You're out of my house At last.'= The translator's final responsibility is to his author. The writer's final responsibility must be to himself.

Notes

Chapter 1 1. Provincial Latin of course evolved into a number of different (Romance) languages, depending on the particular province and its own subsequent history. In Spain, for example, it is clear that "the structure of Spanish . . . is in this respect totally an importation. . . . Spanish is the product of the evolution of the Latin of daily intercourse." Robert K. Spaulding, How Spanish Grew (Berkeley: University of California Press, 1943), 33. 2. Benjamin Lee Whorf,Languuge, Thought,andReality (New York: Wiley, 1956), 252, 257. 3. Donald Keene, Japanese Litemture (New York: Grove, 1955), 4-6. 4. Ibid., 7, 26. 5. Geoffrey Bownas and Anthony Thwaite, The Penguin Book of Japanese Verse (Harmondsworth: Penguin Books, 1964), xliii. 6. Ibid., xlv. 7. Arthur Waley, Japanese Poetry: The 'Uta' (Oxford: Clarendon Press, 1919; reprint ed., London: Lund, Humphries, 1959), 9, 13. 8. James J. Y. Liu, The Art of Chinese Poetry (London: Routledge and Kegan Paul, 1962),40. 9. hid., 4 0 4 1 , 4 5 4 6 . 10. A. V. Gronicka and H. Bates-Yakobson, Essentials of Russian, 4th ed. (Englewood Cliffs, NJ.: Prentice-Hall, 19641, 141-42. 11. Erich Auerbach, ZnWuctwn to Romance Languages and Literature, trans. Guy Daniels (New York: Capricorn, 1961),68-69. 12. Whorf, Language, Thought, and Reality, 265. 13. Thomas Mann, "Letter to a Translator," Delos 4 (1970): 211. 14. "Verse, whatever else it is, is a manipulation of linguistic givens." W. K.

188 Notes Wimsatt, "Foreword," in Versification:Mqjor L a m e Types, ed. W. K . Wimsatt (New York: New York University Press, 1972), xl. 15. Leonard Bloomfield,Language (New York: Holt, Rinehart and Winston, 1933), 127. 16. "There are published descriptions of the auditory quality of the vowels in a

large number of languages. For some of these languages, sets of acoustic measurements are also available." Peter Ladefoged, A Course in Phonetics, 2nd ed. (New York: h u r t Brace Jovanovich, 19821,204. 17. Robert A. Hall, Jr., Introductory Linguistics (Philadelphia: Chilton, 1964),37. 18. Noam Chomsky, Syntactic Structures (The Hague: Mouton, 1957),11. 19. Anne Cluysenaar, Aspects of Litemry Stylistics (New York: St. Martin's, 1976), 21. 20. W. Nelson Francis, "Revolution in Grammar," in Introductory Language Essays, ed. Dudley Bailey (New York: Norton, 19651, 282. 21. Paul A. Oleon, “Transformations," in Bailey, Introductory Language Essays, 177-78. 22. Winfred P. Lehmann, Language (New York: Random House, 1983), 114. See also C. F. Hockett, The View From Language (Athens: University of Georgia Press, 1977), 12: "Language is the most typically human of all man's sign-systems, and

unquestionably the most elaborate sign-system to be found, either among humans or elsewhere." 23. Angus McIntosh and M. A. K. Halliday, Patterns of Language (Bloomington: Indiana University Press, 1967), 5, 14. 24. Hockett, The View From Language, 13, 55. 25. Hall, Introductory Linguistics, 406. 26. Hockett, The View From Language, 119,262. 27. "The sonnet, in its gradual changes through history, could serve as an (ideologically conservative) model for a disquisition on 'poetry as a cultural system' or as a fascinating instance of a poetic form limited to and deeply ingrained in a linguistically demarcated area." Paul Friedrich, The Langwge Pamllax (Austin: University of Texas Press, 1986),84. 28. Burton Raffel, The Development of Modern Indonesian Poetry (Albany: State University of New York Press, 1967),35. 29. Robert Payne, "On the Impossibility of Translation," in The World of Tmnslation (New York: P.E.N. American Center, 1971),361. 30. Hall, Introductory Linguistics, 64,406-7. 31. James Halporn, Martin Ostwald, and Thomas Rosenmeyer, The Metres of Greek and Latin Poetry (London: Methuen, 1963), 4,59. 32. Maurice Bowra, "Foreword," in Bacchylides: Complete Poems, trans. Robert Fagles (New Haven: Yale University Press, 19611, xiii.

Chapter 2 1. Northrop Frye, "Introduction: Lexis and Melos," in Sound and Poetry, ed. N. Frye (New York: Columbia University Press, 1957), xiii, xxv, xxvii.

Notes

189

2. Craig La DriGre, "Structure, Sound, and Meaning," in Frye, Sound andpoetry, 89.

Liu, The Art of Chinese Poetry, 22. La Drihre, "Structure, Sound, and Meaning," 94. An& Lefevere, Tmnslating Poetry ( A m e t e b Van Gorcum, 1975), 20. Ezra Pound, Tmnslations (New York: New Directions, 1954), 207. Catullw, trans. Celia and Louis Zukofsky (London: Cape Goliard, 1969) [unpaginated]. 8. The Poems of Catullus, trans. Horace Gregory (New York: Grove, 1956), 166. 9. Lefevere, Tmnslating Poetry, 26. 10. Lines 2201-5, spelling normalized. 11. Sir Gawain and the Green Knight, trans. B. M e 1 (New York: Mentor, 1970). 3. 4. 5. 6. 7.

116. 12. The Complete Poetry and Prose of Chairil Anwar, ed. and trans. B. Raffel (Albany: State University of New York Press, 19701, 125. 13. Julia C. Lin, Modern Chinese Poetry: An Introduction (Seattle: University of Washington Press, 1972), 7. 14. LiPo and Tu Fu, trans. Arthur Cooper (Harmondsworth: Penguin Books, 19731, 91. 15. Liu, The Art of Chinese Poetry, 38. 16. From Zhou to Mao, trans. Ding Zuxin and Burton M e l , unpublished. 17. Stephen Owen, Traditional Chinese Poetry and Poetics (Madison: University of Wisconsin Press, 19851, 121. 18. Witter Bynner, The Jade Mountain (New York: Knopf, 19291, 148. 19. Robert Kotewall and Norman L. Smith, The Penguin Book of Chinese Verse (Harmondsworth: Penguin Books, 19621, 15. 20. David Hawkes, A Little Primer of Tu Fu (Oxford: Oxford University Press, 19671, 32. 21. bid., ix. 22. Owen, Traditional Chinese Poetry and Poetics, 121, 124, 125. 23. V. V. Levik, "La traduction et la crbation liMraires," in The Nature of Tmnslation, ed. James S Holmes (The Hague: Mouton, 19701,165, 166,168. 24. Roman Jakobson, "On Linguistic Aspecta of Translation," in On Tmnslation, ed. Reuben A. Brower (Cambridge: Harvard University Press, 19591,235-36. 25. Quoted in Hall, Zntmductory Linguistics, 15. 26. The Poems of Saint John of the Cross, rev. ed., trans. John Frederick Nims (New York: Grove, 19681,25. Four Humours (Calcutta Writers Workshop, 19791, 76. 27. Burton -el, 28. Poems of the Late Tang, trans. A. C. Graham (Harmondsworth:Penguin Books, 19651, 22. 29. bid., 24. 30. Bynner, The Jade Mountain, 169. 31. Cooper, Li Po and Tu Fu, 167. 32. The White Pony, ed. Robert Payne (New York: Mentor in.d.11, 193. 33. "The giR of terseness is the least dispeneible literary qualificationof a translator from Chinese." Graham, Poems of the Late Tang, 19. 34. Gems of ChinesePoetry,ed. and trans. Ding Zuxin and Burton M e 1 (Liaoning, China: Liaoning University Press [1986]), 39. 35. Payne, The White Pony, 189.

190 Notes 36. Cooper, Li Po and Tu Fu, 171. 37. Old Friend From Far Away, trans. C. W. Kwock and Vincent McHugh (San Francisco: North Point, 1980). 77. 38. Graham, Poems of the Late T'ang, 24.In a private communication, the translator Ronnie S. Apter speculates "that one might wish to translate in order to stretch English-to bring into it new ideas and new ways of using words, even new grammar!' To the extent that such things might prove to be linguistically possible, the p d u r e might well be of use. My own suspicion, however, is that language is far too exclusively a societal cultural expreeeion-that is, a group phenomenon-for it to be amenable to any individual's efforts. 39. Robert A. Hall, Jr., Sound and Spelling in English (Philadelphia: Chilton, 1961),2. 40. Bloomfield, h g u a g e , 196. 41. Raffel, The Complete Poetry and Prose of Chairil Anwar, 21. 42. The Complete Poems of Chairil Anwar, ed. and trans. Liaw Yock Fang (Singapore: University Education Press, 1970),24. 43. ThePoetry and Prose ofHeinrich Heine, ed.Frederick Ewen (New York: Citadel, 19481,69. 44. The CompletePoems ofHeinrich Heine, trans. Hal Draper (Boston: Sukhrkampl Insel, 1982). 52. 45. Aa reprinted in ThePortable Dante, ed. Paolo Milano (New York: Viking, 1947). 3-4. 46. Dante Alighieri, The Inferno, trans. J. Ciardi (New York: Mentor, 1954), 28. 47. Dante's Inferno, trans. M. Mum (Bloomington: Indiana University Press, 1971),1. 48. The Divine Comedy of Dante Alighieri: The Inferno, trans. A. Mandelbaum (Berkeley: University of California Press, 1980). 2. 49. Unpublished.

Chapter 3 1. Friedrich, The Language Parallax, 86,88,95,96. 2. Paul Fussell, Poetic Meter and PoeticForm, rev. ed.(New York: Random, 1979), 109,126. 3. Keene, Japanese Litemture, 9,10,11-12. 4. Bownae and Thwaite, The Penguin Book of Japanese Verse, xlviii. 5. Anthology of Japanese Litemture: Earliest E m to Mid-Nineteenth Century, ed. Donald Keene (New York: Grove, 1955), 22. 6. Waley, Japanese Poetry, 5. 7. The Monkey's StrawRaincoat, trans. Earl Miner and Hiroki Odagiri (Princeton: Princeton University Press, 1981), 11. 8. Keene, Japanese Litemture, 33,36,37. 9. Earl Miner, Japanese Linked Poetry (Princeton: Princeton University Press, 1979), vii-viii.

Notes

191

10. bid., ix. 11. Back Roads to Far Towns, trans. Cid Corman and Kamaike Susumu (New York: Grossman, 19681, 10. 12. Miner, Japanese Linked Poetry, x. 13. Basho, The Narrow Road to the Deep North, trans. Nobuyuki Yuasa (Harmondsworth: Penguin Books, 1966), 49. See also The Year ofMyLife, trans. Nobuyuki Yuasa (Berkeley: University of California Press, 1960),32. 14. The Genial Seed, trans. Frank Hoff (New York: Groaaman, 1971), 29. 15. Miner, Japanese Linked Poetry, 14. 16. Miner and Odagiri, The Monkey's Straw Raincoat, 49, 51. 17. Ibid., 48. 18. Kenneth Rexroth, One Hundred Poems From the Japanese @JewYork: New Directions, 19561, xiii-xiv, xix. 19. Raffel, The Development of Modern Indonesian Poetry, 14. 20. The Poems of Saint John of the Cross, trans. Willis Barnstone (Bloomington: Indiana University Press, 1968; reprint ed., New York: New Directions, 19721, 21. 21. Ibid., 59. 22. St John of the Cross: Poems. trans. Roy Campbell (London: Harvill, 1951; reprint ed., Harmondsworth:Penguin Boob, 1960), 47. 23. Nims, The Poems of Saint John of the Cross, 25. 24. W e l , Four Humours, 76.

Chapter 4 1. Henri Meschonnic, "Translating Biblical Rhythm," in BiblicalPatterns in Modern Litemture, ed. David Hirsch, Brown Judaic Studies 77 (Decatur, Ga.: Scholars Press, 1984),229,237. 2. Lois Bragg, "Whale-Roadsand Mead-Seats: Four Ways of TranslatingBeowulfn Humanities Education 3:3 (Sept. 1986): 71. 3. Russian Poetry Under the Tsars, ed. and trans. Burton Raffel (Albany: State University of New York Press, 1971), xvii. 4. The approximately century-long metrical morass into which English (as contrasted to Scots) poets fell after Chaucer's death remaim inexplicable. But the new tradition which Chaucar had solidified thrived in Scotland at the same time as it seame to have been lost sight of in England. Early in the sixteenth century Chaucer's p d y once again began to thrive in England. 5. Cent Deur Pdmes Zndodsiens, trans. Louis-Charles Damais (Paris: AdrienMaisonneuve, 1965), 91. 6. Charles Baudelaire, The Flowers of Evil, ed. Marthiel and Jackson Mathews, rev. ed. (New York: New Directions, 1963), 46. 7. Burton W e l , The Forked Tongue: A Study of the Tmnslotion Process (The Hague: Mouton, 1971), 17-19. 8. Martin h e l l , Baudeloire (New York: New Directions [n.d.]), 32. h e l l

192

Notes

adds: "His versification and syntax are in the main traditional. They provide a discipline, but they are at the same time the vehicle of a new vision." bid. 9. bid., 119. 10. Charles Baudelaire, Selected Poems, trans. Joanna Richardson (Harmondsworth: Penguin Books, 1975), 81. 11. Poems: An Anthology, ed. Burton Raffel (New York: Mentor, 1971),170. 12. Raffel, The Forked Tongue, 19. 13. Much the same point is made by James C. Dana, "Rilke's 'Spanish Dance': A Translator's Commentary," The Denver Quarterly 12:l (Spring 1977): 160: "I was a little less exacting in the meter, especially the metrical length of the line. I used a basic pentameter, but allowed four or six feet!' 14. Rainer Maria Rilke, Selected Poems, trans. J. B. Leishman (Harmondsworth: Penguin Books, 1964),25. 15. Kate Flores, "Autumn's Day," inAn Anthology ofGermunPoetryFmm HdMerin to Rilke, ed. Angel Flores (Garden City, N.Y.: Doubleday, 19601,388. 16. Raffel, Four Humours, 61. ~ Agung, 17. H. B. Jaasin, Chairil Anwar, Pelopor Angkatan 46 ( J a l t Gunung 1959), 90.

Chapter 5 1. Robert H. Eli-, Theodore Dreiser (New York: Knopf, 1948;reprint ed., Ithaca: Cornell University Pre8.9, 19701,181. 2. Norbert Wiener, Cybernetics: The Human Use of Human Beings (Boston: Houghton Mifflin, 1954;reprint ed., New York: Avon, 1967), 19. 3. James D. Watson, The Doubk Helk (New York: Atheneum, 1968;reprint ed., New York: Signet, 1969), ix-x. 4. Andree Conrad, '%Editor, Author, Translator," in Translation 74 (New York: P.E.N. American Center, 1974),88. 5. Ibid., 93. 6. Denver Lindley, "The Editor's Problem," in The Craft and Context of TransIrrtion, ed. William Arrowsmith and Roger Shattuck (Austin: University of Texas Press, 1961),159. 7. Harold Clurman, Zbsen (New York: Collier, 1977), 90. 8. Henrik lbaen, Letters and Speeches, ed. Evert Sprinchorn (New York: Hill and Wang, 1964), 144--45. 9. bid., 217-18. The letter is dated 25 May 1883. 10. Grace Hegger Lewis, With Low Fmm Gmcie (New York: Harcourt, Brace, 1955), 84,8-9, 15,21. 11. Fram K&a, ThePenul Colony:Stories and ShortPieces (New York: Schocken, 1948),184. 12. Arthur Waley, The Secret History of the Mongols and Other Pieces (London: Allen and Unwin, 19631,294.

Notes

193

13. Jaroslav Prnsek, ed.,Dictionary of Oriental Litemtures, vol. 3,West Asia and North Africa, ed. Jiri Becka (New York: Basic, 19741,161a. 14. The Rubaiyat of Omar Khayam and Other Persian Poems, ed. A. J. Arberry (London: Dent, 1954),33. 15. Open Secret: Versions ofRumi, trans. John Moyne and Coleman Barks (Putney, Vermont: Threshold Books, 1984), 65. 16. Mystical Poems of Rumi, trans. A. J. Arberry (Chicago: University of Chicago Press, 19681,106. 17. Night and Sleep, trans. Coleman Barks, John Moyne, and Robert Bly (Cambridge, Maas.: Yellow Moon Press, 1981)[unpaginatedl. 18. Ibid. 19. Raffel, The Forked Tongue, 68-69. 20. Musa, Dante's Inferno, xi. 21. Ciardi, The Inferno, 28. 22. Sidney Alexander, "On Translating From Renaissance Italian," in The World of Tmnslation (New York: P.E.N. American Center, 1971), 16. 23. Anthology of Troubadour Lyric Poetry, ed. and trans. Alan R. Press (Edinburgh: Edinburgh University Press, 1971), 2,3. 24. Ibid., 173, 174. 25. Ibid., 189. 26. Ibid., 175. 27. Lyrics of the Troubadours and Trou&res, ed.and trans. Frederick Goldin (New York: DoubledayIAnchor, 19731,209,220n. 1. 28. Ibid., 221. 29. Pound, Tmnslations, 425. 30. Ibid., 11-12. 31. Burton -el, Ezm Pound: Prime Minister of Poetry (Hamden, Conn.: Archon, 1984),62-63. 32. Burton Raffel and Vincent J. Cleary, Why Re-Create? (SanFrancisco: Chandler and Sharp, 1973), 10. 33. Rowland L. Collins, "Introduction," in Beowulf, trans. Lucien Dean Pearson (Bloomington: Indiana University Press, 1965), 27. 34. Ibid., 35. 35. Stanley B. Greenfield,A Readable Beowulf (Carbondale:Southern Illinois University Press, 1982), 29. 36. Ibid., 37. 37. Beowulf and Sir Gamin and the Green Knight, trans. Gordon Hall Gerould (New York: Ronald Press, 19351,10. 38. Ibid. 39. J. R. R. Tolkien, "Prefatory Remarks," in Beowulf. . . ,trans. John R. Clark Hall, new ed. (London: Allen and Unwin, 19501,ix-x. 40. Beowulf, trans. E. Talbot Donaldson, ed. Joseph F. Tuso (New York: Norton, 1975),XV-xvi. 41. Beowulf, trans. Burton -el (New York: Mentor, 19631,xxi. 42. Ibid., 23. 43. Homer, The Iliad, trans. Richmond Lattimore (Chicago: University of Chicago Press, 1951), 331. 44. Christopher Logue, Patrocleia of Homer (Ann Arbor: University of Michigan Press, 1963),11-12.

194

Notes

45. Ibid., 52,53. 46. Frederick R.Rebsamen, BeowulfIs My Name (San Francisco: Rinehart, 1971), 3-4. 47. Ibid., x. 48. Ibid., xi. 49. Ibid., xii. 50. Robert Lowell, Imitations (New York: Noonday, 1962),xi-xiii. 51. Ibid., 148. 52. The Penguin Book of Russian Verse, ed. Dimitri Obolensky (Harmondsworth: Penguin Books, 19621,335-36. 53. Boris Pasternak, Fi& Poems, trans. Lydia Pastemak Slater (London: Allen and Unwin, 1963),57. 54. Lowell, Imitations, xii. 55. Burton Raffel, Robert Lowell (New York: Ungar, 1981),113,122.

Chapter 6 1. Bownas and Thwaite, The Penguin Book of Japanese Verse, Ixxii. 2. Ibid., lxxiii-lxxiv. 3. Kwock and McHugh, Old Friend From Far Away, jacket copy. 4. Ibid., 184-85. 5. Ibid., 188. 6. Ibid., 193,195. 7. Osip Mandelstam, Selected Poems, trans. Clarence Brown and W. S. Merwin (Oxford: Oxford University Press, 1973;reprinted., Harmondeworth Penguin Books, 1977),14-15. 8. Olga Carlisle, Poets on Street Comers (New York: Random, 1968,reprint ed., New York: Vintage, 19701,xiv-xv. 9. Suzanne Maasie, The Living M i m r (Garden City, N.Y.: Doubleday, 1972), unpaginated note. 10. Beowulf, ed. A.J. Wyatt (Cambridge: Cambridge University Press,18941,xiii. 11. Beowulf, trans. Edwin Morgan (Lyrnpne, Kent, England: Hand and Flower Press;reprint ed., Berkeley: University of California Press,19621,vii. 12. Philip Henderson, William Monis (London: Thamea and Hudson, 1967;reprint ed., Harmonhorth: Penguin Books, 1973), 402. 13. George Seferis, Collected Poems, trans. Edmund Keeley and Philip Sherrard (Princeton: Princeton University Press, 1967);C. P. Cavafy, Selected Poems (Princeton: Princeton University Press, 1972). 14. Jorge Luis Borges, Selected Poem, 1923-1967, ed. Norman Thomas Di Giovanni (New York: Delacorte, 19721,d-xxii. 15. A Thousand Years of VietnumesePoetry,ed. Nguyen Ngoc Bich, trans.Nguyen Ngoc Bich with Burton Raffel and W. S. Merwin (New York: Knopf, 1975);From the Vietnumese, trans. B. Raffel (New York: October House, 1968).

Notes

195

16. Yehuda Amichai, "Elegy for a Lost Boy," trans. Burton M e 1 and Noam Flinker, Midstream 155 (May 1969): 43-47; Yehuda Amichai, "The Day Martin Buber Died: A Radio Play," trans. Noam Flinker and Burton Raffel, Midstream 16:s (Oct.1970): 20-31. 17. The Tale oflgofs Men, trans. Sidney Monas and Burton M e l , Delos 6 (1971): 5-15; Nikolai S. Gumilev, Selected Works, trans. Burton M e 1 and Alla Burago (Albany: State University of New York Press, 1972);Osip Emilievich Mandelstam, Complete Poetry,trans. Burton M e 1 and Alla Burago (Albany: State University of New York Press, 1973). 18. Raffel, Russian Poetry Under the Tsars. 19. Gems of ChinesePoetry,ed. and trans. Ding Zuxin and Burton M e 1 (Liaoning, China: Liaoning University Press, 1986;U.S. ed., Hamden, Conn.: Archon, 1987).

Chapter 7 1. The Living Talmud, ed. and trans. Judah Goldin (New Haven, Conn.: Yale University Press, 1955;reprint ed., New York: Mentorm, 1957),24,24n. 2. bid., 22.See Robert Kellogg, "Literature, Nonliterature, and Oral Tradition," New Litemry History 8:1(1977):531-34. See also Gary Snyder, "Foreword," in Songs of Gods, Songs of Humans, trans. Donald L. Phillipi (San Francisco: Northpoint, 1982),vii: "In completely pre-literate society the oral tradition is not memorized but remembered." 3. Marc Slonim, The Epic ofRussian Litemture: From Its Origins Through Tolstoy (New York: Oxford University Press, 1950),10.See also Patricia Arant, "Formulaic Style and the Russian Bylina," Indiana Slavic Studies 4 (1967):7-51,and "Concurrence of Patterns in the Russian Bylina," Journal of the Folklore Institute (1970):8088. 4. SeeJohn M. Foley, "Literary Art and Oral Tradition in Old English and Serbian Poetry," Anglo-Saxon England (1983):183-214. 5. Dmitri S. Mirsky, A History of Russian Literature (New York: Knopf, 1949), 14. 6. ,The Song of Igor's Campaign, trans. Vladimir Nabokov (New York: Vintage, 19601,6. 7. Thais Lindstrom, A Concise History of Russian Literature, vol. 1, From the Beginnings to Chekhov (New York: New York University Press, 1966),11. 8. Obolensky, The Penguin Book of Russian Verse, xxxii. 9. Medieval Russia's Epics, Chronicles, and Tales, ed. and trans. Serge A. Zenkovsky (New York: Dutton, 1963). 137-38. 10. Mirsky, A History of Russian Litemture, 15. 11. Nabokov, The Song of Zgofs Campaign, 29. 12. Hall, Beowulf. . ., xviii. 13. Zenkovsky, Medieval Russia's Epics, 139. 14. Monas and M e l , The Tale of Igofs Men, 5.

196 Notes 15. Let me expressly exclude the intelligent and obviously singable translations of Mark Herman and Ronnie S. Apter. Professor Apter, who is also the author of Digging for the Treasure: Translation After Pound (New York: Lang, 1984,observes in a private communication that while the translator of poetry "may choose to be tied to a syllable-for-syllable translation, or a stress-for-stress translation, he may choose to be free of both. The translator of lyrics [meant to be sung] has no such freedom. He must translate syllable-for-syllable,stress-for-stress (although the stress may be ordained by the music, rather than by the original words). He must crest meaning where the melodic line crests. Also, he muat . . . [ask] can this syllable be held for two beats without sounding silly? Can the tenor get off this syllable in the space of an eighth note and take a catch breath?" 16. See Donald K. Fry, "Caedmon as a Formulaic Poet," in Oral Literature: Sewn Essays, ed. Joseph J. Duggan, Forum for Modern Language Studies 10 (Edinburgh and New York: Scottish Academic Press and Barnes and Noble, 1975),41-61, and "The Memory of Caedmon," in Om1 Tmditwnal Literature, ed. John M. Foley (Columbus: Slavica, 1983),282-93. 17. See The Anglo-Saxon Minor Poems, ed. Elliott Van Kirk Dobbie (New York: Columbia University Press, 1942),xciv. 18. Foley, "Literary Art and Oral Tradition," 206. 19. All we really know is drawn from Bede, A History of the English Church and People, trans. Leo Sherley-Price (Harmondsworth: Penguin Books, 19551,245-48. 20. The Battle of Maldon and Other Old English Poems, ed. Bruce Mitchell, trans. Kevin Crossley-Holland (London: Macrnillan, 1965), 95. 21. A Choice ofAnglo-Saxon Verse, ed. and trans. Richard Hamer (London: Faber, 19701,123. 22. Poems From the Old English, ed. and trans. Burton W e l , 2nd ed. (Lincoln: University of Nebraska Press, 1964),21. 23. See Foley, "Literary Art and Oral Tradition," passim.

Chapter 8 1. Pound, Tmnslations, 207. 2. See Raffel, Ezm Pound, 68: "No poet ever devoted himself more succesefully or more importantly to translation." 3. The Seafarer, ed. I . L. Gordon (London: Methuen, 1960),1-2. 4. bid. 5. Pound, Translations, 208. 6. Ibid., 209. 7. Raffel, Poems From the Old English, 32. 8. Ibid., 33. 9. "Anadyomene," in Pound, Tmnslatwns, 436. 10. John Dryden, Of Dramatic Poesy and Other Critical Essays, ed. G. Watson, 2 vols. (London: Everyman's Library, 1962), 2:28-29.

Notes

197

11. Raffel, The Forked Tongue, 115. 12. bid., 59. 13. Ibid., 51. 14. Horace, The Art o f P o e 4 , verse trans. Burton Raffel, prose trans. James Hynd, notes by David Armstrong, &word by W. R. Johnson (Albany: State University of New York Press, 1974),4. 15. J. C. Catford, A Linguistic Theory of Translation (London: Oxford University Press, 1965),48,27. 16. Hamer, A Choice of Anglo-Saxon Verse, 189. 17. Ibid., 22. 18. The Earliest English Poems, trans. Michael Alexander (Harmondsworth: Penguin Books, 19661,25. 19. Ibid., 75. 20. Ibid., 69. 21. Greek Lyrics, trans. Richmond Lattimore, 2nd ed. (Chicago: University of Chicago Press, 1960),41. 22. Henry Thomton Wharton, Sappho (London: John Lane, 1895),91. 23. Sappho, trans. Willis Barnstone (New York: DoubledayIAnchor, 1965), 111. Barnstone combines this fragment with another in his translation. I have reproduced only the first fragment. 24. Dudley Fitta, uForeword," in Sappho, trans. Mary Barnard (Berkeley: University of California Press, 19581,viii-ix. 25. Ibid., #60. 26. Burton Raffel, "Mary Barnard's Sappho," Hudson Review 18:2 (Summer 1965): 236. 27. Archilochus, Sappho, Alkman, trans. Guy Davenport (Berkeley: University of California Press, 1980), 106. 28. Barnard, Sappho, 102. 29. Lyrics From the Greek, ed. and trans. Burton Raffel (Calcutta: Writera Workshop, 1978), 32. 30. The Greek Anthology, ed. and trans. W. R. Paton, 5 vols. (London: Heinemann, 19161,1:299. 31. Poems From the Greek Anthology, trans. Dudley Fitta (New York: New Directions, 1956),17. 32. Paton, The Greek Anthology, 4213. 33. Fitta, Poems From the Greek Anthology, 69. 34. Selections From the Greek Anthology, trans. Andrew Sinclair (New York: Macmillan, 19671,26,61. 35. Ibid., 17. 36. Ibid., 15. 37. Paton, The Greek Anthology, 4169. 38. Fitts, Poems From the Greek Anthology, 78. 39. Private communication. 40. The Pillow Book of Sei Shonagon, ed. and trans. Ivan Morris (New York: ColumbiaUniversity Press, 1967;reprinted., Harmondsworth: Penguin Books, 19711, 233.There is a striking contrast to this encomium on page 22: "I could actually see the texture of their [several women's] faces. Some of them were not properly powdered, here and there their skin showed through unpleasantly like the dark patches of earth in a garden where the snow has begun to melt."

198 Notes 41. Beowulf, trans. David Wright (Harmondswortk Penguin Books, 1957). 42. Morgan, Beowulf, 76. 43. Donaldson, Beowulf, 49. 44. Pearson, Beowulf, 117. 45. M e l , Beowulf, 92. 46. Ibid., 109-10. 47. Morris, The Pillow Book of Sei Shonagon, 17. 48. M e l , The Forked Tongue, 30. 49. Ibid., 27. 50. T. S.Eliot, To Criticize the Critic (London:Faber and Faber, 19651,133. 51. John Frederick Nims, From Sappho to VaUry (New Brunewick: Rutgers University Press, 1971), 301. 52. M e l , Russian Poetry Under the Tsars, 32. 53. Selected Works of Nikolai S. Gumilev, ed. and trans. Burton M e 1 and Alla Burago (Albany: State University of New York Press, 19721,238. 54. Nims, From Sappho to Val&, 306. 55. M e l , Ezm Pound, 62,72. 56. M e l , Russian Poetry Under the Tsars, 32. 57. Raffel,Ezm Pound, 14. 58. Waley, The Secret History of the Mongols, 184. 59. Ezra Pound, Personae (New York: New Directions [n.d.l), 45. 60. Ewen, The Poetry and Prose of Heinrich Heine, 89. 61. Nims, From Sappho to VaUry, 302,319. 62. Arrowsmith and Shattuck, The Cmft and Context of Tmnslation, 123. 63. RafFel, Four Humours, 14.

Index

Book titles are in italics; all other titles are set in roman and without quotation marks. 'Ranslations not otherwise ascribed are by the author. Material contained in indented quotations has not been indexed. academics, vii-ix, 110,111-15,11821,124-25,130,167,177-79 adaptation, 21,63 adat, 54,85 a t y , 101-10 agglutinative morphology, 8-9, 53-54 Aku (Chairil Anwar), 51-55,83-86; quoted, 55 Alexander, Michael, 168-69;quoted, 168 Alexander, Sidney, quoted, 111 Alighien See Dante Amichai, Yehuda, 136 analytical syntax, 46,155 Andreyev, Leonid N., 132 Anglo-Saxon (Old English), 4, 25-26, 28, 132-33, 135,149-56,159-65, 167-69,174-76,177-78 animal communication, 15 Annu Kareninu (Leo Tolstoy), 108 Anthology of TmubadourLyric Poetry (Alan R. Press), 112 Anwar, Chairil: See Chairil

appropriative translation, 181 approximation, 12-13, 25-26, 29,3233,36-41,43,44,55,67,71,74,79,

89-90,145-48,154-56,159-86 Arabic, 4, 176-77 Arberry, A. J., 104-6; quoted, 104,105 Archilochus, 170 Arrowsmith, William, quoted, 95,185 Ars Poetica (Horace), 167 Asia, 51,63 Asia Society, 134-36 aspiration, 30,32 audience, 110-28,177-78 Auerbach, Erich, quoted, 10 Austro-Polynesian: See MalayoPolynesian Aveling, Elinor Marx, viii bahasa indonesia: See Indonesian Balcon, Le (Charles Baudelaire), 86-90 Ballad of the Army Carts: See Chariot song

200

Index

ballads, 57,75,77-78, 147-48 Bar, Elvire D., 17 Barks, Coleman, 104-5; quoted, 105 Barnard, Mary, quoted, 170-71 Barnstone, Willis, quoted, 75-76, 170 Basho, 67-70; quoted, 68 Baudelaire, Charles, 18,86-90,135, 158;quoted, 86 Beowulf, viii, 108,118-21, 124-25, 132-33,174-76.177-78 Bich, Nguyen Ngoc, 136 Binyon, Laurence, 59,60,184;quoted, 59 Bloomfield, Leonard, quoted, 13,51 Bly, Robert, quoted, 105-6 Borges, Jorge Luis, 133-34 borrowing: See linguistic borrowing Bownas, Geoffrey, 130;quoted, 6-7,65 Bowra, Maurice, quoted, 22 Boyan, 143 Bragg, Lois, quoted, 80 Brodsky, Joseph, 132 Brown, C. C., quoted, 146 Brown, Clarence, quoted, 131-32 Burago, Alla, 136-37,158 Butt, John, 102 byliny, 139 Bynner, Witter, 36,46;quoted, 35,46 Caedmon's Hymn, 149-56 calligraphy, Japanese, 7 Campbell, Roy, 77-79; quoted, 76-77 Candide (Voltaire), 102 Cantos (Ezra Pound), 181-82 Carlisle, Olga, quoted, 132 Carne-Ross, Donald, quoted, 124 cases (syntactical), 10 Catford, J. C., quoted, 167 Catullus, 26-28; quoted, 27 Cavafy, C. P., 133 Chairil Anwar, 20,30-33,52-55,8386,93-94,166,186 change: See linguistic change Chariot Song (Du Fu), 46-48;quoted, 46 Chaucer, Geoffrey, 139 Chaucerian Compromise, 82 Chicago, University of, 135

Chinese, 5,7-8,13,24,33-37,45-50, 131,137,176 Chomeky, Noam, quoted, 14 CMtien de Troyes, viii-ix Ciardi, John, 61,107-10;quoted, 59, 108-9 Cluysenaar, Anne, quoted, 14 Collins, Rowland L., 120;quoted, 118 Collins, Wilkie, 129 Conrad, Joseph, 129 Cooper, Arthur, 47-48; quoted, 46,4849 Coplas del mismo . . . (San Juan de la Cruz), 41-45;quoted, 41 Corman, Cid, quoted, 67 Cornell University, 30,32 Cmfi and Context of Tmnslatwn, The (Arrowsmith and Shattuck), 95 criticism and critics, viii Crossley-Holland, Kevin, 151-55; quoted, 151 Crown, Bonnie R., 134-36 culture: See language as culture Damais, Louis-Charles, 83-86; quoted, 83-84 Daniel, Arnaut, 112-17 Danish, 1 Dante, viii, 1,58-62, 184;quoted, 59, 107,179 Davenport, Guy, quoted, 170 dead languages, 4 Denhavin, Gavrila Romanovitch, 81 dialects, 28-29 Dickens, Charles, 129 diction, 60-62, 78-79, 103,108-10, 119,128,142,152-53,155-56,16869,174 Dictwnnuire &s Synonyms (Elvire D. Bar), 17 Di Giovanni, Norman Thomas, quoted, 133-34 Ding Zuxin, 36,47-48;quoted, 34,47 Divinu Cornmedia, The (Dante), 1,5862,107,110,179 Donaldson, E. Talbot, 175;quoted, 12021 Donne, John, 61,107

Index

201

double-imperfectiveverbs, 9 Draper, Hal, quoted, 58 Dream of the Red Chamber, The: See The Story of the Stone Dreieer, Theodore, quoted, 97 Dryden, John, 165-66; quoted, 166 DU Fu,33-36,46-50,106;quoted, 3334 Dutch, 19,32,51,135

free translation: See expansive translation French, 4,13,16, 18,38-41, 56,83, 83-90,135,158 Freud, Sigmund, 98 Friedrich, Paul, quoted, 63-64 Frye, Northrop, quoted, 23 full translation, 11,38,111-15 Fussell, Paul, quoted, 64-65

Earliest English Poems, The (Michael Alexander), 168 Egyptian, 4 elaboration, 44,47-48, 50, 109-10, 121-25,128 Eliot, T. S., 33,99;quoted, 179 English, 4,8,9,10,14,16,17,18,28, 29-30,31-33,36,39-41,51,53,55, 56-58,82,83,114,147,158,185 epic, 18 errors, 116-17 exact translation: See full translation expansive translation, 110,121-25

Gaul, 3-4 Gawain and the Green Knight: See Sir Gawain and the Green Knight Gembala (Mohammad Yamin), 18-19 Genji (Murasaki), 183 genres, literary, x, 63-79,98-102, 147-48 G e m , 3-4,10,16,17,18,56-58,83, 90-94,103, 135,184-85 Gerould, Gordon Hall, quoted, 120 Gjersten, Frederik, 1 glottal stop, 31 Goldin, Frederick, 113-15;quoted, 114 Goldin, Judah, quoted, 138,139 Gordon, I. L.,quoted, 161 Goase, Edmund, 99 Graham, A. C., 46,50;quoted, 45 grammar: See syntax Greek, 14,22,103,133,169-74,181 Greek Anthology, The, 173 Greek Lyrics (Richmond Lattimore), 169 Greenfield, Stanley B., 119,120; quoted, 119 Gregory, Horace, quoted, 27 Gurnilev, Nikolai Stepanovitch, 158, 181;quoted, 180

facts: See linguistic facts faithfulness: See literal translation Fang, Achilles, 131 Faulkner, William, 101 Federal Writers Project, 131 fidelity: See literal translation Fitts, Dudley, 171-74; quoted, 170, 172,174 Flaubert, Gustave, viii, 102 FZeurs du Ma1 (Charles Baudelaire), 86 Flinker, Noam, 136 Flores, Kate, 92-93; quoted, 92 Foley, John M., quoted, 150 Ford, Ford Madox, 129 Forked Tongue, The (Burton M e l ) , vii, ix, 166 formal translation, 110,111-15, 11721 forms, literary, x, 12,17-21.63-79, 147-48 fortis languages, 30 Francis. W. Nelson, quoted, 15 Franks (tribes of Gaul), 3-4

Haifa, University of, 136 Hall, John R. Clark, 120 Hall, Robert A., Jr., quoted, 13-14, 1718,21-22,50-51 Halliday, M. A. K., quoted, 15 Halporn, James, quoted, 22 Hamer, Richard, quoted, 153,167-68 Hamlet (Boris Pasternak), 126-28 Hardy, Thomas, 99

202

Index

Having Anwar Out of the House (Burton Raffel), 186 Hawkes, David, quoted, 35,36 Hayward, Max, 132 Hebrew, 4,13,136,138,176-77 Heine, Heinrich, 56-58, 184-85; quoted, 56,184 Henderson, Philip, quoted, 133 Herbsttag (Rilke), 90-94; quoted, 90 heroic couplet, 74 History of Myself, A (Theodore Dreiser), 97 Hockett, C. F., quoted, 16,18 Hoff, Frank, quoted, 67 Homer, 89,121-24,169,178-79,181, 182 Hopkins, Gerard Manley, 18 Horace (Quintus Horatius Flawus), 135 Huxley, Aldous, 101 Hyrnne B la Beaut4 (Charles Baudelaire), 158 Ibsen, Henrik, 99-100;quoted, 1,99100 Iliad (Homer), 121-24,169,178-79 Imitations (Robert Lowell), 126,128 imitative translation, 110-11, 126-28 impossibility of translation, 11,67,158 improvisation, 66,67,70 Irn wundemhiinen Monat Mai (Heinrich Heine), 56-58; quoted, 56 indeterminate verbs, 9 indirection, 72-74 Indonesian, 5,8-9, 13,16,17,18-21, 29-33,51-55,59,63,70-74,82,8386,93-94,135,143-48,186 Indo-European, 9,10,12-13,15,22,45 Inferno (Dante), 58-62,107-10 interdisciplinary approaches, vii interpretive translation, 110,115-21 Italian, 1,13,18,58-62,63,107-10, 135 Jacobson, Roman, quoted, 37 Japanese, 5-7,8,13,30,65-70,74, 166,174 Javanese, 24 Jespersen, Otto,177

Jewish tradition, 138-39 Katka, Franz, 106,quoted, 103 Keeley, Edmund, 133 Keene, Donald, 7, 117,166,174; quoted, 5-6,65,66 Kenner, Hugh, 115-17;quoted, 115-16 Kikaku, 67-70;quoted, 68 Kipling, Rudyard, 163 Kotewall, Robert, 36;quoted, 35 Kramer, Aaron, 57-58;quoted, 57 Kubelreiter, Der (Franz Katka), 103 Kwock, C. K., 49-50,130-31; quoted, 49-50 La DriBre, Craig, quoted, 23-24.24-25 Language (Leonard Bloomfield), 13 language as culture, 3-10, 14,16,1721,24-25,63-65,67,72-74,79,85-

86,146,156,157-58,163-64,17479 language, current, 142 language, knowledge of, importance, 102-3,129-34,137,157 language, sexual distinctions in, 8 language, sounds of: See phonology language, spoken (oral), 5 language, written, 5 Latin, 4,10,22,26-28, 82,103,135, 167 Lattimore, Richmond, 121;quoted, 122 Lawrence, D. H., 102,185-86 Lehmann, Winfred P.,quoted, 15 Lefevre, An&, quoted, 25,27-28 Leishman, J. B., 92;quoted, 91-92 lenis languages, 29-30 Levik, Vilgelrn Venyaminovich, 38-41; quoted, 37 Lewis, Grace Hegger, quoted, 100,101 Lewis, Sinclair, quoted, 100-101 lexicon: See vocabulary Li Bau, 106 Lin, Julia C.,quoted, 33 Lindley, Denver, quoted, 98-99 Lindstrom, Thais, quoted, 140 linguistic borrowing, 4,16-17, 51-52 linguistic change, 3-10, 82 linguistic fa&, 10,11

Index linguistics, importance of, viii-ix, 111, 167 Linguistic Theory of Tmnslation, A (J. C. Catford), 167 linked verse (Japanese), 66-70 Li Po: See Li Bau literal translation, 11,39,51,111-15, 14548,185 literary history, 12,17-21,38,51,57, 72-73,79,90,91,156,157-58 literary values, importance of, viii-ix Liu, James J. Y., 131;quoted, 7-8, 24, 33 Lo ferm voler . . . (Amaut Daniel), 113-15 Loeb Library, 171,172,173 Logue, Christopher, 121-24,125,181; quoted, 122-23 Lomonosov, Mikhail V., 81 Looking a t Spring: See Spring Scene Lord of the Rings, The (J. R. R. Tolkien), 175 Lowell, Robert, 126-28; quoted, 126-27 lyrical poetry, 58 Madame Bovary (Flaubert), viii, 102 Malay, 51-52,70-74 Malaya, 52,53 Malayo-Polynesian, 8,12 Malay Sayings (C. C . Brown), 146 Mandelbaum, Allen, 60-61;quoted, 60 Mandelstam, Osip E.,131-32 Mann, Thomas, quoted, 11 Maseie, Suzanne, quoted, 132 Mathews, Jackeon, 86 Mathews, Marthiel, 86 McHugh, Vincent, 130-31; quoted, 4950 McIntosh, Angus, quoted, 15 Meschonnic, Henri, quoted, 80 meters, 19,22,24,25-26, 28-29, 33, 35-36,48,57,60-62,71,74-78,80-

94,1434,150,152-56,159-60, 164-65,173 Middle East, 51 Middle English, 28-29 Milton, John, 61,107,109 Miner, Earl, 66-70; quoted, 66,67,68

203

Mirsky, Dmitri, quoted, 140,141 mistranslation, 115-17, 160-65 Molbech, Christian Knud Frederik, 1 Monas, Sidney, 136,140;quoted, 142 Moonlit Night (DuFu), 33-36 Morgan, Edwin, 174-75;quoted, 13233 morphology, 8-9,41,53 Morris, Ivan, quoted, 177 Morris, William, 132-33 Mosel, James N., 135 Mope, John, 104-5; quoted, 105 Mozart, Wolfgang Amadeus, 177 Muir, Edwin, quoted, 103 Muir, Willa, quoted, 103 Murasaki, Lady, 183 Musa, Mark, 61;quoted, 60,108 musical texts, 143-48 Musim Gugur (Chairil Anwar), 93-94 Nabokov, Vladimir, 141-43;quoted, 140,142 Naim, C. M., 135 names, 176-79 Nee Wen-yei, 4748,50;quoted, 47 New Yorker, The, 131 Nims, John Frederick, 42-45,77-78, 179-80, 182;quoted, 42-43,44,77, 179,180-81,185 nonhuman communication: See animal communication Normans (in England), 4,56,81,82 objectivity, 97 Obolensky, Dimitri, quoted, 127,14041 Odagiri, Hiroki, 67;quoted, 68 ode, 18 Odyssey (Homer), 181 Old English: See Anglo-Saxon Old Friend From Far Away (Kwock and McHugh), 130-31 Olson, Paul A., quoted, 15 oral transmission, 72-73, 138-41 organs of speech, 13-14 Ostwald, Martin, quoted, 22 Owen, Stephen, quoted, 35,36-37 Oxford University, 130

204

Index

pantun, 18,70-74,144-48,156 Paradise Lost (John Milton), ix parenthetical interjections, 43,4445 Pastemak, Boris, 126-28 Patrocleia (Homer), 121 patronymics, 176-77 Payne, Robert, quoted, 21 Pearson, Lucien Dean, 118-19,175; quoted, 118 Penguin Book of Japanese Verse, The (Bownas and Thwaite), 130 Persian, 104 Phaedrus (Plato), 138 phonemic distinctions, 24,81-82,91 phonology, 12,13-14,23-37,38,42, 43,55,57,69,85,158,160,164 phrase. units, 14445,151-52.154 physical bases of language, 13-14 pidgin translation, 50 Pillow Book (Sei Shonagon), 174,177 Pin-yin,33 pitch, 24,33-34,69,145 Plato, 138,171-72;quoted, 172 pluralization, 9,72 Poems Fmm the Greek Anthology (Dudley Fitts), 171 poet manquk See translators as poet Polish, 81 Pope, Alexander, 74,89,90,182 Pound, Ezra, 25-26,112,115-18,15968,18142,184-85;quoted, 25,115, 159,162,163,182,184 prepositions, 10 Press, Alan R., 112-15; quoted, 112-13 Princeton University, 131 pmse translations of poems, viii, 36, 113,120-21,174 prosody, books on, ix, 21 prosody, X, 12,21-22,24-26,28-29, 33,38,57,75-79,80-94,160,164:

See also meters Provenqal, 114 Pushkin, Alexander Sergeyevitch, 70, 81,180 quantitative prosody, 22,82 Racine, Jean, 85 Raffel, Burton, quoted, 20, 29,424,

44,54-55,62,78,81,93,116,11718,121,128,154,166-67, 171,176, 186 Rebsamen, Frederick B., 169;quoted, 124,125 reduplicative pluralization, 9 reduplicative verbs, 46 rhyme, 1,19,20,29,31,32,34,42,44, 48,57,60,61,62,70,71,75-79, 92, 93,144,147-48 Richardson, Joanna, 88-90; quoted, 88 Rilke, Rainer Maria, 18,90-94 Rimbaud, Arthur, 165 River Merchant's Wife: A Letter (Ezra Pound), 117 Roche, Paul, 132 Rome, 3-4 Rosenmeyer, Thomas, quoted, 22 Rumi, Maulana Jalal-uddin, 104-6 Rwlan and Ludmilla (Alexander Pushkin), 180 Russian, 9,81,131-32,136-37,13943,158,176 Saint John of the Cross: See San Juan de la Cruz Salam, Nurdin, 134 sampiran, 70-71,73-74 San Juan de la Cruz, 41-45,75-79 Sanskrit, 4 Sapir, Edward, quoted, 38 Sappho, 169-71 satisfactory translation, 11 Sayers, Dorothy, viii scholars: See academics Schubert, Franz, 56,58 Seafarer, The, 25-26, 159-65, 167-69 Secret History of the Mongols, The (Arthur Waley), 182-83 Seferis, George, 133 Sei Shonagon, quoted, 174 Semangat (Chairil Anwar), 51-55; quoted, 52 sexual distinctions: See language: sexual distinctions in Shakespeare, William, 36,85,99,109, 178 Shattuck, Roger, 116;quoted, 95 Sherrard, Philip, 133

Index Shi Akindo No Maki (Basho and Kikaku), 67-68 Sinclair, Andrew, 172-73; quoted, 172 Sindapanese, 12 Sir Gawain and the Green Knight, 2829 Slater, Lydia Pasternak, quoted, 12728 Slonim, Marc, quoted, 139 Slovo o pulku Igoreve, 140-43,150 Smith, Norman L., 36; quoted, 35 sonnet, 18-21,30-33,63 sounde: See phonology source language, 11-22,25 Soviet Union, 136 Spanish, 13, 15-16,41-45,75-79, 133-34,135 specialization, vii-ix speech organs: See organs of speech sprachgefiihl, 103 Spring Scene (DuFu),48-50; quoted, 48 stanzas, 19,41,42,57,67,75-79,114 Stathos, John, 132 Steegmuller, Francis, 102 Story of the Stone, The, 176 stress (linguistic), 22,81-83, 87, 88,91, 143-45 Sturm, Frank P., 86-88,90; quoted, 86 subjectivity, 97-110 Surabaya, 30 Suaumu, Kamaike, quoted, 67 syllabic prosodies, 81-90,119,143-48 syllabo-tonic verse, 81 Syntactic Structures (Noam Chornsky), 14 syntax, 6-7,12,14-16,37-50,53,55, 57,114-15,158 tanka, 18 target language, 11,25 Tartar invasions (of Russia), 81 Tennyson, Alfred Lord, 100 ten~es,8-9,48 tena rima, 61,62 Texan, University of, 136 Thai (Siamese), 134-36 Thwaite, Anthony, 130,136; quoted, 67,65

205

Tillyard, E. M.W., 124 Tolkien, J. R. R., 120-21,175; quoted, 120,142 Tolstoy, Leo, 108 tones: See pitch tradition, 156 translation, decision points in, x translation, impossibility of: See impossibility of translation translation, nature of, ix translation, varieties of, 110-28 translators as poets, vii, 101-2,17982.184-86 translators as literary practitioners, vii translators as scholars, vii-viii Trediakovsky, Vasily K., 81 Tu Fu: See Du Fu Turnell, Martin, quoted, 87, 88 Tuti Artic (Chairil Anwar), 30-33; quoted, 30-31 Untermeyer, Louis, 184-85; quoted, 184 Urdu, 135 verbs, 9,43 Verga, Giovanni, 102 vers libre, 76, 78 Vietnamese, 134,136 vocabulary, 12,16-17,38,50-62,87, 152-53,155,160-64,167-68,172, 174-76 Voltaire, 102 Vulgar Latin, 10 Waley, Arthur, 103-4,106-7,182-84, quoted, 7,65,103,183 War and Peace (Leo Tolstoy), 108 Watson, James D., quoted, 98 Weiner, Norbert, 97-98; quoted, 98 West Saxon, 149, 150 Wharf, Benjamin Lee, 10; quoted, 4-5 Wharton, Henry, quoted, 169 What Dante Means to Me (T. S. Eliot), 179 WoH, Lucie, 99 words: See vocabulary Wright, David, quoted, 174

206

Index

writers manqu6,106 Wyatt, A. J., 132-33 Yamin, Moharnmad, 19-21,63; quoted, 18-19 Yuasa, Nobuyuki, quoted, 67

Yvain (Chdtien de *yes),

viii-ix

Zenkoveky, Serge A., quoted, 141,142 Zhukovsky,Vasily Andreyevitch, 180,181 Ziegfeld, Dr. Richard, vii Zukofsky, Louis, 26-28; quoted, 26