I

ONCE, IN A Paris bookstore, I turned to the shelves that house literature in translation, and took down some volumes of Wordsworth and Yeats in French. Reading The Prelude’s measured lines, with their Racinian metre and the rhetorical temper so congenial to the French language, I could discover nothing of Wordsworth’s own poetic manner; the laconic, even tedious voice had been imbued with a resonance absent in the original. More than the opinions of critics and writers (Doctor Johnson once stated that poetry ‘cannot be translated’, though his own ‘London’ and ‘The Vanity of Human Wishes’ refute the claim), the experience of reading in translation a work composed in one’s own language, especially a work that has contributed to our conception of that language, illustrates the distance between a translation and the original text.

     And yet, so much of what we receive outside our own immediate literature involves translation that it could be considered the poem’s essential act: either to resist translation or collaborate in its possibilities. Eliot pointed out that Laforgue (whom Pound also translated) had provided him with a useful example not only because he had explored certain modalities of colloquial speech that instructed Eliot in his own work, but also because he could appropriate influence in a more general manner through a foreign language than if he had chosen a model writing in English. The result, in this case, has been that the shaping voice of Modernism comes to us through an alien tradition, French literature in the nineteenth century. And one of the most evident properties of Modernism has been its international character. Eliot is, of course, not the only instance. As George Steiner has so tirelessly noted, the three most brilliant novelists of our time, Nabokov, Beckett, and Borges, have all written in two languages and in two diverse traditions (Borges, perhaps, doesn’t quite conform to this scheme, since he writes in Spanish, only collaborating in the translation; but his knowledge of English literature is so wide, and exploited with such constancy, that it virtually constitutes a second tradition). Finnegans Wake represents an attempt to break down autonomous elements in language, to trace, through complex etymologies, the location of a word’s origin in other languages. It could be said that what these writers share in is an impulse to translate themselves. In producing a work immune to translation or even influence, Joyce wanted his book to be inalterable, in the sense that it couldn’t then become other than what it was: a work composed nominally in English, but which so transfigures and isolates each word that it can’t be comprehended without reference to Joyce’s entire devised language (this is where the novel parallels the principles of structural linguistics).

     But it was Ezra Pound, above all, who made ours an age of translation; the close affinity between The Cantos and Finnegans Wake is reflected in their impulse toward the multilingual. What motivates the inscription of other languages within these texts is less impatience with the limits of English than a desire to reveal those properties common to language itself. It was logical that Pound, whose poetics were situated in the Image, should choose the Chinese ideogram to represent its visible aspect. To ‘Make it New’ implied more than simple innovation; traditions remote in both time and in linguistic character had to be restored and appropriated to our own. Joyce, echoing Vico, elaborated the metaphor of language as a river, and Pound’s intention was to navigate those tributaries which led to its source. In order to renew the language, he selected works which had been crucial to their time, and renovated them in our own.

     Translation, then, has informed the practice of Modernist poetics to such an extent that it could even be considered the dominant motif. This isn’t to propose that the phenomenon of translation is at all recent; the Schlegel-Tieck translation of Shakespeare had such a vivid effect on classic German literature, on its entire dramatic tradition, that it can no longer be regarded as other than an original work, while Hugo’s or Gide’s Hamlet stand in the same relation to the French. And in our own literature, the continuum from, say, Dryden’s to Johnson’s to Lowell’s translation of Juvenal’s Tenth Satire represents a tradition no less coherent than that evident in their other works. The genesis of style, the temper of a mode evolved through several centuries, can be detected in the history of translation. That Pound approached The Odyssey through the sixteenth-century Latin version of Andrea Divus (Canto I) rather than the original, as Pope had done with The Iliad, suggests that his concern was to locate those events in literature which could be recovered with minimal loss, and which could then instruct us in the shape, the texture of our own language, widening its parameters, deepening its expressive capabilities. Not that Pound ever violated the original or made it serve his own interests; his translations are, in Steiner’s words, ‘so immediate to the ways we experience language and objectify emotion that the Latin or Provençal poem is inseparable from the grammar of modernity’.

     Steiner’s observation about grammar is more than rhetorical; language has become in our time the crucial referent in several disciplines, and translation is, in its essence, caught up in this linguistic net. (Foucault has called language the modern episteme, and it is significant that even the word he chooses to represent the concept of a language-web through which each era views the concrete objects that inform its character is morphological, identified with the vocabulary of linguistics; like morphemes and phonemes, Foucault’s episteme lies embedded in the actual properties of speech.) It could be said that, since Wittgenstein, philosophy has become critique of language, and the work of Claude Lévi-Strauss has turned anthropology in the same direction. Because his entire method is based on De Saussurian linguistics, Lévi-Strauss has been able to transpose the model of language onto societies, and so to regard all social transactions as themselves elements in a language that embraces lived experience. This notion is crucial to translation because it releases that act from its confinement within the realm of language, just as literature is released through translation from confinement within its own linguistic borders; in demonstrating that myth involves the classification of natural phenomena in stories, ritual, familial organisa— tion, and prohibitions, Lévi-Strauss proposed that social life could be read in the same way that we would approach a written text. Central to this order is the act of translation from events or things to their articulation in these various ‘languages’, an act which would then explain why so much recent translation has concentrated on ethno-poetics, the poetry of societies without a written literature, rather than on those traditions we associate with our own (Jerome Rothenberg’s Technicians of the Sacred is a crucial document in this respect).

     Lévi-Strauss acknowledges having taken over his linguistic model from Roman Jakobson, whose own work in translation has been important. In his essay ‘On Linguistic Aspects of Translation’, Jakobson distinguishes between three modes: intralingual translation, or rewording, which involves interpretation within the same language; interlingual translation, or interpretation ‘by means of some other language’; and intersemiotic translation, or ‘transmutation of nonverbal into verbal signs’. The implication here is that translation is a component in all language transactions, and that the second mode Jakobson cites, which applies to the translation of literature, is no more than a single term in a complex equation. As our attention to the structure and properties of language becomes more vigilant, Jakobson s conviction that ‘poetry by definition is untranslatable’, that only ‘transposition’ is possible, becomes more difficult to contest. The sheer singularity of a poem resides in its willingness to suspend or violate rules, employing only those elements in the language which cannot be reproduced in some other phrase. This reverts to Jakobson’s first mode, translation within the same language, an event even less possible in poetry than the second, since what determines a poem is its exact nature, the sense that nothing else will do. Valéry has stated this condition: ‘Writing anything at all . . . is a work of translation exactly comparable to that of transmuting a text from one language into another.’ To Valéry, the poet is a translator, ‘who translates ordinary speech, modified by emotion, into “language of the gods”’ (this ‘emotion’ belongs to Jakobson’s third, inter-semiotic mode); translation, then, is the model of the poet’s act, mediates the poem’s composition.

     Such vivid self-consciousness is evident in Valéry’s own work, and in that episode of Modernism which can be traced back through Mallarmé to Baudelaire. (As Eliot noted in ‘From Poe to Valéry’, the consequence of Poe’s having entered the French tradition through Baudelaire’s translation was that he became an author whose influence had no real relation to his original work; what unifies Baudelaire, Mallarmé, and Valéry is their stance towards an oeuvre which existed only in their own translations, Baudelaire’s of the stories and Mallarmé’s prose renderings of the poems.) Mallarmé’s odd treatise on the English language, ‘Les Mots Anglais’, is more important as a discourse on the nature of language, ‘l’étude matérielle des termes eux-mêmes’, than as a grammar; his method of instruction was to compose ‘A Thousand English Phrases to Learn by Heart’, providing his own translations. The result is a text in which his native language appears foreign to the English reader; the language is reconstituted in all its strangeness, a strangeness which vanishes the moment we take it up as a familiar instrument.

     This is what all translation should strive to accomplish: the creation of a language which mimes the character of the original, even as it invents linguistic modalities unavailable to that language in its common use and structure. I recently heard a reading by Yves Bonnefoy where he provided his own translations, insisting that no English translator, however competent in French, could reproduce either the cadence or intent of the poems. Bonnefoy’s command of English was unexceptional, and his accent rendered the English versions incomprehensible at times, but what could be heard through the translations was a radical diction that owed little to either language. Because the poem Bonnefoy devised wasn’t obligated to be plausible in English, it was released from those restraints which impose themselves on such a developed language, and so could become a new text, situated somewhere between English and French.

     There are liabilities implicit in this mode, just as in the others. To begin with, the poem would be confined to translation within those languages known to the author, and we would be limited, obviously, to contemporary works. But this impulse to conceive a mode which belongs to translation alone can reach backwards and revive works composed in some distant era. Much has been written about problems of translation from the classics, and from other eras (Steiner contends that, even when we read an earlier work in our own language, an act of translation is involved); On Translation (Harvard, 1959) includes papers by Renato Poggioli, Nabokov, and Edwin Muir, while Roger Shattuck has edited another collection of essays, The Craft and Context of Translation, with contributions by Kenneth Rexroth, Robert Fitzgerald, and Donald Came-Ross, among others. These volumes are valuable, but their concern is more with translation from those works which have established our tradition, in the widest sense. The best essay I know of on the subject of translation as a metalanguage, a discourse distinct from both the original and the language in which the translation comes to be composed, is still Walter Benjamin’s ‘The Task of the Translator’. This essay, written as a preface to his own translation of Baudelaire’s Tableaux Parisiens into German, elaborates on a number of motifs that recur elsewhere in his writings: the duration of a work’s life and after-life; its relation to the language which sponsors its existence; the transmutations that occur in a work through time. These events become even more significant when the work is a translation, because it is the nature of translation to capture the resonance, the temper, inscribed in the original, a resonance impossible to reproduce except in some other language:

Unlike a work of literature, translation does not find itself in the center of the language forest but on the outside facing the wooded ridge; it calls into it without entering, aiming at that single spot where the echo is able to give, in its own language, the reverberation of the work in the alien one.

This forest, like Baudelaire’s forêt de symboles, is impenetrable; all we can hope to do is listen to the original and establish a resemblance more tonal than exact. If not poetry, then translation should ‘aspire to the condition of music’, in this respect: that it constitute a notation of what cannot be expressed in language.

II. The Penguin Translation Series

THE PROLIFERATION OF literature, in the form of inexpensive editions, armies of competent writers and literary practitioners, and the production of texts that compete among themselves, has been accompanied by a vast increase in the number of translations available. What was once the province of littérateurs and established poets has now, through division of labour, been delegated to translators whose livelihood depends on the demand for literature in translation. In a sense, there’s nothing new about this; authoritative editions of the classics have long existed: the French edition of Ulysses, H. T. Lowe-Porter’s Thomas Mann, the Maude translation of Tolstoy, all of which were achieved in collaboration with the author. What is different is that now such translations have either been replaced or come to be read as if they were themselves original works. Translation is no longer an event, but a trade.

     The implications of this development extend to poetry, where, with the possible exception of Lowell’s Imitations and Robert Bly’s work, no real attempts have been made to produce a literature in translation. Of course, it’s a commonplace that poetry is more difficult to translate than prose, and that each era should provide its own translations in order to register changes within the language; but in our time, translation appears to have become a business transacted in accordance with those economic laws which advanced capitalism dictates. What we see, then, is a duplication of some texts, while others remain untranslated, collaboration being rare within a structure of such intense competitiveness. And since our competence in other languages and literatures has diminished, while recent work in those traditions has become no less complicated than our own, it’s difficult to follow developments within them. In France, we’re required, not only to read an opaque literature, but to be familiar with the critical penumbra that envelops it; so that the translations we receive are chosen more or less at random, and appear divorced from the complex linguistic and social conditions which gave rise to them.

     This problem has been more acute in England than in America, where translation has taken on a polemical character. This seems to me the only way to render translation active; in celebrating Neruda’s charged Surrealism, in writing poems that deliberately echo those qualities absent in our own literature, and in explaining over and over the bias implicit in his translations, Robert Bly has been able to show that criticism is an essential feature of translation. This enterprise has exerted considerable influence on the direction of contemporary American poetry; an awakened interest in the unconscious, the liberation of the image from its more conventional expression in those older poets whose tutor was the English tradition.

     Where, in America, contemporary translations have appeared in small press editions, or in isolated volumes through the larger houses, the market in England has been dominated by the Penguin Modern European Poets series. Penguin has sponsored other approaches to translation as well; there are the foreign-language editions, usually ranging over a specific, often quite expansive period, which provide prose translations, and (to my mind the most successful) George Steiner’s Poem into Poem, a brilliant collection of ‘world poetry’ from all eras and languages translated by well-known modern writers. Like Hans Magnus Enzensberger in his Museum der modernen Poesie, Steiner has assembled translations which are themselves ‘events’, poems that call up their originals before the tribunal of what he calls ‘felt relevance’. Enzensberger has confined his selection (of both translations and original texts) to the modern era, but within that limitation, has filled his museum with objects remarkable in their collective portrait of an epoch; Enzensberger notes in the introduction that ‘What is not itself poetry cannot translate poetry’, and his own translation of ‘The Love Song of J. Alfred Prufrock’ explores the balance achieved between two poems where both are enriched through the translation.

     As far as I know, Poem into Poem is the most ambitious anthology we have now; but it is the Modern European Poets series which seems to me characteristic of the present approach to translation. Though innumerable other translations have been published over the last decade, these thin, inexpensive volumes have become, in some sense, indices of reputation. Among their published or projected works, the Russian list is the most complete: Blok, Akhmatova, Mayakovsky, Tsvetayeva, Zabolotsky, Brodsky, Mandelstam, and their best-seller, Yevtushenko. And to have made available the work of Eastern European poets is a substantial contribution; because these languages are inaccessible to most English readers, the translation of such representative Czech poets as Vladimír Holan, Miroslav Holub, and Vitešlau Nezval, Hungarians (Attila József, Sándor Weöres, Ferenc Juhász, and Janos Pilinsky), and Poles (Rozewicz, Herbert, Vasko Popa) is of crucial importance. In these translations, what we receive is closer to information than to literature (though the Hungarians appear to have been so well-translated that their work possesses a resonance and verve in English); their intent is to inform Western audiences about the situation of the writer in East Europe, which comes over in the poems’ thematic subject. What is lost, of course, is the manner in which language, through reticence, linguistic dislocations, or silence, enacts a lived condition; but even to relate an aspect of this other world justifies their publication. About the translations from Swedish (Tomas Transtromer, Gunnar Ekelöf), Greek, and Hebrew (Abba Kovner, Yehuda Amichai), I can’t comment at all, knowing neither the languages nor the traditions involved. It is in the selections from the German, French, and Italian poets, though, that real questions arise about both the works chosen and the translations themselves.

     What effect is produced when we read poems in translation without having any knowledge of the original? As I’ve suggested, we can receive what I would call sociological information, through the nature and even appearance of the translated text; or, if the translation aspires to become another poem, modelled ‘after’ the original, we can approach it as we would an autonomous work. We have to be reconciled to these limitations in order to approach those volumes in the series devoted to poets whose language is unfamiliar (even so, it is possible casually to evaluate such translations; I know no Russian, but Richard McKane’s Akhmatova seems lifeless and repetitive in comparison with the versions of Max Hayward and Stanley Kunitz, while what I’ve seen of George L. Kline’s translations of Brodsky is impressive in English). It’s difficult to decide whether the uniformity of these volumes, the resemblance in format, subject, and even tonality, is the result of what I’ve called the international character of Modernism or of the mode of production. It would appear to be both, since neither condition can be illuminated without reference to the other.

     In what form does this international character emerge? Here, too, there are two reciprocal terms: the tendency to compose a work in such a way that no translation is possible, the motivation being to reveal those properties in language which are irreducible, and have no other name. This has been the situation with poets like Celan and Ungaretti, whose lives in our disastrous epoch led them to the liminal, exasperated language evident in their work, untranslatable because their experience itself remains untranslatable, and has to be related in a half-articulate language. To cite two simple examples, Ungaretti’s famous Mattina, reads:

M’illumino
d’immenso

and Celan’s Einmal:

Ems und Unendlich,
vemichtet,
ichten.

Ingeborg Bachmann, in Enzensberger’s Museum, has translated Ungaretti’s poem as ‘Ich erleuchte mich/ durch Unermessliches’, which carries over the shape of the original, even if the cadence and music are lost; Patrick Creagh, who translated the Penguin edition, quoted the poem in his introduction, but only to demonstrate its ‘untranslatable’ nature. In Celan’s case, Michael Hamburger has given a sensitive account of how he came to translate, with Celan’s own intervention ‘ichten’ as ‘ied’; Hamburger’s version reads:

One and infinite,
annihilated,
ied.

Joachim Neugroschel, Celan’s American translator, rendered this as ‘dieing,/ were I’ing’. Both pose adequate solutions, articulating the word within a word that Celan invented, where the ‘I’ serves as a verb. But whether or not these two fragments can be reconstructed in English, their essence lies in the rhyme, which a translation can only replace or imitate; embedded in their own grammar, these poems, like exiles in a remote land, suffer when forced to live in some other language.

     In a time when millions of Europeans have been uprooted, or seen those who spoke their language systematically exterminated, the impulse to protect from distortion or loss whatever is written down appears in the form of concealment; meanings have to be hidden, language has to become encrusted in the text so that it can’t be dislodged or appropriated. Where cultural disruption has been most intense, as in East Europe, this hermeticism is evident even in the poem’s arrangement on the page: spare, unrhymed lines, a diction ambivalent and worn.

     This condition, if not its exact expression, is communicated in the volumes devoted to Bobrowski and Bienek, Celan, Abba Kovner and Nelly Sachs; but what scope is intended in the series? It appears that ‘modern’ has been taken to mean the twentieth century; J. B. Leishman’s out-dated Rilke translations appear beside writers now in their forties, like Günter Grass, while Aleksandr Blok, a poet of the Revolution, shares the series with Iosip Brodsky, still in the midst of his career. The danger implicit in this ambition to embrace both innovators like Apollinaire and their successors is that no coherent portrait can emerge; Apollinaire, Ponge, Guillevic, and Tzara represent moments in the achievement of French poetry during our century, but without Valéry, Breton, Desnos, or Eluard (to simply name those who come to mind), is it possible to comprehend the sheer variousness and range which Modernism unleashed?

     It would be to no purpose to complain more about the inclusion or exclusion of names on the Penguin list; what I question, aside from the relative merits of the translations themselves, is the format in which these works have come to us. To begin with, I would argue that no translation from a Romance language should be published without a facing text of the original. This doesn’t mean that all educated readers are conversant with German, Italian, French, and Spanish; but that, however limited one’s knowledge of these languages is, there exists between them a shared grammar and origin, enough so that the mere presence of the original casts an aura over the translation, makes explicit the relation between two disparate texts. The effect produced by reading volume after volume of poems in translation without recourse to the original is weariness; the suspicion takes over that these are commodities, mass-produced and with a slight enough difference to distinguish them, but not enough to retain the real distinctive qualities on which their value depends.

     To choose just one example: George Kay’s edition of Montale conforms to the other volumes in the Penguin series; it contains a competent preface, and includes poems from all of Montale’s work. In comparison with the New Directions edition of Montale, though, the problems that plague Penguin’s own series become evident; Glauco Cambon, who edited the New Directions book, has assembled sixteen translators, some of them eminent poets, others well-known scholars like Mario Praz. He writes in his introduction: ‘In some instances it seemed advisable to offer more than one translation of the same poem. One flame can be refracted through different crystals, one music transcribed for different instruments.’ With the Italian text before us, and, in some cases, a variant translation, the tentative nature of the enterprise is clarified; a single instance will serve to illustrate this advantage: Montale concludes ‘Notizie dall’Amiata’ with the image, ‘i porcospini/ s’abbeverano a un filo di pietà’. Cambon has chosen two English versions; that of Irma Brandeis, which reads, ‘the porcupines/ will sip at a thin stream of pity’, and Lowell’s ‘imitation’, ‘The porcupine sips a quill of mercy’. George Kay’s rendering, in the Penguin edition, is adequate and more or less literal: ‘Porcupines/ drink long at a trickle of pity.’ But to have a close translation, and then, once the line is understood, be rewarded with Lowell’s rich interpretation, with its echoes of his own ‘Skunk Hour’, renews and amplifies Montale’s words.

     In a recent number of Letteratura, Montale noted that ‘lyric poetry is untranslatable and almost incommunicable’; perhaps this has come to be so now. Still, the life of a text can revive and be heard through some other voice, be made to recapitulate some distant, other form. I would side with Renan, who said once, ‘A work that has not been translated can only be considered half-published’.