But the relation of the poem to its translation is not that of an object to its shadow, nor even that of an object to its reflection in a mirror. What that relation really is, I will define in a moment.
To understand the relation of the translator to the poem, one must understand the relation of the translator to the original poet.
Now if the notion that the poem-in-translation pre-exists and must be uncovered is my first false but necessary proposition, my second false but necessary proposition is this: Craft becomes becoming. By this I intend that the translator should feel himself or herself to be the poet, reborn into another language. Though the thesis is magical, the process is not: it is gradual. First the translator begins, bit by bit, to discover the poet’s verbal and ideational habits; then, little by little, the translator acquires the poet’s obsessions. Next the translator is able to take in the ruling obsession of each single poem. Next the translator is certain he or she inhabits the poet, walking up and down inside the maze of the poet’s convictions, conditions, tics, and itches. And then at last the translator turns into the poet. Not that the translator has “merged” with the poet; but the one has genuinely become the other.
As with the translator, so with the translation. The poem’s translation is not the poem’s shadow or reflection. The poem and its translation are two separate artifacts, each equal to the other; and not only “equal” in the sense of being “alike,” but each having become the other.
We should be able to put two poems before a reader, the original and its translation, and say, “Here are two poems. They are the same poem. Which was translated from the other?” A better question would be: “Which was written first?”—As if the poem preexisted not only in translation for the translator, but in conception for the poet; and as if both the poet and the translator had transcribed from the same source, one in one language, the other in another language.
The only justification for translation is the promise that when you hear the poem in translation, you hear the poem itself.
Perhaps it would be helpful if I paused here to summarize my three false propositions, if only to show even more emphatically how false they really are:
First, that the poem in translation is already there, hidden in the language of translation, waiting to be let out, an imperative, imminent, immanent, immutable form ready for release, and, when released, instantly recognizable as both primary and authentic.
Second, that the translator, in becoming the poet, assumes all the authority of the poet over the poem.
Third, that the translation of the poem is the poem, as much as the poem itself.
Now here is a summary of the ways in which these falsehoods are important and useful for the translator:
One. If you do not believe the poem-in-translation is already there, you will never find it.
Two. If you do not believe that you have full authority over the poem, its form and its meaning, you will have no authority over it.
Three. If you do not believe you can achieve the poem itself, you will be in possession only of a fuzzy shadow and a cracked mirror.
Now you will protest that all these falsehoods and their fine corollaries are only an apparatus to give courage to the fainthearted translator; and I agree; and the moment I agree, my apparatus collapses, and I am left cowering before the poet and the poem.
I am left cowering especially before the Yiddish poem. Most of the time it is no more difficult to translate from Yiddish into English than from French into German; but those times when it is more difficult, it is so much more difficult that the translator despairs. French to German, or German to Italian, or Italian to Russian, brings with it a consistency of theology, hence of culture, hence of role and artifact. But Yiddish to English means a crossing-over from Jewish concepts to Christian concepts, or at best to a secularized sensibility. And whereas prose, whether essay or fiction, can give us a glossary or footnotes, an explanatory parenthesis or an extended paraphrase, the poem is relentless in requiring an equivalent word or phrase, compacted within the compass of three syllables or half a line.
What good then are Ideal Passions about the Power of the Pre-Existent Poem?
A translator at work is embrangled—a word derived from brawl and wrangle—in a thousand questions of diction, gait, suggestion, as opposed to transcription of meter and line length, tonal renderings, fidelity versus sacrifice, transmutation of form; head-cracking marvels and experiments of craft.
What use then are Ideal Passions about the Power of the Pre-Existent Poem?
I have found in an old box in the attic a great helter-skelter heap of papers. In sorting them out, I came upon twenty-two sheets relating to a single line of a single poem. These twenty-two pages are the ones that turned up; for all I know, there may have been many more. The line was the first line, and also the title, of a poem by Dovid Einhorn: geshtorbn der letster bal-tfile—“the last bal-tfile is dead.” And if I fail at this moment to translate bal-tfile, the reason for it will soon be clear.
The Einhorn poem was among the earliest in the various groups of poems I had translated as a contributor to an anthology of Yiddish poetry, and I came to it as a novice. One of my many errors at that time was to see the translator as a being in thrall to the editor. If the editor offered a suggestion, I took it as an irreversible command. But meanwhile, behind the scenes, I developed a kind of translator’s cunning. It wasn’t the cunning of brains; I couldn’t outwit the editor, who was smarter than I was. But I could outcry him. So, in letter after letter, I raised clamorous laments, I pleaded, I implored, I whined and I wheedled. And as I wheedled the editor toward what I conceived to be the poem’s needs, I discovered at last that the poem, all on its own, could make unreasonable demands: for example, it was the poem’s assumption that, quite apart from the translator’s being in thrall to the editor, English ought to be in thrall to the poem. Or, in other words, that a pretty good, workable English equivalent was all that was requisite, rather than an exactly nuanced representation.
Very gradually, through a series of feverish letters, the line evolved—from the editor’s suggested line, to the poem’s “English equivalent” line, to that stage where the translator was ready to assume moral authority over the poem. In the final stage the translated line became, not a line of a translated poem, but the line of a poem.
The original contained the word “Shekhina,” which presents no difficulty to the English reader. It is a word both in the English dictionary and in the vocabulary of Western philosophy. But the perplexity lay in the term bal-tfile. For bal-tfile the editor proposed “prayer leader.”
I dispatched the following moan:
If it has to be “prayer leader,” it has to be “prayer leader.” Only this: my spirit drops at the thought of that thin phrase. I have been trying all week with real despair to get at some oblique way of suggesting the role without naming it. The chief trouble with “prayer leader” is that it isn’t poetry. I’ve tried it on the line and it looks grotesque: “The last prayer leader is dead.” It trivializes an awesome idea. It lacks even the smallest redolence of the original. It’s empty-sounding. To one who knows nothing of synagogue practice, it illumines nothing; to one who knows everything, it points to nothing—who will guess bal-tfile from “prayer leader”? Isn’t this a case of “correct” translation resulting in falseness, in violation? “The last prayer leader is dead” sounds to me exactly as fake, as flat, and as silly as an equally data-ridden term would sound, e.g., “The last underpaid secondary cantor is dead.” It moves the poem out of majesty and into personnel. But for the moment I can offer only wails, no solutions.
That was written in April. In August, I find another letter still embroiled in geshtorbn der letster bal-tfile. Apparently I had just discarded the phrase “singer in the pulpit.”
“Singer in the pulpit” [I wrote], though metrically nice (which is why it lured me), wears churchly robes, and is hardly bal-tfile. “Singer before the Ark” came next—at least i
t describes a synagogue—but the bal-tfile is usually not much of a singer, and used in the line it has too many accented syllables anyhow. “Reader of the Law” would suggest that the original is bal-krie rather than bal-tfile. You will say that the obvious thing to settle on, then, is “prayer leader,” which is accurate and neutral enough to come out not entirely Quakerish or Christian Scientist. But it is too bland, I think, just because of its neutrality, so I have taken the risk and stayed with “reader of the Law.” It is, as “translation,” wrong, but in English it is more right than any other alternative. Or so it seems to me now.
To someone not familiar with synagogue practice and personnel, “reader of the Law” is completely in context and holds the poem together. And this would also be true of one familiar with the synagogue, even though he might think bal-krie. If he does think that (though a reader of a poem should stick with the poem, and not try to translate back in his head), not an iota of violence is done to the poem anyhow.
(Einhorn’s shade rises before me and says, “Well, as a matter of fact, I was going to say bal-krie, and it was only an accident I said bal-tfile.”)
At this juncture in the struggle, you observe, I was at that second stage of error where I believed not that the poem is a law over the translator— that would mean “prayer leader”—but that the poem is a law over English, that what is suitable in English will have to do, no matter how mistaken in substance. I was unsympathetic to the poet and was quite willing to call up the poet’s ghost in order to get his approval for a workable English, even if it made him recast his poem. When I insisted to the editor that “not an iota of violence is done to the poem,” I was clearly an advocate of doing violence.
That was August. In September the editor answered as follows:
I agree that “prayer leader” is no good.
So what had I accomplished? I had wheedled him out of one mistake into a new mistake, this one of my own coinage.
By the next month we were skirting bal-tfile, letting it lie fallow, and were now embrangled in the word “dolor.”
I had rendered a stanza this way:
And soundless on the steps of the Ark
the abandoned Shekhina rests,
her head bowed down in dolor,
black as night her dress.
The editor wanted “Jewish Spirit” instead of “Shekhina”; he wanted “grief” instead of “dolor.” I answered:
As for “Jewish Spirit,” isn’t the Shekhina here a concrete figure, like Heine’s occasional Virgin? The poet will have his Astarte, no matter what. To keep faith with the ballad-feel of the poem, I’ve indulged in an old ballad-word, dolor.
Apparently I managed to wheedle out “Shekhina,” but not “dolor,” because my next letter is still at work.
I guess I did choose “dolor” for its archaic feel, so you mustn’t object that it’s archaic; I meant to reinforce the ballad-quality. But if you prefer another word, I can part with “dolor.” What would you think, though, of “sorrow” instead of “grief”? Like “dolor,” “sorrow” carries out all those open vowels of the rest of the line: “head,” “bowed,” “down,” whereas “grief bites the line off rather too quickly, almost as though the stanza were ready to end too soon. Do let your own preference rule, however.
So he let me keep “sorrow.”
Very soon afterward, though, we were back at bal-tfile. Now I was writing hopelessly:
So what’s to be done? Capitulate to literalness, and remove the phrase from poetry and into data? One last-ditch idea, which I throw down on the page in desperation: how about a still more reckless literalness? How about a direct and wholesale translation from the original, how about “master of prayer”? At least it sounds suitably ancient, at least it doesn’t sound Protestant. It has rather a Buberian dignity, a bit of authority, a drop of majesty: “The last master of prayer is dead.” It comes out, if not poetry, a bit closer to poetry. But what I put under the head of dignity, authority, and majesty (not that the person of the bal-tfile has all that; it’s the liturgy I’m thinking of)—what I put under that head, you may pronounce pretentious.
Well, if it has to be “prayer leader,” it has to be “prayer leader.”
Nearly a year later—but I have no documents to show how this came about—I had abandoned the ordeal of the imprecise precision, I had abandoned my trust in English as offering a solution of workable equivalence, and the opening stanza read as follows:
The last to sing before the Ark is dead.
Padlocks hang in the house of the Jews.
The windows are boarded, and shadows
huddle in shame in the pews.
“Pews” seems to me now very bad. But what of “The last to sing before the Ark is dead”? Will that do for gesbtorbn der letster bal-tfile) Has the Pre-Existent Poem been uncovered? If seventy translators went into seventy separate rooms, would they all come out word-for-word with this very line? Or would they all come out with “prayer leader”?
What I did not understand then was that I was not wheedling the editor, but educating myself; that I was not exhorting the editor, but beginning dimly to perceive the terrible complexities of the craft of translation. My trials with gesbtorbn der letster bal-tfile reveal the problem of translation at its most elementary and primitive stages—a tyro’s tale. For a long time I did not comprehend that a translator, though continuing to quail before the idea of translation, must nevertheless not be afraid of the poem that awaits; that the translator must dare to be equal master of the poem together with the poet. I did not sympathize with Einhorn because I did not yet know that I was obliged to become Einhorn. I did not have authority over the poem because I did not believe that it was already there; I thought I had to jerry-build it myself, in various makeshift ways. I did not yet see that the poem had a blueprint of its own, a meticulous blueprint as singular as the whorl on a fingertip; and that what I had to do was not look for the ink to reproduce the print, but look for the inexorable lines of the print itself.
By the time I had acquired some experience—by then I was concentrating on the poet H. Leivick—I had learned to trust the doctrine of the Pre-Existent Poem, I was a believer, it seemed to me I was becoming Leivick. With a poem called Tate-legende I was surely Leivick, and one of my letters to the editor records how that extraordinary realization opened itself out:
Meanwhile, as you suggested, I’ve gone ahead with Tate-legende [Father-Legend]. An extremely affecting poem, clear-eyed, sinuous, unsentimental. I toiled over it with a kind of calculating joy—I have about sixteen pages of crowded work-sheets, filled with calculated alternatives: I imagine I keep saying “calculate” because I took risks here and there to seize the tone. I am just now too much devoted to the poem to tell whether it all works; but you will tell me that.
One very, very still night, coming to the words yingl du mayner, yingl du mayner, I ikh bin dayn tate der roiter [oh my child, my child, I am your red-haired father], I all at once felt Leivick’s father’s ghost enter me. Through the ribs and throat.
You can hear in that last paragraph how “calculation” and “risk” suddenly fly away, replaced by becoming.
Published in Prooftexts, vol. 3 (Baltimore: Johns Hopkins University Press, 1983), pp. 1–8
S. Y. Agnon and the First Religion
Shmuel Yosef Agnon, the 1966 Nobel winner for literature, was born one hundred years ago, in Galicia, Poland, and died in Jerusalem in 1970. Not long after his death, I wrote a story about Agnon, a kind of parable that meant to toy with the overweening scramble of writers for reputation and the halo of renown. It was called “Usurpation” and never mentioned Agnon by name. Instead, I pretended he was still alive, not yet a laureate: “It happens that there lives in Jerusalem a writer who one day will win the most immense literary prize on the planet.” I referred to this writer as “the old man,” or else as “the old writer of Jerusalem”—but all the while it was Agnon I not so secretly had in mind; and I even included in my story, as a solid an
d unmistakable clue, one of his shorter fables: about why the Messiah tarries.
To tell the truth, this midrashic brevity (God knows where I came upon it) was the only work of Agnon’s I had ever read. Nothing could have tempted me to look more extensively into Agnon, not even the invention of a story about him: enchanted by the dazzlements his great name gave off, my story was nevertheless substantially blind to the illuminations of his pen. I could scarcely blame myself for this. For decades, Agnon scholars (and Agnon is a literary industry) have insisted that it is no use trying to get at Agnon in any language other than the original. The idea of Agnon in translation has been repeatedly disparaged; he has been declared inaccessible to the uninitiated even beyond the usual truisms concerning the practical difficulties of translation. His scriptural and Talmudic resonances and nuances, his historical and textual layers, his allusive and elusive echoings and patternings, are so marvelously multiform, dense, and imbricated that he is daunting even to the most sophisticated Hebrew readers. What, then, can a poor non-Hebraist possibly make of an Agnonic masterwork when, willy-nilly, it is stripped of a quarter or a half of its texture and its substance, when the brilliant leaves are shaken off the spare, bare, naked-toed trunk? A writer in monolingual America, confined to writing and reading wholly in English, will clearly have no Agnon other than the Agnon who has been Englished. If the prodigal Agnon can be present only in Hebrew, to read him in any other tongue is to be condemned to paucity. The Hebrew prince is an English-language pauper.
So, drawn almost exclusively to the lustiness of literary blue blood, unwilling to see it ransacked and pauperized, it is no wonder that I have kept my distance from the translated Agnon.
Metaphor and Memory Page 22