SUSAN BERNOFSKY
Revision isn’t the first thing that comes to mind when we think about creativity and artistic production, but it is nonetheless a crucial part of the writing process. Occasionally a revision radically alters a book’s conceptual framework (as when Franz Kafka decided to have his protagonist Josef K. “arrested” rather than “captured” in the opening sentence of The Trial); more often, it’s a matter of fine-tuning a work’s nuances and voice. Many authors known for the richness of their style rework virtually every sentence: The manuscripts of Flaubert and Proust are thickly crabbed with excised and inserted lines of text, and Nabokov is famous for the index cards he belabored with both ends of the pencil. Shirley Hazzard, to cite a more recent master of the perfectly chiseled sentence, did such extensive revisions on her novel The Great Fire that only about one-third of the published text was original to the first draft. That effortless-sounding prose, those spot-on rhythms came about through sustained labor.
In translation, similar principles apply. Although we strive to produce translations that look as though they hatched perfectly formed from the translator’s skull, generally a great deal of reworking is required. Before I explore some of the issues involved in revising translations and techniques for addressing them, let me provide a quick overview of what the revision process looks like in my case. I tend to revise quite extensively, putting my translations through what usually winds up being at least four drafts. The first of these is intentionally sloppy and quickly executed; it is meant to be discarded but at the same time to serve as a seedbed for ideas worth preserving. This “sketch” of the text gets typed into the computer as quickly and with as little thought as I can manage. After this, I produce a painstakingly meticulous second draft, still working at the computer, but with frequent pauses to consult multiple bilingual and German-language dictionaries. I google German phrases to confirm their usage and look up English words too, in both Webster’s and the Oxford English Dictionary, to make sure a word’s range of meaning makes it the best choice in a given context. Roget’s International Thesaurus (indexed, not in dictionary form) helpfully sorts words by categories, grouping together associated nouns, verbs, and adjectives; and the new historical thesaurus embedded in each of the OED’s entries provides first-usage dates for each synonym given. When translating an older author like Robert Walser, I make a point of verifying the senses in which a word was used a hundred years ago—if it was in use at all. Recently I resisted the temptation to use the word “brinkmanship,” which would have fit handily in one particular Walser sentence, because when I looked it up I found it was coined only in 1956 (when Adlai Stevenson used it in reference to then Secretary of State John Foster Dulles). It just wouldn’t have made sense to invoke that bit of history in a story dating from 1905.
When I finish the second draft of my translation, I print it out, because the next round of revision involves paper and pencil. I’ve found that no matter how carefully I revise onscreen, sentences look different on the page. At this third-draft stage, I try to avoid looking at the original text as much as possible—the point of this draft is to ensure the English-language text works on its own terms. I read the text aloud to myself, since a surprisingly large number of problems that the eye overlooks reveal themselves to the ear. Inevitably I wind up referring to the original, because trouble spots in the English can often be explained by some difficulty in or with the original text. The fourth, and with any luck final, draft is mainly a last read-through to check for any bits that stick out as unpolished, unclear, or stylistically discordant.
This is a procedural road map for revision; it describes the form the work takes, not the work itself. The word “revision” derives from the Latin revisere (“to look again”), and it’s important to remember that revision is about looking: scrutiny and inspection, a critical looking. The problem I most often encounter when coaching younger translators is a willingness to be too easily satisfied with a solution to a problem, too quickly convinced that a line is “good enough.” In this deadline-driven world, the temptation to accept a line and move on is enormous. In literary terms, however, a line is almost never good enough, and often even a line that is good enough might be made even better. The true translator is a person who sees this circumstance as a cause for celebration rather than despair. There is rarely a single perfect solution to any given translation problem, and so the process of revising involves trying out dozens of potential solutions until one of them begins to shimmer in that peculiar way that marks it as the best possible choice.
The art of revision is generally learned through trial and error; in other words, through practice. Another way to study it is to examine how truly great translators revise their work. Unfortunately, it isn’t so easy to find examples without heading to the archives to study translators’ papers. One of my favorite examples of a beautifully revised translation happens to be in German. It comes from August Wilhelm Schlegel’s 1799 translation of The Merchant of Venice, Der Kaufmann von Venedig. Schlegel was one of the greatest German translators of all time, and his translations of seventeen of Shakespeare’s plays were instrumental in laying the foundations for German Romanticism. This master of rhythm, tone, and nuance revised his translations so heavily that his manuscripts contain, for example, five completely different versions of this one simple line: “Sir, I would speak with you” (IV.ii.12, spoken by Portia’s handmaid Nerissa disguised as a man). It doesn’t sound like a line that would give its translator much trouble. But Schlegel seems to have found it quite troublesome.
Here are Schlegel’s five versions of the line:
Herr, ich muß mit euch sprechen. (Sir, I must with you speak.)
Ich muß euch sprechen, Herr. (I must [to] you speak, sir.)
Herr, laßt euch etwas sagen. (Sir, let [to] you something be said.)
Ich wollt’ euch etwas sagen, Herr. (I would like [to] you something to say.)
Herr, noch ein Wort mit euch. (Sir, a word with you.)1
All of these are perfectly correct translations of Shakespeare’s line. The first two are both pretty good, the third and fourth somewhat less so, and the fifth is splendid. To begin with, Schlegel translates the line following the path of least resistance: “Herr, ich muß mit euch sprechen” (Sir, I must speak with you). He follows the structure of the original line as precisely as the German language allows, putting the verb “sprechen” (to speak) at the end only because of standard German syntactical conventions. With this “muß” (must), Schlegel gives the verb “would” an emphatic reading: I would speak with you, I must. That works fine, and Schlegel could have declared himself satisfied with this version. But there is something not quite satisfying about the rhythms of the line. Sir, I would speak with you. Herr, ich muß mit euch sprechen. The line has shrunk from three to two stressed syllables, Shakespeare’s iambs are now trochees, and the German line gets mumbly around the middle, trailing off into a weak feminine ending in “sprechen.” In fact, after “Herr,” the only naturally accented syllable of the line is the first syllable of “sprechen.”
Next Schlegel tries putting the “Herr” at the end of the sentence and cutting the grammatically optional “mit” (with), giving him: “Ich muß euch sprechen, Herr” (I must speak [to] you, sir). This is already quite an improvement. Starting the line with “ich muß” (I must) gives the “muß” an accent it didn’t have in his first version. Suddenly the line has three real stresses again, making it sound stronger. It’s pretty good, good enough. Schlegel could have stopped here. But there’s still something not quite perfect about the line. The new stolidly iambic rhythm means there’s no caesura to set off the “Herr” at the end: the line just rolls from one end to the other without a rhetorically appropriate pause, and shifting the “sir” to the end makes the request less polite.
Schlegel’s next attempt is the version “Herr, laßt euch etwas sagen”—literally “Sir, let something be said [to] you,” but really meaning “allow me to say something to you.” Rhythmically, this is an improvement over the version Schlegel started with: the “laßt euch” sharpens the caesura after “Herr.” But the line is now too long—it’s gained a syllable—and the speaker is no longer explicitly present. In place of the speaker’s stated desire to converse, we are left with the notion of allowing the message to be uttered, which is semantically weaker.
Looking at this series of revisions up to here, it would appear that Schlegel is getting frustrated with the difficulty of the line and has started trying out experimental solutions (such as dropping the subject); this doesn’t quite work. Nor does his fourth version (“Ich wollt’ euch etwas sagen, Herr”—I would like to say something [to] you, sir): the line is still too long, the “sir” impolitely at the end. But having experimented with various rhythms and the possibility of a sentence without an “I” prepares Schlegel to pull off a sort of quantum leap of revision when he writes: “Herr, noch ein Wort mit euch” (Sir, a word with you). The rhythm of this line perfectly captures the cadence of the original, matching it stress for stress, and the tone of the request is also appropriate. By allowing himself flexibility in the line’s semantic content and above all by listening to the rhythm of the words, Schlegel became able to hear the line in German.
The shift of approach that gives Schlegel’s final version of the line its strength involves the addition of a word that has no direct antecedent anywhere in the original: the flavoring particle “noch.” This specifically German part of speech is a variety of adverb we don’t have in English. The function of German flavoring particles in their sentences is less semantic than tonal, though all these words do have individual meanings of their own (“noch,” for example, means “in addition,” “still,” or “yet”). But here the main role of the word “noch” is to “soften” the sentence, making the demand for “a word” sound a bit less forceful, a bit more respectful. It also, as it happens, fills out the rhythm of the line, providing just the unstressed syllable Schlegel needs to match Shakespeare’s rhythm beat for beat. The key to this successful revision, then, involves falling back on a strength of the German language to express in both sense and tone something that was said quite differently in the English original.
What can be learned from watching Schlegel revise? Often it is only by revisiting a sentence again and again that we can get our brains to stop repeating the thought patterns that brought us to solutions previously tried and rejected, opening up access to a new set of possibilities. I found myself remembering Schlegel’s revision process last year when I was faced with the particularly difficult—though on the surface simple—concluding sentence of the story “New Year’s Page” by Robert Walser. Since “New Year’s Page” was also the final piece in the collection Microscripts, the sentence had to be strong enough to end a book on.
“New Year’s Page” concludes with an observation on the relationship between endings and beginnings. In my third or fourth draft, the final sentences looked like this: “When a year stops, another instantly commences, as if one were turning the page. The story keeps on going, and we see the beauty that lies in connectedness.” The final phrase of the last sentence was my attempt to render the line “und man sieht die Schönheit eines Zusammenhangs.” The word “Zusammenhang”—literally “context,” or “connection” in a sense not involving human relationships—proved surprisingly recalcitrant. For the very first draft I had tried “and we see the beauty of a context,” a quite literal rendering of the original, but it didn’t make rhythmical sense as the ending of a sentence, much less of a story or book. What I wanted was a phrase that would work in the manner of Wallace Stevens’s lines “I do not know which to prefer, / The beauty of inflections / Or the beauty of innuendoes, / The blackbird whistling / Or just after.” For a long time I played with variations on this theme, such “the beauty of contextuality” and “the beauty of contiguity.” But all these variants sounded too academic and tricky. So I tried other approaches: “the beauty of having a context”; “the beauty given us by context”; “the beauty of things in context”; “the beauty of things in their contexts.” None of these phrases was right; it just didn’t seem plausible that someone would have chosen to write them in this way. So I let go of the word “context” and tried out other possibilities: “and we see what is beautiful about contiguity”; “and we see what is so beautiful about contiguity”; “and we experience the loveliness/beauty of contiguity.” But “contiguity” wasn’t really the right choice, since it refers more to the continuousness or close proximity of objects to one another, and not so much to the fact that, taken together, they form a context or whole. So I tried “and we experience the beauty of continuation” (similarly problematic). And then: “the beauty of connection.” But the problem with “connection” is the same as with “the beauty of connectedness”: in each case, human connection is implied, rather than the things that make up a world joining together to constitute a context within which humans exist and act. So I tried yet another approach, one that would refocus attention on the things: “the beauty of linking things together”; “contextual beauty”; “the beauty of putting things in context”; “the beauty of things linked together”; “the beauty of things conjoined.” I liked these a bit better, but they all ended too abruptly; none felt dramatic enough to serve as a final phrase for the story.
It wasn’t until I’d puttered about in this sentence for several weeks that it occurred to me I could solve the problem by recasting the sentence entirely. This was like the moment of inspiration that prompted Schlegel to insert a German flavoring particle into a line of Shakespeare. I realized that the main difficulty with the word “context” was that it wasn’t a rhythmically appropriate word to end the sentence on; if I could rearrange the final part of the sentence, I thought, maybe the cadence might be differently balanced. So I decided to try translating the “man sieht” (“one sees”—which I’d rendered without much forethought as “we see”) differently so as to turn “context” into the subject rather than the object of its phrase; then it would no longer have to conclude the English sentence. I tried “The story keeps on going, and the beauty of a context appears to us” and “the beauty of a context reveals itself.” These were all near misses. Then it occurred to me to try the passive voice: “and the beauty of a context is revealed.” Immediately it was obvious how much better this version was. For one thing, the story (and book) now ended on the suggestive, luminous word “revealed,” with a stress on the final syllable to underscore the finality. For another, the phrase was now ambiguous in the same way as Walser’s German, with the words “a context” referring both to contexts in general and to the particular context at hand.
Both Schlegel’s translation of the line by Shakespeare and mine of a sentence by Robert Walser aspire to the same thing: achieving that unity of sense and sound that characterizes all good writing. This doesn’t necessarily mean matching the original author’s work syllable for syllable, rhythmically or otherwise. But the translation does have to find a rhythmical identity and integrity that will convince readers they are encountering a genuine piece of writing. When we experience a text as “well written,” it will never be with the thought that the author first decided what she wanted to say and then looked around for the words to express it. Ideally, the “what” of a statement arrives on the page together with the words to embody it—sense accompanies sound, sound accompanies sense. This is what Walter Benjamin describes in his essay “The Task of the Translator” as an organic relationship: the language of an original text, he writes, is like the skin of a fruit that has grown together with it, while that of the translation is like royal robes draped about it. When people complain about “translationese,” it’s generally because they feel this sound/sense unity has not been achieved.
In my translations, this unity tends to arise during the process of revision, which is where the real work of writing occurs. Blocking out the rough contours of a sentence or a paragraph is a preparatory exercise to hearing the text’s heartbeat in the cadences of its phrases. To immerse oneself fully in the work of translation is to become a medium, transcribing a text that exists only as a sort of phantasm in the translator’s imagination: the text that is just like the original but written in a different language. Revising means listening to a potential text, hearing it amid all the rhythmical detritus of inadequate versions. With each successive draft, the text draws closer to the ideal form it will inhabit when its transformation is complete. The process of repeatedly subtracting whatever isn’t working, replacing it with stronger material, is difficult to grasp, describe, and teach. In the end, it is a matter of learning to calibrate dissatisfaction, to judge when a sentence can still be improved on and when a solution—perfect or imperfect—should be left to stand. The best translators are particularly suspicious of the intermediate drafts of their work, of their own ability to produce “good enough” translations.
And the point of all this dogged labor and persistence? To give the impression of effortless-sounding rightness. Although the vision in revision implies something visual, revising has less to do with something seen than with something heard: the text’s voice. Voice is the crux of all translating. Hearing it happens on a noncognitive level, but approaching the text cognitively while listening can help. Are the dominant vowels in a passage bright, soft, or dark? What about the consonants? Sharp and jagged, or sonorous and smooth?
Syntax too requires attention, as different languages have different ways of assembling sentences. Sometimes the best translation is one whose syntactical structure bears little resemblance to that of the original. At the same time, it is important to be conscious of the order in which information arrives. Every sentence is a journey that begins with a particular phrase or image and takes the reader somewhere. So what does the itinerary of a particular sentence look like, and where does it lead? When I am revising, I pay particular attention to endings: the endings of phrases, sentences, and paragraphs. In writing poetry, it goes without saying that a line’s final word carries particular weight; the same holds true in prose. I often work to sharpen my sentences by ending them on a strong note, even if this means deviating from the syntax of the original.
Style can go soggy in translation. It is important to counteract this softening trend whenever feasible. This might involve exaggerating certain stylistic features—in fact, I believe that emphasizing and underscoring a text’s characteristic attributes is crucial to good translation, a way of turning up the volume on a key aspect of a sentence or phrase to solidify the writer’s voice in the translation. I put this technique into practice in the final sentence of my translation of Hermann Hesse’s novel Siddhartha (Modern Library, 2006), concluding a scene in which Siddhartha’s companion Govinda realizes that Siddhartha has attained enlightenment. My first draft of the sentence looked like this:
Deeply he [Govinda] bowed, down to the earth (ground), before the one motionlessly sitting there, whose smile reminded him of everything that he had ever loved in all his life, everything that had ever in all his life been dear and holy to him.
Tief verneigte er sich, bis zur Erde, vor dem regungslos Sitzenden, dessen Lächeln ihn an alles erinnerte, was er in seinem Leben jemals geliebt hatte, was jemals in seinem Leben ihm wert und heilig gewesen war.
This sentence in German has a feeling of balance and suspension, in large part thanks to the chiasmus at its center that unfortunately doesn’t work so well in English (Hesse writes the equivalent of: “in his life ever … ever in his life”). In revising the sentence, I looked for ways to compensate for the weakened chiasmus and create the sense that all the searching and frantic motion we have seen throughout the novel has given way to stillness and peace. “Holy” is not the final word of the German text. Because of German syntactical rules, the last words of Hesse’s book are the equivalent of “had been.” But it seemed to me that ending the sentence (and book) on the resonant word “holy” would underscore the peace that lies at the end of Siddhartha’s quest. I chose to make “holy” more emphatic by attaching “to him” only to “dear” and not to the unit “dear and holy.” In addition, when I revised I doubled the verb “bowed” to give the sentence a slower cadence and air of finality. Revised, the novel’s last sentence reads: “Deeply he bowed, bowed to the very earth, before the one sitting there motionless, whose smile reminded him of everything he had ever loved in all his life, everything that had ever, in all his life, been dear to him and holy.”
A more dramatic example of turning up the volume in the revision process can be seen in my various revisions of Robert Walser’s 1925 story “Letter to Edith.” This story takes the form of a letter that the narrator is writing to his ladylove, and eventually it becomes clear that he is drinking as he narrates, his sentences becoming loopier and loopier. By story’s end, he sounds thoroughly inebriated: “Ich wankte in eine Konditorei und trank im Wanken sogar noch Kognak. Zwei Musiker spielten mir zuliebe Grieg, aber der Gastwirt erklärte mir den Krieg.” When I was first starting out as a translator twenty-five years ago, I translated draft after draft of this dizzying sentence. One of the early ones sounded like this: “I swayed now into a pastry shop and while reeling drank even more cognac. Two musicians played Grieg for my sake, but the proprietor declared war on me.” Even this imperfect version of the sentence is already somewhat humorous—the narrator is swaying and reeling, drinking, listening to music, and having war declared on him. I puzzled over this war for quite some time. I did understand that part of the reason Walser used the word “war” in the sentence was that the word “Krieg” rhymed with the name of the composer Grieg, but I wasn’t so happy with how the words “declared war” sounded. When I revised the story a year or two later for publication in the volume Masquerade and Other Stories (Johns Hopkins University Press, 1990), I inserted an extra phrase to make the passage “rhymier”: “I swayed now into a pastry shop café and, reeling, if I may, put away some cognac. For my benefit two musicians played Grieg, but the proprietor declared war on me.” Even after the book was published, this passage continued to bother me; I knew it wasn’t quite good enough, and when I wound up writing about this story in an article on Walser’s style several years later,2 I took advantage of the opportunity to revise the sentence one last time.
I had learned some things as a translator meanwhile—including a great deal about just how much freedom a translator really is required to take in order to translate in a way that deserves to be thought of as “faithful.” Revisiting Walser’s drunken sentence, I began to think about the extent to which the rhyme is actually constitutive to the sentence; the semantic value of the word “war” is less important than the fact that the narrator has suddenly started speaking in rhymes in the middle of a prose text. The rhymes exist to provide an index of his drunkenness, and that’s where the humor comes from: he’s so drunk he can hardly walk, and since the way you walk on the page is by writing, the rhymes represent a sort of spinning in circles. So I looked harder at this war declaration—what did it mean? Probably what the proprietor was actually doing in the scene described was taking the soused narrator off to one side and asking him to either pipe down or leave. That might be described, with paranoid exaggeration, as a war declaration, but was the word “war” itself crucial? Maybe not.
Time is the best medicine for translations. If you are able to lay a text aside long enough to forget exactly what you wrote and what the original looked like in the first place, you can revise with a fresh eye. This is what my fresh eye came up with: “I swayed now into a pastry shop café and reeling, if I may, put some cognac away. To please me, Grieg was played by two musicians, but the proprietor now brought out his munitions.” I am delighted with the word “munitions,” even though it took me literally years to come up with it. It isn’t a synonym for “war,” but it comes from the same semantic field and it also, even more importantly, captures the drunken rhyme. It’s helpful that “munitions” is a polysyllabic word with a two-syllable rhyme; this makes it stand out in its sentence, which helps produce the sense of comic emphasis.
In the end, all translation is transformation. It just isn’t possible for a text to work in its new language and context in exactly the same way it worked in the original. When you create a translation of a literary work, you are creating a new set of rules for the text to operate by. This is what revision is for. Only by revisiting a text again and again, doubting and testing the strength of each of its sentences, can we produce translations that merit consideration as works of literature. And yes, somewhere along the line the original text must be forgotten. It takes a certain amount of pluck—not to mention aesthetic sense and the ability to write well in English—to let go of an original long enough to allow oneself to fully imagine the English words that will take its place, but without this no fully realized translation is possible.
Notes
An earlier version of “Translation and the Art of Revision” was delivered as the keynote address at the Fourth Biannual Graduate Student Translation Conference at the University of Michigan in Ann Arbor, April 24, 2010.
1. The drafts showing Schlegel’s revisions are quoted in Michael Bernays, Zur Entstehungsgeschichte des Schlegelschen Shakespeare (Leipzig: Hirzel, 1872), 238.
2. Susan Bernofsky, “Unrelenting Tact: Elements of Style in Walser’s Late Prose,” in Tamara S. Evans, ed., Robert Walser and the Visual Arts, Pro Helvetia Swiss Lectureship 9 (New York: City University of New York, 1996), 80–89.
