Translating As Saying

Lawrence Rosenwald


In this essay, Rosen­wald offers some crit­i­cal reflec­tions on an impor­tant claim made by Mar­tin Buber and Franz Rosen­zweig about their trans­la­tion of the Hebrew Bible: that the trans­la­tion was essen­tial­ly an oral text rather than a writ­ten one. He sets out the sig­nif­i­cance of that claim, and the ways in which the text val­i­dates it, but also inves­ti­gates the aspects of oral­i­ty that are absent from the text, and the sig­nif­i­cance of those absences. Final­ly, he con­sid­ers how Buber’s 1958 record­ing of the text for West Ger­man radio actu­al­izes the bib­li­cal oral­i­ty that the trans­la­tion aspires to but does not entire­ly accomplish.

Click here for a pdf of this article.

“A Jewish way of saying things” is a phrase of many meanings, each brought out by stressing one of its components. The most tempting to stress is “Jewish,” since doing so promises to differentiate between what is Jewish and what is not. The one stressed in the present essay, though, is “saying,” the goal of the essay being to get a sense of some modes of Jewish spokenness and of spokenness more generally.

There are four central texts here: the Hebrew Bible as understood by Martin Buber and Franz Rosenzweig, who see—or rather hear—in it an essentially oral text; the splendid, controversial translation they produced, aiming above all at representing that orality; their essays in explanation and defense of that translation, especially Rosenzweig’s “Scripture and Word”; and Buber’s 1958 recording—Buber’s “way of saying” in a strict sense—of some of the great passages from that translation. In my judgment, it is in Buber’s recording that the biblical orality aspired to in the translation is finally accomplished. Translation is made complete.


Rosenzweig’s “Scripture and Word” is the most oracular of the essays collected by Buber in 1936 for Scripture and Translation, but also the most precise account of the Bible’s orality as Buber and Rosenzweig understood it. “Every word is a spoken word,” it begins, and after embroidering that claim turns to the Bible in particular, arguing that this book more than any other book must not become entirely a text, that in it and not just near it the word, the spoken word, must be retained. 1 1 Martin Buber and Franz Rosenzweig, Scripture and Translation (Indiana University Press, 1994), tr. Lawrence Rosenwald with Everett Fox, p. 40. Page numbers for subsequent quotations from this work will be given in parentheses in the text. Sometimes I have silently modified my own translation, either because I have come to understand the passage differently or because the modified translation works better in the present context. Rosenzweig nods in passing to the spokenness of Protestant Reformer Martin Luther’s sixteenth-century translation of the Bible and its reliance on “the spoken speech of the people” (42), then turns to the very different spokenness of his and Buber’s translation; he locates it in their translation’s freedom from the “chains” of logical punctuation, their attention to the drawing of breath.

The distribution of breath-renewing silences follows the inner order of speech, which is only occasionally determined by its logical structure, and which for the most part mirrors directly the movements and arousals of the soul itself in its gradations of energy and above all in its gradations of time. . . . Sentences that in unambiguous logic are distinct and so separated by periods—say, Cain’s appalling answer, “I do not know. Am I my brother’s keeper?”—are by the rendering of the vital, breathing course of speech brought together into a single movement, and thus given their full horror, previously half covered over by the logical punctuation. (42)

In the essay’s most analytic section, Rosenzweig distinguishes between his and Buber’s attention to the “inner order of speech” and the attention provided by the biblical cantillation signs, the te’amim, which he calls a “comprehensive logical analysis of the text” (43). He means by “logical” both praise and blame. He notes that Rashi interprets Genesis 1:1 in a way distinctly at odds with the interpretation suggested by the te’amim, and takes courage from his example, seeing in it a justification for the process he and Buber have gone through: figuring out how the “distribution of breath-renewing silences” is to be made to correspond to the “movements and arousals of the soul itself”—i.e., to something different from what ”logic” or syntax might demand.

To be sure, there is something strange here, in Rosenzweig’s setting the spokenness of the translation against the te’amim. However limited the logical, syntactical analysis the te’amim offer, they are in their nature designed to make the text into speech in the strict sense of the phrase; it is through them that those who leyn, who chant the Torah from the scroll, can know what pitches, rhythms, disjunctions, conjunctions the reading voice is to enact. Why does Rosenzweig refuse to acknowledge as much? Whatever his motive, his refusal is characteristic of him, and of Buber; other aspects of the surprising tension between the intended orality of the translated text and the actually spoken and heard orality of the original will surface in what follows.

The translation is then meant to be word not scripture, breathed not punctuated, organic not mechanical, mikra - i.e., read aloud - and not text perceived only by the eye. The chief means of achieving this goal is the division of the text into lines, lines understood as corresponding to breaths and to the inner order of speech. This division is the most immediately striking feature of the Buber-Rosenzweig translation. Consider as an example the passage that Rosenzweig’s comments call our attention to:

Kajin sagte es Habel, seinem Bruder.
Aber als sie dann auf dem Felde waren,
stand Kajin auf wider Habel seinen Bruder and schlug ihn tot.
ER sprach to Kajin:
Wo ist Habel dein Bruder?
Er sprach:
Ich weiß nicht. Bin ich meines Bruders Hüter?
ER sprach:
Was hast du getan!
Horch, das Blut deines Bruders schreit zu mir aus dem Acker.
Und nun,
verflucht seist du hinweg vom Acker,
der den Mund aufmachte, das Blut deines Bruders von deiner Hand zu empfangen.

[Kayin said to Hevel his brother./ But then when they were out in the field/ Kayin rose up against Hevel his brother and he killed him./ HE said to Kayin:/ Where is Hevel your brother?/ He said:/ I do not know. Am I my brother’s keeper?/ HE said:/ What have you done!/ Hear, the blood of your brother cries out to me from the ground./ And now,/ cursed be you from the ground/ which opened up its mouth to received your brother’s blood from your hand.] 2 2Das Buch Im Anfang (Berlin: Schocken, n.d.), tr. Martin Buber with Franz Rosenzweig, pp. 19-20. I adapt Everett Fox’s beautiful translation, inspired but not determined by Buber and Rosenzweig’s work, so as to make it correspond more precisely to the German. See Fox’s In the Beginning: A New English Rendition of the Book of Genesis (New York: Schocken, 1983), pp. 18-20.

The line divisions are all the more conspicuous because in the original edition the lines are almost alone on the page: no marginal glosses, no footnotes, no enveloping commentary, not even chapter and verse indications (the range of chapters and verses covered by the page is indicated, but only at the bottom). All we see is the lines as Buber and Rosenzweig have divided them, in their suggestive, uneven, expressive movement and pattern.

Reading that translation, silently or aloud, we experience something of what the translators sought to represent, that movement of breath from which speech emerges, on which speech depends, by which some of speech’s meanings are created. Consider as examples the separating out by breath and line of the painful transitions and indications “he spoke,” and “and now”; the linking of Kayin’s profession of ignorance to his troubling question; and the longer lines with their implicit accelerando, not “Kayin stood up against his brother/ and struck him dead,” but the two clauses linked by the breath and by the mind moving irresistibly from the first to the second, as if standing up and striking were not two actions but one.


But something crucial connected with speech is absent from the Buber-Rosenzweig translation. We can begin describing it by returning to Rosenzweig’s point about Luther’s translation, that it draws on the spoken speech of the people. This is no metaphor; Luther when translating went from the Wartburg into the streets of Eisenach to listen to people speak, on the street and at home and in the market, men and women and children, “looking them on the mouth” so as to capture their language for his translation. 3 3 Luther as quoted in Philip Schaff, History of the Christian Church (Peabody: Hendrickson, 1996), vol. VII, ch. 4, sec. 62. “The spoken speech of the people,” “looking them on the mouth”: both phrases suggest a different sense of orality than what we experience in the Buber-Rosenzweig translation: not breath that underlies speech but the speech that the breath makes possible.

Robert Frost’s phrase “sentence sound” best defines the mode of orality in question, a mode indispensable to some of Frost’s great lines: “And to do that to birds was why she came,” for example. 4 4 For “sentence sound” see The Collected Letters of Robert Frost (Cambridge: Harvard University Press, 2014) Donald Sheehy et al. eds., p. 233. “And to do that to birds was why she came” is the last line of “Never Again Would Bird Song Be the Same.” The line is unsurpassable partly because it is so sayable; it purifies but does not purge ordinary speech, as English iambic pentameter, the meter of that line, purifies but does not purge the rhythm of that speech. Appropriately enough, the Frost-influenced poet and translator Robert Bly in Eight Stages of Translation sees one of those eight stages as being the production of sentence sound in the translated text. “The aim is not street language, not slang as such nor the speech rhythms of half-educated people, but rather the desperate living tone or fragrance that tells you a person now alive could have said the phrase. Robert [Frost] believed in such rhythms and wrote of them brilliantly; he called the fragrance ‘sentence sound.’” 5 5 Bly, Eight Stages of Translation (Boston: Rowan Tree Press, 1983), p. 25.

Nothing of this sort is audible in the Buber-Rosenzweig translation, which is attentive to breath but not to speech or voice or tone. It is in fact uncannily toneless on the page. (My native speaker friend and colleague Jens Kruse once said to me that the translation seemed to his ear to have “almost no tone at all.”)

Nor could tone in this sense have been their goal; their enterprise as translators was in fact programmatically at odds with “the spoken speech of the people.” The spoken speech of any people, the sentence sound of any people, is rooted in place and time and identity. The quality of sentence sound is accordingly hard to translate, because whatever possesses sentence sound is beautifully but inextricably so rooted, and how can the translator uproot it and transplant it successfully? When such transplantation of idiom succeeds—say in Hillel Halkin’s translations of Sholem Aleichem’s Tevye stories and Motl the Cantor’s Son, or the late Sacvan Bercovitch’s translation of the same author’s “The Pot”—it feels like magic.

Buber and Rosenzweig approached the question of idiom differently. They believed, as Rosenzweig wrote, that “there is only One Language.” 6 6 Rosenzweig, Jehuda Halevi: Fünfundneunzig Hymnen und Gedichte, Rafael Rosenzweig ed. (The Hague: Nijhoff, 1983), p. 3, my translation. That is a visionary and inspiring belief. But acting on it risks minimizing the differences between the languages we actually speak. Acting on it in their Bible translation committed Buber and Rosenzweig, moreover, to finding equivalents in German for everything they discerned in Hebrew. They accepted this obligation because they believed that inevitably equivalents would be found: “there is no linguistic peculiarity of one language that cannot be found contained, at least in embryo, in every other language, even if only in idioms, in nurseries, in jargons.” 7 7 ibid. To find them they ransacked the long history of the German language, all of its periods, all of its places. 8 8 Albrecht Schaeffer praised the translation’s “Germanness in a higher sense . . . the supratemporality of its poetical language.” Schaeffer, “Bibel-Übersetzung: Aus Anlass der neuen von Martin Buber and Franz Rosenzweig,” Preussische Jahrbücher 225:1 (July 1926), p. 73, my translation. That gave them an incomparable treasure-house of possibilities but led them away from the spoken speech of the people—which in Luther’s case meant not the people in general but the people of Eisenach in the early sixteenth century, and in every case means the people of a particular place at a particular time. Siegfried Kracauer’s invidious comparison of the Buber-Rosenzweig translation with Wagner’s libretti is largely false, but the two dictions have in common breadth of vocabulary and indifference to idiom. 9 9 Kracauer, “Die Bibel auf Deutsch,” in Das Ornament der Masse (Frankfurt, 1963), p. 180.

To be sure, there are two senses of idiom in play here. There is what is fluent, unmarked, unremarkable, in easy accord with the intuitions of a native speaker. Buber and Rosenzweig are right to reject idiom in that sense as a translator’s constraint; it is at odds with both fidelity and poetry. But idiom in another sense is of high value, because a voice is audible in it. Buber and Rosenzweig got rid of idiom in both senses.

The reader will have noted that the account so far offered has little to say about the Jewishness of the translation’s way of saying things. That silence is nearly inescapable if one stays inside the text of Buber and Rosenzweig’s account of their translation, which almost never presents the translation’s special features, its orality among them, as Jewish. 10 10 On this see my “Between Two Worlds: Martin Buber’s ‘The How and Why of Our Bible Translation’,” Jewish Studies Quarterly 14:2 (2007), pp. 144-51. Rosenzweig’s great essay “Scripture and Luther,” for example, considers Luther’s great Bible translation, does it honor, but also shows its philological and theological limitations. We as readers want Rosenzweig to link these differences to the differences between Christians and Jews. But he never does; he refers to “our time” and “this man” rather than “our Jewish community” or “a Jewish reader.”

The only essay by either translator to consider the Jewishness of the translation as central to its nature is Buber’s “The How and Why of Our Bible Translation,” written in Palestine around 1938. He situates the genesis of the translation in his experience of a Berlin Stammtisch:

In the years before the Great War I had been a member of a Berlin Stammtisch, which met once a week over a glass of wine . . . The relation between the German and Jewish members . . . was one of openhearted camaraderie. But some of the Jewish members seemed sometimes to themselves like excerpts of a Hebrew book read in Luther’s translation; it seemed to them that what counted in the camaraderie was not the Urtext of their being but the translation, the translation that was more beautiful than true. (208-09)

Even here, though, Buber makes no specific connection between the Jewishness of the translation and its orality; the connection is rather between its Jewishness and its fidelity.

At one point he comes close to making the more specific connection. The difference between Luther’s enterprise and theirs, he writes, is that Luther’s was to “make available to perception whatever seems to ‘practice’ [the logos], i.e., to ‘practice’ Christ,” whereas theirs was “but to free the real, spoken, and speakable word that lies caught in Scripture, and to let it sound again in the world” (215). But he does not draw out that association, between the spokenness of the translation and its Jewishness, and though he hints at the influence of Luther’s Christian theology on his translation, he is silent on his own Jewish one. He leaves the question bewilderingly, temptingly, infuriatingly open.

Mar­tin Buber reads from Das Buch Im Anfang, tr. Mar­tin Buber with Franz Rosen­zweig, record­ed for West Ger­man Radio in 1958.


In 1958, Buber recorded passages of the translation for West German Radio. 11 11Martin Buber liest aus der Heiligen Schrift Israels (Heidelberg: Christophorus Wort, 1958). Buber seems to be reading chiefly from the final version of the translation, the one now published by the Deutsche Bibelgesellschaft; when that version differs from the earlier Schocken edition in my possession, published in Berlin though without an indication of date, it is almost always the former version he reads. But his reading is not identical to the former version either. Clearly Buber’s Jewish way of saying things involves a degree of improvised freedom from the text he has before him! What the translation banished, the recording restores.

There is of course something preposterous in seeing the translation and the recording as comparable. The translation was a sensation; the reviews were abundant and passionate, and whether positive or negative they made clear the importance of the enterprise. 12 12 See Rosenwald, “On the Reception of Buber and Rosenzweig’s Bible.” The recording is known to few, made no waves. It is not a great performance. Sometimes it feels flat, and sometimes there is a strained, over-willed quality in the passages where Buber is seeking a prophetic grandeur.

But the comparison is necessary, because in performing the translation Buber restores to it what it was meant to offer but systematically excludes: a voice, an idiolect, a performance, a place. Every text when spoken takes on a “local habitation and a name,” and in more than one way. Buber’s voice is a recognizably male voice, high baritone in register. He was born in Vienna and lived there for three years; his German is Austrian, an origin manifested in, among other things, the fact that in his mouth the final g of heilig is sounded as /k/ rather than as the frontal /ch/ of ich. But his German is also a high Austrian German, not strongly marked by Viennese characteristics. 13 13 My thanks to my colleague Thomas Hansen, of the Wellesley College Department of German, for giving me a precise sense of Buber’s idiolect.

It is also a Jewish German, most notably in that in words like gingen (“went,” third person plural past tense) the /g/ of the final syllable is sounded after the velar nasal of /ng/. This is non-standard in German but standard in Yiddish. Buber grew up hearing both languages in the home of his grandfather Solomon in Lvov; the crossover is hardly surprising but vividly characteristic. (Gershom Scholem’s German, which one can hear in a wonderful recording of his called Die Erforschung der Kabbala, never exhibits this striking trait of Buber’s.) 14 14 Scholem, Die Erforschung der Kabbala (Berlin: supposé, 2006:1967). Buber’s idiolect makes the eerily toneless German of the translation a recognizably Jewish way of saying things.

The reading is also an interpretation, an individual’s particular and willed way of saying a particular set of things. This is of course inevitable; a written text is, when one is preparing to perform it, like a musical score. It offers some constraints and guidelines: the words, the punctuation, the paragraphing, the lineation. Some texts offer more: rhythm, dialect, volume, tempo, silences. But no textual notation system specifies every aspect of performance, so inevitably something is left to the micro-compositional decisions of the performer, whether Glenn Gould playing Bach or Charles Dickens declaiming his own Oliver Twist.

These general reflections are of special importance here. That is partly because of Buber and Rosenzweig’s passionate view that the Bible itself is not only text but also voice and word, partly because of their tenacious labor to make the translation be that as well. But it is also, and crucially, because of the way that the translation is not what its makers desired it to be; it is less spoken, in certain senses, than translations about which no such claims are made, concerning which no such ambitions are felt. In these contexts, Buber’s choices and practices as reader take on a representative importance.

Here are some of them. The tetragrammaton is indicated in the text by singular pronouns in small capitals, masculine pronouns whenever gender is indicated—Da rief SEIN Bote, then HIS messenger called (Genesis 22:11). Buber always speaks the Name more loudly than he does the words around it, and more slowly, and at a somewhat higher pitch—this regardless of who the speaker is, whether Abraham or Jehoshaphat or the false prophets telling Jehoshaphat and Ahab what they want to hear. In general Buber’s speech is slow, in general he observes the line divisions. Sometimes, though, the pauses within lines are as long as those between them, e.g., in the single line in which God indicates to Abraham whom he is to sacrifice: “take your son, your only one, whom you love, Yitzchak,” the pauses between epithets being as long as the pause between the end of the line and the beginning of the next, “And go to the land of Moriah.” Sometimes the performance is almost pedantic, as if he were seeking to help pupils taking dictation, for example in his pronunciation of darhöhung, “therehighing,” an etymologically based rendering of a word often translated “sacrifice”; Buber sounds the second /h/ as well as the first, slowing the word to an extreme degree.

Most of the time Buber’s reading has as little of the tone of ordinary speech as the translation does. But not always. Isaac’s speech to Abraham, for example—“Here are the fire and the knife,/ but where is the lamb for offering?”—has a child’s colloquial tone to his father, confident and almost querulous. Jezebel consoling Ahab, after Naboth has refused to sell him his vineyard, has the bracing cheerfulness of a wife consoling her husband after, say, their son’s soccer team has lost a close game. The tempo is hardly ever fast, but becomes sometimes poignantly, unbearably slow. Sometimes the performance is at odds with the textual punctuation. Isaiah’s imploring command to God, “send me,” has an exclamation point in the text but nothing corresponding to it in Buber’s performance.

In all these ways, Buber’s recording restores the elements of spokenness that are missing from the putatively spoken translation. The translation makes the recording possible, of course; but the recording makes the translation complete.

Is that spokenness distinctively Jewish? The question is as pressing, and as challenging, and as unresolved by Buber’s own commentary, as the more general question considered earlier, about the Jewishness of the spokenness of the translation’s text.

In my judgment, the answer is “yes,” but in a way as distinctive and unusual as the translation’s own. Often we think of Jewish ways of saying things as ways of saying by which Jews talk with Jews—“two Jews talking,” in Ruth Wisse’s phrase. 15 15 Wisse, “Two Jews Talking: A View of Modern Yiddish Literature,” Prooftexts 4:1 (1984), 35-48. Tevye the Dairyman with Sholem Aleichem, the narrator with Hersh Rasseyner in Chaim Grade’s “My Quarrel with Hersh Rasseyner,” Shulames and Solomon in Kadya Molodowsky’s tsum meylekh shloyme kumt di herlekhe shulamis, Everett Gendler and Art Green in their recent exchange of views over Gendler’s Judaism for Universalists. 16 16, consulted July 24, 2016. Buber’s performances cannot be so understood.

In 1958, when he made the recording, the German-Jewish community in which the translation came into being had largely been murdered. Gershom Scholem pointed out as much in his speech at the event celebrating Buber’s completion of the translation, in 1963: “For whom is this translation intended and whom will it influence? . . . The Jews for whom you translated this are no more.” 17 17 Scholem, The Messianic Idea in Judaism (New York: Schocken, 1971/1995), Michael Meyer tr., 318. One can hear bewilderment as well as disdain in those lines; why, he seems to be asking, would anyone want to complete such a translation? German-speaking Jews were mostly “no more,” and surely many German-speaking non-Jews had been actively or passively among their killers. Whatever the translation was, it was not at its completion an instance of two Jews talking.

But was it in fact for Jews that Buber made the translation? If Buber’s way of saying is Jewish, it is to a significant extent a Jewish way of saying things to Gentiles. One might go further and characterize it as a Jewish way of saying things to Gentiles that has been defiantly imprinted on the German language. The background, or the interlocutor, is the argument made in the 30s by German antisemites that German Jews wrote inauthentic German. An anecdote told by Marcel Reich-Ranicki can stand for the whole loathsome discourse. In 1936, Reich-Ranicki came across an article in a Nazi magazine. The goal of the article was to denounce the German of Heinrich Heine, even—or especially—in two of his most popular poems, “The Loreley” and “The Two Grenadiers.”

Both, argued the author, were representative of Heine’s inadequate and shallow German and his “unshed Yiddish.” Witness, wrote another Germanist at the time, even the first line of the “Loreley,” ich weiß nicht, was soll es bedeuten. A German man would have written, ich weiß nicht, was es bedeuten soll.

Buber lived through that time and those absurd but lethal attacks. His recording is among other things an assertion of Jewish presence in the German language even after the Shoah. “Here I am,” says Buber, “hineni.”

Which makes sense. The account of the genesis of the translation quoted earlier situates it in a context in which Jews are speaking chiefly to non-Jews. “We had only an indirect way to proclaim our truth,” Buber writes: “a faithful translation of scripture.” But “proclaim our truth” to whom? Clearly to the “German,” i.e., the non-Jewish, members of Buber’s Stammtisch, risking whatever reactions those members might have. Not to other Jews; Jews are the proclaimers, not the audience. In its genesis as in its completion, then, and with all the complex peculiarities of its relation to spokenness, the Buber-Rosenzweig translation as written or as spoken was in significant measure a Jewish way of saying things to non-Jews.

When writing “Scripture and Word,” in 1925, Rosenzweig was incapable of speech, paralyzed by the amyotrophic lateral sclerosis that would kill him four years later. He could communicate his thoughts only by writing, and that only indirectly, indicating to his indefatigable, faithful, generous helpers, by means of a movement of his eyelids, the next letter or word in the sentence he was creating. What writer ever needed more to claim that a text was in itself speech, given that for Rosenzweig speech could come only by way of writing? It feels almost blasphemous to challenge him.

That said, however, I like to imagine that Buber and Rosenzweig would have had some sympathy with my reflections on their work. Few translators responded as generously as they did to their critics, for one thing. Few cared as much about spokenness, few have had such an acute sense of how easily spokenness is lost even when it is being sought, and few would be more open to acknowledging that calling spokenness into being requires every means necessary.

Rosenwald, Lawrence. “Translating As Saying.” In geveb, June 2020:
Rosenwald, Lawrence. “Translating As Saying.” In geveb (June 2020): Accessed Jan 17, 2021.


Lawrence Rosenwald

Lawrence Rosenwald is the Anne Pierce Rogers Professor of English at Wellesley College, where he is also the Director of the Peace & Justice Studies Program.