Ancient Israelite coins.

In Genesis 31, Jacob and his father-in-law Laban end a long quarrel by piling up stones.

Jacob has worked fourteen years for two wives, six more for his flocks, and watched his wages change ten times. He is fleeing in the night with his family and his herds. Laban catches up. They argue. They reconcile. They build a stone heap to mark the border between them, and they name it.

Laban calls the heap of stones Yegar Sahadutha. Jacob calls it Gal-Ed (Genesis 31:47). The phrases mean the same thing — “the heap of witness” — but they are in different languages. Lavan speaks Aramaic. Hence the famous line in the Haggadah, “My father was a wandering Aramean [an Aramaean oppressed my father]” (Arami oved avi.) Jacob speaks Hebrew. This is the first appearance of Aramaic in the Bible, and it appears precisely where two men end a dispute by naming the same stone in two different tongues; one language will define the Written Torah, the other the Oral Torah—the Babylonian Talmud is written in Aramaic (atop a Hebrew Mishna foundation).

The great Hasidic master Levi Yitzchak of Berditchev reads this verse and makes a bold claim: God knew that His people would one day live in Babylon. He knew they would pray and quarrel and bury their dead in Aramaic for centuries. So God planted Aramaic inside Genesis as an anchor for an exile not yet arrived, so that when the exile came, the sacred text would already be reaching toward the language its readers would need. God, so to speak, licenses translation, or at least allows some vernaculars to attain their place in Scripture itself.

This is a theory of translation, openly theological. The Berditchever says the translatability of Torah is a property of revelation, not a human add-on. Translation is covenantal foresight made textual. The stone heap at the border is not a poetic flourish in the Genesis narrative. It is the architecture of revelation showing itself in miniature: a single object, two names, one history that will require both.

Walter Benjamin, writing in Berlin in 1923 as the preface to his German translation of Baudelaire, spent the most famous essay in modern translation theory reaching toward something the Hasidic master simply had.

Benjamin’s essay is called “The Task of the Translator” (Die Aufgabe des Übersetzers). It is the most cited document in the Western theory of translation.

Benjamin opens by rejecting the premise every theory of translation begins with: that translation exists for the sake of the reader who cannot read the original. No real work of art, Benjamin insists, is made for an audience, and translation is not made for readers either. Translation is the afterlife (Fortleben) of the work — the form a text takes when it passes beyond its own moment into the long weather of linguistic time. A translatable work is one whose meaning was always reaching past the language that first held it. Translation is how that reaching gets done.

Benjamin’s deepest claim, steeped in Lurianic imagery, is that all natural languages are fragments of a greater whole, shards of a vessel. German Brot and French pain point at the same loaf, but their modes of meaning — Art des Meinens — do not coincide. Brot carries the weight of black bread and Lutheran hymns. Pain carries the weight of bakeries and douleur. The languages do not match. They supplement each other. They gesture, between them, toward what Benjamin calls reine Sprache, pure language: a messianic tongue in which all modes of meaning would finally be reconciled.

The translator’s task is not to communicate content but to let these modes echo against one another, to allow the foreignness of the original to disturb the receiving tongue. The German poet Hölderlin, translating Sophocles by bending German almost to breaking under the weight of Greek, is for Benjamin the highest achievement of the form. Translation is a redemptive labor performed on language itself. Laban’s Yegar Sahadutha is not just another way of describing a stone-heap. It is somehow effecting the work of pure language.

Both Levi Yitzchak and Benjamin locate translatability as an intrinsic property of the text rather than something the translator imports. Both see translation as participating in something larger than communication. But the Hasidic reading has three things Benjamin’s lacks: a community, a calendar, and a God who plans ahead. For Levi Yitzchak, Torah calls for translation because God foresaw where His people would live. For Benjamin, a great work calls for translation because its meaning exceeds any single language.

Benjamin’s account requires a messianic horizon — a future reconciliation toward which all translation gestures. But it cannot supply a community for whom that reconciliation is real. The translator in Benjamin’s essay works alone. He bends his native tongue to honor a foreign one, and this bending is its own justification. No synagogue, no congregation, no community of readers waits for the verse to be carried over. Only the translator, the text, the asymptote of pure language, and silence.

Six centuries before Benjamin, six centuries after the destruction of the Second Temple, Jews who could not follow Hebrew gathered in synagogues across Babylonia to hear the Torah read aloud. The reader chanted the Hebrew. Beside him stood another man, the meturgeman, the translator. After each verse, the meturgeman rendered the Hebrew into Aramaic, the spoken tongue of the people in the room. He did this from memory. He was not allowed to read his translation from a scroll, because the rabbis worried that a written translation would compete with the Hebrew original for authority (Bavli Megillah 32a; Yerushalmi Megillah 4:1). The translation had to remain secondary, oral, alive in the mouth of the man who delivered it.

The integration of the meturgeman into the liturgical event is so complete that the Mishnah’s structure of Torah reading is built around him. The reader may chant no more than one verse at a time before pausing — the pacing of the service is calibrated to the translator’s delivery (Megillah 24b; Mishnah Megillah 4:4). They are not two sequential acts, reading and then translation. They are a single liturgical event in two registers, simultaneous in the same room, before the same congregation.

This institution rests on a claim Benjamin would not concede. Benjamin says translation is not for the reader. The meturgeman exists because the listener is constitutive.

And yet the targum is not transparent. The most authoritative of the Aramaic translations, the work attributed to a sage called Onkelos, is interpretive. When the Hebrew says “and God descended” onto Mount Sinai (Exodus 19:20), Onkelos refuses to let God descend; he writes “and God revealed Himself” (ve-itgali Hashem). When the Hebrew gives God a face, Onkelos gives God a presence. Onkelos chafes against the literal.

This is precisely Benjamin’s Art des Meinens. Onkelos has identified that Hebrew’s mode of meaning, when it speaks of God, includes an anthropomorphic register that Aramaic cannot reproduce without distortion of the theology. As if to say, only Hebrew is holy enough to risk misunderstanding, to risk saying the apparently blasphemous.

Onkelos, in other words, is doing what Benjamin would describe as the highest form of translation: letting the foreignness of the original disturb the receiving tongue, refusing to pretend that Brot and pain mean the same thing. The difference is that Onkelos is doing it in front of a congregation that holds both languages at once. Benjamin’s Hölderlin is performing this labor for a German readership that cannot read Greek. Onkelos is performing it for a Babylonian congregation that has just heard the Hebrew chanted thirty seconds earlier. The covenantal setting changes the work without changing its theory.

So far the comparison is favorable to both traditions. Benjamin gives us translatability as a property of the work. The Hasidic master gives us translatability as a property of revelation. Benjamin gives us pure language as messianic horizon. The Targumic tradition gives us a covenantal community already living inside the play between languages. The two accounts seem complementary.

But here is where it breaks.

A passage in the Babylonian Talmud tells the story of a sage named Yonatan ben Uzziel, the greatest student of the great Rabbi Hillel, who lived in the last decades before the Common Era (Sukkah 28a; Bava Batra 134a). Yonatan composed the Aramaic translation of the Prophets — Isaiah, Jeremiah, Ezekiel, the rest. The Talmud claims he received the translation from the last prophets themselves, from Haggai and Zechariah and Malachi, who had received it in turn at Sinai (Megillah 3a). Translation, in this account, is not a later supplement to revelation. It is part of revelation, transmitted through the same chain.

When Yonatan finished the translation and prepared to release it, the land of Israel shook for four hundred parsa by four hundred parsa, a square the size of a country. A heavenly voice tore through the trembling and asked: who is this who has revealed my secrets to humankind? (Megillah 3a). Yonatan stood up and answered. I am the one. I did this not for my own honor, nor for my father’s house, but for the honor of Heaven, that dispute should not multiply in Israel.

The voice fell silent. Yonatan, emboldened, sought to translate the third and final section of the Hebrew Bible — the Writings, which include Psalms, Proverbs, Job, and the Book of Daniel. The voice came forth a second time and said: enough. The reason given is that the ketz, the messianic end, is hidden in the Writings, and translation cannot carry it (Megillah 3a, with Rashi ad loc. on Daniel as the locus of the ketz).

Benjamin imagines pure language as the horizon translation gestures toward. The Talmud says: some texts must not be translated precisely because their translation would bring that horizon too close. What Benjamin calls the messianic vocation of translation, the Talmud calls the messianic prohibition on translation. The eschaton is not where translation arrives. It is where translation must stop.

Daniel himself is told, twice, to seal his words: and you, Daniel, shut up the words and seal the book until the time of the end (Daniel 12:4); go your way, Daniel, for the words are shut up and sealed until the time of the end (Daniel 12:9). The angel does not tell Daniel that his prophecy is too holy to be communicated. The angel tells Daniel that the time is sealed. The content that translation cannot carry is a number that the people must not be able to compute.

This is a remarkable theory of translational limit. It says: translation is dangerous not because the source is too sacred but because the source contains information that, made available, would reorder the listener’s relation to time.

Daniel translated into Aramaic on the floor of the synagogue would let the am ha’aretz read the date of the end. The community would either be paralyzed by the proximity of redemption or shattered by its delay. Either way, the covenantal structure that holds Jewish life across exile depends on the end of history remaining unknown.

In 1925, Martin Buber and Franz Rosenzweig began the most ambitious Jewish translation project of the twentieth century: a German rendering of the Hebrew Bible — they called it a Verdeutschung, a “Germanizing” — that would let the Hebrew remain audible as Hebrew through the German. They were Benjamin’s contemporaries. Rosenzweig had read “The Task of the Translator.” Buber knew Hölderlin’s Sophocles. They were attempting, as Jews, the program Benjamin had described in theory, applied to the text Benjamin had written about only obliquely.

Their method was extraordinary. They restored the Hebrew root structure to the German by inventing compound words that mimicked Hebrew morphology. They preserved the Hebrew’s repetitions where Luther had smoothed them away. They rendered the Tetragrammaton not as Herr but as Er, Du, Ich — He, You, I — depending on the pronominal stance of the verse, refusing the German theological convention that had stabilized for four hundred years. They were trying, in Benjamin’s exact vocabulary, to let the Hebrew’s mode of meaning disturb the German. They were trying to make a German Bible in which Brot would still feel like lechem.

Rosenzweig died in 1929, halfway through. Buber finished the project in exile, in Jerusalem, in 1961. The completed Verdeutschung is one of the great achievements of modern Jewish letters and one of the most difficult Bibles ever produced in any language.

It is also a monument to a vanished world. The German-Jewish readership for whom Buber and Rosenzweig had begun the work — the community whose ear they were training, whose Hebrew they were trying to keep audible inside German — was murdered before the translation was finished. Buber completed in Jerusalem a Bible meant for Frankfurt. The Verdeutschung became, in the most exact Benjaminian sense, an afterlife: a translation whose original readership had passed beyond the language that first held it, the work outliving the community that called it forth. Levi Yitzchak’s God planted Aramaic in Genesis because He foresaw Babylon. Buber and Rosenzweig planted Hebrew inside German for a Babylon that would not survive to read it.

In Genesis, the heap of witness has two names. Laban names it in Aramaic. Jacob names it in Hebrew. They eat together on the stones (Genesis 31:46). Laban returns to his country (Genesis 31:55), and Jacob continues toward the land that has been promised. The stone heap stays. It does not belong to either man. It belongs to the border between them, which is also the place where their argument ended.

But Jacob did not rest there. At the ford of the Jabbok, he wrestled alone with something he could not name. He was wounded, and he would not let go, and he asked the figure for its name, and the figure refused (Genesis 32:25-31). Jacob named the place Peniel, the face of God, and limped away into the dawn.

Wrote this piece with help from Yochai (my AI chevruta). Will be rolling out the beta in next couple weeks.