A targum (an Aramaic translation of Scripture) is a translation that does not come alone: hardly ever is it left unattended by its parent text, the Hebrew Bible. While it may play, it is always supervised, its game subject to specific rules. A targum is not supposed to ever leave home and strike out on its own. The reasons for this peculiar and probably unique conception of translation as one part of a bilingual text are to be sought in contemporary rabbinic views on how to read and translate the Hebrew Bible.

To translate or not to translate a holy text is not an easy question. The answer depends on the view of how, if at all, such a text may be translated, whether indeed it is possible to adequately translate it, all the while minding the danger that a successful translation tends to usurp the position of the original. To defend the first Greek translation of the Bible, known as the Septuagint, an apologetic myth explained its miraculous accuracy visà- vis the original, thereby stating the claim of the translation's divine inspiration. For the Jewish philosopher Philo of Alexandria, the truth of the translation meant that it stood on a par with the original. Those who master both Greek and Hebrew, he claimed, "would admire and reverence them both as sisters, or rather as one and the same both in their facts and in their language; considering these translators not mere interpreters but priests and prophets to whom it had been granted in their honest and guileless minds to go along with the most pure spirit of Moses." Sometime later, the Talmud described how the earth shook in astonishment when Yonatan ben 'Uzziel first dared to air his Aramaic translation of the Prophets, with the translation of the Writings forbidden to him by a softly spoken divine decree.

The defense of either Greek or Aramaic translation is the flip side of the view that any translation is impossible without the text suffering significant loss. It is true that a certain sensitivity to scriptural translation is indeed manifest in many statements that are scattered over early rabbinic literature. The skeptic's view of translations is aptly captured in the following famous statement ascribed to R. Yehudah b. 'Ila'i: "R. Yehudah said, 'Whoever interprets a verse plainly is a liar, whoever adds something to it a blasphemer and a reviler'" (B. Megillah 32b). While R. Yehudah's remark raises the bar for translation, as he emphasizes interpretative fidelity between a plain translation that loses meaning and a rich one that adds some, his words still leave room to embark on a translation, albeit precious little. Unsurprisingly, some voices would advocate an ever-stronger position on the translatability of the Holy Writ. The Amoraic source Sefer Torah (1.6) espouses a downright negative view on translation when it compares the day the famous Septuagint was penned to the desert day on which the Israelites in the absence of Moses molded the golden calf, that symbol of idolatry par excellence.

But two factors mitigate the skeptic's view on scriptural translation, and both of these are born of multilingualism. By the early rabbinic period translations were a fact of life, both in the Diaspora and in Roman Palestine. Under Roman and Sassanid rule, the vast majority of Jews spoke Aramaic or Greek. As is clear from quotations and manuscript evidence, Greek translations had long gained a foothold in Jewish societies, including rabbinic circles, which is exemplified by Rabban Shimon ben Gamaliel's permission to write the Holy Writ in Greek (M. Megillah 1:8). Even the myth of its inspired origins, which started with the Letter of Aristeas, had gained acceptance among the early rabbis. However the suspicion of scriptural translation, rabbinic opinion had to accommodate a tradition that had already been firmly embraced by the rabbinic predecessors.

Not only had multilingualism made translation inescapable, the early rabbis sometimes voiced the belief that things were not lost but rather won in translation, as they embraced multilingualism as the manifold expression of God's language. The notion of the Torah as a multilingual text occurs in several sources, perhaps most notoriously in the statement that every single word that God spoke "split into seventy languages" (B. Shabbat 88b). Another example relates to Moses's speech on what the Israelites should do upon entering the Promised Land, namely to erect stones on which to inscribe God's teaching "most distinctly," which can also be read as "well explained" (Deut. 27:8). In Mishnah Sotah (7:5) this story is taken up as follows: "and they wrote on them all the words of the Torah in seventy languages, as it is written, 'well explained.'" The Torah found full expression in a multitude of translations.

As far as we know, Greek was the first target language of scriptural translation, but Aramaic followed relatively soon. Some Aramaic translations appear among the Dead Sea Scrolls, although what is today known as "targums" are products of the rabbinic period, beginning in the first centuries CE, when some anonymous but erudite Jews—later named as Onkelos and Jonathan—cultivated Aramaic scriptural translations of the Torah and the Prophets for oral dissemination. Under Christian rule in Palestine, these were soon followed by the Palestinian targums to the Torah and even later by Aramaic translations of the Writings. All of these translations are widely regarded as a translation sui generis, which earned them the moniker "targum," which simply means "translation" in Hebrew but as a technical term came to denote "Jewish Aramaic Bible translation" in modern scholarship. What made the targum different were the guidelines it came with, and these guidelines above all highlight the absolute necessity to distinguish between the Scriptures and their translation, with the latter always playing second fiddle. In the Talmud, 'Ulla prohibited the recitation of a written translation, for "they should not say that the targum is written in the Torah" (B. Megillah 32b). The distinction between the written Torah and the oral translation is designed to safeguard the unassailable position of the original; it became the hallmark of all targums.

At this point we see how the rabbinic movement eventually reconciled the positive, cautious, and skeptical views on translation. The careful distinction between the written text and its oral interpretation is the ingenious resolution, perhaps at the risk of stifling interpretation, of the dangers inherent in the practice of translation. Crucial is not the distinction, but the decision to tie in translation with the preeminent Hebrew text on which it would forever depend. Targum would forever be framed as a counterpoint to the Hebrew recitation. This central construction has apparently been carried over from contemporary Halakhah into liturgical practice, when the former stipulated that Torah and targum should be recited by two distinct persons, alternating verse by verse, with the Torah read from a scroll and the targum declaimed by heart. The interpreter should not be the senior of the Hebrew reader, neither in age nor in standing. And in the end, the Hebrew could be recited singly, not so the targum.

The evidence is there for all to see. The targums handed down to us were never meant to be an independent text, a translation in their own right; instead, they point to the Hebrew original, which the manuscripts almost always included in their text. Only a small minority of manuscripts have no Hebrew source text. The majority of textual witnesses present a running text in which Hebrew and Aramaic text alternate verse by verse (sometimes with other translations added); others have Hebrew and Aramaic in parallel columns (often with a smaller script for Aramaic), or on facing pages, or with an abbreviated Hebrew text (a few lemmata) followed by the complete translation for that verse; all of these basic formats, on which variations occur, signal the priority of the Hebrew text and that the targum should be read against that text, whilst no one should arrogate biblical status to any targum.

Even the grammar of many targums reveals the presence of the Hebrew original underneath its text. As long ago as 1864, Abraham Geiger observed how Onkelos's anxiety brought about many Hebraisms, a view confirmed by many authors since. The literal aspects of the translation so closely emulate the Hebrew that the Aramaic has a distinctly translational feel about it, the direct result of a strategy to carefully reproduce all the building blocks and boundaries of the biblical verse. The anonymous translators responsible for these targums—teachings often had name tags, but texts remained anonymous—mapped the Hebrew text to their Aramaic translation with utmost precision. The two translations that came to be seen as authoritative, Onkelos (to the Torah) and Jonathan (to the Prophets), correlated virtually every single element in the original text with its new, translucent overlay, which by explicit design never quite obscures the original text. The targum translates and simultaneously refers to its source text. Grammar and translational structure betray the targum as a transparent overlay. It goes without saying that this targumic foil frequently shows its own colors, not despite all the ostentatious fidelity to the Hebrew original, but because of it. Plain translation would not convey biblical meaning, as R. Yehudah bar 'Ila'i had spelled out so vividly. Often very subtle changes indicate an exegetical direction, for which the very first word of the Torah, bereshit, may serve as an example, since Onkelos translates this word with be-kadmin "in olden days," thereby studiously avoiding any statement on what came first. Targum Neofiti, our only complete Palestinian Aramaic translation of the Torah, agrees with Onkelos but adds a second translational equivalent, "in olden days, in wisdom . . ." This example illustrates two common characteristics of the targums: substitutions and pluses that steer the meaning of the original text in new directions. Sometimes the true significance of these subtle changes only emerges when we consider their parallels in ancient Jewish exegesis. While all targums share certain characteristics, they can be quite dissimilar to one another. The so-called Palestinian Targum shares many translational aspects with Onkelos and Jonathan but weaves far more aggadic material into its text. Other targums, such as Targum Song of Songs and Esther II, almost transform the meaning of translation, taking interpretation to new extremes and pushing the very boundaries of what a translation is; they may follow the original verse boundaries and order, but their relationship to the Hebrew becomes apparent only after careful exegetical study of their text. Some of these latter targums, such as Targum Chronicles, may reflect the new realities of medieval Europe, where Aramaic no longer served as anyone's vernacular and the use of targum evolved accordingly. Its traditional role of a linguistic and, to a lesser extent, interpretative repository in the talmudic period received more and more emphasis. Medieval sources cite targum as a prep for Talmud study since its language was considered to be very similar to that of Onkelos. Although unmentioned, knowledge of Onkelos and Talmud would also have lent mystical creativity good services, since the Zoharic corpus was written in what may be termed "cod Aramaic." Gradually, targum occupied the position of an authoritative commentary to be perused by biblical scholars. By this time, the child had escaped its original confines: targumic manuscripts without any Hebrew appear, and make sense now that they no longer function as translations, but as linguistic preparation for Talmud study and commentaries on the Scriptures, just as Rashi, with whose commentary they would soon be accompanied, and more often than not replaced altogether. New pastures beckoned when the study of the targums took on a new impetus among Christian Hebraists, who appreciated the way the targums emulate the Hebrew "truth" and frequently elucidate obscure passages; moreover, a new christological use of the targums emerged, with polemical or missionary interests never far away. Our only complete manuscript of the Palestinian Targum to the Torah was thus preserved in a monastery for those who converted from the old faith to the new.