4

Ancient Translations of the Old Testament

The previous chapter described our most important texts of the OT in Hebrew. The present chapter deals with witnesses to the ancient text of the OT in a variety of languages other than Hebrew. These translations were made from Hebrew texts at a much earlier time than the Masoretes worked, so they are important witnesses to earlier stages in the history of the text. The Greek Septuagint is particularly significant because it is a witness to the Hebrew text before it was standardized. However, using translations as textual witnesses presents some real challenges.

When we use a translation for textual criticism, we are actually interested in the Hebrew text from which it was created—its parent text or source text. But the translator has converted that Hebrew text into a new language system with different syntax, lexemes, cultural differences, and social norms. Some translators try to maintain as much correspondence to the original Hebrew text as possible, while others make many adjustments in order to communicate the sense of the text to their readers. Therefore, before we can use one of these translations as a witness to the Hebrew text, we must first retrovert the translation back into Hebrew. And in order to do that, we must understand as much as possible about how the translator worked and what kinds of changes he might have made in the text. After our survey of the Greek Septuagint, Aramaic targums, Syriac Peshitta, and Latin Vulgate (in approximate chronological order), we will discuss this challenge and how we can respond to it.

Greek Septuagint

History/Origin

The Septuagint is the most important of the ancient versions because it is the earliest known translation of the OT and because it reflects more significant deviations from the MT than all other textual witnesses combined.1 The name refers sometimes to the original translation of the OT into Greek and sometimes to the entire collection of later Greek translations and revisions that compose a distinct canonical shape. It is abbreviated LXX (for the seventy scholars who supposedly made the original translation of the Pentateuch) or G (for Greek).

When the people of Israel were scattered at the time of the Babylonian exile in the sixth century BCE, many found their way to Egypt and became permanent residents there. By the middle of the fourth century BCE, when Alexander the Great conquered the known world, there was a noticeable population of Jews concentrated in Alexandria, on the Mediterranean coast.2 They spoke Koine Greek, the common language, which had developed from Classical or Attic Greek and had become dominant because of increased Hellenization and the influence of Alexander’s conquest.

An ancient document called the Letter of Aristeas explains in great detail how the Greek OT came to be. It tells the story of King Ptolemy II of Egypt who instructs his librarian in Alexandria to collect all the books in the world, even a translation of the Jewish law. At the request of the king, Aristeas goes to Jerusalem and asks Eleazar the high priest for assistance. Six worthy men from each tribe (a total of seventy-two) travel to Egypt and are well received. They are isolated on an island and given everything that they need for their work. After only seventy-two days, they complete their translation, and when their individual renderings are compared, they agree on every word! Although this account was accepted as historical in antiquity,3 many of the details are undoubtedly only legends.4 Nevertheless, certain historical facts may be gleaned from the letter. These include the location (Alexandria, Egypt), the time (during the reign of Ptolemy Philadelphus, 285–246 BCE), the initial extent (the Pentateuch), the parties involved (the Jews), and the nature of the effort (an officially sanctioned effort aimed at synagogue worship and instruction).5 The translation was probably not made at the request of a royal librarian, but to serve the needs of the Jewish community living in Alexandria who needed a copy of the Scriptures in their own language.6 Following the Pentateuch, the rest of the OT was translated in stages between the third century BCE and the first century BCE. Sometimes more than one translator worked on a particular book. There is evidence that later translators used the Greek Pentateuch as a kind of dictionary in order to determine Greek equivalents for certain Hebrew words. Scholars refer to this initial translation as the Old Greek (OG) in order to distinguish it from the later translations and revisions that became a part of the Septuagint tradition.

Those who received and used this original translation (OG) were aware of obvious differences from the standardized Hebrew text. Therefore, some sought to “correct” the Greek text toward the protomasoretic text so that the two would correspond more fully. Others created new Greek translations in an attempt to represent the Hebrew text more faithfully than the original translation had done. The evidence for these revisions and translations is fragmentary and consists of quotations preserved in the church fathers or talmudic literature, fragments found in the Cairo Genizah, remnants in Origen’s Hexapla (see below), and marginal notes in Septuagint manuscripts.7 In addition, distinguishing between revisions and new translations is sometimes difficult.

Kaige

One of the earliest of these translations/revisions is called kaige for its characteristic translation of the Hebrew word גם (also) or וגם (and also) with the Greek phrase και γε (kai ge). The kaige revision also wrote out the tetragrammaton (God’s covenant name, YHWH) in the Paleo-Hebrew script rather than translating it into Greek. It dates to the late first century BCE or the early first century CE, and constitutes a revision of the Old Greek translation in the direction of conformity with the protomasoretic text. This revision is sometimes called kaige-Theodotion because of its shared readings with Theodotion (see below).

In time, the Septuagint came to be adopted by the Christian churches. Since it was often used in debates between Christians and Jews, it came to be viewed with suspicion by the latter. This led, in the course of the second century CE, to the production of three rival Greek translations that each bore a different relationship to the original Septuagint.

Aquila (Siglum α)

The earliest of these rival versions (ca. 150 CE) was produced by Aquila, a Jewish proselyte and disciple of Rabbi Akiva. The most noteworthy characteristic of his version is its extreme literalness. He translates the same Hebrew word with the same Greek word, even if the word is not really appropriate in the context, and preserves the same word order as the Hebrew text. This sometimes obscures the meaning of the text in Greek.8 Aquila’s translation is so literal that it is often not understandable without knowledge of the Hebrew text upon which it is based.9 It may have been intended as an aid to exegesis (perhaps for those with a limited knowledge of Hebrew) rather than as a useful translation.10

Symmachus (σ)

Symmachus was a Samaritan who converted to Judaism and worked at the end of the first century CE.11 He probably based his work on that of Aquila, but he is much more versatile in his use of vocabulary in order to communicate clearly in Greek. Symmachus adapts Hebrew idioms to Greek usage and does not always use the same Greek word to translate each occurrence of a particular Hebrew word. He may have been attempting to avoid the absurdities of Aquila’s version and to create something much more readable.12

Theodotion (θ)

The third rival version of the second century CE was produced by Theodotion.13 He came from Ephesus in Asia Minor and was also a convert to Judaism. Theodotion worked at the end of the second century CE and produced a version located between Aquila and Symmachus in terms of formal correspondence to the Hebrew text (see fig. 4.1 below). He often leaves difficult Hebrew words and constructions untranslated. Some of Theodotion’s distinctive readings were known long before he lived. Therefore, it is likely that he was updating the kaige revision mentioned above. For this reason, and due to confusion and uncertainty of the relationship, some scholars reject this as a distinct translation and refer to it as kaige-Theodotion.14

Origen’s Hexapla

The next major stage in the history of the Greek OT was a logical one given the preceding history. At the end of the second century there were (at least) four competing Greek versions of the OT. Origen, one of the most important theologians in the Eastern church, was born in Alexandria, Egypt and was active in the middle of the third century CE. Aware of differences between the Greek and Hebrew texts, he set out to bring order and understanding to the confusing array of competing textual witnesses and to produce an edition that would account for those variations.15 His work resulted in the massive volume known as the Hexapla. When originally completed, the Hexapla ran to some six thousand pages in fifty volumes.16 In his Letter to Africanus, Origen writes that his aim was primarily apologetic. He created the Hexapla so that “in discussion with Jews, Christians do not quote passages not to be found in their Scriptures and so that Christians, in turn, could also use what was to be found in Jewish manuscripts even though not in their own.”17

The Hexapla was a six-column work in which the existing Greek versions could be compared with the Hebrew text that was current in Origen’s time. The arrangement was as follows. The first column was the Hebrew text. The second column was a transliteration of the Hebrew text in Greek letters. The third column contained Aquila’s version, the fourth Symmachus’s, the fifth Origen’s own revision of the Septuagint text, and the final column Theodotion’s version. Origen’s fifth column contained obelus symbols in the text to mark readings that were found in the Greek version but not the Hebrew, and asterisk symbols for omissions from the Greek that were in the Hebrew. This revision of the Greek text was so significant that it “dominated the subsequent history of the [Septuagint].”18

Figure 4.1

Character of the Greek Versions

Hyper Literal				Less Literal
Aquila	Theodotion	Symmachus	Old Greek	Lucianic
Early 2nd CE	Late 2nd CE	Late 2nd CE	2nd BCE > 1st CE	Late 3rd CE
• Lexical stereotyping • Hebrew etymology • Obscures meaning	• Abdicates when encountering difficulties	• Adapts to Greek idiom • Uses most appropriate lexeme	• Varies in degree of correspondence to the Hebrew	• Supplies implied information • Corrects difficulties and errors

In spite of the importance of Origen’s work, today we only have fragments of partial copies.19 It can be consulted in the classic edition by Frederick Field.20

Lucian (GLuc)

The most important revision after the Hexapla is traditionally associated with Lucian of Antioch, who was martyred in 312 CE.21 This revision was a stylistic update of an existing Greek text that was not Origen’s edition in the fifth column of the Hexapla. Like Theodotion, it contains distinctive readings that were known long before Lucian lived in the fourth century. We call this earlier Greek text proto-Lucian. The Lucianic revision tends to fill in gaps in the Greek text (in comparison with the Hebrew MT), adds clarifying elements, and corrects grammatical difficulties.22 It is a full text, and less woodenly literal than the previous translations and revisions (see fig. 4.1).

Figure 4.2

Textual History of the Greek Versions

Only the complexity of the textual history of the Greek versions is certain. This diagram illustrates one possible reconstruction.
Source: Karen H. Jobes and Moisés Silva, Invitation to the Septuagint, 2nd ed. (Grand Rapids: Baker Academic, 2015), 49.

Therefore, there were five stages in the development of the Septuagint as a group of Greek translations. First, there was an original translation (Old Greek) of the Pentateuch and then the rest of the OT. Second, there were early revisions of the Greek text (proto-Lucian and kaige-Theodotion). Third, the translations/revisions of Aquila, Symmachus, and Theodotion were completed. Fourth, Origen included the preceding work along with his own revision in the Hexapla. Finally, Lucian completed a new revision (see fig. 4.2). The Septuagint “was produced by many people unknown to us, over two or three centuries, and almost certainly in more than one location. Consequently, the Greek OT does not have the unity that the term the Septuagint might imply.”23

Character

We have already described some aspects of the character of the Septuagint tradition in the preceding description of its origin and development. The translation was created by various translators with different backgrounds and training. Perhaps some of them knew Hebrew better than others. They translated differently in different genres such as legal or narrative. And they adjusted their approach according to their purpose. For example, in a liturgical text they might translate in an archaic style, avoiding colloquial language. But in educational texts they were motivated to communicate clearly and to smooth out difficulties.24 This means that we should consider each book of the Greek Septuagint on its own without assuming that its character is consistent throughout.

Sometimes, the difference between the Septuagint and the MT is limited to minor differences in spelling or in the reading of one Hebrew letter. The Greek translators were working from unvocalized Hebrew texts and were continually required to make interpretive decisions. The Greek translation represents the way that the translator understood his Hebrew source text, and quite often there were ambiguities. The vocalization tradition in the Septuagint commonly differs from that of the Masoretes hundreds of years later.

In addition, in a few places the Septuagint tradition witnesses to major differences from the MT. For example, the Greek version of Jeremiah is one-eighth shorter, and the contents of the book are in a different order.25 Emanuel Tov discusses twenty-three of these major differences in the Greek tradition, including longer or shorter texts and different arrangements.26

Our early manuscript evidence for the Septuagint consists of early fragments and papyri. We have discovered two fragments of the Septuagint at Qumran. 4QLXXDeut contains only a few words of Deut. 11:4. 4QLXXLeva contains a free translation of Lev. 26:2–16.27 Papyri from the second century BCE to the fourth century CE contain parts of books from the pre-Hexaplaric period.28

Our most complete manuscripts of the Septuagint are codices from late antiquity into the medieval period. Until the ninth century CE, they were written in large capital Greek letters (uncials) without any word division (scriptio continua). From the ninth century to the sixteenth century they were written in a common cursive script (minuscules).29 The three most important codices are Vaticanus, Sinaiticus, and Alexandrinus:

Codex Vaticanus (B)

Codex Vaticanus is our most important witness to the text of the Greek Septuagint before the revisions. It is from the fourth century CE and contains the complete text of the OT.30

Codex Sinaiticus (א)

Also from the fourth century, Sinaiticus was discovered at St. Catherine’s Monastery in the Sinai Peninsula in Egypt.31 It contains parts of Genesis, Numbers, Joshua, Judges, 1 Chronicles, Ezra-Nehemiah, Esther, and most of the Prophets and Poetic books.32

Codex Alexandrinus (A)

This fifth-century codex is almost complete; it lacks some of 1 Samuel and the Psalms. It was originally from the library in Alexandria, from which it gets its name. Because it is a mixed text—some of it close to Vaticanus and some Hexaplaric—Ernst Würthwein urges caution when using it for textual criticism.33

Evaluation of Usefulness

The Septuagint is one of our most important witnesses to the text of the OT for several reasons. First, unlike the Qumran scrolls, it is complete and contains the text of the entire OT.34

Second, because the original Old Greek translation was made so early, it witnesses to a Hebrew source text before the standardization of the Hebrew tradition. Until the discovery of the Dead Sea Scrolls at Qumran, we could not be sure that the differences between the Septuagint and the MT really indicated variant readings rather than changes that arose in the transmission and translation of the Septuagint. We now know that many of the differences in the Septuagint were present in Hebrew texts of the Second Temple period. Tov writes, “When turning to the background of this situation, the assumption is unavoidable that the Hebrew scrolls used for the Greek translation were valuable, authoritative, and sometimes more ancient than [(proto-)MT].”35 Therefore, the Septuagint is one of our most important witnesses to the text of the OT. It has a greater number of textual variants than all other translations put together.36

Natalio Fernández Marcos argues that the Septuagint is so important that we dare not privilege the MT. He writes, “the Greek Bible contains genuine, textual and literary variants from the Hebrew to the extent that we have to respect both traditions [i.e., the Septuagint and the MT], without trying to reduce or adjust one to the other.”37

Editions

Göttingen

The best critical edition of the Septuagint is the Göttingen edition.38 This is an eclectic text that attempts to reconstruct the earliest form of the Septuagint translation. It also provides readings of the revisions and other edited forms in the Greek tradition. There are two apparatuses: the first contains readings from Greek texts such as uncials, papyri, and minuscules; the second apparatus contains other Greek versions such as the revisions, including the Hexapla. So far, twenty-three volumes of the series have appeared.39

Rahlfs (1935)

Knowing that the Göttingen edition would not be complete for the entire Bible for many years, Alfred Rahlfs had the foresight to publish an editio minor.40 This is a smaller edition which covers the entire text of the OT, but with a less complete critical apparatus. It is an eclectic text based on Vaticanus, Sinaiticus, and Alexandrinus, but it rarely departs from Vaticanus. Würthwein evaluates the text as reliable due to Rahlfs’s careful scholarship and good intuition, but the critical apparatus is inadequate, especially since it was published before the discovery of the Qumran scrolls.41 This edition was revised and improved by Robert Hanhart in 2006.

MT-LXX Parallel (CATSS)

The major Bible software programs—BibleWorks, Accordance, and Logos—have differing versions of a module that presents an aligned text of the Hebrew MT and the Greek Septuagint.42 The Hebrew and Greek texts are arranged in order to indicate each translation segment, allowing the user to see easily each Hebrew word or phrase and its corresponding Greek translation.43 In addition, the text contains notes and retroversions from Emanuel Tov that explain some of the differences between the Hebrew and Greek.

New English Translation of the Septuagint (2007)

The NETS is a new translation of the Septuagint into English.44 The English is intentionally patterned after the NRSV English translation of the Hebrew so that the two versions can be compared to get a sense of how the Septuagint and MT differ.45 It is available online.46

For Further Study

Introductions

Fernández Marcos, Natalio. The Septuagint in Context: Introduction to the Greek Version of the Bible. Translated by Wilfred G. E. Watson. Boston: Brill, 2000.

Jobes, Karen H., and Moisés Silva. Invitation to the Septuagint. 2nd ed. Grand Rapids: Baker Academic, 2015.

Tov, Emanuel. The Text-Critical Use of the Septuagint in Biblical Research. 3rd ed., revised and expanded. Winona Lake, IN: Eisenbrauns, 2015.

Concordance

Hatch, Edwin, and Henry A. Redpath. A Concordance to the Septuagint and Other Greek Versions of the Old Testament. 2nd ed. Grand Rapids: Baker, 1998.

Lexicons

Lust, J., E. Eynikel, and K. Hauspie. A Greek-English Lexicon of the Septuagint. 2nd ed. Stuttgart: Deutsche Bibelgesellschaft, 2003.

Muraoka, Takamitsu. A Greek-English Lexicon of the Septuagint. Leuven, Belgium: Peeters, 2009.

Aramaic Targums

History/Origin

Following the Babylonian exile, when Aramaic was the lingua franca in Palestine and in the East, Jews slowly began losing their knowledge of Hebrew. By the turn of the era, they were increasingly unable to understand the Torah and the Prophets when they were read aloud in the synagogue. Rather than create a translation that would replace the Hebrew text (as the Greek Septuagint was intended to do), the rabbis instead elected to accompany the reading of the Hebrew text with translation in Aramaic. These Aramaic translations were given the name “targum” (תרגום), a word that originally referred to a translation in any language but came to be used specifically for biblical translations in Aramaic.47

The original intention and official directive were that the targum would be only oral. One reader would stand at the front of the synagogue and read the Hebrew text from the open scroll. Then a second man would translate into Aramaic. Because all translation involves interpretation of what the source text means, the translators took the opportunity to interpret the OT according to the theology and practices of rabbinic Judaism. The two different languages kept the text and the interpretation distinct and allowed the translators to be quite creative in their renderings while the original Hebrew text remained intact. Therefore, the targum functioned not only to render the linguistic meaning of the biblical text but also to provide a guide to its contemporary meaning. For many, the Aramaic translation became their scripture, for it was all that they could understand.48 Later these translations came to be written down (or perhaps some were originally composed in written form). Targums were created for each of the books of the OT except for Ezra and Daniel. One explanation for these exceptions is that Ezra and Daniel contain significant sections of Aramaic, so perhaps they did not require translation.49

Our earliest evidence of targums, albeit only fragmentary, comes from the Qumran scrolls in the Judean Desert. 4QtgLev comes from the second to first centuries BCE, and 4QtgJob and 11QtgJob come from the first century CE.50 We do not know when targums were first created, but Jewish tradition associates the beginnings of this process of giving an oral rendering of the Hebrew Scriptures with the time of Ezra in the postexilic community (Neh. 8:8).

Character

Flesher and Chilton argue that the base translations of the targums are quite literal and follow the Hebrew original text closely. They write, “In the vast majority of their passages, Targums present a translation that recreates anywhere from 85–100 percent of the original’s linguistic information—in particular, its grammatical information.”51 But interwoven throughout this literal translation are additions, updates, and harmonizations that reflect the theology and views of mainstream, classic rabbinic Judaism and the world of Oral Torah. The rabbis taught (and still teach) that when God gave the Written Torah to Moses at Mount Sinai, he also gave him the authoritative interpretation of the Torah, or Oral Torah, which was to be passed down through the generations.52

Targums make a number of characteristic adjustments to the text. Sometimes, a targum fills in exegetical details. In Gen. 4:8 the Hebrew text says that Cain spoke to his brother Abel and then rose up to kill him. But what was his motivation? Targum Neofiti contains a lengthy expansion at this point in which Abel and Cain get into an argument about the nature of divine justice and whether there is punishment for the wicked. As the argument becomes more intense, Cain rises up and kills his brother.53 Targums also update the text in accordance with specific religious practices. For example, in the Hebrew MT, Exod. 34:26 commands the reader not to “boil a young goat in its mother’s milk.” But Targum Neofiti translates, “you shall not boil, and you shall not eat flesh with milk, mixed together.” This rendering significantly widens the scope of the law to include all meat mixed with dairy, reflecting the rabbinic dietary rules (kashrut).54 Targums also update geographical references, remove anthropomorphisms in which human imagery is used of God, and attempt to rehabilitate the reputation of characters that the text casts in a bad light.55 If the Hebrew text is ambiguous and could mean two different things, sometimes a targum will simply translate both possibilities (i.e., a double translation). Very frequently, targums also attempt to maintain God’s transcendence by using the idea of his Memra (or word) as a mediator between himself and humanity. For example, in Hosea 7:13 where the Hebrew MT has “they rebelled against me,” Targum Jonathan translates “they rebelled against my Memra.”

The targums are divided into two groups, Palestinian and Babylonian, reflecting the regions where they were used. The Palestinian targums to the Pentateuch come from the late second and early third centuries CE. They are written in Jewish Palestinian Aramaic (JPA), which was used in northern Palestine. The Palestinian targums contain significant creative and interpretive expansions of the biblical text.

Targum Neofiti (TN). MS Vatican Neofiti is the only complete Palestinian targum of the Pentateuch. It was discovered in 1956 in a manuscript dating from 1504 or slightly later.56 It has probably been altered in the course of scribal copying since its original production.57

Fragment Targum II, III (TF). This targum is fragmentary and does not include the entire Pentateuch, but not because some of it was lost. Rather, only selections of the Hebrew text were ever translated and then grouped in collections.

Targums from the Cairo Genizah. The oldest known manuscript remains of the Palestinian targums were discovered in the Cairo Genizah. These remnants from more than seventeen different manuscripts evidence far fewer scribal errors and emendations than Targum Neofiti.58

The Babylonian targums were written in Jewish Literary Aramaic (JLA), a dialect used in Palestine between 200 BCE and 200 CE.59 The two most important of these translations, Targum Onqelos and Targum Jonathan, have relatively fewer expansions and additional material in comparison with the Palestinian targums. They were originally composed in Palestine and were then edited and used in Babylonia.

Targum Onqelos (TO). Targum Onqelos is a complete translation of the entire Pentateuch and was the official targum of Babylonia. It was created in the first, third, or fifth century CE.60 It follows the plain sense of the Hebrew text but has many exegetical elements in poetic sections.61 Unlike the Palestinian targums, Onqelos was copied and transmitted in a way analogous to how the Masoretes transmitted the MT, with absolute precision even in the preservation of unusual readings.62

Targum Pseudo-Jonathan (TPs-J). This is a complete manuscript of the Pentateuch that integrated elements from Targum Onqelos. It was originally, incorrectly, identified as Targum Jonathan. When this was corrected in the eighteenth century, its name was changed to Pseudo-Jonathan.63

Targum Jonathan (TJ).64 This targum is ascribed to Jonathan ben ʿUzziel, a pupil of Hillel the Elder. It was created at the end of the first century CE in Palestine and revised in the fourth century in Babylonia. It generally resembles Onqelos and contains the text of the Former and Latter Prophets.65

The Writings, or third part of the Jewish biblical canon, do not neatly fall into the Palestinian or Babylonian categories. They are written in a third, recently recognized dialect of Aramaic called Late Jewish Literary Aramaic (LJLA). These targums were composed at different times and by different people and vary greatly in their character. Table 4.1 contains a summary of the major targums.

Table 4.1
Significant Aramaic Targums

Pentateuch	Onqelos (Babylonian)
	Neofiti (Palestinian)
	Pseudo-Jonathan (Babylonian)
	Fragment Targum (Palestinian)
	Fragments in Cairo Genizah
Prophets	Targum Jonathan (Babylonian)
Writings	Various unaligned targums available (except for Daniel and Ezra-Nehemiah)

Evaluation of Usefulness

The use of the targums for textual criticism is complex because of their paraphrastic and interpretive character. Although the basic translation might be quite literal, the many expansions and harmonizations make it difficult to reconstruct the Hebrew source text. In order to do this, we must evaluate the translation character of each targum and learn to distinguish between readings that are present in the underlying Hebrew and those that have been inserted by the translator in the service of his interpretive and theological agenda.

However, Tov argues that all of the targums reflect the medieval form of the MT anyway.66 As we mentioned above, by this time the MT was highly standardized. This means that even aside from the challenges of their expansive nature, the targums typically do not contain significant variant readings distinct from the MT.

Editions

Sperber

The standard edition of the Babylonian targums is the four-volume work by Alexander Sperber, which was reprinted in one volume in 2004.67 It contains Targum Onqelos (Pentateuch) and Targum Jonathan (Former and Latter Prophets), but the Writings are not included in the one-volume edition. Textual variants are presented in two apparatuses. The upper apparatus contains only variants in vocalization. The lower apparatus lists all textual deviations from the manuscripts upon which the edition is based. Sperber has been criticized for his selection of manuscripts and the critical apparatus. Würthwein, for example, states that the edition “can no longer be recommended without reservation” for these reasons.68 However, in the preface to the edition, Gordon writes that Sperber has represented the consonantal text accurately, and most of the errors concern vocalization.69

Other Targum Editions

Other significant editions of Babylonian and Palestinian targums include the following:70

Clarke, Ernest G., ed. Targum Pseudo-Jonathan of the Pentateuch: Text and Concordance. Hoboken, NJ: Ktav, 1984.

Díez Macho, Alejandro, ed. Neophyti 1: Targum Palestinense Ms. de la Biblioteca Vaticana. 6 vols. Madrid: Consejo Superior de Investigaciones Científicas, 1968–79.

Klein, Michael L. The Fragment-Targums of the Pentateuch according to Their Extant Sources. 2 vols. Rome: Pontifical Biblical Institute Press, 1980.

———. Genizah Manuscripts of Palestinian Targum to the Pentateuch. Cincinnati: Hebrew Union College, 1986.

McNamara, Martin J., ed. The Aramaic Bible: The Targums. 22 vols. Collegeville, MN: Liturgical Press, 1987–2007.

The multivolume Aramaic Bible is a helpful English translation, providing notes on the text and the rationale for certain renderings.

For Further Study

Introductions

Beattie, D. R. G., and M. J. McNamara, eds. The Aramaic Bible: Targums in Their Historical Context. Sheffield: Sheffield Academic Press, 1994.

Flesher, Paul V. M., and Bruce Chilton. The Targums: A Critical Introduction. Waco: Baylor University Press, 2011.

Lexicons

Cook, E. M. A Glossary of Targum Onkelos according to Alexander Sperber’s Edition. Studies in the Aramaic Interpretation of Scripture 6. Leiden: Brill, 2008.

Jastrow, Marcus. A Dictionary of the Targumim, the Talmud Babli and Yerushalmi, and the Midrashic Literature. 2 vols. 1903. Reprint, Peabody, MA: Hendrickson, 2006.

Syriac Peshitta

History/Origin

Syriac is a late dialect of Aramaic used in ancient Edessa (in modern southeast Turkey) and the surrounding area. Edessa was a significant city on the trade routes, populated with a mixture of Greeks, Persians, Jews, and other minorities in addition to locals. The majority of the population was pagan, and about 10 percent was Jewish. It was so cosmopolitan that it was called the Athens of the East.71 Perhaps because of its location, Edessa became a major center of the early Eastern church. It was most likely there that the Bible, both OT and NT, was translated into Syriac and called the Peshitta, meaning “simple” or “obvious.”72 This name distinguished it from other Syriac versions such as the Syro-Hexapla, which had a very idiosyncratic and l iteral style.

We do not know the exact date or circumstances of the translation. The OT must have been translated before the fourth century, when the Syriac church fathers Aphrahat and Ephrem quote from it. Most scholars place it in the first or second century CE. Internal evidence suggests that the translators were somehow connected to both Judaism and Christianity. The translation was made from Hebrew and sometimes reflects Jewish exegetical traditions.73 This suggests that the translators were Jewish since it would be very unusual for non-Jews to know Hebrew. On the other hand, the Peshitta was used and transmitted in the Christian church, and it appears to be uninterested in rabbinic dietary laws and other Jewish theological emphases. Therefore, Michael Weitzman has argued that it was created by nonrabbinic Jews who converted to Christianity.74

The Peshitta was rejected by Jews because, like the Septuagint, it had become the Bible of the church. In addition, its particular renderings did not support dominant rabbinic interpretations.75 But by the fifth and sixth centuries, new Syriac versions were being created (especially of the NT) because the Peshitta was not considered to be sufficiently “Christian” or sufficiently reflective of the Greek text traditions.

Character

In our discussion above, we mentioned that the Greek Septuagint is more literal or formally correspondent to its Hebrew source text(s), whereas the Aramaic targums are sometimes more expository and contain interpretive expansions. The Peshitta is more similar to the Septuagint in this regard in that it is essentially a literal translation, rendering each Hebrew word or phrase into Syriac without much additional comment. This approach suggests that the Peshitta was intended to serve as a replacement for the Hebrew OT rather than as a kind of commentary.

It is likely that, like the Septuagint, the Peshitta was created by a number of different translators. They often add or omit conjunctions or prepositions in order to help the text flow smoothly. They simplify complex syntax, remove linguistic oddities, and choose Syriac words that seem appropriate in the context rather than attempting complete consistency. A few books in the Peshitta demonstrate some interest in supporting a particular theological agenda.76

We have a number of Peshitta manuscripts from the fifth century CE. One of the oldest, containing most of the Pentateuch, is from 464 CE and was written in a monastery in Lower Egypt. Our earliest manuscript containing the entire OT is Codex Ambrosianus (abbreviated in the Leiden edition as 7a1) from the sixth to seventh centuries CE.77

Evaluation of Usefulness

Although the Peshitta is fairly literal in character, it is not as significant as the Septuagint for textual criticism because its Hebrew source text is quite similar to the MT. Because it was created in the first centuries of the Common Era, the Hebrew text was already becoming quite standardized at that time. Therefore, Tov groups the MT, targums, Syriac Peshitta, and Vulgate in what he calls the “MT+” group.78

However, the Peshitta does reflect more textual variants than the targums or the Vulgate. In addition, the Peshitta, like the Septuagint, frequently reflects a different vocalization tradition than the MT. It also frequently agrees with the Septuagint and targums against the MT. Sometimes these agreements are simply “polygenesis,” which means that the three versions were translated in the same way by coincidence, but in other cases agreements against the MT may be evidence that the versions are witnessing to an actual textual variant.

Editions

International Organization for the Study of the Old Testament: Peshitta Institute, ed. The Old Testament in Syriac according to the Peshitta Version. Leiden: Brill, 1972–.

The first critical edition of the Peshitta is currently underway by the Peshitta Institute at the University of Leiden, Netherlands. This is essentially a diplomatic edition based on Codex Ambrosianus.79 Fourteen volumes have appeared in the series thus far.

For portions of the OT not yet completed in the Leiden edition, the following noncritical editions are helpful:

David, Clemens Joseph, ed. The Syriac Bible according to the Mosul Edition. 3 vols. Piscataway, NJ: Gorgias, 2010.

Syriac Bible: Reprint of the 1826 Edition of the Peshitta by S. Lee. [London]: United Bible Societies, 1979.

For Further Study

Introductions

Brock, Sebastian P. The Bible in the Syriac Tradition. Gorgias Handbooks 7. Piscataway, NJ: Gorgias, 2006.

Dirksen, Peter B. “The Old Testament Peshitta.” In Mikra: Text, Translation, Reading & Interpretation of the Hebrew Bible in Ancient Judaism & Early Christianity, edited by Martin Jan Mulder, 255–97. 1988. Reprint, Peabody, MA: Hendrickson, 2004.

Weitzman, Michael P. The Syriac Version of the Old Testament: An Introduction. Cambridge: Cambridge University Press, 1999.

Concordance

Borbone, P. G., ed. The Pentateuch. Vol. 1 of Concordance, which is part 5 of The Old Testament in Syriac. Leiden: Brill, 1997.

Lexicons

Payne Smith, R., and J. Payne Smith. Compendious Syriac Dictionary: Founded upon the Thesaurus Syriacus of R. Payne Smith. Winona Lake, IN: Eisenbrauns, 1998.

Sokoloff, Michael. A Syriac Lexicon: A Translation from the Latin, Correction, Expansion, and Update of C. Brockelmann’s “Lexicon Syriacum.” Piscataway, NJ: Gorgias, 2009.

Vulgate

History/Origin

Although Greek had been the lingua franca of the ancient world throughout the Second Temple period, the advance of Rome and its military and cultural power meant that Latin eventually became the dominant language of politics, literature, and religion. Once again, there was a need for new translations as one language gave way to another in society.

Our first evidence of Latin translations of the Bible comes from the second century CE. Christian writers wrote and quoted Scripture in Latin, which they seem to have drawn from known Latin translations. Evidence indicates that there were multiple fragmentary and sometimes contradictory Latin versions during this early period.80 We use the name Old Latin (siglum OL) or Vetus Latina to refer to these original translations and to distinguish them from the Vulgate. They were generally based on the Greek Septuagint.81 Thus their importance is greater for helping to determine the original text of the Greek OT than of the Hebrew.82

As the Western church grew, and copies and revisions of the Old Latin multiplied, there was a need for one authoritative Latin version. Eusebius Hieronymus, whom we commonly know as Jerome, lived a life of asceticism and study and had a significant background in Greek and Latin literature. In a significant departure from his Christian contemporaries, he believed that if there was to be a truly accurate version of the OT, it should be based on the original Hebrew text (hebraica veritas). As he traveled through Palestine and settled in Bethlehem, he discussed the Hebrew text with Jewish teachers and became quite proficient in the language. Because of this, he was a careful, accurate translator with sensitivity to Hebrew philology and the differences between Hebrew and Latin.83 Between 390 and 405 CE, Jerome made a fresh translation of the OT based on the Hebrew text. It was eventually called the Vulgate, meaning “common one.”

At first, Jerome’s work met with a hostile reception, partly because it was not based on the Greek Septuagint and partly because Jerome was receiving translation help from Jews who were accused of misinterpreting the Scriptures.84 However, by the beginning of the seventh century CE the Vulgate was accepted on an equal basis with the Old Latin, and by the eighth and ninth centuries it superseded the Old Latin as the more accepted version. The Vulgate was adopted by the Roman Catholic Church as its official Bible at the Council of Trent in 1546.

Character

Among the ancient versions of the OT, the Vulgate is unique because it is the product of one man’s work rather than multiple translators. In addition, we have Jerome’s own statements about his translation philosophy, method, and the challenges that he faced. Even so, the character of the Vulgate is not uniform throughout. Rather, the approach varies from book to book.85

Jerome translated the Hebrew text closely and carefully, but he also translated according to his own theological convictions. Some of his renderings demonstrate an eagerness to bring out christological connections in the OT text. Whenever the language was ambiguous or left room for a christological interpretation, Jerome took the opportunity to translate in a way that supported that nuance. According to one scholar, his “Christian faith proved the fundamental exegetical guide.”86 At times, he used the same language as the Old Latin translations in order to leave church tradition intact.87

Evaluation of Usefulness

The Vulgate is important as a witness to the Hebrew text, but its importance is lessened by interference from the Old Latin throughout long periods and from the Septuagint.88 Even more significantly, because the Hebrew text was highly standardized by the fourth century, Jerome’s Hebrew source text was very similar to what would be the MT. This means that the Vulgate does not contain many important textual variants. It is important for the history of exegesis, especially when compared with Jerome’s commentaries on the Bible, but it is of limited value for textual criticism.

Editions

Pontifical Abbey of St. Jerome-in-the-City, ed. Biblia Sacra iuxta latinam Vulgatam versionem. Rome: Vatican Press, 1926–.

The most comprehensive critical edition of the Vulgate was commissioned by Pope Pius X in 1907 and has been in preparation by Benedictine monks. It is an eclectic edition89 and comprises eighteen volumes.

Vercellone, C., ed. Biblia Sacra Vulgatae editionis Sixti V. et Clementis VIII. Pontt. Maxx. iussu recognita atque edita. Rome, 1861.

In the nineteenth century, Pope Clement VIII commissioned an official version of the Vulgate for the Roman Catholic Church. It is not a critical edition, but it is influential.

Weber, Robert, et al., eds. Biblia Sacra iuxta Vulgatam versionem. Stuttgart: Deutsche Bibelgesellschaft, 2007.

There is an editio minor version of Biblia Sacra available in one volume. While it contains less information in the apparatus, according to Tov it actually represents a better text.90

For Further Study

Kedar, Benjamin. “The Latin Translations.” In Mikra: Text, Translation, Reading & Interpretation of the Hebrew Bible in Ancient Judaism & Early Christianity, edited by Martin Jan Mulder, 299–338. 1988. Reprint, Peabody, MA: Hendrickson, 2004.

Roberts, Bleddyn J. The Old Testament Text and Versions: The Hebrew Text in Transmission and the History of the Ancient Versions. Cardiff: University of Wales Press, 1951.

Figure 4.3

The Witnesses to the Old Testament Text

Source: Shemaryahu Talmon, “The Old Testament Text,” in The Cambridge History of the Bible, vol. 1, From the Beginnings to Jerome, ed. Peter R. Ackroyd and Christopher F. Evans (Cambridge: Cambridge University Press, 1970), 195. Reproduced with permission.

Other Ancient Versions

The scope of this book precludes any detailed discussion of other ancient versions of the OT. Information about these versions—Coptic, Ethiopic, and Armenian—can be found in various sources.91 These are all “daughter” translations made from the Septuagint, and therefore they are only of indirect usefulness for establishing the text of the Hebrew Bible.92 The Arabic translation of Saadia Gaon (882–942 CE) was based on the Hebrew text. It is the last of the ancient translations and the first medieval translation.93

Figure 4.3 contains one view of the history of the OT text.

Translation and Textual Criticism

In the introduction to this chapter, we mentioned that the ancient translations of the Bible are significant because they witness to the state of the Hebrew text long before the Masoretes wrote their codices in the Middle Ages. As we have surveyed these different translations, we have noted that they each have their own textual history and character. In some cases, there were multiple translators for each version, each with their own idiosyncrasies and different approaches, working in different parts of the canon. We must understand the implications of these distinctives for textual criticism and think about how to overcome them if we are to avoid errors in identifying and weighing textual variants.

It is important to remember that when we use an ancient translation as a textual witness, our goal is actually to uncover the Hebrew source text behind the translation (see fig. 4.4).

We begin by comparing the version (e.g., the Greek Septuagint or an Aramaic targum) with the MT. In figure 4.4, the version is in a triangle to indicate that it is in a language other than Hebrew. In making this comparison between the version and the MT, we are seeking to discover whether the version contains any variant readings—readings different from the MT—that might help us establish the earlier Hebrew text. Therefore, in comparing the version and the MT, we are actually attempting to uncover the Hebrew source text of the version and to make the comparison between it and the MT (labeled line 2).

Figure 4.4

Translations and Their Hebrew Source Texts

The challenge in uncovering the Hebrew source text of the version is that changes were introduced in the translation process (labeled line 1). First, the two languages have different syntax, vocabulary, and cultural associations. For example, as a Northwest Semitic language, Hebrew verbs occur in various binyanim (or stems) indicating transitivity, mood, and lexical meaning. Greek does not have any binyanim, so a Greek translator must decide how to render those nuances in a completely different language structure. On the other hand, Syriac does have binyanim quite similar to Hebrew, but while Hebrew verbs indicate aspect, Syriac verbs indicate tense. Therefore, a translator of the Peshitta needed to find new ways to represent the verbal nuances in a tense system.

Translators also introduce unintentional errors into their translations. Sometimes they misread the Hebrew text, much like scribes (see chap. 6). Sometimes they read the Hebrew text correctly but misunderstand or misinterpret its meaning, so that when they translate, they create a new reading.

Finally, translators introduce intentional changes in the translation process. Sometimes they make minor adjustments to “improve” the syntax, such as adding conjunctions or prepositions or harmonizing all of the verbs so that they agree with the context. Sometimes translators attempt to fix perceived grammatical errors in the Hebrew or remove theological problems. We saw in our discussion above that the targums attempt to maintain God’s transcendence and remove anthropomorphisms. Additionally, sometimes translators are very rigid in their word choice, consistently using one word in the target language to render every occurrence of a particular Hebrew word.94 Other translators may freely use multiple words in the target language to render the same Hebrew word, depending upon what they perceive it to mean in the context.

Figure 4.5

The Circular Nature of Translation Analysis

When we make our comparison between the version and the MT, how do we know which differences developed in translation (i.e., line 1 in fig. 4.4) and which actually represent a variant reading in the underlying Hebrew (line 2)? For the purposes of textual criticism, the former—differences in the version due to translation—are irrelevant. The only way to distinguish between these two possible causes of a difference between the translated version and the MT is to understand the translator’s usual behavior. For example, say we know that throughout the translation the translator freely adds words, but almost never omits words. If we encounter a situation where a word exists in the MT but not in the version, it is very likely that the word was not originally in the translator’s Hebrew source text, for we know that the translator does not usually leave things out.

Normally, we would learn the translator’s usual behavior by comparing the translation with its source text. As we compare each word or phrase in the target text with its corresponding segment in the source text, we note similarities and differences and come to an understanding of what the translator typically does or does not do in a given situation. But you may have already guessed where we get into trouble: we do not have the source text of any of these ancient versions! Therefore, our only recourse is to use the MT as a substitute source text for the purposes of comparison. But the MT is certainly not the source text of any of these versions: it was created 700–1000 years later! This creates a circular problem. We are using the MT as a substitute source text . . . in order to create a profile of the translation . . . in order to determine which differences in the version were independent of the MT. How can we break this impasse (fig. 4.5)?

Figure 4.6

A Recursion Model for Translation Analysis

Gideon Toury addresses this problem by suggesting that one should work in a “helical” (or spiral) pattern.95 In other words, we do not simply oscillate within this circular pattern, but we move forward as we gather more data. As an alternative analogy, we suggest the idea of recursion from computer programming. In recursion, the solution to a problem depends on solutions to subordinate problems. The computer “calls” itself. It runs, generates data, runs again with that data as its starting point, runs again, and so on.96 We can think about identifying the translator’s technique and the nature of the source text in the same way. We begin with a hypothesis (i.e., using the MT as a substitute source text), make comparisons, note the nature of differences, and then allow our conclusions about those differences to be the starting point of more comparisons. As we continue our work, we are developing a profile of the translator’s tendencies and then constantly refining that profile as we make judgments about the nature of his source text (see fig. 4.6).

It is clear, then, that we must consider the translator’s approach to his task as a significant variable when using an ancient translation as a textual witness. Fortunately, there are some practical ways that we can proceed through these challenges. First, we can rely on the work of others. There have been numerous studies on the translation technique of various versions of different books of the Bible.97 The text critic may consult these works and learn the general approach of the translator, the kinds of specific adjustments that he typically makes or does not make, and the relationship between his work and other versions. In addition, after much careful study, this detailed analysis puts scholars in a position to identify actual, significant textual variants.

Second, careful work with a concordance is an important first step in establishing correspondences between Hebrew and the version, especially when dealing with lexical equivalents. Let’s look at an example. In Hosea 9:1, the Hebrew word אֶתְנָן (prostitute’s fee) corresponds to the Greek δόματα (gifts) in the Septuagint. In English, these do not seem like they are exactly the same. Does the Greek version (G) have a reading that is different from that of the MT or not?

G		MT
You have loved	ἠγάπησας	אָהַבְתָּ	You loved
the prostitute’s fee	δόματα	אֶתְנָן	gifts
on	ἐπὶ	עַל	upon
all	πάντα	כָּל	every
the threshing floors	ἅλωνα	גָּרְנוֹת	threshing floor
of grain.	σίτου	דָּגָן	of grain.

The question we need to ask is: What is the typical Greek equivalent for Hebrew אֶתְנָן in the OT?

The easiest way to conduct this kind of search is to use Bible software. For example, in BibleWorks98 or Accordance,99 we can use a “Parallel Hebrew and LXX” module to conduct a search for the Greek equivalents of the Hebrew word אֶתְנָן. The search reveals that the most common Greek equivalent for אֶתְנָן is μίσθωμα (agreed upon price). Hosea 9:1 (our passage) is the only occurrence of δόμα (gift) corresponding to אֶתְנָןa:100

1	אתנן	αὐτός
1	אתנן	ἀρχαῖος
1	אתנן	Εθνανa
1	אתנן	δόμα
1	אתנן	Μισθός
3	אתנן	καίb
3	אתנן	ὁ
7	אתנן	μίσθωμα
aThis is a proper noun.     bThe search results depend upon the way that the database has been tagged. These results contain some false hits, including a conjunction and the article.

Because the use of δόμα as a translation for אתנן is unusual and the translator could have used μίσθωμα, we might conclude that either the Greek translator had something other than אתנן in his Hebrew source text or that he has made an intentional adjustment here for sense. Perhaps he chose δόμα because, in his mind, this word brought out the sense of the Hebrew as he understood it.

At this point, it might be instructive to conduct a reverse search of the Greek word δόμα in Hosea 9:1. The results of this search indicate that the Greek translators frequently use δόμα to translate the Hebrew word מַתָּנָה (gifts) or other general words for “gift” rather than technical terms like אֶתְנָן (prostitute’s fee). We conclude that this is probably a case in which the Greek translator simply made a translation decision. Perhaps he was avoiding the prostitute metaphor here. Or, perhaps he is intentionally linking Hosea’s statement in 9:1 to the Lord’s gifts that he had given to Israel. In any case, it seems likely that the translator thought that the Greek word δόμα made more sense in this context even though it was not the usual word that he might have used to translate אֶתְנָן. It is less likely that there was a variant reading in the Greek translator’s Hebrew source text. The Greek still has the same basic idea as the Hebrew, and the difference is a semantic nuance brought out in the choice of a specific Greek word to represent the Hebrew word.

Bible software is a very powerful way to conduct these kinds of searches. However, there is a fairly steep learning curve with each software platform, and there are sometimes multiple ways to conduct the same type of search. This case from Hosea 9:1 is merely an example of the kinds of work that we can do to get behind the translation and to investigate what the translator was doing and how he may have handled his source text.

An alternative to Bible software are concordances to the various versions. For example, the most important is the exhaustive concordance to the Septuagint by Edwin Hatch and Henry A. Redpath.101 One begins by searching for a Greek word in the Septuagint. The concordance indicates every occurrence of that word, and each is labeled with the underlying Hebrew word to which it corresponds. These data indicate whether that particular lexical correspondence is frequent and typical or not. An appendix in the back of Hatch and Redpath is organized according to the underlying Hebrew words. Each word has a list of Greek words that translate it, which can in turn be investigated in the concordance. While the use of a concordance is considerably slower than using Bible software, for many it might be a good place to start because of its intuitive layout, simple mechanics, and lower initial investment.