The name "Septuagint" designates the ancient Jewish-Greek translation of Hebrew Scripture. Septuaginta means “seventy” in Latin (usually indicated as “LXX”) and this name derives from the tradition that the very first Greek translation, that of the Torah (Pentateuch), was translated by seventy-two elders, six from each tribe. The number of seventy-two translators was subsequently rounded off to seventy. The story of the miraculous creation of the translation (thirty-six pairs of translators working in separate cells yet producing identical renderings in seventy-two days) is first represented in the Jewish-Hellenistic Epistle of Aristeas § 301–7 and expanded in later sources, especially Epiphanius, On Weights and Measures (fourth century c.e.). At the same time, the tradition in rabbinic literature, especially Soferim 1.5, of there being five translators of the Torah, one for each book, is more realistic than that of seventy-two (seventy) translators.
2. Nature and Content
The translation of the Torah into Greek was soon to be followed by that of the other books of Hebrew Scripture. However, the first translation was so dominant that its name, the "Septuagint," was ultimately attached to these other translations as well. The various translations differ greatly among themselves and the name "LXX" ultimately came to designate a group of many translations of different nature that represent different approaches and were produced at different times. Most translation units reflect the original Greek translations (the “Old Greek”), while some reflect anonymous later revisions, for example 2 Kings (4 Kingdoms in the LXX) and the Song of Songs. Daniel of the “LXX” contains a revision by Theodotion. These internal differences among the various translations in the collected Greek Scripture texts existed already in antiquity, and consequently modern editions of the “LXX” are of equally mixed character. When analyzing books of the LXX, one has to take this variety into consideration.
The collection of Greek Scripture contains Greek versions of all the books of Hebrew Scripture (the Hebrew "canon"). In addition, it contains Greek versions of Hebrew books such as Baruch and Sirach that were not included in the collection of Hebrew Scripture. It also includes writings originally written in Greek (e.g. 1–4 Maccabees), so that strictly speaking the “LXX” is not only a collection of translated works. All these Greek books, most of them translations from Hebrew and Aramaic, were accepted as authoritative (sacred) by the Alexandrian Jewish community and later by all the Jews of the diaspora and Palestine. Some scholars claim that even the writings that are not included in the Hebrew canon, such as Baruch, Sirach, and the Wisdom of Solomon were considered authoritative in Palestine at a certain period. Otherwise, so goes the argument, they would not have been accepted in Alexandria. All the books that are not included in the collection of Hebrew Scripture have been rejected by traditional Judaism, and are therefore traditionally named sefarim hitzoniyim (“outside books”, “external books”). For traditional Judaism the contents of the entire collection of Greek Scripture have no binding force. In Christianity the approach to the basis for modern translations of Hebrew Scripture (the “Old Testament”) is ambiguous. By and large, modern Christian and scholarly biblical translations are based on the Masoretic Text of Hebrew Scripture. However, occasionally these translations include readings from the LXX, while Jewish translations are strictly based on MT only. The JPS translation mentions some details from the LXX in footnotes. However, in scholarship the LXX is approached differently. It is increasingly accepted as an ancient source of Hebrew Scripture of equal status to MT, together with the Samaritan Pentateuch and several Qumran scrolls.
The books in the LXX are arranged differently from their position in Hebrew Scripture (“canon”). In the latter collection, the three large divisions are Torah, Prophets, and Writings, reflecting different stages of the gradual creation of Hebrew Scripture. On the other hand, the books of Greek Scripture are arranged according to their content (Torah and historical books, books of poetry and wisdom, and prophetical books, followed by the books of the New Testament [NT]). Within each group, the sequence of the books differs from Hebrew Scripture. For example, in Greek Scripture, Ruth (one of the Five Scrolls, included among the Writings in Hebrew Scripture) follows the book of Judges since its story took place "in the days of the Judges" (Ruth 1:1). Often the names of the books differ from their counterparts in Hebrew Scripture (e.g. Samuel–Kings are named 1–4 Kingdoms in the LXX).
Just as Hebrew Scripture is known in various forms, Greek Scripture (the LXX) is transmitted in various ways (scrolls, manuscripts, etc.). Hebrew Scripture is known mainly from the Masoretic Text (MT), the forerunners of which are found among the Dead Sea Scrolls dating from the third century b.c.e. until the second century c.e. The LXX, likewise, is known from ancient leather and papyrus scrolls and codices, among them several early copies found near the Dead Sea. The most reliable complete texts of the LXX are the codices B (Vaticanus), A (Alexandrinus), and S (Sinaiticus), from the 4th–5th centuries c.e. With the aid of these codices and some earlier witnesses, modern editions reconstruct the early form of the LXX, although admittedly such reconstructions are always tentative.
According to the Epistle of Aristeas, the translation of the Torah was initiated by Ptolemy II Philadelphus (reigned 285–246 b.c.e.). This date is probably correct, while most other details in this Epistle may be fictive. The translations of the Prophets and Writings were completed by the middle of the first century b.c.e. The grandson of Ben Sira knew the translation of the Prophets and part of the Writings in 132 or 116 b.c.e. according to different computations of the date of his Greek translation of Ben Sira.
5. Jewish Origin and Christian Use, Revisions of the LXX
The Jewish origin of the LXX is described in the Epistle of Aristeas, rabbinic literature, and various additional sources. Its Jewish nature is reflected in its terminology and exegesis. Several Hebrew words were preserved in the LXX in their Hebrew or Aramaic form (at the time of the translation, Aramaic was more commonly spoken by Jews than Hebrew). Some Hebraized Greek words in the LXX probably reflected the spoken language of the Alexandrian Jews, such as sabbata (Hebrew shabbat and Aramaic shabta’), and pascha (Hebrew pesach, Aramaic pascha’). The word holokautoma ("whole-burnt offering") was probably coined by the translators to reflect the special meaning of the ‘olah offering. Further, the Greek Torah made a distinction between two types of "altar" (mizbeah), a Jewish one rendered thysiasterion, and a pagan altar rendered bomos. The Aramaic Targumim likewise distinguished between the Jewish madbeha’ and the pagan ’agora’ (literally "heap of stones"). This distinction derived from the translators' wish to differentiate the Jewish religion from that of the non-Jews.
Jewish exegesis is visible wherever a special interpretation of the LXX is known also from rabbinic literature. Such exegesis reveals the Palestinian background of at least some of the translators. For example, the "second tithe" in the LXX of Deut. 26:12 (MT shenat ha-ma‘aser, "the year of the tithe," read as shenit ha-ma‘aser, as if, "second, the tithe") represents the rabbinic term ma‘aser sheni ("second tithe"). See further § 9 below.
The LXX translation was a Jewish venture, created for Jews and probably also for Gentiles. It was used by Jews in their weekly ceremonial reading from Scripture and served as the base for the philosophical-exegetical works of Philo and the historical-exegetical writings of Josephus. However, the central position of the LXX in Judaism did not last for a long period. It was soon recognized that the LXX often differed from the Hebrew text that was current in Palestine from the second-first centuries b.c.e. onwards and that was later to become the Masoretic Text. These differences were not to the liking of the Pharisaic (proto-rabbinic) circles, and soon a trend developed to replace the LXX with new translations. These new translations adapted the Old Greek translation to the Hebrew text then current in Palestine. They also changed the wording of the original translation when it represented the source text imprecisely (see below, § 8). Because of their revisional character, the translations that were produced after the Old Greek translation are usually named “revisions.”
The dislike of the LXX by the Jews became even stronger when the Greek writings of early Christianity (the "New Testament") based themselves, quite naturally, on the LXX (for Christianity: the Greek translation of the “Old Testament”). The LXX influenced the NT at various levels. Many of the terms used by the LXX translators became part and parcel of the language of the NT. For example, christos, originally a Greek rendering of the word mashiah ("the anointed") became the central name of Christ. Furthermore, the NT quotes the LXX frequently, and some of its theological foundations are based on the wording of passages in the LXX. For example, the idea of the birth of the Messiah to a virgin is based on the Greek translation of Isa. 7:14 "Behold, a virgin shall conceive . . ." (MT speaks of an ‘almah, a "young woman"). At an early stage the belief developed that this translation was divinely inspired and hence the way was open for several Church Fathers to claim that the LXX reflected the words of God more precisely than the Hebrew Bible. Christianity held on to the LXX as Holy Scripture until it was replaced by the Vulgate translation produced by the Church Father Jerome (created around 400 c.e.). In the Russian and Greek Orthodox churches, the LXX is still considered sacred.
6. Difficulties in Translating the Hebrew and Aramaic Texts
In the modern world, we are accustomed to translations of literary compositions, and it is hard to imagine that at one time no such translations existed. Indeed, in antiquity, cross-cultural enterprises such as the translation of Hebrew Scripture into Greek were a rarity. In fact, the LXX embodies the first major translation from an oriental language into Greek and it was the first written translation of Hebrew Scripture. Therefore, the translators had to overcome many problems. They also had to devise procedures for translating grammatical features of the Hebrew language (for example, infinitive absolute constructions like halokh halakhti, literally “to go I went”) with equivalent systems in Greek since such traditions had not yet developed.
When trying to analyze the Hebrew and Aramaic words, the translators could not resort to any tools such as dictionaries or any other sources of lexical information. They had to rely on their living knowledge of these languages as well as on exegetical traditions relating to words and contexts. We may assume that the translators were guided by such traditions when a specific rendering is found also in other sources. Thus a qesitah (a monetary unit of unknown value) is rendered in Gen. 33:19, Josh. 24:32, and Job 42:11 as “lamb” in the LXX, the Aramaic Targum Onkelos, and the Vulgate. This explanation is also reflected in Gen. Rabba 79:7. By the same token, the identification of difficult words was often guided by the context. Such a procedure frequently was little more than guesswork, especially in the case of rare and unique Hebrew words. Thus the translator of Isaiah rendered the rare word nesheph (“twilight”) differently in each of its three occurrences, based on its context (Isa. 5:11 [“late”], 21:4 [“soul”], 59:10 [“midnight”]). Often the translators derived the meaning of a rare word from its root (“etymological renderings”) or from its meaning in postbiblical Hebrew or in Aramaic, when they should have turned to its meaning in biblical Hebrew. In rare cases they left words without translation, representing them with Greek characters, for example barqanim (briers?) represented as “borkonnim” in Judg. 8:7, 16.
In light of these remarks, it causes no surprise that the translators often did not understand the Hebrew, and therefore sometimes created unusual renderings. For example, in Gen. 47:31 the LXX reads “And he swore to him. And Israel bowed over the top of his staff.” The background of this unusual Greek translation is that the translator misunderstood the vowels of the unvocalized hmth as ha-matteh (MT correctly reads ha-mittah, “…Then Israel bowed at the head of the bed.”). Unusual or not, the LXX rendering is quoted as such in the NT (Hebrews 11:21).
7. The Greek Language of the LXX
The LXX was written in the Hellenistic dialect of the Greek language, named koine, i.e. the dialect that was in general use (“koine”) by those who spoke and wrote in Greek from the fourth century b.c.e. onwards. Research into the language of the LXX is important, since it forms the largest literary source written in this dialect. However, the study of the language of the LXX is complicated because of its many lexical and syntactic Hebraisms that are described in § 8. Until the end of the 19th century many scholars asserted that the language of the LXX differed from the other witnesses of the koine because the Egyptian Jews spoke a special Greek dialect (“Jewish Greek”) that contained many elements deriving from the Hebrew language. There were even those who claimed that the language of the LXX and the NT (together: “biblical Greek”) was the language of the “Holy Spirit.” Deissmann’s investigations at the beginning of the 19th century made it clear that many words that were previously considered as characteristic of biblical Greek were in fact indigenous to Hellenistic Egypt. This conclusion was based on the many Hellenistic Greek papyri discovered in Egypt. It was thus the Hellenistic Egyptian “couleur locale” that made the language of the LXX appear different from other documents written in koine Greek. However, this view is only partially true, since the influence of the Hebrew on the language of the LXX should also be taken into consideration. The high level of adherence by the translators to the Hebrew created new meanings and usages that can only be explained against the Hebrew background of the LXX. Thus the standard rendering of shalom by eirene created a new shade of meaning in 2 Sam. 11:7 where “David asked … how the war was going” (… u-li-shelom ha-milhamah) was rendered as “… and the peace of the war” (eis eirenen tou polemou).
When the Greek translators could not express a Hebrew word adequately with an existing Greek word, they sometimes coined new words (“neologisms”). Examples are the verb sabbatizo (“to keep Sabbath”) and proselytos (“proselyte”) for Hebrew ger (“stranger,” understood in its postbiblical meaning as “someone who joined the religion of the Israelites”).
8. Translation Character
The first translators had to develop translation styles. The general approaches of translators are usually expressed as ‘‘literal”, “wooden”, “stereotyped”, “faithful”, or “careful” and their opposites, “free”, “contextual”, or when exceedingly free, “paraphrastic.” Between these two extremes many gradations and variations may be discerned, from extremely paraphrastic (to the extent that the wording of the parent text is hardly recognizable) to slavishly faithful. Life in contemporary Egypt may have offered some guidance in the development of a translation system. Some scholars thus assume that at the time of the translation a dragoman, a translator, was always available when needed for the translation of commercial or legal documents. According to some scholars this dragoman provided the example for the development of literal translations chosen by those translators who opted for a word-for-word rendering. For example, the characteristic Hebrew phrase in Gen. 11:10 “Shem was 100 years old” (literally: Sem was a son of one hundred years) was translated into Greek as “a son of.” In natural Greek a more appropriate phrase would have been chosen.
Word-for-word renderings reflect the translator’s respectful approach to Hebrew Scripture expressed by their faithfulness to details. Free renderings adapt the translation to the context, the Greek language, or the culture of the readers of the translation. However, free renderings, even very free ones, did not necessarily lack such respect. Translators who chose a free approach for their translations considered that the best way to transfer the word of God to another language and cultural environment was to adapt the source text to the new environment.
The books of the LXX are characterized by different translation styles that often appear within books of similar content. The reason for these differing styles is unclear. Thus, the versions of Jeremiah, Ezekiel, and the Minor Prophets are rather literal while the translation of Isaiah was free and at times very free. Similar differences are visible within the Hagiographa, where Psalms is presented in a very literal Greek version, while the translations of Job and Proverbs are very free and at times paraphrastic.
Analysis of the level of freedom and literalness in the translators’ approaches forms a key element in our understanding of them and their use as an ancient document in the study of Hebrew Scripture. In short, the argument runs as follows. If a translator represented his Hebrew text faithfully in small details, we would not expect him to insert major changes in the translation. Therefore, when we find major differences between the LXX and MT in relatively faithful translation units, they must reflect different Hebrew texts. These differing Hebrew texts are of central importance to our understanding of Hebrew Scripture. On the other hand, if a translator was not faithful to his parent text in small details, he also could have inserted major changes in the translation.
Most of the books of Hebrew Scripture were rendered into Greek in a relatively faithful way, while some are characterized by very literal renderings. In books of the latter type we can more easily assess the nature of the deviations from MT. Some books, however, were rendered freely. These units (see especially Joshua, Esther, and Daniel) pose special challenges since in these cases it is more difficult to assess the nature of the Hebrew text behind the LXX.
9. The World of the Translators
Many renderings reflect the cultural environment of the translators, which consisted of elements of both the Palestinian and Egyptian societies. The Egyptian background is visible in some local technical terms (e.g. the nogsim [“taskmasters”] in the story of the Israelites in Egypt in Exod. 3:7 and elsewhere, were rendered by ergodioktai, “literally: “those who speed up the workers,” known from Egyptian papyri; the Hellenistic division of cities into nomoi (districts) is reflected in the LXX of Isa. 19:2). The translators also updated elements relating to Egypt. Thus Joseph’s Egyptian name, Zaphenath-paneah, was changed in the translation to its Egyptian form, Psonthomphanech. It was probably the translator, and not his Hebrew parent text that rendered Goshen in Gen. 45:10 and 46:34 as “Gesem of Arabia” (Arabia was the name of the desert land east of the Nile). Palestinian background is reflected in Jewish-Palestinian exegesis. For example, in Exod. 22:19 yohoram (JPS “shall be proscribed”) is expanded by the LXX to “shall be destroyed by death,” as in B. Sanh. 60b where the verb is explained as “shall be stoned” by analogy to Deut. 17:2–5. See further § 5 above.
When analyzing individual renderings, the translator’s focus is on their linguistic and exegetical background and on the ideas behind them. Even in fixed and seemingly frozen renderings one sometimes recognizes the translator’s ideas. Thus the translator of the Latter Prophets, who usually rendered YHWH tzeva’ot (literally: "the Lord of armies") as kyrios pantokrator ("the Lord omnipotent") must have had a certain view of the Hebrew phrase. For him, tzeva’ot included not only a body of “angels” or “armies” but also encompassed everything in the universe. Some translators distinguished between the Jewish religion and that of the non-Jews in their terminology when rendering mizbeah (see above). Likewise, they distinguished between the cultic and non-cultic use of bamah ("high place"), lehem ("bread"), and minhah ("gift, sacrifice").
The translators often added religious background to verses in Hebrew Scripture. This phenomenon occurs especially in Esther and Proverbs. Probably the most characteristic feature of the LXX of Esther is the addition of a religious background to a book that lacks the mentioning of God’s name in MT. See further Isa. 5:13 "Therefore my people go into exile without knowledge" (NRSV), to which the LXX added "of the Lord." Likewise, Targum Jonathan often identified "knowledge" with "the Torah" (Isa. 28:9 etc.)
In several places, the translators interpreted the context as referring to the Messiah. Thus MT "A star rises from Jacob, a scepter comes forth from Israel" in Num. 24:17 is interpreted in the LXX as "A star shall come forth out of Jacob, and a man shall rise out of Israel.” A similar interpretation is reflected in the Aramaic Targumim.
In other instances, the translators avoided a physical depiction of God. Thus in Num. 12:8 "and he beholds the likeness of the Lord" has been rendered as referring to the "glory of the Lord."
10. The LXX and Textual and Literary Analysis of Hebrew Scripture
The LXX was translated from a Hebrew text that differed, often greatly, from MT. This is not surprising, since in antiquity many differing copies of the Hebrew Scripture text were in circulation. Some of these differences are minor, while others involve a whole paragraph, chapter, or even book. All these copies contain “Scripture.” In our analysis of Hebrew Scripture, we ought to supplement the data of MT with valuable information included in the LXX, some Qumran scrolls, and the Samaritan Pentateuch.
The small differences between the LXX and MT are recorded in the critical text editions of MT and the commentaries to the individual books. These details consist of small changes inserted by scribes and mistakes made in the course of the copying of the manuscripts. However, other differences may involve a whole chapter or sometimes a complete book. Such large differences are concentrated in specific biblical books and they may be summarized as follows:
Genesis: genealogies, chronological data, especially in chapters 5 and 11.
Exodus: the second account of the building of the Tabernacle in chapters 35–40.
Numbers: sequence differences, pluses and minuses of verses.
Joshua: significant transpositions, pluses, and minuses, especially in chapters 5, 20, 21, 23, 24.
Samuel–Kings: many major and minor differences, including pluses, minuses, and transpositions, involving different chronological and editorial structures. See especially 1 Sam. 1–2, 16–18, all of 1 Kings, 2 Kings 17, 21.
Psalms: Additional Psalm (Psalm 151).
Jeremiah: major differences in sequence, much shorter text throughout the book.
Ezekiel: slightly shorter text, different arrangement of chapter 7.
Proverbs: differences in sequence in chapters 24–31, different text.
Daniel and Esther: completely different books, including the addition of large sections, wrongly described as “apocryphal.”
Chronicles: “synoptic” variants, that is, readings in the Greek translation of Chronicles agreeing with the LXX in the parallel texts.
When the evidence from the LXX is analyzed together with that of MT, we often gain insights into the different composition stages of Scripture books. In some cases, the LXX contains a compositional layer that may have preceded that of MT (for example, Joshua and Jeremiah). In other cases, the Hebrew text underlying the LXX rewrote MT, often like a Midrash (for example, 1 Kings, Esther, and Daniel). In yet other cases, the relation between the two texts cannot be determined easily (for example, the chapters from Genesis, 1 Samuel, and Proverbs).
Bickerman, Elias. “The Septuagint as a Translation.” Pp. 167–200 in idem, Studies in
Jewish and Christian History. Leiden: E. J. Brill, 1967.
Dines, Jennifer M. The Septuagint. London/New York: T.&T. Clark, 2004.
Fernández Marcos, Natalio. The Septuagint in Context: Introduction to the Greek
Versions of the Bible. Leiden: E. J. Brill, 2000.
Jellicoe, Sidney. The Septuagint and Modern Study. Oxford: Clarendon Press, 1968.
Seeligmann, Isaac L. “Problems and Perspectives in Modern Septuagint Research,”
Textus 15 (1990): 169–232.
Silva, Moisés and K. H. Jobes. Invitation to the Septuagint. Grand Rapids, Mich.: Baker
Swete, Henry B. An Introduction to the Old Testament in Greek. 2d ed.; Cambridge,
Cambridge University Press, 1914.
Tov, Emanuel. “The Septuagint.” Pp. 161–88 in Mikra, Compendia Rerum Iudaicarum ad
Novum Testamentum, Section Two, I. Ed. M. J. Mulder; Assen–Maastricht and
Philadelphia: Fortress Press/Van Gorcum, 1988.
Idem, “The Nature of the Large-Scale Differences between the LXX and MT S T V,
Compared with Similar Evidence in Other Sources.” Pp. 121–44 in The Earliest
Text of the Hebrew Bible. The Relationship between the Masoretic Text and the
Hebrew Base of the Septuaginta Reconsidered. Ed. A. Schenker; SCS 52;
Atlanta, Ga.: Scholars Press, 2003.
Idem, The Parallel Aligned Text of the Greek and Hebrew Bible (division of the CATSS
database, directed by R. A. Kraft and E. Tov), module in the Accordance
computer program, 2002 (with updates 2003 – ) and the Logos computer program, 2004 (with updates 2005 – ).