We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 10 June 2011
Fifty years ago, Charles C. Torrey, writing about Esther, asked on the pages of this journal, “Why is there no Greek translation of the Hebrew text? Every other book of the Hebrew Bible, whatever its nature, has its faithful rendering (at least one, often several) in Greek. For the canonical Esther, on the contrary, no such version is extant, nor is there evidence that one ever existed.” It is common knowledge that the extant Greek versions of Esther, both the longer Septuagint text and the shorter A-text, are textually distant from the Hebrew Masoretic version. Indeed, the distance is so great that when a passage in the Complutensian edition (5:1–2) does correspond to the Masoretic text, Robert Hanhart confidently labels it as “newly translated.” His characterization seems justified in this case; the two verses required a new translation because the original Septuagint text had been removed, along with the apocryphal addition D, and put at the end of the book in accordance with the Latin tradition. Hanhart correctly states, “It is improbable that such an intervention, which sacrifices the inner coherence of the Greek text to the benefit of the Masoretic text, belongs to old Greek tradition,” indicating “a scholarly re-working according to the Masoretic text in the period of the Renaissance”; his confidence, however, rests on the fact that scholarly literature contains nothing about a Greek Esther that resembles the Masoretic text.
1 Torrey, Charles C., “The Older Book of Esther,” HTR 37 (1944) 1–40CrossRefGoogle Scholar, esp. 1.
2 The usual Septuagint text is called the B or β-text, in contrast to the rare, shorter version, the A or α-text; see Torrey, “The Older Book of Esther,” 5; and Moore, Carey A., Esther (AB 7B; Garden City: Doubleday, 1971) lxiiGoogle Scholar.
3 Hanhart, Robert, ed., Esther (Septuaginta 8.3; Göttingen: Vandenhoeck & Ruprecht, 1966) 42Google Scholar.
4 The additions in the Septuagint text are divided into chapters that are labeled alphabetically.
5 Ibid.
6 The term Slavonic denotes the type of languages used by medieval Orthodox Slavs, chiefly in religious texts, but also in more secular documents. The closely related and interchangeable variants—described by such compound epithets as Serbian (Church) Slavonic, Moldavian (Church) Slavonic, and Novgorod (Church) Slavonic—are all based on the original Old Church Slavonic written language elaborated by Constantine—Cyril and his brother Methodius in 863–85, and further codified between 890 and 973 in Bulgaria and later in various areas of Serbia and Rus' (the East Slavic realm up to about 1400). Note that Old Church Slavonic refers only to a small group of early manuscripts (and possibly graffiti). The modern Orthodox printed church books are in Synodal Slavonic, loosely codified in eighteenth-century Russia.
7 The manuscripts are written in Slavonic, with slightly varying details that allow us to locate most of the copies in the western lands that after 1400 belonged to the Lithuanian Commonwealth, in modern terms the Ukraine or Belorussia. The most striking idiosyncracy of the Slavonic Esther is the lexicon; this aspect, however, is of minor interest for the question of a Greek Vorlage.
8 For a typology of Slavic Hebrew Bible codices, see Mathiesen, Robert, “Handlist of Manuscripts Containing Church Slavonic Translations from the Old Testament,” Polata knigopisnaja 7 (1983) 3–48Google Scholar; and idem, “The Typology of Cyrillic Manuscripts,” in Michael S. Flier, ed., American Contributions to the Ninth International Congress of Slavists: Kiev, September 1983, vol. 1: Linguistics (Columbus: Slavica, 1983) 193–202.
9 We have ample evidence showing older and newer versions for the Gospels and Psalms, and also for much of the Pentateuch, Isaiah, Daniel, Ruth, and probably the Song of Songs. Although much work remains to be done in sifting out variants that have been considered as diagnostic for time and place, there is a reasonable comparative framework for evaluating whether a translation belongs to the period before or after about 1300.
10 Apparently even before 1800 an anonymous glossator (unfortunately not dated in our source, see Arximandrit Leonid, “Svedenija o slavjanskix rukopisjax postupivshix iz knigoxranilishcha Svjato–Troickoj Sergievoj Lavry v biblioteku Troickoj duxovnoj Akademii,” Chtenija obshchestva istorii drevnej Rossii [St. Petersburg: Tipografija Imperatorskoj Akademii Nauk, 1883] 4. 2) commented in the margin of a late fourteenth-century manuscript: “Neither Vulgate nor Septuagint, but a very precise translation from the Hebrew—an odd matter.”
11 Sobolevskij was so certain of his hypothesis that he published only brief and rather offhand statements (chiefly in an addendum to his book on translations, Perevodnaja literatura Moskovskoj Rusi XIV–XVII vekov [=Sbornik Otdelenija Russkogo Jazyka i Slovesnosti 74; St. Petersburg: Imperatorskoj Akademii Nauk, 1903] 433–36) that do not constitute a thorough analysis. For the history of the controversy and the major arguments see Altbauer, Moshé and Taube, Moshe, “The Slavonic Book of Esther: When, Where, and from What Language was it Translated?” Harvard Ukranian Studies 8 (1984) 304–20Google Scholar; and Lunt, Horace G. and Taube, Moshe, “Early East Slavic Translations form Hebrew?” Russian Linguistics 12 (1988) 147–87CrossRefGoogle Scholar.
12 Meshcherskij, Nikita A., Izdanie teksta drevnerusskogo perevoda ‘Knigi Esfir’ (Acta Universitatis Szegediensis, Dissertationes slavicae 13; Szeged: Tudomanyegyetem, 1978) 131–64Google Scholar. The faulty edition is inconsistent in editorial principles and teems with errors of every kind, but it gives the general sense of the shape and quality of the Slavonic book of Esther.
13 Moshé Altbauer, Professor Emeritus of Slavic of the Hebrew University in Jerusalem, obtained photocopies of the oldest manuscript (Moscow MS [Lenin Library] Troice-Serg. 2) and a fifteenth-century copy (Vilnius ms [Academy Library] 52), which he generously made available to us. More recently the Harvard University Library obtained microfilms of the second-oldest manuscript (St. Petersburg ms [Public Library] Q.I.2), two late fifteenth-century copies (Warsaw ms, Zamojski Library 105; St. Petersburg ms [Public Library] Q.I.838), and the 1499 Bible commissioned by Archbishop Gennadius of Novgorod (Moscow ms [State Historical Museum] Sin 915). We have also consulted several other copies in Russia. The evidence points to a single ancestral copy, made almost certainly in the “West Russian Lands,” probably no earlier than the mid-fourteenth century. We believe the translations may well have been made considerably earlier.
14 By comparing the reconstructable mid-fourteenth-century ancestral Vorlage of all extant Slavonic copies to the Masoretic text, we can identify forty-six omissions and thirty-three additions (beside the ten glosses defining the Jewish months in Greco–Roman terms). On the whole, the one hundred and sixty-seven verses remain intact, although the semantic equivalence is often tenuous and sometimes lost. We assume a reasonable translation from Hebrew to Greek, some textual modifications in the course of successive copyings of the Greek, a reasonable translation from Greek to Slavonic, and then a rather faulty transmission of the Slavonic version, probably with secondary editorial intervention to “improve” the translation of the extant text. A Russian critic, Anatolij A. Alekseev (“Russko-evrejskie literaturnye svjazi do 15 veka,” in Wolf Moskovich, et al., eds., Jews and Slavs [St. Petersburg/Jerusalem: Nauka, 1993] 1. 51–52) apparently assumes that the original Slavic text was itself obscure and faulty; see Lunt and Taube, “Early Slavic Translations,” 169 n. 29, 172 n. 43, 173 n. 46.
15 Dan 8:2; 9:1; Gebhardt, Oscar, ed., Graecus Venetus (Leipzig: Brockhaus, 1875)Google Scholar.
16 The distinguished theologian-philologist Ivan E. Evseev was confused by the difference between translation of names, such as Artaxerxes, and the details of transcription equivalents; for him, the Slavonic Axaъveros- was full proof of a direct translation from Hebrew to Slavonic and he brushed aside Sobolevskij's contentions that we might expect *Axašъveroš- from a Slav who knew Hebrew. Moreover, he indignantly rejected the notion that there ever could have been another Greek text. Two, hallowed by age and tradition, already existed, and to propose a third “would go off beyond the bounds of scholarship” (quoted in Lunt and Taube, “Early Slavic Translations,” 149).
17 Esth 1:10. Unless otherwise indicated, Slavonic evidence is cited according to our forthcoming edition.
18 Such an inconsistency appears in the name of the first eunuch in Esth 6:2 as vigъxvana (in the accusative) but in 2:21 as bixvan ъ, with b in spite of our expectation that *βιγθαν should yield the Slavonic *vigfanъ. We hold that this is a purely mechanical confusion between the Cyrillic B and Ъ; the evidence is ubiquitous. Furthermore, the second of these important eunuchs is not the expected *Feresъ (from *θερες,), but Vafes ъ, in Esth 2:21 and Vaverosa (in the accusative) in 6:2. Obviously the Hebrew ו (“and”) was carried over as an integral part of the name; such a misunderstanding can belong only to the translator from Hebrew. Compare Exod 6:19 in The Ostroh Bible, 1581 [Winnipeg: St. Andrews College, 1983] facsimile, where Omusi is derived from LXX Ομουσει, in turn, derived from ישומו [“and Mushi”].) The same uncertainty is apparent in two of the exotic names in Esth 1:10, which were further distorted by copyists. These failures to maintain a high level of transcription show a less than perfect knowledge of Hebrew on the part of the translator, but they do not, in our opinion, constitute proof that the translator rendered the Hebrew directly into Slavonic.
19 Evseev (“Zametki po drevneslavjanskomu perevodu sv. pisanija,” Izvestija Imperatorskoj Akademii Nauk 8 [1898] 339–44, esp. 341–42), invoking the late glosses, generally minimizes the significance of the names. Meshcherskij (“K voprosu ob izučenija perevodnoj pis'mennosti Kievskogo perioda,” Učenye zapiski Karelo-Finskogo pedagogičeskogo instituta 2.1 [1956] 198–219, esp. 209) vaguely speaks of “the living Hebrew pronunciation of that [?] time.” There were Jews in eleventh-century Kiev who knew Hebrew, but it is pointless to speculate about the details of their Hebrew pronunciation. Alekseev's remarks about these spellings (“Perevody s drevneevrejskix originalov v drevnej Rusi,” Russian Linguistics 11 [1987] 8–9; and idem, “Russko-evrejskie Literaturnye,” 51–52) show a lack of familiarity with the relationship of Hebrew and Septuagint names and introduce random elements from other Jewish-Slavic contacts that are irrelevant to this discussion.
20 Moore, Esther, 1.
21 We leave it to our readers to decide how many categories are represented in the two Greek versions and how they correspond to the Masoretic text.
22 See Sobolevskij, Perevodnaja, 435. The insertion of an extra i (“and”) before the phrase iže ixъ o nemb (“those of them around him”) changes the modifier into a new category of guest. Another superfluous i stands before the verb stvori (“made”) at the beginning of this excerpt; the i could be translated “even, indeed.” The confusion of wording in the verse shows a history of careless scribes and contradictory editorial intervention.
23 It occurs in the Bible only in Esth 1:3, 6:9 and Dan 1:3. In Esth 6:9, the Septuagint also has ἔνδοεος (“esteemed”). In Daniel, it uses ἐπίλεκτος, (“chosen”), while the variant redaction ascribed to Theodotion transcribes φορθομμειν/πορθομμειν.
24 Although confused and self-contradictory, Ivan Rozhdestvenskij (Kniga Esfir', v tekstax evrejskom—masoretskom, grecheskom, i slavjanskom [St. Petersburg: n.p., 1885]) provides the earliest explanation of stranam ъ; see Altbauer and Taube (“The Slavonic Book of Esther,” 307) and Lunt and Taube (“Early Slavic Translations,” 153) where the notes address Alekseev's murky speculations of 1987 (“Perevody s drevneevrejskix”). In 1993, Alekseev (“Russkoevrejskie literaturnye,” 53) muddies the waters even more.
25 The interpretation of initial sa as a prefix and respelling the word as sotrap- or sъrap- and then strapam ъ is a process supported by variants in manuscripts of Daniel. The final step, replacing Slavonic p with n, is easy, since the word satrap ъ, is extremely rare in Slavonic; in the chronicle of George Syncellus it is glossed, eže sutʻ voevody, “i.e., generals” (see Lunt and Taube, “Early Slavic Translations,” 174 n. 51) and in the chronicle of George the Monk it is rendered vojevoda and starěišina vojevodam ъ, “chief general.” One of the many lexical puzzles of Slavonic Esther is the lack not only of vojevoda (“general”) but of the otherwise ubiquitous knjazъ, “prince, ἄρχων” and vladyka, “ruler”—all three widely used in early Slavic to translate a variety of Greek terms.
26 Hanhart, Esther, 187.
27 See Istrin, Vasilij M., Aleksandrija russkix xronografov (Moscow: Universitetskaja Tipografija, 1893) 350Google Scholar, 356 n. 8; we are grateful to Francis J. Thomson for locating the Greek original of the sermon and providing the third variant.
28 Note that Emanuel Tov (“The ‘Lucianic’ Text of the Canonical and Apocryphal Sections of Esther: A Rewritten Biblical Book,” Textus: Annual of the Hebrew University Bible Project 10 [1982] 1–25, esp. 8) considers the phrase ἐποίησέ με ζῆν (“he made me live”) in one of the additions to the Greek A-text as evidence for a Hebrew original text with a Hiphil form.
29 The equivalent of םיאצמנה in Esth 4:16 is textually more complex but can also be explained by Greek mediation, in our opinion.
30 Compare, for example, John 10:8 vъsi elikože ixъ pride (πάντες ὅσοι ἦλθον; “all as many [of them] as came”) and Gen 39:22 vsja … elikože ixъ vъ temnici (πάντας ὅσοι ἐν τῷ δεσμωτηρίῳ; “all as many [of them as (were)] in prison”).
31 Compare Acts 20:26.
32 Compare LXX Esth 5:13 καὶ ταῦτά μοι οὐκ ἀρέσκει (“and this does not please me”) with ἀρκέσει (“suffice”) in one Greek manuscript and Old Latin non sufficiunt mihi. Similar examples are found in Num 22:34, Joel 17:16. Incidentally, בוט ךלמה לע םא (“if it please the king”) occurs seven times in the Masoretic text; Slavonic Esther has six different equivalents. The agreement of Esth 7:3 and Esth 8:5 on ašče carevi udobrju (“if I appear good to the king”; with a verb that seems inappropriate) is clearly a secondary interpretation.
33 See for example, Num 19:17 and Deut 4:39. Susan Daniel (“Expressions with Head, Mouth, Heart in the Septuagint Translation of the Bible,” in Mosheh Bar-Asher, et al., eds., Hebrew Language Studies: Presented to Professor Zeev Ben Hayyim [Jerusalem: Magnes, 1983] 161–72, esp. 167 [Hebrew]) calculates the proportion of διάνοια to καρδία for בל at about 1:2. In other words, out of seventy-four occurrences of διάνοια, thirty-nine correspond to בל or בבל (see Hatch, Edwin and Redpath, Henry A., A Concordance to the Septuagint [Oxford: Clarendon, 1879–1906]Google Scholar).
34 Note that רה (“turn”) occurs in the Masoretic text only in these two verses of Esther. LXX Ezek 9:4, 6 has σημεῦον for ות (“mark”). For Esth 2:12, we suggest ἐν τῳ ἐλθεῦν (τὸ) σημεῦον (“in the coming of the signal”) as the intermediary reading.
35 The codex Venetus has ΚΑΙΡΟϒϹ for ΚΛΗΡΟϒϹ in Esth 9:26 (Hanhart, Esther, 204).
36 The Greek versions also suggest that the translators were puzzled by the Hebrew phrasing in Esth 6:12. The Septuagint wording λυπούμενος κατὰ κεφαλήν involves the word “head,” but is incoherent (something like “grieving as to head”). The A-text has a single word, ἐσκυθρωπωμένος (“of sad countenance”).
37 Moore (Esther, 72) accepts the view that Haman is instantly being treated as a condemned criminal whose head is covered as a sign of his guilt.
38 In Judg 18:7, διατρέπω is the first member of a doublet, διατρέπω ἤ καταισχύνων (“perverting and shaming”) that translates the Masoretic text's םילכמ (“shame, humiliate”). In Dan 1:10, where the Septuagint has μήποτε ἴδη τὰ πρόσωπα ὐμῶν σκυθρωπά (“lest he see your faces gloomy”), Theodotion has the doublet διατετραμμένα καὶ ἀσθενῆ (“perverted and ill”). Hebrew has םיפען (“angry, ill-humored, sullen”), but since the context implies undernourishment as the cause of this negative state, translators reach for more suitable epithets. In Dan 1:13, the Masoretic text and the Septuagint have no adjective, but Theodotion paraphrases ἐὰν φανῆ ἠ ὄψις ἠμῶν διατετραμμὲνη (“whether our faces will look perverted”).
39 Moore (Esther, 72) enumerates the proposed suggestions to emend the Masoretic text of Esth 7:8; we venture to propose ןלפנ (“fell”) instead of ופח (“they covered”).
In any case, evidence for a Greek intermediary with συνὲπεσεν is found in 1 Kingdoms 1:18, where τὸ πρόσωπον αὐτῆς οὐ συνὲπεσεν ἔτι (“and her face no longer fell”) is a free rendering of 1 Sam 1:18 רןע הל ןיה אל הינפן (literally, “and her face was no longer to her,” meaning “her face was not sad any more”). Here one Slavonic tradition also translated freely, i lice eja ne izmêni sja k tomu (literally, “and her face did not change thereafter”; Warsaw MS BOZ 105), although the 1581 Ostroh Bible has a more literal wording: i lice eja ne ispade k tomu (“and her face did not fall thereafter”).
40 Compare Vulgate 7:7 arboribus consitum (“planted with trees”) and 7:8 nemoribus consito (“planted with groves”).
41 Note that dъnešbnii also appears in Esth 4:11 הימיפה רצחה לא (“into the inner court;” LXX: εἰς τὴν ἐσωτέραν) and Esth 5:1 הימיבפה ךלמהחיב רצחב (“in the inner court of the royal house”).
42 See Moore, Esther, 7.
43 Compare also Ibn Ezra ad locum, who cites the pair ץוחמ/חיבמ (“within/outside”).
44 Moore (Esther, 7) characterizes Esth 1:6 as “a difficult and corrupt verse, which in the Hebrew is syntactically unrelated to the preceding material.”
45 In Old Church Slavonic, βύσσος simply was borrowed. The spellings in later copies of old texts vary wildly, and the term often is replaced by words whose proper meaning is “purple” or “crimson.”
46 See Grossfeld, Bernard, The First Targum to Esther (New York: Hermon, 1983) 81Google Scholar n. 16.
47 Moore, Esther, 30 n. 20.
48 Hannah's words in Slavonic 1 Kingdoms 2:1 utverdi sja serdce moje (“my heart is affirmed”) are easily derived from the Septuagint ἐστερεῶθη ἡ καρδία μου (“my heart is established”), but the connection with the יבל ץלע (“rejoiced”) of 1 Sam 2:1 is tenuous.
49 A second possibility is a confusion between ןמא (“nurture”) and ץמא (“make firm, strengthen”). A parallel may be found in Ps 17:18; the Old Church Slavonic (normalized) Jako utvrъdiše sę pače mene is a literal translation of ὅτι ἐστερεῶθησαν ὑπὲρ ἐμὲρ ἐμὲ (literally, “for they were established more than I,” that is, they are stronger). The Masoretic text (Ps 18:17) has יכממ ןצמא (“for they were strong[er] than I”).
50 Grossfeld, First Targum to Esther, 55.
51 Torrey, “The Older Book of Esther,” 36.
52 Moore, Esther, 70.
53 Zimmermann, Frank (Biblical Books Translated from the Aramaic [New York: Ktav, 1983] 91Google Scholar) offers an extremely conjectural hypothesis that the supposed original Aramaic meant “so that distress be not put in the palace of the king”—a sense very close to the Slavic clause—involving a misreading of לכיה (“palace”).
54 Haupt, Paul, “Critical Notes on Esther,” AJSL 24 (1907–1908) 151Google Scholar.
55 See for example, the (Jewish) Spanish translation (Biblia en lengua Española traduzida palabra por palabra de la verdad Hebrayca por muy excelentes letrados vista y examinada por el officio dela inquisicion [Ferrara: n.p., 1553]) toda forteza, and the Yiddish (Pentateuch with the Five Scrolls in [Judeo] German Tongue Written a Long Time Ago and Now Printed in Augsburg [Augsburg: n.p., 1544]) al di šterkung.
56 Moore (Esther, 51) has “in this condition.”
57 As it stands, the text has vraždu i, “enmity and” instead of “die, which (is)”; we emend to vražu jaže. Without emendation, this phrase is meaningless, and it ruins the story. It has given rise to much discussion; see Altbauer and Taube, “The Slavonic Book of Esther,” 308; and Lunt and Taube, “Early East Slavic Translations,” 152–53, with notes.
58 Note that the Slavic adjective is derived from žrebii (“lot”), which appears in Esth 3:7 for the normal Hebrew term לךןכ (“lot”).
59 The fourth example of rastvoriti is in Esth 4:8, where the Masoretic text has מרימשהל. The fifth occurrence of the Hebrew רימשה, in Esth 3:6, is translated by the more appropriate pogubiti, (“cause to perish”) which corresponds to רבא in seven other passages.
60 Perhaps rastvoriti contributed to a further corruption in Esth 3:13 in the two oldest copies where rastvoriti i izbivati i pogubiti (“dissolve and kill and cause to perish”), correspond to only two verbs, rastvoriti i izbayiii (“dissolve and save”), a devastating shift of meaning.
61 Another instance of consistency in error is vlasjanica i sukbno (“hair shirt and homespun”) instead of “sackcloth and ashes” in Esth 4:1 and 4:3; there is good reason to ascribe this normalization to a Slavic editor.
62 An attested parallel with older razoriti replaced by newer rastvoriti is found in the earlier rather than the later redactions of the apocryphal Life of Moses, which we believe was translated from Hebrew in about 1400.
63 The Babylonian Talmud (b. Meg. 17–18), in discussing the obligation of reciting the scroll of Esther on Purim, raises the question of whether recitation in Greek fulfills the obligation. In view of the general awareness of the Septuagint's textual deviations from the Masoretic text (see the discussion in b. Meg. 9), we must conclude that a suitable Greek text—not the Septuagint, but more faithful to Masoretic text—was known to the third-century rabbis. In the 1547 Polyglot Pentateuch, printed in Constantinople, which contains the Hebrew Masoretic text along with the Aramaic Targum and the Ladino and Judeo-Greek translations of the Pentateuch, all in Hebrew characters, the title page promises not only the five books of Moses, but also the Five Scrolls, a combination that is customary in printed Hebrew Pentateuchs. Unfortunately, none of the copies of this 1547 volume contain any of the scrolls.
64 We disagree with the common Russian point of view that Old Church Slavonic books circulated in Rus' before the official conversion in 988, and that there was a high level of learning, with lively schools of translators by ca. 1050. In particular, we reject the repeated declarations (for example, by Alekseev, “Perevody s drevneevrejskix”) that Slavic bookmen in Rus' made translations from Hebrew before the fifteenth century; no credible evidence has been presented to support this myth. For some details and further references, see Altbauer and Taube, “The Slavic Book of Esther”; Lunt and Taube, “Early East Slavic Translations”; Horace G. Lunt, “On Interpreting the Russian Primary Chronicle: the Entry for 1037,” Slavic and East European Journal 32 (1988) 251–64; and idem, “The Language of Rus' in the Eleventh Century: Some Observations about Facts and Theories,” Harvard Ukranian Studies 12/13 (1988/1989) 276–313.
