The Septuagint and Egyptian Translation Methods

The Septuagint and Egyptian Translation Methods James K. Aitken Abstract: Theories on the identity of the Septuagint translators have been built upon assumptions regarding the evidence from multilingual Egypt. While evidence of translation activity in Egypt is limited, there are a few surviving translations from Demotic into Greek that can shed much light on the Septuagint translation practice. Similar approaches to Greek register, lexical consistency, transliteration, inflection of noun phrases, polysemous prepositions, interpretive renderings and literary embellishment are found in both Egyptian translations and the Septuagint. From this conclusions can be inferred on the social position of the Jewish translators in Egypt. The contents of Mr. Grey’s manuscript are of a nature scarcely less remarkable than its preservation and discovery: it relates to the sale . . . of a portion of the Collections and Offerings made from time to time on account, or for the benefit, of a certain number of MUMMIES, of persons described at length, in very bad Greek.1 Thomas Young’s words reflect the excitement of finding a Demotic text with an extant Greek translation in the early years following the decipherment of Hieroglyphs and accordingly of Demotic. His excitement is coupled with surprise at the contents in this particular case of the transfer of legal ownership over the rights of caring and making offerings to mummies. Amidst his enthusiasm regarding this translation, however, is a swift denigration of the 1. Thomas Young, An Account of Some Recent Discoveries in Hieroglyphical Literature, and Egyptian Antiquities. Including the Author’s Original Alphabet, as Extended by Mr. Champollion, with a Translation of Five Unpublished Greek and Egyptian Manuscripts (London: John Murray, 1823), 60. -1- 2 XV Congress of the IOSCS: Munich, 2013 Greek as “very bad.”2 Similar judgements of Greek written by Egyptians, the “grammatical blunders”3 of the “uneducated”4 class, continued well into the twentieth century. The restitution of the status of the Greek of Egyptian translations is as much called for as it is in the case of biblical Greek. Furthermore, the importance of such Egyptian translations for contextualizing Septuagint Greek has been largely ignored, although it sheds much light on the nature of the Greek and the translation technique adopted. 1. THE SEPTUAGINT AND TRANSLATION IN ANTIQUITY It is generally recognized that the first translations of the Septuagint, namely of the Pentateuch, were undertaken in Egypt. 5 Beyond this we know little about the context of the translation or the identity of the translators. One of the few resources we have is the nature of the translations themselves and the translation technique therein, and it is from that technique that theories have often been derived regarding the social context or religious presuppositions of the translators. Any theory is predicated, however, on the nature of translation in antiquity and the extent to which the Septuagint is distinctive in its method. It is regularly presented as a unique enterprise, a literary translation unprecedented in scale for its time.6 Brock has done the most to place the translation within the context of other translation activity in antiquity, and justifiably has influenced 2. His words are (without precise reference) noted by Rachel Mairs, “κατὰ τὸ δυνατ ν: Demotic-Greek Translation in the Archive of the Theban Choachytes,” in Beyond Free Variation: Scribal Repertoires in Egypt from the Old Kingdom to the Early Islamic Period (ed. Jennifer Cromwell and Eitan Grossman; Oxford: Oxford University Press), forthcoming, who also compares them to those of P. W. Pestman, The Archive of the Theban Choachytes (Second Century B.C.): A Survey of the Demotic and Greek Papyri Contained in the Archive (Leuven: Peeters, 1993), 333. See too the criticism of such assessments by Marja Vierros, Bilingual Notaries in Hellenistic Egypt: A Study of Greek as a Second Language (Collectanea hellenistica 5; Brussels: Koninklijke Vlaamse Academie van België voor Wetenschappen en Kunsten, 2012), 225. 3. B. P. Grenfell and A. S. Hunt, New Classical Fragments and Other Greek and Latin Papyri (Oxford: Clarendon Press 1897), 46. 4. James H. Moulton “Grammatical Notes from the Papyri,” Classical Review 18 (1904): 106–12, 151–55 (151). 5. E.g., Jan Joosten, Collected Studies on the Septuagint: From Language to Interpretation and Beyond (FAT 83; Tübingen: Mohr Siebeck, 2012), “Historical Milieu,” 185–239. 6. Cf. Sebastian P. Brock, “The Phenomenon of the Septuagint,” OtSt 17 (1972): 1136 (11); Tessa Rajak Translation and Survival: The Greek Bible and the Jewish Diaspora (Oxford: Oxford University Press, 2009), 1. Author, Title 3 subsequent studies.7 He stressed the unparalleled nature in the Hellenistic world of such a translation, but also pointed to translations in various cultures where the style varied, noting that legal texts tend to be literal while literary ones more free.8 For Brock the Septuagint was a compromise between the two. This led on to a discussion of the religious significance of literal translation, especially as taken up by the school of Aquila. In his later article he explored further the relation between literal and free translations, examining literary accounts of translation processes.9 He drew comparison with the senatus consulta documents in the east, where the Greek follows closely the Latin, and he reached the conclusion that legal and religious texts are “literal,” typified by the senatus consulta in which the Greek directs the reader back to the original in Latin. 10 In Brock’s work some attention is given to Egyptian translations, if only briefly. For examples of religious texts he notes two instances of translation from Greek into Egyptian (the Canopus decree of 238 B.C.E. and the Rosetta Stone of 196 B.C.E.),11 one freer than the other. He also draws attention to one literary text where both the Greek and Demotic survives, the legend of Tefnut, 12 although in this case the translation is more a literary rewriting of what has been preserved in Demotic (see below § 3.1). Iamblichus is called upon to justify that there was a reluctance to translate religious texts in Egypt, 13 although how far literary rationalisation should be used to account for actual practice is debatable. What we find then in Brock are some conclusions regarding translation technique derived from a large range of sources, and only brief comparative mention of Egyptian translations without detailed analysis. Prior to Brock, Bickerman and Rabin had also examined the Septuagint within the context of translation in antiquity, although neither play much of a role in Brock’s discussions and both seem to be largely rejected in recent 7. Brock, “Phenomenon.” 8. Brock, “Phenomenon,” 17. 9. “Aspects of Translation Technique in Antiquity,” Greek, Roman and Byzantine Studies 20 (1979): 69–87. 10. “Aspects,” 74. 11. It remains debated as to the original language of the Rosetta Stone. 12. “Phenomenon,” 17–19. 13. “Aspects,” 75–76. The picture is now far more complex than Brock allows. Ian S. Moyer, Egypt and the Limits of Hellenism (Cambridge: Cambridge University Press, 2011) discusses in detail the complex relationship between Greek and Demotic. See too Nikolaos L. Lazaridis, Wisdom in Loose Form: The Language of Egyptian and Greek Proverbs in Collections of the Hellenistic and Roman Periods (Leiden: Brill, 2007). 4 XV Congress of the IOSCS: Munich, 2013 scholarship.14 The weakness of Bickerman’s and Rabin’s studies lies in their adoption of Kahle’s theory of an oral translation tradition, on the model of the Targum,15 and in their conjectures as to how a translator would behave, without sufficient citation of evidence. Nevertheless, they did aim to place the Septuagint within Egyptian translation practice, and although some of their presuppositions can be questioned, their work has perhaps been undervalued in recent years. Bickerman explicitly stated that the features of the translation can be explained by the traditional art of translation as known at the time. He sought to compile evidence of translators in Egypt, 16 and, building upon the insights from the papyri that the language is the vernacular and not a peculiar dialect, aimed to place the translation within the broader social landscape. 17 Believing the theory of extempore oral translation, he attributed the work to the “dragoman,” itself an anachronistic term from the Ottoman court. Rabin reinforces this oral nature of translation by suggesting, erroneously, that Greek society did not go in for written translation as such, preferring rewriting. 18 He assigns to the dragoman certain practices with little evidence beyond occasional reference to Phoenician and Punic inscriptions, and once to Demotic. 19 His conjectures, though, should not obscure the valid attempt to explain the translation in the Egyptian context of translation. His suggestions too of features in the Septuagint that can be accounted for as the work of a professional translator should also be considered: following word-lists to avoid mistakes (15), word-for-word and clause-for-clause translation (16-17), use of prepositions (18) transliteration (22),20 occasional good renderings (23), and “intentional” barbarisms (28). Rabin largely follows Bickerman on these phenomena (including “Semitisms” of syntax), but further suggests that omission of parts of text, words, or phrases derives from the model of the dragoman technique.21 To him such changes would be strange for one respecting a sacred text, but would make sense in an oral business translation where the 14. Elias J. Bickerman, “The Septuagint as a Translation,” Proceedings of the American Academy for Jewish Research 28 (1959): 1-39; Chaim Rabin, “Translation Process and the Character of the Septuagint,” Textus 6 (1968): 1–26. 15. Bickerman, “The Septuagint,” 8; Rabin, “Translation Process,” 23. 16. Bickerman, “The Septuagint,” 14 n. 27. 17. Bickerman, “The Septuagint,” 11–12. 18. Rabin, “Translation Process,” 19. Assuming that there was no model for the translation, he believed the idea had to come from the synagogue Targum (20). 19. E.g., Bickerman, “The Septuagint,” 14: “professional dragomans who generally clung to the letter.” 20. Cf. Rabin, “Translation Process,” 24. Oddly, most examples are given from LXX with very few from other translations. 21. “Translation Process,” 23. Author, Title 5 main purpose is to convey the message. As Bickerman, Rabin builds his argument from supposition, and is thus criticised by others, especially for a lack of coherence in the features identified and for still relying on a theory of a Jewish specialized (Hebraic) Greek.22 It remains true that there is little comparable evidence from Egypt to study translation of the time, but there are sufficient data to save us from moving as far afield as Brock did, or from resorting to the speculation of Bickerman and Rabin. There is considerable evidence of multilingualism in Egypt, where languages played complex roles within society including diglossia between Middle Egyptian (Hieroglyphs) and Demotic. More significantly some examples of translation survive that can be examined. Egyptian translations are the obvious place to begin in contextualising the Septuagint, and they do confirm the phenomena identified by Bickerman as typical of Egyptian translators, even if his explanations and characterizing of the translators require reconsideration.23 The importance of specifically Egyptian translations is that they come from the same region and same time period (Ptolemaic) as the Septuagint, and they represent translation from a different (non-Indo-European) language family (Afroasiatic, i.e. Hamito-Semitic) into Greek. There are a number of grammatical affinities between Hebrew and Egyptian Demotic that differentiate the languages from Greek. It may be, following Brock, that different genres reflect different types of translation style, but such simple labels as literal or free are not reflective of the complexity of translation styles. We need to examine the minutiae of the translation methods to gain a greater understanding of the techniques employed. In the case of the Septuagint we might well have a mix of genres within the Pentateuch,24 and we do not know how the first translators viewed the text or understood their task. It is not self-evident that they were translating literature. 2. TRANSLATION IN MULTILINGUAL EGYPT With the arrival of Greeks as the new colonizers and settlers in Egypt, Greek soon became the language of administration and the means by which anyone, including native Egyptians, could advance in society. For the first generation 22. See, e.g., Albert Pietersma, “A New Paradigm for Addressing Old Questions: The Relevance of the Interlinear Model for the Study of the Septuagint,” in Bible and Computer. The Stellenbosch AIBI Conference. Proceedings of the Association Internationale Bible et Informatique ʽFrom Alpha to Byte.ʾ University of Stellenbosch 17-21 July, 2000 (ed. Johann Cook; Leiden: Brill, 2002), 337–64 (343–44). 23. On a cultural level this is what J. Mélèze Modrzejewski aimed to do: “‘Livres sacres’ et justice lagide,” Acta Universitatis Lodziensis, Folia Juridica 21 (1986): 11–44 24. Recognized by Brock, “Phenomenon,” 20. 6 XV Congress of the IOSCS: Munich, 2013 under Ptolemy I and Ptolemy II the administration functioned largely in the indigenous Egyptian Demotic language,25 but during the reign of Ptolemy II the number of Greek documents started to outstrip Demotic, even when it is clear that much of the Greek documentary sources were being written by native Egyptians.26 Demotic continued to hold a special place among the Egyptian priests and was used alongside Greek. Egypt became, even more than before, a multilingual environment, where translation, both oral and written, was the everyday norm.27 Translators are referred to in various documents, although they reveal little about their working practices, 28 and at times written translations would not always have been needed, when oral translation was also possible. Bilingual archives illuminate the situation in which persons within the one archive seemed to be competent in both languages. 29 We have evidence of bilingual individuals,30 such as Dionysios son of Kephalas, an Egyptian and 25. Dorothy J. Thompson, “Literacy and Power in Ptolemaic Egypt,” in Literacy and Power in the Ancient World (ed. Alan K. Bowman and Greg Woolf; Cambridge: Cambridge University Press, 1994), 67–83 (71–72). 26. Willy Clarysse, “Egyptian Scribes Writing Greek,” ChrEg 68 (1993): 186–201; Josh Sosin and Joseph G. Manning “Paleography and Bilingualism. P. Duk. inv. 320 and 675,” ChrEg 78 (2003): 202–10. 27. For summaries of the multilingual situation in Egypt, see Thompson, “Literacy and Power”; B. Rochette, “Sur le bilinguisme dans l’Égypte gréco-romaine,” ChrEg 71 (1996): 153–68; Arietta Papaconstantinou, “Introduction,” in The Multilingual Experience in Egypt, from the Ptolemies to the Abbasids (ed. Arietta Papaconstantinou; Farnham/Burlington, VT: Ashgate, 2010), 1–16; Vierros, Bilingual Notaries, 29–34. 28. Cf. W. Peremans, “Les ἑρμηνεῖς dans l’Égypte gréco-romaine,” in Das römischbyzantinische Ägypten: Akten des internationalen Symposions 26.–30. September 1978 in Trier (ed. G. Grimm, H. Heinen, and E. Winter; Mainz: Philipp von Zabern, 1983), 11– 17; B. Rochette, “Traducteurs et traductions dans l’Égypte gréco-romaine,” ChrEg 69 (1994): 313–22; B. G. Wright, “The Jewish Scriptures in Greek: The Septuagint in the Context of Ancient Translation Activity,” in Biblical Translation in Context (ed. F. W. Knobloch; Bethesda, Md.: University Press of Maryland, 2002), 3–18; Trevor V. Evans, “The Court Function of the Interpreter in Genesis 42:23 and Early Greek Papyri,” in Jewish Perspectives on Hellenistic Rulers (ed. T. Rajak, S. Pearce, J.K. Aitken, and J. Dines; Berkeley, Calif.: University of California Press, 2007), 238–52. 29. Willy Clarysse “Bilingual Papyrological Archives,” in Arietta Papaconstantinou (ed.), The Multilingual Experience in Egypt, from the Ptolemies to the Abbasids (Farnham: Ashgate, 2010), 47–72. See too E. Bresciani and R. Pintaudi, “Textes démotico-grecs et gréco-démotiques des ostraca de Medinet Madi: un problème de bilinguisme,” in Aspects of Demotic Lexicography (ed. S. P. Vleeming; Leuven: Peeters, 1987), 123–26; Katelijn Vandorpe, “Archives and Dossiers,” in The Oxford Handbook of Papyrology (ed. Roger S. Bagnall; Oxford: Oxford University Press, 2009), 216–55 (240–42). 30. Moyer, Egypt, 31 n. 115. Author, Title 7 priest of the local Ibis cult. He became Hellenized, having been recruited as an infantry man into a Ptolemaic garrison, and as the head of a mixed family was apparently competent in both languages. 31 One Apollonios too appears to be bilingual and possesses texts along with his brother in both languages. A Ptolemaic letter records the words of a woman writing to her husband who is learning Egyptian to teach it to Greek slave boys learning Egyptian medical techniques (reflecting both a woman who is bilingual and bilingual education at the time): Discovering that you are learning Egyptian writing, I am happy for you and for myself, because now when you come to the city you will teach the slave-boys in the establishment of Phalou. . . the enema-doctor, and you will have a means of support for old age. (UPZ I 148 second century B.C.E.)32 The ability to write in both languages might also be possible, although assistance might have been on hand. In one letter (third century B.C.E.) a Greek man speaks of how he has learned Egyptian, perhaps to partake of the Egyptian speciality of interpretation of dreams: Ptolemaios to Achilleus, greeting. After writing about the. . ., it seemed good to me to inform you also about the dream, so that you may know in what way the gods know you. I have written below in Egyptian, so that you may understand correctly. . . (there follows a Demotic description of the dream) (P.Cair.Goodsp. 3) Despite the wide contact between the two languages and despite the scribal class of Egyptian priests learning Greek alongside Demotic in temples, the influence of Demotic Egyptian upon Greek is difficult to detect. Mayser only finds, for example, 23 loan-words from Egyptian in Greek, and some of these are doubtful,33 indicating how significant the few Egyptian loan-words are in the 31. Texts in Katelijn Vandorpe, The Bilingual Family Archive of Dryton, His wife Apollonia and Their Daughter Senmouthis (P. Dryton) (Collectanea hellenistica 4; Brussels: Comite Klassieke Studies, Subcomité Hellenisme, Koninklijke Vlaamse Academie van België voor Wetenschappen en Kunsten, 2002). Discussion in Naphtali Lewis, Greeks in Ptolemaic Egypt: Case Studies in the Social History of the Hellenistic World (Oxford: Clarendon, 1986), 88–103; Vandorpe, “Archives,” 226. 32. See R. Rémondon, “Problèmes de bilinguisme dans l’Égypte lagide,” ChrEg 39 (1964): 126–46. 33. Edwin Mayser, Grammatik der griechischen Papyri aus der Ptolemäerzeit (4 vols; Leipzig: Teubner, 1906–1934), 1:35–40; noted by G. Mussies, “Egyptianisms in a Late Ptolemaic Document,” in Antidoron Martino David oblatum. Miscellanea 8 XV Congress of the IOSCS: Munich, 2013 Greek Pentateuch.34 The reason for the lack of influence between the languages has been attributed to the care of the priests to preserve their linguistic heritage.35 Nonetheless, Clarysse has provided a survey of possible features in Greek that indicate interference.36 Identifying any Egyptian influence can be difficult when we are still learning about the history of Koine, and some features could be attributed to internal developments within the language and need to be compared with the standard post-classical Greek.37 As much for Egyptian documents as for the Septuagint, the possibility has to be considered of a change in the history of language as well as contact-induced change. In his study of Egyptianisms Mussies is careful with his evidence in contrast to earlier editors. He examines documents that are known to have been translations, stated in their superscriptions. He considers whether the grammatical feature has developed into modern Greek, indicating that it may well be a change in the language, or whether it is attested in only certain localities, suggesting it arises from language contact in the region rather than a universal development in the language. He does cautiously, therefore, suggest some possible Egyptianisms, although without further research on Koine caution has to be exercised.38 Even though focus here will be on translation technique, inevitably some conclusions will be drawn from possible linguistic interference. 3. EGYPTIAN TRANSLATIONS The evidence for actual written translations is not as diverse as we might expect. Rajak rightly surmises that translation was such an everyday activity, both ubiquitous and small scale, that we encounter so few explicit references to it. 39 The type of translation material can be divided into four categories, the third and fourth being the ones of concern here. 3.1. LITERARY TRANSLATION Papyrologica (ed. E. Boswinkel, B. van Groningen, and P. Pestman; Papyrologica Lugduno-Batava 17; Leiden: Brill, 1968), 70–76 (70). 34. ἄχι “reed-grass,” θίβις “casket,” and οἰφ , an Egyptian measurement. 35. Cf. Mussies, “Egyptianisms,” 71. 36. Willy Clarysse, “Egyptian Scribes,” 197–200. 37. Trevor Evans, “Complaints of the Natives in a Greek Dress: the Zenon Archive and the Problem of Egyptian Interference,” in Multilingualism in the Graeco-Roman Worlds (ed. Alex Mullen and Patrick James; Cambridge: Cambridge University Press, 2012), 106–23. 38. As Evans, “Complaints,” strongly advocates. 39. Translation, 138. Author, Title 9 There are few cases of translated literary texts, and where they do exist we hardly ever have the original source from which it was translated. Some translations can nevertheless be identified, but in most cases appear to be reworkings or expansions rather than pure translations. Where there are both Demotic and Greek versions, one is not a direct descendant of the other but diverges considerably. Manetho is perhaps our earliest example, although his translation of Egyptian chronicles is very much a rewriting of them. 40 We have some Demotic sources with a Greek equivalent, such as the Dream of Nectanebo (Ptolemaic period), the Oracle of the Potter, The Oracle of the Lamb, the Demotic Chronicle, the Praise of Imouthes (second century C.E.) and the legend of Tefnut.41 These works are mostly preserved in later Roman manuscripts, but might well reflect earlier translations of Demotic works, some from the Ptolemaic period. These literary translations can have interpretive value for Septuagint studies in terms of comparative prophecy, as shown by van der Meer, building upon suggestions by van der Kooij. 42 There are also important sociocultural implications for the role of translation in Egypt and the place of GraecoEgyptian literature, a neglected topic. Given that the Demotic “originals” and the Greek versions differ considerably, however, they cannot be directly compared as regards translation technique. 3.2. PARTIAL TRANSLATIONS There are numerous Demotic-Greek bilingual texts where the text in one language receives a summary in the other, or includes an inserted section, a signature, a registration, archival summary or subscription in the other language. Depauw surveys some 367 bilingual texts from Ptolemaic Egypt in this class. 43 40. Garth Fowden, The Egyptian Hermes: A Historical Approach to the Late Pagan Mind (Cambridge: Cambridge University Press, 1986), 52. 41. See on all these texts Moyer, Egypt, 31. Cf. Stephanie West, “The Greek Version of the Legend of Tefnut,” JEA 55 (1969): 161–83. 42. Arie van der Kooij, “The Old Greek of Isaiah and Other Prophecies Published in Ptolemaic Egypt,” in Die Septuaginta – Texte, Theologien, Einflüsse; 2. internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D), Wuppertal 23.–27.7. 2008 (ed. Wolfgang Kraus and Martin Karrer; WUNT 252; Tübingen: Mohr Siebeck, 2010), 72–84; Michaël N. van der Meer, “Visions from Memphis and Leontopolis. The Phenomnon of the Vision report in the Greek Isaiah in the Light of Contemporary Accounts from Hellenistic Egypt”, in Isaiah in Context. Studies in Honour of Arie van der Kooij on the Occasion of his Sixty-Fifth Birthday (ed. M. N. van der Meer, P. van Keulen, W. van Peursen, B. ter Haar Romeny; VTSup 138; Leiden: Brill, 2010), 281–316. 43. Mark Depauw, “Bilingual Greek-Demotic Documentary Papyri and Hellenization in Ptolemaic Egypt,” in Faces of Hellenism: Studies in the History of the 10 XV Congress of the IOSCS: Munich, 2013 As these are only partial translations and summaries they do not afford suitable comparative material. 3.3. TRANSLATIONS WITH NO EXTANT VORLAGE We have occasional references to the production of translations between Demotic and Greek where the actual documents are no longer extant, but where the formulaic opening states that it has been translated (with the wording ἀντ γραφον συγγραφῆς . . . μεθηρμηνευμ νης Ἑλληνιστ or similar).44 In the archive of the Theban choachytes (responsible for making offerings at tombs) we have frequent reference to such translations, but only two have the Greek and Demotic preserved (P.Choach.Survey 12 and 17), among a total of 18 Greek documents. Such texts are important since we know they are translations and therefore that some of the features could have been generated through the translation process.45 They are in that sense a more reliable source than a Greek document whose author we can only infer was a native Egyptian 3.4. TRANSLATIONS WITH VORLAGE PRESERVED The best evidence is those few cases where we have both the Demotic original and the Greek translation. In the past it has been difficult to study such translations, when both the Demotic and the Greek have not always been published. Where published, they would usually not be found together, the Demotic in a volume on Demotic and the Greek in a volume on Greek. This also corresponded to the distribution of the manuscripts themselves, each being archived in a different University collection.46 There are two good examples from the archive of the Theban choachytes, one dated to ca. 145 B.C.E. (P.Choach.Survey 12)47 and known since the nineteenth century, and one to 136 Eastern Mediterranean (4th Century B.C. –5th Century A.D.) (ed. Peter Van Nuffelen; Studia Hellenistica 48; Leuven: Peeters, 2009), 113–46. 44. Some examples are noted by Anna Passoni Dell’Acqua, “Translating as a Means of Interpreting: the Septuagint and Translation in Ptolemaic Egypt,” in Die Septuaginta Texte, Theologien, Einflüsse. 2. Internationale Fachtagung veranstaltet von Septuaginta Deutsch (LXX.D.) Wuppertal 23. –27. 7. 2008 (ed. W. Kraus, M. Karrer. and M. Meiser; WUNT 252; Mohr Siebeck: Tübingen 2010), 322–39 (323 n. 6) but a fuller listing is given below, n. 54. Note too P.Giss. I 39, Thebes 205–181 B.C.E.; and later, the muchdiscussed P.Oxy. XLVI 3285 (150–200 C.E.) 45. Initial observations on these texts were made by W. Peremans, “Notes sur les traductions de textes non littéraires sous les Lagides,” ChrEg 60 (1985): 248–62. 46. See Vandorpe, “Archives,” 226. 47. The Greek P.Choach.Survey 12B was published as UPZ II 175a (with corrections by Pestman, The Archive, 74-76). The Demotic of 12 is also referred to by its Author, Title 11 (P.Choach.Survey 17).48 The latter has received renewed interest with the republication of the Demotic and Greek versions side by side.49 Choachytes were Egyptian priests (“libation-pourers”) responsible for the care of burials and longer-term performance of the cult of the dead.50 The archive records legal transfers of the rights to offer such provisions, an apparently lucrative enterprise. Pestman’s earlier work on this archive contained many insights and observations on the translations, but it is a survey, that is to say a description of the papyri, rather than an edition.51 From the early Roman period a Demotic sale document with five Greek translations has been found at Soknopaiou Nesos (in the Fayum) and recently received a full publication too (CPRXV, 2 and 3 and 4).52 Each of the Greek translations from Soknopaiou Nesos reflects a slightly different reading of the original. B.C.E. 4. TRANSLATION TECHNIQUE IN EGYPTIAN TRANSLATIONS Where both the Demotic and Greek versions of a text have been preserved codicological as well as translational differences can be observed. The Demotic is understandably written in Egyptian fashion with the Egyptian brush across the individual manuscript reference as P.Berlin 3119, and the Greek by P.Lond. I 3 (F. G. Kenyon, Greek Papyri in the British Museum, vol. 1 [London: The British Museum, 1893], 44–48). 48. The Greek P.Choach.Survey 17B was published as UPZ II 177 (corrections in Pestman, The Archive, 91). The Demotic of 17 is also known as P.Berlin 5507 and P.Berlin 3098, and the Greek as P.Leiden 413. 49. Rachel Mairs and Cary J. Martin, “A Bilingual ‘Sale’ of Liturgies from the Archive of the Theban Choachytes: P. Berlin 5507, P. Berlin 3098 and P. Leiden 413,” Enchoria, Zeitschrift für Demotistik und Koptologie 31 (2008–2009): 22–67. 50. A brief summary of this office is given in Dorothy J. Thompson, Memphis Under the Ptolemies (2d ed.; Princeton: Princeton University Press, 2012), 145–46. 51. Pestman, The Archive. His publication of some of the texts does not contain the relevant ones for our purposes here: P. W. Pestman, Il Processo di Hermias e altri documenti dell’archivio dei choachiti (P. Tor. Choachiti): papiri greci e demotici conservati a Torino e in altre collezioni d’Italia (Catalogo del Museo egizio di Torino. Serie prima, Monumenti e testi 5; Turin: Ministero Per i Beni Cultruali e Ambientali, Soprintendenza al Museo Delle Antichità Egizie, 1992). 52. M. Schentuleit, “Die spätdemotische Hausverkaufsurkunde P. BM. 262: Ein bilingues Dokument aus Soknopaiu Nesos mit griechischen Übersetzungen,” Enchoria, Zeitschrift für Demotistik und Koptologie 27 (2001): 127–54; cf. M. Schentuleit, “Satabus aus Soknopaiou Nesos: Aus dem Leben eines Priesters am Beginn der römischen Kaiserzeit,” ChrEg 82 (2007): 101–25. The original publication of the Greek was P.Lond. (P. BM.) II 262. 12 XV Congress of the IOSCS: Munich, 2013 breadth of a wide scroll, while the Greek is in vertical columns written with a reed pen, as is to be expected by the second century B.C.E. The transfer from one linguistic domain to another is visually present in the change in writing style. The translations are introduced by a note indicating that they are indeed translations, calling them “renderings” or “copies” of the original: 53 [ἀντ γρ αφον συγγραφῆς Αἰγυπτ ας με]θηρμηνευμ ν[ης Ἑλληνισ]τὶ κατὰ τὸ δυ[ν]ατ ν. (P.Choach.Survey 17B = UPZ II 177, Thebes 136 B.C.E.) “a rendering of an Egyptian document that has been translated into Greek as far as possible.”54 That they are “renderings” of the originals means that the translators are left unnamed, silent witnesses to their work, rarely given the credit. 55 This reflects the wider situation in Egypt whereby scribes often identified themselves by name at the end of the document, but copyists did not.56 In this respect, the Septuagint is not unusual in its silence on the names of the translators, even if it fits in with Jewish anonymity in literature. While we would not necessarily expect the translators of the Septuagint to name themselves, it does throw into sharp relief Aristeas’s agenda in choosing to name the translators (Aristeas 4750). The author in seeking to identify them by name elevates the status of the translation from a mere “copy” to a self-standing literary work. The apology κατὰ τὸ δυ[ν]ατ ν “as far as possible” is so standard in these translation documents that it must be seen as a formula deriving from 53. In referencing these versions, A signifies the original Demotic text and B the Greek translation. 54. The almost identical formula is found in P.Tebt. I 164, Kerkeosiris 105 B.C.E.; P.Giss. I 36 II, Pathyris 145-116 B.C.E.; PSI V 549 (C.Ptol.Sklav. I 16), Oxyrhynchos 41 B.C.E.; BGU III 1002, Hermopolis 55 B.C.E.; BGU XVI 2594, Chennis 8 B.C.E.; CPR XV 2 and CPR XV 3, Soknopaiu Nesos 11 C.E.. See too, for translation from Latin, BGU VII 1662 ἔ[θετο] Ῥωμ[α]ι[κῇ διαθ ]κῃ ̣ ̣ ̣ ̣ [μεθ]ερμηνευ[μεν]ης leg. -μ νῃ) ἑλληνιστὶ κατὰ τὸ δυνα[τὸν ] (cf. BGU I 140, 119 C.E.; CPR I 51, 198-211 C.E.). P.Choach.Survey 12B (UPZ II 175 a, 1-2) has been reconstructed as κατὰ δ [ναμιν] (UPZ II 175a 1; Pestman, The Archive, 333), although obviously it is not beyond doubt. The translation of ἀντ γραφον by “copy” or even “transcript” (LSJ 154) does not convey the sense that these are original renderings, and hence I have opted for the gloss “rendering.” My thanks to Patrick James for noting this and suggesting a back-formation such as “rescript” might be appropriate. 55. There appears to be one exception to this naming: Ἡρακλε δης μ[ε]τε [λ]η φα εἰς ἀναγρ αφ ν (SB 1.2051, Thebes 117 B.C.E.). 56. Pestman, The Archive, 329; Mairs and Martin, “Bilingual ‘Sale,’” 57 n. 90. Author, Title 13 literary conceit.57 It probably has little force other than convention, although, since these are legal contracts, it could also be a defence against any charge of misrepresentation. Comparison to the grandson’s apology in his Preface to the translation of Ben Sira (ll. 21–22) can easily be made,58 although we may hesitate drawing parallels between these documentary texts and the literary and rhetorical preface of the grandson. Some of the specific features of the Egyptian translations can easily be summarized, reflecting characteristics both typical of translation interference and induced by contact with the Demotic language. Greek documents written by Egyptians in general tend to show a lack of connective particles, 59 perhaps governed by a similar lack in their Demotic mother tongue but also reflecting the decline in the use of particles in the post-classical period.60 The connector in Demotic is as simple as waw in Hebrew. The translations reflect a high degree of equivalence in that translation equivalents and vocabulary are largely consistent throughout, and the word order is faithfully maintained. This would imply that there were established sets of equivalents or a common practice among bilinguals, perhaps common in oral use before the translations were produced. 4.1. OMISSIONS In the Choachyte archive there are sections where the translator omits or abbreviates phrases, indicating the relative importance of the information for the target audience. Thus, the Greek does not reproduce the full dating formula at the beginning, excluding mention of the ruling Pharoah, and it omits legal rulings at the end, content simply to write καὶ τὰ ἄλλα τὰ κοινά (“plus all the other usual clauses”)¸ as well as the names of the witnesses.61 Frequently we see the expressions ὁμο ως or ὡσα τως “likewise” to indicate additional occupants of tombs. The names of the tombs appear to have been key, and not the date, scribal authority or seller’s obligations. As Mairs and Martin note, P.Choach.Survey 17 does not even repeat the Greek word for tomb, merely recording ἄλλος or ἕτερος “another one.” 57. So P. Fewster, “Bilingualism in Roman Egypt,” in Bilingualism in Ancient Society: Language Contact and the Written Word (ed. J. N. Adams, Mark Janse, and Simon Swain; Oxford: Oxford University Press 2002), 220–46 (232). Cf. Peremans, “Notes,” 252; Pestman, The Archive, 333; Rajak, Translation, 136. 58. Cf. James K. Aitken, “The Literary Attainment of the Translator of Greek Sirach,” in The Texts and Versions of the Book of Ben Sira: Transmission and Interpretation (ed. J.-S. Rey and J. Joosten; JSJSup 150; Leiden: Brill, 2011), 95–126 (101–2). 59. Clarysse, “Egyptian Scribes,” 199. 60. Evans, “Complaints,” 111. 61. Mairs and Martin, “Bilingual ‘Sale,’” 50, 56. 14 XV Congress of the IOSCS: Munich, 2013 4.2. LEXICAL CHOICE The vocabulary of the translation is standard Koine of the time, containing the typical specialized vocabulary of administration that had developed under the Ptolemies.62 Since these documents are lists of names of the deceased in tombs we particularly find titles of occupations, naturally using the standard terms for the time. A section of the tomb is called the παστοφ ριον (P.Choach.Survey 17B.25) “chamber assigned to παστοφ ροι ‘priests’” (LSJ 1346), a term appearing in Koine onwards. The very same term is applied in the Septuagint to the priests’s chamber in the temple at Jerusalem (e.g., Jer 42[35]:4; Ezek 40:17; 1 Chr 28:12) and later by Josephus (Jewish War 4.9.12). Other words attested only in Koine and later are σιδηρουργ ς “iron-worker, smith” (P.Choach.Survey 17B.7), λογε α (“collection,” P.Choach.Survey 12B.7), a term restricted to documentary papyri, καρπε α “profit” (12B.17, 19; 17B.4 [reconstruction], in distinction from the classical terms κ ρδος or καρπ ς), and, in a sense only known in Koine, γεωμ τρης, “land-measurer” (17B.11). κατ γαιος “underground chamber” (17B.24, 28) has a much longer history, but in its specification of a part of a tomb (in the neuter plural) may be new (in contrast to the classical τ φος inter alia, which is used in relation to κατ γαιος, 17B.24).63 It has been well documented how much the Septuagint draws upon the standard Koine of the time.64 It appears that both sets of translations use the same register. 4.3. LEXICAL CONSISTENCY Lexical consistency, rendering the same Demotic word by the same Greek word, is a feature of these translations, as much as it is a feature of the translations from Hebrew into Greek. This can be seen both in consistency across the two Choachyte translations, such as the choice of καρπε α “profit” in both (12B.17, 19, 21; 17B.4, 25, 30), as well as within any one of the translations. Repeated terms such as κατ γαιος “underground chamber” (17B.24, 28), ὑποβρ χιος “drowned” (17B.8, 16, 22, 24, 29), and the expression τὰ προσπ πτοντα “the revenue” (17B.25, 30) display the regular choice by one translator.65 62. See Thompson, “Literacy and Power,” 76–77. 63. The word is also known, if in its classical sense, to the Septuagint translators (see Gen 6:16; PsSol 8:9). 64. E.g., John A. L. Lee, A Lexical Study of the Septuagint Version of the Pentateuch (SBLSCS 14; Chico, Calif.: Scholars Press, 1983). 65. Mairs, “κατὰ τὸ δυνατ ν,” provides tables of all the regular equivalents with their Demotic counterparts. Author, Title 15 Consistency is a marked feature of most of the Septuagint translations, and the use of a consistent religious vocabulary has been observed in such terms as δικαιοσ νη, διαθ κη, νομ ς, ἔλεος, δ ξα, and προσ λυτος.66 Such consistency would suggest there was an agreed method of translation. This could have been enshrined in word lists from which the translators worked, but equally could suggest an oral stage preceding the written during which decisions would have been made regarding the most apposite translation choice. 67 At times consistency comes at the price of elegant renderings, as in the following example from the Choachyte archive: ... [καὶ ἐ]δεξ μ[ην] παρ σου | τὴν το τω[ν τιμὴν ἐκ πλ ρους] ἄνευ παντὸς [nb n pɜ tɜ, 17A, l. 11] [ὑπ]ολ γου. (P.Choach.Survey 17B.32–33) “And I have received from you the price of these in full without any remainder/deduction.” While the use of πᾶς to denote “any” in Greek rather than “all” or “every” is permissible, it is very rare and therefore unexpected here. The translation choice presumably arises from rendering the Demotic nb, which as a definite does mean “all, every,” and therefore the Greek πᾶς serves as a suitable counterpart to it. However, the Demotic nb can also function as an indefinite “any,” frequently reinforced as here by n pɜ tɜ “at all [lit. of the earth].”68 The motivation here seems to be lexical consistency corresponding to the Demotic. Similar oddities are known in the Septuagint where the desire for consistency overrides good sense. 4.4. INCONSISTENCY While lexical consistency is a prominent feature of the translations from Demotic, there is inevitably some inconsistency, although it is not extensive. Thus in P.Choach.Survey 17 the obscure Demotic term ḥsy (“blessed”) is translated by ὑποβρ χιος “drowned” (8, 16, 22, 24, 29) and once by ἁσιῆς (12B.22; cf. ἑσιῆς in the PGM).69 This is not mere randomness on the part of the 66. Anneli Aejmelaeus, “The Septuagint and Oral Translation,” in XIV Congress of the IOSCS, Helsinki, 2010 (ed. Melvin K. H. Peters; SBLSCS 59; Atlanta, Ga.: SBL, 2013), 5–13 (8). 67. So, Aejmelaeus, “The Septuagint and Oral Translation,” 8–10. 68. The Demotic Dictionary of the Oriental Institute of the University of Chicago (CDD) (ed. Janet H. Johnson; Chicago: The Oriental Institute), N: 56 [Online: https://oi.uchicago.edu/research/pubs/catalog/cdd/]. 69. The meaning of the terms is discussed in detail by Mairs and Martin, “Bilingual ‘Sale,’” 60–67. 16 XV Congress of the IOSCS: Munich, 2013 translator, however, since transliteration has its own purpose (see below). Translators seemed to vary in their choice whether to transliterate such Egyptian technical terms or to find a Greek equivalent. In P.Choach.Survey 12 the Demotic šty.w is once rendered by λογε ας (“collections,” 2=770), but more frequently by καρπειῶν (“profits,” 17–18, 19–20, 21). An interesting case of inconsistency is P.Choach.Survey 17 (6=19) where Demotic pɜ bsnṱ “the smith,” elsewhere rendered correctly by σιδηρουργ ς (ll. 7 and 20), is in this case translated by πορθμε ς, “ferryman” (19).71 This particular variation has puzzled the editors, who tentatively suggest that either a smith could also be required to work as a ferryman, which is quite imaginable, or the translator knew this gentleman personally and that he had two occupations. 72 Neither explanation is entirely satisfactory. 4.5. TRANSLITERATIONS The frequency of transliterations in these documents is noteworthy. It seems likely that common priestly titles are translated, as we see in such equivalents προφ της and παστοφ ρος, while more obscure or recherché ones are transliterated. It is possible that some transliterations have already been adopted as loan-words, but even so it is still a choice on the part of the translator to choose an Egyptian-sounding word when a Greek equivalent existed. The most extreme case noted by Schentuleit is the title and name combination of nb w b ḥry šy wɜḏ-wr Nɜ-nfr-ỉr-šty.t transliterated as νεβου πει ῥισῆι γ του Νεφορσ τει.73 We find in the Choachyte archive the Demotic tomb name tɜ ḥ.t n Nbwnn transliterated as Θυναβουνούν (P.Choach.Survey 12.3=8), in this case undeclined, and the titles θυπατεστ μει (12B.20)74 declined in the dative and τση (12B.22; Demotic tɜ s.t “seat”),75 again undeclined. Mention has already been made of the term ḥsy (“blessed”) transliterated in P.Choach.Survey 12 as ἁσιῆς and found in other documents as well (e.g., UPZ II 180a). A half-way position is encountered in the term χοαχ της where the elements of the Demotic, wɜḥ mw 70. In references, the first number given is the line of the Demotic papyrus, and the second is the line from the Greek. 71. Mairs and Martin, “Bilingual ‘Sale,’” 51. 72. Mairs and Martin“Bilingual ‘Sale,’” 51 73. Ibid.; 2007, 108. CPR XV 1 (3 B.C.E.) a translation of P.BM 262. 74. The reading corrected by Pestman, The Archive, 76. 75. This is the reading corrected by Pestman, The Archive, 76. Author, Title 17 (lit. “pourer of water”), are rendered into Greek.76 The Roman Soknopaiou Nesos contracts are more extreme in the extent of their transliterations. For Rabin transliterations were a classic case of Flashar’s Verlegenheitsuebersetzung (translations of embarrassment), lending to the Septuagint “the appearance of a rough draft, with words pencilled in for later reconsideration.” 77 It seems more likely that such variation is not down to an ignorant translator or one not capable of finding suitable equivalents, especially in places where they are easily at hand, but functions to serve the translator’s literary stratagem. Transliteration of titles is acceptable for readers and speakers familiar with Egypt, namely the Greek-Egyptian peoples who would consult the legal documents.78 It thus serves as a case of code-switching that maintains Egyptian identity and expresses the Egyptian nature of the legal issues through the medium of the Greek language.79 In the Septuagint Pentateuch transliterations are primarily for institutions or for realia that have no obvious equivalent in Greek,80 in a similar fashion to the transliterations of the Egyptian documents. We thus find such renderings as: σαβεκ (Gen 22:13), γομορ (Exod 16:16), μαννα (Num 11:16), and in a later book οἱ ναθινιμ (2Esdr 2:70). Each either has a religious significance or denotes a particular object. Proper nouns are a separate category that requires further investigation. It is noteworthy, nonetheless, how many names are transliterated in the Pentateuch and not adapted to Greek grammar. In one verse (Gen 41:45), nevertheless, we find two names written as Egyptian, one not adapted to Greek grammar (Ψονθομφανηχ) and one that is (Πετεφρῆς “Potiphar”; cf. Gen 37:36; 39:1; 41:50).81 For the early Ptolemaic period in Egypt, the practice of writing names varied, often being translated, but equally in the reigns of Ptolemy I and Ptolemy II transcribed without adaptation to Greek (especially in Upper 76. Such neologisation puzzled early editors. Young originally read the Greek as χολχ της, while Kenyon remained sceptical over the reading χοαχ της (Greek Papyri, 44–45). 77. Rabin, “The Character,” 24. 78. Both Schentuleit, “Hausverkaufsurkunde,” 135–37, and Mairs, “κατὰ τὸ δυνατ ν,” propose code-switching as a determinative force. 79. Cf. Sang-il Lee, Jesus and Gospel Traditions in Bilingual Context: A Study in the Interdirectionality of Language (BZNW 186; Berlin: de Gruyter, 2012), for a similar argument for the use of Aramaic forms in the Gospels. 80. See comments of Joosten, Collected Studies, 234. Cf. Emanuel Tov, “Loanwords, Homophony and Transliterations in the Septuagint,” Bib 60 (1979): 216–36. 81. Cf. Joosten, Collected Studies, 168 n. 36; Harl, La Genèse, 276. It was common in papyri for Egyptian names ending in consonants, when declined, to follow the first declension pattern (ending in -ις or –ης; Pestman, The Archive, 485), as Πετεφρῆς except that this name appears to be of the so-called “mixed declension.” 18 XV Congress of the IOSCS: Munich, 2013 Egypt).82 It seems that the Septuagint translators followed this early Egyptian practice and remained more conservative than their Egyptian counterparts. 4.6. WORD LISTS It is natural to find in such legal documents as these Demotic-Greek translations lists of words, denoting both parties involved and the property being assigned to them. Below is a brief example, which continues for another two lines beyond this citation: [σο εἰσιν] οἵ τε [τ ]φοι καὶ <τὰ> κατ γαια <αὐτῶν> καὶ οἱ ἐν α[ὐτοῖς κα]ὶ | <οἱ> ὑποβρ χιο[ι καὶ <τὰ> παστοφ ρι]α κα[ὶ οἱ] ἐν το τοις [κα]ὶ οἱ προγεγραμμ[ νοι] καὶ | 30<οἱ> προσ κο[ντες αὐτοῖς καὶ αἱ καρ]πεῖαι αὐτῶν κ[α]ὶ τὰ προσπ π[τον]τα . . . (17B.28–30) [Belonging to you] are both the [to]mbs and <their> underground chambers, and those in t[hem an]d <the> drowne[d and <the> priestly cham]bers an[d those] in them, [an]d the aforementio[ned] and <those> that com[e to them and their pr]ofits a[n]d reve[nu]es . . . In Greek it was not necessary in such lists to use a copula (κα ) all the way through, while here the Greek uses it throughout even though it is translating a Demotic document in which there is no copula at all. A remarkable comparative example is found in Greek Hosea, where the dating formula in the opening lists the name of kings, each connected by κα . Λόγος κυρίου, ὃς ἐγενήθη πρὸς Ωσηε τὸν τοῦ Βεηρι ἐν ἡμέραις Οζιου καὶ Ιωαθαμ καὶ Αχαζ καὶ Εζεκιου βασιλέων Ιουδα καὶ ἐν ἡμέραις Ιεροβοαμ υἱοῦ Ιωας βασιλέως Ισραηλ. (Hos 1:1) A word of the Lord that came to Hosee the son of Beeri in the days of Kings Ozias and Ioatham and Achaz and Hezekias of Ioudas and in the days of King Ieroboam son of Ioas of Israel. (NETS) What makes this example from the Septuagint all the more striking is that the Hebrew (‫ )עזיה יותם אחז יחזקיה‬also practises asyndeton here without a waw.83 82. This is documented by Brian Muhs, “Language Contact and Personal Names in Early Ptolemaic Egypt,” in The Language of the Papyri (ed. T. V. Evans and D. D. Obbink; Oxford: Oxford University Press, 2010), 187–97. 83. I am grateful to Jan Joosten for this point and for drawing my attention to Hosea. Author, Title 19 4.7. INFLECTION IN NOUN PHRASES Such lists in both papyri and the Septuagint can also lead to breach of concord, which has already been a recognized feature of Greek papyri.84 One notable aspect, the placement of a nominative in apposition to a noun in a different case (especially in lists), has been documented in the Septuagint too. Thackeray described it as “drifting into the nominative,” 85 which is a gentle if imprecise designation for a phenomenon that requires some explanation. Its appearance in 2 Esdras has been subjected to investigation by Wooden, who then compares it to other places in the Septuagint and New Testament, with passing mention of the papyri.86 Comparison is easy to draw. It is striking, though, that it is not a mere feature of papyri, as implied by Moulton, but specifically of those papyri known to be translated by Egyptian bilinguals. 87 Thus, we find the phenomenon a number of times in the archive of Hermias, a notary in the town of Pathyris in the Ptolemaic period.88 In a noun phrase, especially in a list of peoples, the phrase initial element is inflected and the remainder default to the nominative. We see this in the names of the agreeing parties in contracts drawn up by Hermias: καθʼ ἣν ὁμολογεῖ | Νεχθανοῦπις Πατσεοῦτος Π ρσης . . . συνκεχω|ρηκ ναι Πετεαρσεμθεῖ Πανοβχο νιος καὶ τοῖς | ἀδελφοῖς Πετεσουχος καὶ Φαγωνις καὶ Ψεννησις (P.Grenf. 2 25.4 –7, 103 B.C.E.)89 an agreement by which Nechthanoupis son of Patseous the Persian agrees . . . to have granted to Peteharsemthes son of Panobchounis and to the brothers Petesouchos and Phagonis and Psennesis In the following list only the name of the first member, Peteharsemtheus, is inflected correctly in the dative. The rest of the names follow in the nominative. 84. E.g., Moulton, “Grammatical Notes,” 151. 85. H. StJ. Thackeray, A Grammar of the Old Testament in Greek According to the Septuagint (Cambridge: Cambridge University Press, 1909), 23. 86. R. Glenn Wooden, “Interlinearity in 2 Esdras: A Test Case,” in Septuagint Research: Issues and Challenges in the Study of the Greek Jewish Scriptures (ed. Wolfgang Kraus and Glenn R. Wooden; SBLSCS 53; Atlanta: SBL, 2006), 119–44 (133–43). 87. See Mussies, “Egyptianisms,” 72; Martti Leiwo, “Scribes and Language Variation,” in Grapta Poikila, vol. 1 (ed. Leena Pietilä-Castrén, Marjaana Vesterinen; Papers and Monographs of the Finnish Institute at Athens 8; Helsinki: Finnish Institute at Athens, 2003), 1–11 (5–6). 88. Gathered in Vierros, Bilingual Notaries. It is the same archive in which Grenfell noted grammatical blunders and from which some of Moulton’s examples are drawn. 89. See Vierros, Bilingual Notaries, 141. 20 XV Congress of the IOSCS: Munich, 2013 ἐδ νεισεν Πετεαρσεμθεὺς Νεχο του | Π ρσης Πετεαρσεμθεῖ καὶ Πετεσουχος (P.Grenf. 2 27.3-4, 103 B.C.E.) Peteharsemtheus son of Nechouthes a Persina lent to Peteharsemtheus and to Petesouchos. Vierros in describing this phenomenon says that the nouns appear as if they were placed in parentheses.90 Comparable examples are found in the Septuagint: ἐπὶ τοὺς υἱοὺς τῶν Ἀσσυρίων ἐπέθετο, ἡγουμένους καὶ στρατηγοὺς τοὺς ἐγγὺς αὐτῆς ἐνδεδυκότας εὐπάρυφα, ἱππεῖς ἱππαζομένους ἐφʼ ἵππων· νεανίσκοι ἐπίλεκτοι πάντες. (Ezek 23:12)91 She applied herself to the sons of the Assyrians, governors and commanders near her, wearing fine purple, horsemen riding upon horses. They were all elite young men. (NETS) Οὐκ ἐχωρίσθη ὁ λαὸς Ισραηλ καὶ οἱ ἱερεῖς καὶ οἱ Λευῖται ἀπὸ λαῶν τῶν γαιῶν ἐν μακρύμμασιν αὐτῶν, τῷ Χανανι, ὁ Εθι, ὁ Φερεζι, ὁ Ιεβουσι, ὁ Αμμωνι, ὁ Μωαβι, ὁ Μοσερι καὶ ὁ Αμορι. (2 Esdr 9:1) The people of Israel and the priests and the Leuites were not separated from the peoples of the lands with their things put far away, in reference to the Chanani—the Heththi, the Pherezi, the Iebousi, the Ammoni, the Moab, the Mosri and the Amori. (NETS)92 In both these examples, the dative is followed by nominatives. That the translator could write such lists according to classical grammar is shown by 2 Esdr 19:18 where there is a similar list, but all the names are Graecized gentilics and in the correct case.93 This does not stop the translator again later resorting to a list with nouns in the nominative: καὶ ἐν Ιερουσαλημ ἐκάθισαν ἀπὸ υἱῶν Ιουδα καὶ ἀπὸ υἱῶν Βενιαμιν. – ἀπὸ υἱῶν Ιουδα· Αθαια υἱὸς Αζαια υἱὸς Ζαχαρια υἱὸς Αμαρια υἱὸς Σαφατια υἱὸς Μαλεληλ καὶ ἀπὸ υἱῶν Φαρες. (2 Esdr 21:4) and in Ierousalem lived some of the sons of Iouda and some of the sons of Beniamin. Of Iouda’s sons: Athaia son of Ozia son—he being a son of Zacharia—he being a son of Amaria—he being a son of Saphatia—he being a son of Maleleel, and some sons of Phares. (NETS) 90. Vierros, Bilingual Notaries, 141 91. Noted by Wooden, “Interlinearity,” 140–41. See too Ezek 23:7; Zeph 1:12. 92. Given the incongruity of the syntax, NETS presents the nominative as in apposition to the dative through the use of the dash. 93. Wooden, “Interlinearity,” 134. Author, Title 21 Once more, the translator is elsewhere able to render genealogies correctly (7:15; 13:4).94 Wooden, who also described them as parenthetical interjections,95 accounts for this phenomenon as interference from the source text in which morphemes are rendered at a visual but not a grammatical level. 96 The presence in 2 Esdr 9:1 of the Hebrew preposition ‫ ל‬on the first element but not the subsequent ones could in part explain why only the first is in the dative. Similarly, in Demotic the oblique cases were preceded by n, but this was not placed before those in apposition—therefore also an explicable factor in not using the dative.97 It seems unlikely that we should view it as economy in translation,98 since there is little saved in rendering the nominative rather than the dative. Rather than being a direct result of the translation task, such incongruence probably arises from features within the spoken languages of the translators. Demotic as Hebrew had no case marking system (only markers for feminine and plural) in contrast to the five cases of the Greek system. As a result those trained to write Greek were probably drilled in the cases, such that the nominative as the subject case may have been indoctrinated deeply in second language education.99 The nominative would then have been favoured to denote semantic subject or agent even in structures where Greek would use a different case, and only the initial part of the phrase, which was functionally more important, was deemed vital for inflection. Evidence from the later ostraca of the Narmouthis archive confirms how Egyptian scribes handled case endings.100 As well as providing further examples of this appositional use of the nominative, they show how nouns were abbreviated and thereby the need obviated for marking the case ending. The construction thus arises from language contact and is hence inevitable in translation, and even if it is not a specific feature of translation,101 it is encouraged by the use of prepositions in the source languages. As confirmation that this is a specific language-contact induced change, Mussies observes first that this use of the noun is not represented in modern Greek, and therefore cannot be shown to be an internal development. The 94. Wooden, “Interlinearity,” 140. A further example of grammatical incongruence n 2 Esdras is 13:24–26. 95. Wooden, “Interlinearity,” 138. 96. Wooden, “Interlinearity,” 134. 97. Mussies, “Egyptianisms,” 72. 98. So Wooden, “Interlinearity,” 135. Mayser, Grammatik, 2.3:192, likewise calls it “Breviloquenz des Tabellenstils.” 99. Vierros, Bilingual Notaries, 139. 100. Fewster, “Bilingualism,” 238-40. 101. As much is implied by the examples attested in the New Testament, discussed by Wooden, “Interlinearity,” 141–42. 22 XV Congress of the IOSCS: Munich, 2013 situation is different with participles (cf. LXX Am 2:6-7) since an indeclinable absolute form of the participle operates in modern Greek.102 Second, Mussies finds it is only restricted to texts from Egypt and Syria (e.g., OGIS 611, Hauran), whose vernacular contained no case categories. To this we may add Hebrew speakers, whose native language equally lacked case categories. 4.7. POLYSEMOUS PREPOSITIONS Stereotyped translation of prepositions can lead to odd uses in Greek, where interference can be prominent. Husson has shown how Greek ὑπ as a calque of Egyptian ḫr adopts a range of senses from the Demotic.103 Most notable is the rendering ὄνος ὑπὸ οἴνου, denoting not the surprising “donkey under wine” as might be expected in Greek, but “in charge of” the wine, following Demotic ḫr. In P. Choach.Survey 17B there are two odd uses of prepositions, although interference from Demotic is not conclusive. Of course, prepositions are notoriously difficult to handle, owing to the fact that the case system is changing in Greek, affecting in turn the use of prepositions. Nevertheless, in similar manner to the stereotyped equivalents for prepositions sometimes seen in the Septuagint, we might have the same phenomenon in the Egyptian translations. ἥμ[ισ]υ τῶν [παρὰ] | Ἀμενώθου ο[ἰ]κοδ μου A half of those with Amenothes, builder... Demotic: nɜ rmt.w n Imn-ḥtp (P. Choach.Survey 17.5 = 13-14) This use of παρ with genitive is frequent in the document, although normally one would expect the meaning “from” with the genitive, whereas here it appears to mean “with.”104 The second century is too early to see a definitive decline in the use of the dative in favour of the genitive, but very occasionally one can find the genitive use of παρ for the dative meaning “by, beside” (LSJ sv A.III). In the Demotic the simple preposition n is polysemous covering the meanings “in, through, with, by means of.” 105 The meaning “with” is the obvious sense in the 102. Mussies, “Egyptianisms,” 72. In this he argues against Ludwig Radermacher, Neutestamentliche Grammatik: das griechisch des Neuen Testaments im Zusammenhang mit der Volkssprache (Tübingen: Mohr Siebeck, 1911), 86–87; James H. Moulton, A Grammar of New Testament Greek (2 vols.; Edinburgh: T&T Clark, 1929), 1: 12, who see it as a development of the popular language. 103. G. Husson, “ὙΠΟ dans le grec d’Égypte et la préposition égyptienne ḫr,” ZPE 46 (1982): 227–30. 104. Cf. Mairs and Martin, “Bilingual ‘Sale,’” 51. 105. CDD N:3. Author, Title 23 Demotic, although the Greek translation appears to have opted for “through” or “by means of.” In P. Choach.Survey 17 (8=29) we find the translation of the preposition as follows: ... κα[ὶ οἱ] ἐν το τοις “and those in them” Demotic: ỉrm nɜ nty ḥtp.w ỉrm˭w “and those who rest along with them” This is a slightly odd case. Demotic ỉrm “with” here indicates that the phrase refers to those who rest with the dead. If we translate the Greek as ἐν “in,” then the reference must be the tombs, in which they rest. 106 The Demotic preposition ỉrm is unambiguous unless preceded by n, and therefore there is no obvious solution to this one. While these examples remain ambiguous, Husson has shown that such interference is possible, and we find comparable examples in the Septuagint. The stereotyped rendering of Hebrew bēth, for example, gives rise to the expressions ἐν δυνάμει βαρείᾳ (“with a heavy force”; 4 Kgdms 18:17; contrast Isa 36:2 μετὰ δυνάμεως πολλῆς) and ὤφθη άούδας . . . ἐν τρισχιλίοις ἀνδράσιν (“Judas was seen . . . accompanied by 3000 men”; 1 Macc 4:6).107 4.8. SUBTLE RENDERING At times translators were sensitive to the nuance of a word and did not opt for a mere default rendering. P.Choach.Survey 17 (7=23) translates the Demotic pɜ swnw “doctor” by the Greek ταριχευτ ς “embalmer.” Other passages where the Demotic term for doctor appears also indicate that the role could include that of “embalmer,” suggesting that this translation may well be suitable.108 In an Egyptian context the priests included a class of doctors who no doubt would have also been embalmers thanks to their medical knowledge. It is possible that even in Greece one duty of physicians was also to embalm, since some anatomical knowledge would have been needed for embalming, 109 although how much is not clear. In Alexandria there developed, however, a sophisticated knowledge of human anatomy, and these Alexandrian doctors might well have used their knowledge for embalming too. The translator has therefore chosen an appropriate equivalent for the context of a Choachyte archive. 106. Mairs and Martin, “Bilingual ‘Sale,’” 52. 107. Examples from Thackeray, Grammar, § 91. 108. Pestman, The Archive, 30 n. b. 109. See B. Brier and R. S. Wade, “Surgical Procedures during Ancient Egyptian Mummification,” ZÄS 126 (1999): 89–97. Cf. S. R. Driver. The Book of Genesis (5th ed.; London: Methuen, 1906), 395. 24 XV Congress of the IOSCS: Munich, 2013 This translation equivalent will also strike a note for those familiar with Septuagint Genesis. In Gen 50:2 there are two occurrences of ἐνταφιαστ ς denoting someone trained for preparing a body for burial or embalming, a near synonym of ταριχευτ ς.110 It is clear that the individuals are responsible in the passage for embalming, since they are assigned the task of handling Jacob’s dead body, although in the passage it is a translation of the Hebrew word for “physician” (participle ‫)רפא‬. Even though they were actually embalming the body, they might also have been physicians such that the LXX translators were probably aiming to introduce a word more suitable to the Egyptian context.111 4.9. LITERARY EMBELLISHMENT It is clear that the translators of the Demotic were educated and occasionally showed some attempts at literary embellishment rather than slavishly following the words or structure of their source text. As illustration, we may note a few examples of the use of particles where there is no equivalent in Demotic: [σο εἰσιν] οἵ τε [τ ]φοι καὶ <τὰ> κατ γαια <αὐτῶν> (P.Choach.Survey 17B.28) οὔτʼ ἐγὼ [ο]ὔτʼ ἄλλος (P.Choach.Survey 17B.34) ἐὰν δ τις (no connective in Demotic) (P.Choach.Survey 12B.26; 17B.35) μ ν . . . δ (P.BM 262, l. 3, 6) The sporadic use of particles in the Septuagint can be compared to these examples.112 ὅ τε Αδαμ καὶ ἡ γυνὴ αὐτοῦ (Gen 2:25) οὐκ ἐκκλινοῦμεν οὔτε εἰς ἀγρὸν οὔτε εἰς ἀμπελῶνα (Num 21:22) ἐὰν δ . . . (Gen 44:22; equivalent of Heb waw) Ἡ μὲν φωνὴ φωνὴ Ιακωβ, αἱ δὲ χεῖρες χεῖρες Ησαυ (Gen 22:22)113 110. For full discussion of the Septuagint term, see James K. Aitken, “Context of Situation in Biblical Lexica,” in Foundations for Syriac Lexicography III (ed. J. Dyk and W. van Peursen; Piscataway, NJ: Gorgias Press, 2009), 181–201 (193–95). On the Egyptian and Greek terms, see Thompson, Memphis, 145–46. 111. Adolf Deissmann, Bible Studies: Contributions Chiefly from Papyri and Inscriptions to the History of the Language, Literature and the Religion of Hellenistic Judaism and Primitive Christianity (Edinburgh: T&T Clark, 1901), 120–21; Marguerite Harl, La Bible d’Alexandrie, 1: La Genèse (Paris: Cerf, 1986), 315. 112. For discussion of the use of particles and their significance in the Greek Pentateuch, see James K. Aitken, “The Characterisation of Speech in the Septuagint Pentateuch,” in The Reception of the Hebrew Bible in the Septuagint and the New Testament: Essays in Memory of Aileen Guilding (ed. David J. A. Clines and J. Cheryl Exum; Sheffield: Sheffield Phoenix, 2013), 9–31. 113. The particle μ ν only appears seven times in Genesis, the majority in a standard use with conditionals. Author, Title 25 Despite the decline in the use of particles in post-classical Greek, they would have been taught in the educational system and used by scribes as a marker of their education. 5. CONCLUSIONS The features identified in the Egyptian translations can all be paralleled in the Septuagint. To some extent they are universal characteristics of translations, but the similarities are more than that. They reflect a method of close adherence to the source text in word order, lexical consistency, phrasing, and parataxis. At the same time the translations display a degree of freedom, with occasional variation, alternation between translation and transliteration, literary embellishment, and the rare interpretative rendering. This balance between consistency and formal equivalence on the one hand, and a degree of freedom on the other is a marker of the Septuagint as much as the Egyptian translations. The Egyptian translations have been described as functional rather than an exercise in elegant composition,114 and the same can be said of the Septuagint, although occasional elegance can be found in both. The Greek can be described as the Greek of the speakers of the time, especially when it has become a “world” language. Features are not to be dismissed as poor Greek, but rcognized as the standard Greek of speakers and writers from the region, both Jewish and Egyptian. In this brief survey we have had the opportunity to consider translations from the second century B.C.E. and later from the early Roman period. Despite the time difference of two centuries there seems little difference in the practices of the translators from each period, and there does not seem to be the sort of development towards literalism that Brock surmised. It is an overstatement to think of the translation of the Septuagint as a unique or unprecedented exercise, since the method of translation is the same as that used for documentary translation. There is no reason to doubt that the methods seen in the Choachyte archive in the second century were practised a century earlier, when translation was already a necessary daily activity in Egypt. It is important to note the Egyptian translators, even while rendering documenttary texts, are not the mythical dragomans interpreting without sense or concern for the content. Their method appears to be the same as the Septuagint translators, even if the natures of the respective sources present their own particular responses. Bickerman and Rabin looked for the Septuagint translators among the ordinary translators of Egypt, and in this they were right, falling short 114. Mairs and Martin, “Bilingual ‘Sale,’” 57. 26 XV Congress of the IOSCS: Munich, 2013 only in their reconstruction of an oral dragoman. In opposition to the dragoman theory van der Kooij has sought to locate the translators among learned scribes, familiar with the text and content they were translating and therefore able to introduce interpretation.115 His theory is built upon the ad hoc nature of the dragoman, translating from the first a new text. There is perhaps too great a binary opposition drawn between the putative dragoman and the generalized scribe-translator, when the two might be found in the one person. Within the one translation activity there can be two levels of operation: on the one hand the translation skill and method of rendering into Greek, and on the other the understanding of the source text. These are two separate issues. To compare the Septuagint to the Egyptian translations is to understand the translation technique and, on the level of grammar, morphology, and lexical choice, the translation method applied. It is conceivable that working in a Greek administrative environment the translators learnt their task alongside their Egyptian counterparts. Even within such an environment translation was not merely ad hoc, since, as we have seen, the Egyptian translators were familiar with regular translation equivalents and context-sensitive renderings. That the Septuagint translators were also familiar with their source texts and knew reading traditions is not contrary to this—their knowledge of Hebrew reading traditions and interpretation is a separate issue from their style of Greek writing. They learnt their Greek skills among the administrative class, but as much as for any other translator of the time, that need not have been their only source of income. One translator named Apollonios, for example, seems to have traded in a range of goods (P.Cair.Zen. I 59065), while another, one Limnaios, was also a simple goatherd (P.Cair.Zen. III 59394).116 Our Septuagint translators could well have been school teachers, prayer-house officials, or even scholars, but their methods and probably some of their income came from everyday translation. The dragoman is an Ottoman invention, while the scribe-translator is more appropriate for the Temple setting; perhaps we should simply call them by the term that does exist in Egypt: hermēneis “translators.” 115. This has been argued in a series of publications. See, e.g., Arie van der Kooij, The Oracle of Tyre: The Septuagint of Isaiah XXIII as Version and Vision (VTSup 71; Leiden: Brill, 1998), 112–23. 116. On these see Evans, “Court Function,” 249–50.