A Genius for Mentorship: A Forum in Honor of Ben Wright on his 65th Birthday
~ presented by his students ~
The Translation of the Torah in Alexandria and the Relevance of the Rosetta Stone
Around 150 BCE, an Egyptian Jew composed a fictional letter supposedly a century old between two Gentile courtiers. The setting was the Alexandria of Ptolemy II Philadelphus, king of Egypt and son of one of Alexander the Great’s generals. The letter purported to describe how the Torah came to be translated into Greek. Gold-lettered Hebrew scrolls had been brought from Jerusalem by 72 Jewish scholars well-versed in Hebrew and Greek. Under the generous patronage of the king, the translation proceeded briskly in a beach house by the Lighthouse of Alexandria. The finished product was read aloud before the king and the Jewish community there. The community accepted it as authoritative, while the king placed a copy in the Library of Alexandria.
This letter is conventionally called the Letter of Aristeas after the courtier who is supposed to have written it. The translation is called the Septuagint, after the 72 scholars who produced it. One of the points “Aristeas” is making is that the Septuagint fully replaces the original Hebrew text and stands independently. There is no doubt that, as an historical matter, for many Mediterranean Jews, and later for Christians, the Septuagint was consulted as authoritative in its own right. But was this the original purpose of the translation, 100 years before Aristeas?
Albert Pietersma and Ben Wright have challenged the historicity of the account of Aristeas, and rightly so. They have produced a contrary model called the “interlinear paradigm.” An interlinear text puts the same text side-by-side with a translation of it. It is important to note that Pietersma and Wright treat the interlinear paradigm as a metaphor, without requiring the historical existence of such an artifact. Rather, they observe that while the translators clearly knew Greek grammar very well, they were almost totally unconcerned with good Greek style. As a result, the Septuagint reads as a very un-Greek book. This is because, in the terms of translation scholar Gideon Toury, the Septuagint is a “formal” translation. This means that the Greek represents the Hebrew one word at a time, in frequent violation of the ordinary word order in Greek. In contrast, a “dynamic” translation would match the Hebrew only on a sentence-by-sentence level, allowing it to flow with the canons of Greek style.
This formal translation technique produces a text that is sometimes difficult or impossible to understand without referring to the Hebrew. Based on this fact, Pietersma and Wright argue that the Septuagint was in fact originally intended only to be used to gain access to the Hebrew. The most likely context would have been educational, a tool for the use of the Jewish community and emphatically not for the purpose of inclusion in the famous Library.
Tessa Rajak, more provocatively, has suggested the purpose is political and resistant. Accepting the thesis of the Letter of Aristeas that the translation project was the result of a direct commission from King Ptolemy II Philadelphus rather than a Jewish initiative, she argues that the translators deliberately produced a “bad,” un-Greek text that enabled Greek-speaking Jews to resist Greek cultural mores. Rajak’s thesis is overly dependent on the purported historical reliability of Aristeas, which is undermined by the letter’s clearly apologetic features, such as the king’s lavish praise of the wisdom of the Jewish translators. Most importantly, there is no evidence anywhere else that a translation of anything was included in the Library. Nonetheless, Rajak highlights the fact that the translation of the Septuagint would have been an ethnic event: that is, an event with implications for the ethnic identity of Judeans in Egypt and their relations with both Greeks and Egyptians.
I have been considering another possibility. It may be that formal, word-for-word translation was a colorless choice in Ptolemaic Egypt and its sphere of influence, because it was simply how almost everyone translated. I believe the famous Rosetta Stone, which contains the same passage in hieroglyphic Egyptian, demotic Egyptian, and Greek, offers an opportunity to expand the repertoire of ancient translation technique. The Rosetta Stone, also called the Memphis decree, is one of several multilingual decrees produced by synods of Egyptian priests. Taken together, they show that the formal, word-for-word equivalence known from the Septuagint was also the accepted method for the translation of decrees from Greek to demotic Egyptian in the early dynasty.
The earliest decree extant is known as the Canopus decree (238 BCE). It exhibits a significant degree of formal equivalence, but it is crucial to determine which of the three texts was composed first, and which were translated from it. According to demotic scholars R.S. Simpson and Stefan Pfeiffer, it appears to have been composed in Greek, and then translated into demotic and hieroglyphic Egyptian. Because the hieroglyphic script was literally understood as the speech of the gods, it exhibits less interference from the Greek. In Toury’s terms, the hieroglyphic translation is more dynamic. I believe this is a specific choice made by the priestly translators, precisely on account of the sanctity of the script.
The demotic, “secular” script has no such protection. Therefore I find it significant that the translators, who demonstrate their ability to use dynamic style in the hieroglyphic text, opt for a formal technique in the demotic. First, the Greek is constructed in long, multi-clause sentences that are consonant with good Greek style. Egyptian dialects prefer shorter sentences, yet the demotic text clumsily uses one relative conjunction after another to mimic the longer Greek sentences. Second, the Greek text has been rendered formally, breaking the Greek up into words and very short phrases and rendering these individually. Unlike Greek, with its case endings which control whether a word is subject or object, demotic Egyptian relies on sentence syntax to establish the relationships between words. This means that a fully isomorphic translation here would be utter nonsense.
Nonetheless, even when the rules of demotic grammar require a departure from the Greek word order, the translation has been undertaken, if not quite word-by-word, then certainly phrase-by-phrase. As a common example, the Greek style of the decrees prefers to put the possessive before the thing possessed (e.g., “the king’s land”), while in demotic the possessive must follow the thing possessed (“the land of the king”). This difference in word order accommodates Egyptian grammar while achieving maximum formal correspondence. In this context it is important to remember that while the Septuagint sometimes shows excessive formal correspondence, it rarely outright violates the rules of Greek grammar.
Nonetheless, there are occasional passages in the demotic that either omit (“minuses”) or expand upon (“pluses”) the base Greek text, usually having to do with matters of priestly rank and benefits. This is especially striking in a text that presents both texts one on top of the other and could be checked by anyone literate in both languages. Evidence from papyri indicates that this was a class somewhat larger than has generally been supposed, as I have argued elsewhere. This also demonstrates that formal translation per se was not a strategy of resistance for these Egyptian priests. When they wanted to make a different point in the demotic, they simply wrote a different text.
I believe the Canopus decree, which was composed not many years after the beginning of the translation of the Septuagint, provides evidence that isomorphism was the standard translational mode in the third century BCE in Egypt. This means that “bringing the reader to the original,” in Pietersma and Wright’s phrase, was understood as the basic purpose of any translation, absent countervailing pressures such as the sanctity of the hieroglyphic script. Such a purpose could conceivably be pedagogical, but the manifestly non-educational purpose of the Canopus decree urges caution in making such a supposition. It need not be surprising that Jewish and Egyptian translators shared assumptions about how to translate into and out of Greek. The Egyptian priests were themselves highly assimilated to Greek culture, and there is no reason to suppose they were not also the authors of the Greek text. This means that both sets of translators shared a cultural milieu and an operating methodology.
Wright and Pietersma have argued for a succession of periods of Septuagintal use: an initial period when it was used in a manner dependent upon the Hebrew, and a later period, evidenced by Aristeas and, later, the Jewish philosopher Philo, when it had become an independent text. Similarly, later decrees such as that of Memphis exhibit less formal correspondence and greater variation between texts. The independence of a translation from its original seems to have been a cultural trend in the second century BCE, both among Judeans and Egyptians.
Once again, a focus on the Septuagint as an ethnic event provides an interesting example. The category of the ger in the Hebrew Bible often occupies a twilight territory between being an Israelite insider and an outsider. A ger can be a non-Israelite who nonetheless wishes to participate in the Passover ritual, and who takes on certain ritual obligations—chief among them, for males, circumcision (Exod. 12:48), but also dietary restrictions (Exo. 12:19)—to be able to enter this liminal state. On the other hand, the word ger can refer to Israelites living outside of Israelite territory (e.g., Deut. 23:7). In either case, the meaning is “sojourner,” and the precise specification is provided by the context. The patterns of translation used in the Septuagint for ger are illuminating for how the translators viewed this critical boundary category, and for the relationship of the Greek to the Hebrew.
Most famously, ger is often translated proselytos, always with reference to the first meaning above, of a non-Israelite. However, the translators make an internal differentiation between a ger who is not an Israelite but is with Israel, and the reverse case, where a ger is an Israelite but is living somewhere else. In the latter case, ger is invariably translated with paroikos. This very marked differentiation of something the Hebrew treats as one phenomenon, that of the “stranger” or “sojourner,” indicates a strong ethnic differentiation between someone who adheres to a certain number of Jewish ethnic markers but who is still not considered fully Jewish, and someone whose genealogy is Judean. It indicates that to the Septuagint translators, there was a difference between observing all the practices of Jewish culture and religion, and being a Jew, and they wanted to communicate this with their translation.
The differentiation of proselytos and paroikos means that the Septuagint does not simply have a subservient relationship to the Hebrew original, as Pietersma and Wright have argued. In this instance, the Septuagint actually controls the interpretation of the Hebrew, since if one referred back to the original one would find ger in both cases; the Septuagint’s interpretation would divide these into two different kinds of “sojourners.” Even if the translators’ intention is interlinear—and the strong formal correspondence of all the ger-passages supports this—the relationship between the Greek and the Hebrew was from the beginning dialectical, not unidirectional.
The interlinear paradigm is a fruitful and provocative model which has fueled my own and others’ reflections, yet the purpose and impact of formal translation as a literary phenomenon, and of the Septuagint as an ethnic phenomenon, require further work. The questions I have raised about it are only a consequence of our being embedded in a larger conversation. I want to take this opportunity to thank Ben Wright, both for his scholarly contributions which have inspired my own developing thoughts, and for his personal generosity in discussing both with me. It is an honor and a pleasure to dedicate this essay to him.
Stewart Moore, Jewish Ethnic Identity and Relations in Hellenistic Egypt: With Walls of Iron? (Leiden: Brill, 2014).
Stefan Pfeiffer, Das Dekret von Kanopos (238 v. Chr.): Kommentar und historische Auswertung (München: K.G. Saur, 2004)
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-6 Conference; Proceedings of the Association Internationale Bible et Informatique “From Alpha to Byte”; University of Stellenbosch 17-21 July, 2000 (Johann Cook, ed.) (Leiden: Brill, 2002), 337-64.
Albert Pietersma and Benjamin G. Wright, “To the Reader of NETS,” in A New English Translation of the Septuagint (Albert Pietersma and Benjamin G. Wright, eds.; Oxford: Oxford University Press, 2007), xiv-xx.
Tessa Rajak, Translation and Survival: The Greek Bible of the Ancient Jewish Diaspora (Oxford: Oxford University Press, 2009).
R.S. Simpson, Demotic Grammar in the Ptolemaic Sacerdotal Decrees (Oxford: Griffith Institute, Ashmolean Museum, 1996).
Gideon Toury, Descriptive Translation Studies and Beyond (Amsterdam: John Benjamins, 1995).
Benjamin G. Wright III, “Translation as Scripture: The Septuagint in Aristeas and Philo,” in Septuagint Research: Issues and Challenges in the Study of the Greek Jewish Scriptures (Wolfgang Krause and R. Glenn Wooden, eds; Leiden: Brill, 2006), 47-61.
Stewart Moore received his Ph.D. in Hebrew Bible from Yale University in 2014. He has taught at Yale University, Amherst College and Fairfield University. He is the author of Jewish Ethnic Identity and Relations in Hellenistic Egypt: With Walls of Iron? (Leiden: Brill, 2014).