PANEL PRESENTATION OF A NEW ENGLISH TRANSLATION OF THE SEPTUAGINT
The following presentation was given by Albert Pietersma to The Xth Congress of the International Organization for Septuagint and Cognate Studies, meeting in Oslo, Norway July 31-August 1, 1998.
A New English Translation of the Septuagint and the Other Old Greek Translations Traditionally Included Under That Title (NETS) is sponsored by the IOSCS and will be published by Oxford University Press, New York. No volumes have as yet appeared. It was launched effectively in 1995 with its Statement of Principles (Now Appendix 1 in the Manual).
The NETS Translation Manual (Uncial Books, Ada, Michigan) was published in 1996. All basic NETS materials, including its recently completed general introduction (To the NETS Reader), may be accessed on the NETS Web Page http://ccat.sas.upenn.edu./nets/ (Webmaster Jay Treat). English speaking scholars from three continents (Africa, Europe, North America) have been engaged as translators. An electronic discussion list facilitates easy and frequent communication.
The aim of NETS is twofold: (1) to create a faithful translation of the LXX and (2) to create a tool for synoptic use with the NRSV for the study of the Greek and Hebrew Bible texts.
To provide a rapid overview of the NETS project, I draw on certain sections from the general introduction, a fuller version of which may be found on the NETS homepage.
NETS AND THE NEW REVISED STANDARD VERSION.
Ancient texts have been translated from time immemorial, and the need for such work continues. What is not always clear is the precise reading-public a translation should presuppose. This is perhaps especially true for biblical literature. Writing specifically on the topic of Bible translations, Nida and Taber envisaged no fewer than three such audiences.It is usually necessary to have three types of Scriptures: (1) a translation which will reflect the traditional usage and be used in the churches, largely for liturgical purposes (this may be called an "ecclesiastical translation"), (2) a translation in the present -day literary language, so as to communicate to the well-educated constituency, and (3) a translation in the "common" or "popular" language, which is known to and used by the common people, and which is at the same time acceptable as a standard for published materials.
NETS is addresses primarily the reading public identified in Nida and Taber's second grouping, namely, a biblically well-educated audience, on the assumption that it is most probably this audience that has a more than passing interest in biblical traditions other than their own and may furthermore be expected to have an appreciation not only in differences of content but literary character as well.
Given the fact that NETS has been based on the New Revised Standard Version (1989), its character can be said to derive also—in part at least—from the NRSV. That an existing English translation of the Hebrew Bible should have been used as a point of departure for NETS may need some explanation. Why not, one might object, simply translate the Septuagint as Thomson and Brenton had done, i.e. without any overt dependence on any English translation of the Hebrew? The answer to this question is based on considerations of both principle and practicality. I begin with the considerations of principle.
Though it is patently true that the Septuagint in due time achieved its independence from its parent text and that it at some stage shed its subservience to its source, it is equally true that in its inception it was a Greek translation of a Semitic original. That is to say, the Greek had a dependent and subservient linguistic relationship to its Semitic parent. More particularly, for the vast majority of Septuagint books this linguistic relationship can best be conceptualized as a Greek inter-linear translation of a Hebrew original. Be it noted immediately, however, that the term "interlinear" is intended to be nothing more than a visual aid to help us conceptualize the linguistic relationship that is deemed to exist between the original and its rendition into Greek. In other words, "interlinear" is a metaphor and as metaphor it points not to the surface meaning of its component parts but to a deeper, less visually accessible, linguistic relationship of dependence and subservience. It should be emphasized further, that the deeper linguistic reality, which the metaphor attempts to make more tangible, is in no way contingent on the existence of a physical, interlinear entity at some point in three centuries BCE.
Looked at from a different perspective, NETS is presupposing a Greek translation which aimed at bringing the reader to the Hebrew original rather than bringing the Hebrew original to the reader, to paraphrase Sebastian Brock. Consequently, the Greek's subservience to the Hebrew may be seen as reflecting its aim.
NETS has been based on the interlinear paradigm for essentially three reasons. First, this paradigm best explains the "translationese" Greek of the Septuagint with its strict, often rigid quantitative equivalence to the Hebrew. As Conybeare and Stock (and others) noted nearly a century ago, Septuagintal Greek is often "hardly Greek at all, but rather Hebrew in disguise," especially in its syntax. Secondly, the interlinear paradigm of Septuagint origins makes it legitimate for the NETS translator to draw on the Hebrew parent text as an arbiter of meaning, when appropriate. Differently put, the interlinear paradigm recognizes that unintelligibility of the Greek text qua Greek text is one of its inherent characteristics. Thirdly, the interlinear paradigm safeguards the Greekness of the Septuagint by not having to assume meanings (syntactic as well as lexical) not attested outside of the translated corpus.
Thus the Committee decided to focus on the most original character of this anthology, namely, that of interlinearity with and dependence on the Hebrew; or, from a slightly different perspective, that Septuagint which comprises stage one of the history of the Greek Bible. Or again, NETS aims to focus on what the translator evidently thought the text to mean, as inferred from the linguistic information his text provides.
Once NETS had decided on its aim and focus, a methodological directive seemed unavoidable. If NETS was to render into English the Greek half of the Hebrew-Greek interlinear posited as the paradigm, its English text might then best be made "interlinear to" a modern English translation of the current Hebrew text. Differently put, since NETS was to reflect the original dependent relationship of the Greek upon the Hebrew, one could do no better than to begin by basing NETS on an existing English translation of the Masoretic Text.
But if Septuagint origins can best be understood in terms of the interlinear paradigm, it follows that, as is characteristic of interlinears, one should read this original Septuagint with one eye on the parent member of the two-some, namely, the Hebrew. Thus what this Septuagint says, and how it says it, can only be understood in its entirety with the help of the Hebrew, even though the precise nature of dependence on the Hebrew may vary from book to book, chapter to chapter, and verse to verse. This interlinearity with and dependence on the Hebrew may be termed the Sitz im Leben of the Septuagint, in contradistinction to its history of interpretation, or better, its reception history. From the NETS perspective these two aspects of the Septuagint are in principle distinct even though in practice they may turn out to be indistinguishable.
In the light of what has been said, it is therefore appropriate to think of NETS along the lines of the Goettingen Septuagint: just as the Goettingen editors attempt to establish the original form of the Greek text and in so doing draw on the Hebrew for text-critical leverage, so NETS has availed itself of what leverage the Hebrew can provide in establishing the meaning of the Greek. Moreover, just as the form of the original text differed from its later textual descendants, so what the original translator thought his text to mean differed, in principle, from what later interpreters thought the text to mean.
In addition to the dictum of principle, there also emerged a very practical consideration for basing NETS on an existing English translation of MT. In the our view, central to the raison d' être of a new translation of the (original) Septuagint—i.e. a translation of a translation—is its synoptic potential. That is to say, users of such a translation, especially in light of the interlinear paradigm, should be able to utilize it to the greatest degree achievable (within set parameters) in a comparative study of the Hebrew and Greek texts, albeit in English translation. This aim could best be realized, we believed, if English translations of the Hebrew and the Greek were as closely interrelated as the two texts themselves dictate or warrant, both quantitatively and qualitatively. In other words, ideally the user of NETS would be able to determine not only matters of longer or shorter text and major transpositions of material, but also questions of more detailed textual, interpretational, and stylistic difference. Needless to say, we harbor no illusions about this goal having been fully reached.
Given the above decision, essentially two options were open: (1) we could first translate the MT into English and then use this translation as the basis for an English translation of the Greek, or (2) we could use an existing English translation of the MT as a point of departure. Clearly the latter route recommended itself as being the more practical and economical one. It was, furthermore, difficult for us to see how the work of the committees of scholars that produced the major English translations of the Hebrew could be improved upon to the degree that the extra effort would be justified. History may prove us wrong here.
NETS AS REVISED NRSV
Two considerations have guided us in choosing an English version as a base text for NETS: (1) general compatibility of translational approach with that of the LXX and (2) widespread use among readers of the Bible. The New Revised Standard Version, based as it is on the maxim "as literal as possible, as free as necessary" (Preface x), we thought to be reasonably well suited to NETS purposes on both counts. Consequently, throughout those Septuagint books which have extant counterparts in Hebrew (or Aramaic), NETS translators have attempted to retain the NRSV to the extent that the Greek text, in their understanding of it, directs or permits. At the same they have done their best to keeps concessions to the NRSV from compromising the Greek.Our desire to enable the reader to make use of NETS in synoptic manner with the NRSV, has been second only to our commitment to giving a faithful rendering of the Greek original. In fact, NETS may be said to have two competing aims: (1) to create a tool in English for the synoptic study of the Hebrew and Greek texts of the Bible and (2) to give as faithful a translation of the Greek as is possible, both in terms of its meaning and in terms of its mode of expression. Since these are competing aims, translators have often, especially on the expression side, been called upon to do a balancing act.
TRANSLATING A TRANSLATION
Translating an ancient text can only be described as a profoundly difficult and audacious undertaking. Not only do translators have to contend with the natural gulf that exists between languages and with the absence of the original authors, but they also suffer from the lack of native speakers of the ancient languages, who might be persuaded to give some much needed help. Consequently, what the modern translator of an ancient text is trying to do is something like starting up a one-way conversation, or a monologue that passes for a dialogue.The difficulties of the undertaking are certainly not decreased when one undertakes to translate an ancient translation into a modern language. If translating, of whatever stripe, is an act of interpreting rather than a simple transferring of meaning, a Greek translation of a Hebrew original can be expected to reflect what the translator understood that text to mean. The end result is therefore inevitably to some degree a commentary written at a specific historical time and place by an individual person, whose understanding of the Hebrew will often have been at variance with our own, though at times perhaps equally viable.
As suggested by the interlinear paradigm, much of the Septuagint is, moreover, a translation of a special kind. Therefore, whereas a translation that replaces the original typically "solves" the problems of the original, in an interlinear translation these may simply be passed on to the reader. In fact new problems may often be created because of the translator's inherent preoccupation with representing as much as possible of the linguistic detail of the original—which is not to say that the interlinear type of translators of the LXX were unconcerned about making sense, but simply that the interlinear language-game of the ancient translator has added an extra dimension to the problems faced by the NETS translator. The notion of Sitz im Leben, introduced above, comprises, among other things, certain realities of the source language, Hebrew (or Aramaic). Just as inappropriate as accusing the interlinear translator of lacking concern for making sense would be to saddle him with inadequate knowledge of Greek, since his use of Greek is determined by the aim he wishes to achieve, namely, that of leading the reader to the parent text.
The paradigm of Septuagint origins as an interlinear text, in contradistinction to the Septuagint as a free-standing, independent text now calls for a further distinction, namely, that between its constitutive character on the one hand and its reception history on the other. The distinction is important because it demarcates, in principle, two distinct approaches to the Greek text. That is to say, one can either seek to uncover the meaning of the Greek text in terms of its constitutive character (i.e. in terms of its interlinear dependence on the Hebrew), or one can try to render the meaning of the text from the perspective of its reception history (i.e. in terms of its independence and self-sufficiency). The difference between the two may be simply illustrated. Though the entire Greek language community of 3-1 centuries BCE would agree that Greek DYNAMIS sometimes means "host/army" but at other times means "might/strength," which component of meaning was right for which context might well be a matter of dispute. From the perspective of the Septuagint text as an independent, self-sufficient entity, context is recognized as the sole arbiter of meaning. That is to say, should the context speak of military might, DYNAMIS would be translated by "army," but if the (Greek) context be about bodily strength instead, DYNAMIS would be rendered by "strength." On the other hand, from the perspective of the Septuagint as a dependent, subservient entity, one could not agree that context is the sole arbiter of meaning. What if context should admit either reading and thus fail to steer the reader into one direction or the other? In that case, based on our diglot model, the Hebrew parent text would be the arbiter in the dispute. Should the underlying Hebrew have TsBA ("army, war, warfare"), Greek DYNAMIS should be understood as "host/army," but if the Hebrew be OZ ("strength, might") instead, DYNAMIS would have to be understood as meaning "might/strength."
Perhaps the most obvious examples of Septuagintal dependence (as opposed to independence) are cases in which, due to the ambiguity inherent in Greek grammar, only the syntactic relationships (e.g. subject or object role) of the Hebrew can guide the English translator to what the Greek text means. Thus a sentence such as TO PAIDION EIDEN might mean either "the child saw" or "(s)he saw the child."
The distinction between the text as an independent entity or the text as a dependent entity is, therefore, not only a valid one in terms of the NETS paradigm but, in our view, is an important methodological stance for translators of the (original) Septuagint, with frequent practical consequences for NETS. Differently put, one can either treat the LXX as though it were an original or one can treat it as a translation of that original. Though both are worthy undertakings in their own right, NETS perceives them as different in principle.
Constitutive character is a figure for socio-linguistic realities. As such it includes both what—judging from the language used—the text overtly means and what at times resulted covertly from the model used by the translator. Again, inherent in the model of the LXX as an interlinear rendering is the word-by-word manner of translating, including structural words (articles, prepositions, conjunctions). Also to be expected in an interlinear translation are standard and stereotypical equations between Hebrew and Greek words, including structural words. For these reasons and more, though the LXX is in Greek, there is also much that is decidedly un-Greek.
HOW NETS DETERMINES WHAT THE GREEK MEANS
Simply put NETS has been governed by five lexical guidelines, which can be made to apply as well, mutatis mutandis, to the grammar of Septuagint Greek, and all of them are implicit in or concordant with the posited interlinear paradigm of Septuagint origins. (1) Greek words in the LXX normally mean what they mean in Greek of that time (statistically the vast majority of the lexical stock belongs here); (2) the precise nuance of Greek words is sometimes arbitrated by the Hebrew parent text (see the DYNAMIS illustration above); (3) some Greek words, when they are used rigidly as uniform renderings of the corresponding Hebrew words, fit poorly into some of the contexts in which they stand—these may be dubbed stereotypes (see e.g. "will" [THELHMA] for NRSV's "desire" in Ps 1:2); (4) some Greek words in the LXX have been selected by the translator solely because of their perceived connection with (a) Hebrew morpheme(s)—these may be called isolates (see e.g.BY [oh please!] = EN EMOI = "in/with me" in 1Rgns 1:26 et al.); (5) some Greek words in the LXX have Hebrew meanings, i.e. the chief meaning of the Hebrew counterpart has been transferred to the Greek which has then become part of the living language—these may be labeled calques (see e.g. BRYTH = DIATHHKH= "covenant" throughout the LXX, but "will, testament" in extra-biblical Greek).Even though the title of the NETS project speaks of the literature as a body, namely, the Septuagint, members of this anthology show considerable diversity, the interlinear model not withstanding. Thus, Greek translations range all the way from highly literal to very free. Moreover, on a scale extending from the prototypical translator, who acts as a mere conduit for his author, to the prototypical author, who composes everything from scratch, Septuagintal writers would be seen scattered along most of its baseline. One finds not only full-fledged authors (e.g. 2 Maccabees and Wisdom of Solomon) who composed their works in Greek, but also bona fide translators who in varying degrees attempted to approximate our prototypical translator. Thus one might note, for example, Ecclesiastes as the most prototypical translator (being very literal) and Job as the least prototypical (being very free). Needless to say, a Joban translator must be labeled part author. NETS introductions to individual books or units are designed to give some detail on the nature of specific translations.
What has been noted in the preceding paragraph highlights a number of facts. First, though the paradigm basic to NETS is that of the Septuagint as an interlinear text, it does not follow that all interlinear texts are equally literalistic. In orther words, interlinearity is not contingent on degree of literalness. Second, there are within the translated corpus exceptions that prove the rule, such as Job, Proverbs, Isaiah, and Esther in part. Third, those books originally composed in Greek, such as 2-4 Maccabees and Wisdom of Solomon, by virtue of not being translations are not governed by the NETS paradigm.
NETS: ITS CHARACTER AND EXTENT
Though NETS is based on the NRSV, it is not intended to be the NRSV-once-over-lightly but rather a genuine representation of the Greek, reflecting not only its perceived meaning but also, to the extent possible in an English translation, its literary nuggets as well as its infelicities, pleonasms, problems, and conundra.One scarcely expects literary beauty and rhetorical flourish from an interlinear text, since that was not its purpose. In fact, it would make little sense to accuse an interlinear translator of lack of literary sense. When literary beauty occurs it is the exception that proves the rule. Consequently, NETS readers would be remiss in expecting literary elegance in the English. That would have required, from our perspective, a different Greek. Since the Septuagint, with a few exceptions, was not originally composed in Greek and often uses unidiomatic Greek, a fully idiomatic translation into English can scarcely be justified. NETS is consequently more a translation of formal correspondence than one of dynamic equivalence. All in all, what readers can expect is a reasonable facsimile of the (original) Septuagint such as it is, including many of its warts.
The reason for the NETS approach is integral to the NETS aim: that of reflecting the Septuagint's constitutive character and of attempting to capture the incipit of the history of interpretation of the Greek Bible. Implicit in this aim has been a concerted effort not to make the Greek text say more than is strictly warranted by the linguistic evidence, but to leave such elaboration to later stages of exegesis or eisegesis, as the case may be.
GREEK TEXT
Since NETS claims to be a translation of the Greek text as it left the hands of its respective translators—or a "Goettingen Septuagint in English form"—it stands to reason that NETS has been based on the best available (critical) editions. That is to say, where available, NETS has used the Göttingen Septuagint; Margolis has been deemed best for Joshua, and Rahlfs' manual edition is used for the remainder of the books. In the event that new and improved critical editions appear during the life of the project, the Committee is committed to using these, if at all possible. But since no edition, no matter how carefully and judiciously executed, can lay claim to being the definitive text of the Greek translator, NETS translators have from time to time sought to improve on their respective base texts. Just how much will have been changed, varies with the quality of the edition used. All such deviations, however, have been meticulously noted.The editions used thus reflect the current state of the discipline.
Albert Pietersma
Oslo, August 1, 1998