Timothy Michael Law's episode of Septuagint Sessions with Benjamin Wright is a gem. I highly recommend it to anyone interested in the Bible or Translation, not only Septuagint scholars.
Wright is an editor of the New English Translation of the Septuagint. You can read about the project, including its clear and helpful preface "To the Reader of NETS" here. He says in the video that they began with the question, "How do you translate a translation?" This led them to Descriptive Translation Studies a la Gideon Toury, which emphasizes the contextual purpose of every act of translation.
Wright and his colleagues have attempted to discern how the Septuagint in its original production was intended to meet a need in its target culture. Certainly the LXX later began to serve as Scripture for the community of Greek-speaking Christians. Wright, however, argues that the translators adopted an "interlinear paradigm," rendering Hebrew idioms and syntax woodenly in Greek, which shows that the LXX was subservient to the Hebrew original. It provided access to the scripture for a Greek-speaking target culture; it was not itself intended to be scripture. Wright argues that the Greek LXX is a) "not a composition," and was b) "not intended in its earliest form to be an independent replacement for that Hebrew text."
The NETS project decided to "give a sense of the Septuagint" by preserving the unusual features of the Greek that result from its close adherence to the underlying Hebrew. "Stilted" Greek became stilted but not ungrammatical English, while the idiomatic parts of the Greek became idiomatic English.
A good example is the use of "man man" in Leviticus. Notice how "'ish 'ish," which is an ordinary construction that means "any man" in the Hebrew, became the odd phrase "andri andri" in the Greek. The NETS translators have rendered this "man by man," which preserves the stilted Greek construction.
In the conversation, Wright makes two distinctions that I consider to be important and insightful, but that raise further questions. First, he argues that our descriptive analysis of a text should distinguish between production and reception. Often, analysis of biblical translations have erased the distinction between the work of the translator and the translation's effect or role in the target culture. A good example of this is the recent attention to the KJV in its 400th anniversary. The KJV translators are considered literary geniuses because their translation came to exert a tremendous literary influence on society. It is more difficult to establish, however, what the translators themselves hoped to accomplish. Some have argued that Tyndale and his followers attempted to create an everyday/common style, and others have argued that the original goal was to achieve literary beauty.
We may disagree about what effect the translators hoped to achieve, and we may disagree about the influence that a translation comes to have on a culture. It is important to remember that these are separate debates. One could argue, though, that the "translator's intent" is as elusive (and irrelevant?) as the "author's intent."
The second distinction that Wright makes is between interpretation and exegesis within a translation. He concedes that all translation is "interpretation," but suggests that some translations go beyond interpretation to "exegesis." Wright distinguishes between the translator's "normal mode of operation" and "theologically-driven decisions in a particular text." If the translators always do a certain thing, you cannot say that they have done that thing intentionally in a particular text for theological reasons.
Wright asks, "how do we deal with exegesis; how do we identify purposeful, theologically-driven translation from the normal habits of the translator, which could be interpretive of course but not necessarily exegetical?"
This is an excellent question, and a major challenge to the work I have been doing in theological constraints on translation. It seems to me that the more obvious exegetical decisions help us perceive the underlying assumptions and goals that have influenced the translation on a systemic level. I would argue that the distinction between interpretation and exegesis is not one of kind, but one of degree—or perhaps of relative subtlety.