When I was a teenager I discovered the existence of Rashi, the famous medieval Jewish commentator on the Bible (among other books). Lacking the Hebrew skills necessary to read him, I acquired the only English translation I could find, an edition of the Pentateuch that contained an interlinear translation of Rashi. I liked the idea of it; I would not only be able to read Rashi, but also learn Hebrew along the way.
That project failed quickly and miserably. The truth is that I found the interlinearity to be a nightmare. Typographically I was going from left to right and then right to left all while going right to left. The English translation was almost impossible for me to parse. I could get nothing by lining up the Hebrew and English. Rashi went back to the shelves for several years.
This incident came to my mind as, over the last few months, I’ve been looking into the translation of the Torah into Greek, that is, the production of the Septuagint (or at least the first five books of it) in the third century BCE. As part of his own research into the nature of the Septuagint, Professor Benjamin Wright III has produced a new English translation of the Septuagint and a series of papers that apply theories of translation, particularly those dealing with interlinear translation, to the Septuagint.
The basic problem with the Septuagint is that it is in bad Greek. Who would have used such strange Greek, and for what purpose? Wright asserts that the later myths that developed about this translation, found especially in The Letter of Aristeas (and then used by Philo and Josephus) are historically worthless. The only information that we really have about the Septuagint is the Septuagint itself, and this information has to be delicately teased from the translation.
Wright points out that there are two basic models of translation. One attempts to render sense to syntactical units, be they paragraphs, sentences, or even clauses. The other, in which the idea of interlinearity comes into play, goes word for word, with rendering larger syntactical units comprehensible a secondary concern. This is why I had so much trouble with my attempts to read Rashi – the translation went according to each Hebrew word rather than the larger sense he was trying to make.
The Septuagint, Wright argues, largely follows the word-by-word approach. I think that the later, ancient Jewish readers realized and were also bothered by this: Aristeas, Philo, and Josephus all go to great lengths to assert how “exact” the Greek translation is, which might be a polite way to acknowledge and justifying its peculiarity. Why, though, would anyone translate in this fashion? Wright states that the reason must have been to attempt to draw people into the Hebrew original. Alone, the Septuagint is incomprehensible. It is really meant to be read alongside of the Hebrew, not replace it. Written by and for Jews, it is either a primer or an interpretive guide of some sort. It is interesting to note in this regard that I know not a few people today who claim to have learned Greek, Latin, and Aramaic by reading the weekly Torah readings in those languages.
“Interlinearity” was a scribal practice in antiquity, as we know from several artifacts that write the same “text” in multiple languages (e.g., the Rosetta Stone). But this leads to a technological issue for the Septuagint. To my knowledge all of the earlier fragments of the Septuagint are only in Greek. For a model of interlinearity to work, one would need two different scrolls open at the same time. Who would have bothered?
It is always dangerous to try to infer the nature of a reading community or experience from a text. What we do gain in this case, though, is an insight about the nature of the translation technique of the Septuagint, and, more generally, greater awareness about the kinds of trade-offs that all translators must make.