blog.kennypearce.net

Preserving Ambiguity in Translation

I'm studying Plato's Parmenides in a graduate seminar this semester. It is rather a baffling text, and there is a wealth of secondary literature which contains little consensus on anything. Today, as I was reading Constance Meinwald's guidebook to the dialog, I came across an issue in the translation of the text which I think is relevant to a number of discussion about Bible translation that I've had on-blog, and thought I would share. The issue is one of preserving a (probably intentional) ambiguity in the original in translation, and thus with the degree of interpretation done by translators, and the degree left up to readers of the translation.

What is usually referred to as "part 2" of the Parmenides consists of a series of deductions from contradictory hypotheses. The hypotheses in question are stated in the Greek as hen estin (137c4, etc.) and hen me estin (160c1-2, etc.). The 'standard' translation (that is, the one included in the Complete Works, ed. John M. Cooper) by Mary Louise Gill gives two very literal translations. For the first hypothesis, in the main text Gill has "it is one" and in a footnote gives the alternate translation "one is." For the second hypothesis, Gill has "one is not" and doesn't give an alternate translation. This ambiguity applies to a huge number of statements throughout the dialog and seems to be intentional (more on that in a moment). Meinwald, citing Jowett, suggests (p. 30) that we can preserve that ambiguity if, instead of translations of the form "if the one is" we give translations of the form "on the hypothesis of the being of the one."

Now, if we did the latter in a Bible translation, many people who, like me, favor highly literal translations, would be up in arms about all the words we're adding. However, what I want to point out is that the more litaral translations are more interpretive than the looser ones. That is, because there are multiple possible literal translations of this particular phrase which have substantial differences in meaning, a literal translation requires the translator to pick one of those as the correct interpretation for inclusion in the main text, and thereby leaves fewer readings open in the English than are available in the Greek. That translator is here doing the interpreting and not leaving it to the reader.

To make matters worse, this probably isn't a case where we're simply not sure what Plato means, but a place where he is being ambiguous on purpose. The reason seems to be that the actual thesis of the historical Parmenides was simply "it is." He denied that "it" (that which is) was divided (DK28B8 line 22), and therefore, by implication, that it was plural. So the real hypothesis of Parmenides is that "that which is, is one," or, to put it in better English, "only one thing exists." But Plato's dialog is concerned with the theory of forms, and therefore he seems to use the phrase to mean "Oneness Itself exists." Thus the ambiguity seems to be necessary in order for the phrase to meet both the dramatic needs of the dialog (i.e. to be spoken by Parmenides) and the needs of Plato's philosophical purpose. This seems to be precisely the reason Meinwald embraces Jowett's translation.

To return to the question of the literalness and degree of interpretation of translations, it seems here that the less literal translation turns out to be more accurate. This result will be unsurprising to regular readers of the Better Bibles Blog, where such cases are on display regularly. What I really want to call attention to, though, is that the less literal translation actually involves less interpretation on the part of the translator and leaves more to the reader. This, as I understand it, is the main reason for those (again including myself) who favor more literal Bible translations. You will hear us say "I want a translator to tell me what it says, not what it means." While a certain degree of interpretation on the part of a translator is absolutely necessary, I do agree with that statement. However, as it turns out, there are some cases, such as this one, where that principle ought to cause us to lean toward a less literal translation. How about that?