In packing up my books to prepare for a recent move, I had ample opportunity to peruse the various tomes in my collection. One which caught my eye again was The Word of God in English: Criteria for Excellence in Bible Translation, by Leland Ryken. I read this book years ago, before I began properly studying Greek and Hebrew, and found it very convincing at the time. So I flipped through some chapters to see how well it had weathered time and my growth in knowledge of the languages. While there are certainly aspects of his argument which I would quibble with, one idea central to Ryken’s argument is quite intriguing: literary style is (or should be) part of a full translation theory.
On translation theory…
I’m not big into translation or translation theory. Of course, it is interesting, and I have various opinions on it, some more some less informed. My primary interest in Greek (and Hebrew, and, really, any language I have studied) to this point has been reading Greek as Greek, not translating it into English. Obviously, translation is an unavoidable part of the process, but I have not really been concerned with the theory of how to best translate between languages, or, the more narrow question of how to best translate literature between languages (I think it important to hold these two as distinct, though interrelated, questions). That is where I am coming to this post from: the position of an interested meddler.
For this post I would posit two general approaches to translation that seem to have been active in recent times. One is what I will call the “text-centered” approach. These approaches grew out of sustained trial and error in an environment where there was a continual mix of literacy, translation, and engagement with biblical languages. Their aim is to convey features of the text across languages.
The second approach I will call the “reader-centered” approach. This is meant to describe the set of approaches which have grown out of a mix of trial and error in pioneering Bible translation work in non-European languages (often languages lacking a tradition of literacy at all) and the ideas from various linguistic theories. They are “reader-centered” in that they are concerned with directly conveying the meaning of a text in one language (L1) into another language (L2) in the forms which that meaning would usually be conveyed in L2.
Obviously, this is a simplistic representation of the state of things, and there is much overlap, but it is a simple framework to grasp the basic parameters of the “translation theory” debates.
The place of “style” in Bible translation
I want to briefly explore the notion of style as a criterion for Bible translation. The fact of the matter is, the NT is rather uneven when it comes to “style.” Even a casual read of the Greek NT will make plain that there are radical differences in Greek style throughout the NT. Mark writes quite differently from Luke, who writes differently than Paul, and John, and so forth. As a general rule, English translations completely ignore style differences and flatten them out. All the books speak with the same voice, when translated. I wonder if this is self-evidently the best approach?
Let’s take the Gospel According to Mark, for example. In Mark, the default way that clauses/sentences are connected to each other is with the conjunction “καί.” When students are learning Greek, they learn that this conjunction should usually be translated “and,” if it is translated. “If it is translated” is an important caveat, because English translations vary quite dramatically in how they handle this conjunction. A “text-centered” aproach will default to translating it with “and.” A “reader-centered” approach will default to conveying the idea of connectivity between clauses through whatever set of means the L2 usually does this. As English generally prefers variation in how we connect clauses together, according to English style conventions, a translation of Mark should avoid habitually translating καί “and” as that does not communicate to English readers in terms of the norms of English. Fair enough. This decision, though, is not without its consequences.
One of those consequences is that all the writers of the Bible tend to sound the same. Obvious differences in Greek style are steam-rolled into the same “normal English style.” This is especially true if you pair a decision such as “don’t render καί with ‘and’ but vary the conjunctions according to English conventions” with a decision that often goes hand in hand: aim for short(er) sentences in English because (supposedly) English readers don’t like to read long sentences and would rather read a text that is written at the 6th grade level then one which has long sentences.
Example of narrative from Mark and from John
A brief example comparing how the NIV and ESV render two comparable length narrative sections in Gospel of Mark and Gospel of John. I choose the NIV and ESV because they are the poster children of the two respective approaches outlined above. Both have their obvious strengths and their obvious weaknesses. I am merely illustrating a small way that their translation theories influence the way that the style of writing comes through.
Mark 2.1-12: healing of the paralytic
In Greek καί is used to connect clauses 13x, which is most of the clauses in the story that are not conveying direct discourse (that is, recording what people are saying). The NIV uses “and” for 4 of these (generously counting one verse whose syntax they rather drastically transform), and these are all cases of a complex sentence such as “take up your mat and walk” or “this amazed everyone and they praised God.” The preferred strategy for the NIV is to use nothing in English which corresponds to the καί in the Greek (the use of no conjunction is called “asyndoten” in Greek grammars).
The ESV, by contrast, uses “and” for 12 of the occurances of καί, only omitting the καί in v. 9: “rise and take up your bed and walk”, which is rendered “rise, take up your bed and walk” (there are a variety of textual witnesses which do not have this καί, so the text used by the ESV may not even have it).
Is this a big deal? Well, at this level, not really. Both the NIV and ESV approaches are really quite clear in English and neither is hard to understand. The ESV has the advantage here in more clearly portraying Mark’s Greek style, which is rather choppy and abrupt. He loves to connect clauses with καί, even when it is completely unnecessary to do so from the point of view of Greek. I might even suggest a better way to render this in English, rather than just “and,” would be to use the common lower-register storytelling phrase “and then….”
The clear advantage of the ESV-esque approach for conveying style differences between authors comes when we consider a passage from the Gospel of John.
John 4.1-12: the woman at the well
Here a similar length narrative passage from John, also containing conversation, can serve as a helpful comparison point. This passage tells a completely different story. John uses καί to join main clauses just 6x, which is a sizable minority of clauses in the passage. Both the ESV and NIV use “and” for each one of these instances. They are all instances where καί is joining two halves of a sentence into one complex sentence, which usually requires “and” in English, and 4 of them occur in dialogue. The ESV also uses “and”, probably poorly, as a rendering of the particle δέ at the beginning of v. 4, which the NIV better handles with “now.”
What this brief snapshot is illustrative of is that John does not default to using καί whenever he needs to join clauses together. If anything, John seems to default to using asyndoten (no conjunction) to join clauses together. The NIV, following “English style,” makes regular usage of various English conjunctions even when John defaults to his zero-conjunction connection. For example, v. 7 has no conjunction in Greek, and the NIV inserts the temporal adverb “when,” changing what is a short independent Greek sentence connected with no conjunction (big surprise) to the next short independent Greek sentence into a complex English sentence. This is fine, and the NIV is consistent in this approach. However, my point here is that it illustrates a broader approach that flattens out the style of the writers.
In the NIV, both Mark and John write with the same Greek style. They regularly make use of varied conjunctions–like English putatively prefers–and tend towards a certain length and complexity of sentences. The issue, from the perspective of this post, is that Mark and John have quite distinct writing styles, when you read the Greek. Further, much of their style differences are not hard at all to bring out in English translation, if that is the goal.
Does style really matter?
This is a small and short case study about two of the Gospels and how they join their independent clauses together. Examples of stylistic differences could easily be multiplied between the Gospels. There are even further stylistic differences if we are to include the Epistles. The NIV represents an approach to translation which has decided that all Greek writers of the Bible should write with the same style–a style that is strikingly similar to a lower-register English newspaper or paperback romance. Apparently, a “render-centered” approach to translation tends to envision all readers as only being able to handle one writing style. I’m not knocking the NIV, or any other translation, but am asking the question for myself: does it make sense to completely ignore style in translation?
It may make sense to take the NIV-like approach of imposing a fairly uniform style when you are making a pioneering translation of the Bible into a language without a tradition of literacy. That imposes a lot of difficult restrictions. The Wycliffe Bible translators, and other translation theorists whose work has become central in Bible translation theory debates, are largely operating in this sort of setting. However, that situation is nothing like making a translation into English, or many other languages, with a rich tradition of literacy and a reading populace of people who are regularly engaging with a variety of styles of language use. When this is the case, is there really a compelling reason that all the different writers of the NT should be put into the same straight-jacket, which does not fit many of them particularly well?
Lately, thoughts on style in translation have taken a different turn for me not only in re-finding this book, but also in my more expansive Greek reading. Simply put, not only does the NT attest to a variety of styles inside its pages, but it also sits within a strong literate tradition with a wide variety of styles. Compared to many of the pseudepigraphical works, for instance, the NT writings range from having a solid style to sounding downright literary. By contrast, as I am reading through Xenaphon and Lucian, or looking at Josephus, the writings of the NT come across as rather rough, for the most part. Given that we have a wide array of possible styles and registers in English (two concepts which I am largely conflating in this post, but could be helpfully dealt with separately, in a more detailed investigation), shouldn’t there be more effort to position the Bible, and more particularly each individual author and book, stylistically within the spectrum of English communication and literature?
To the concerns raised by many linguistic-oriented notions of translation which put the experience and understanding of the reader as central, I would still suggest that style is a relevant question. Part of what every reader brings to the exchange with the Bible is a rich set of styles/registers for English communication which we tend to impute different meanings to. Do we really need to try to make the voices of the NT, which speak in a marked array of different styles/register, to speak with one voice? After all, should the choice of style imply meaning, as well?