Interpretation versus Translation -- Competition or Teamwork?

Recently, on the Bible translation list, A. J. Birch asked a very
excellent question:

In terms of definitions, what are the differences between translation and interpretation?

The real answer to this is generally surprising--interpretation must occur before translation and they are very highly integrated.

The question is analogous to asking, "What is the difference between 'construction' and a 'house'?" Many people probably think the question is analogous to "What is the difference between a ranch style and tudor style house?" But it's not.

Translation and interpretation are certainly related but they don't exist along the same axis. Note that in the process of interlingual communication each must happen. But it's not that a translator can remove interpretation to some degree or another and hope to achieve a quality translation. They are distinct but very highly integrated. And let me see if I can illustrate that, but before I do let me deconstruct a popular assumption.

It is generally understood, and wrongly (though seemingly justifiable), that interpretation is something that is added to translation. It's assumed that somehow translation is a somewhat mechanical process involving associating destination language words (and idioms) with source language words (and idioms). The translation process (say, for the New Testament) is viewed as:

1. Look at Greek word
2. Look up Greek word in one or more lexicons and record potential
destination language glosses.
3. Layout a draft of the sentence using the glosses.
4. Style the English without changing the meaning.

That's not how accurate translation works. To quote Eugene Nida:

Professional translators are usually so concerned with the meaning of a text that they seldom give much thought to the grammatical structures of source or receptor languages, because their task is to understand texts, not to analyze them. If, as already mentioned, translators thoroughly understand a source text, they do not need to worry about whether to use nouns, verbs, and adjectives in a particular order so as to represent the meaning. These decisions are made almost automatically.

I would probably change the word understand to the word synthesize to more clearly constrast what has to happen in the mind of a translator.

Accurate translation must consider the meaning of the paragraph, the smallest unit of cohesive communication. This is true since cognitive processes select words within a contextual fabric constructed by at least three things: interlocutor assumptions (assumptions of those involved in the communication event), the synthetic chunking of the text processed so far (which, in the author's mind, creates the paragraph structure), and the linear processing of the current chunk of text (this later is what is commonly and popularly called 'context', but 'context' is much, much larger than even the text itself). What am I saying here? That mentally grasping the context (which includes interlocutor assumptions) is prerequisite to assigning a specific sense (selected from the potential senses) to a particular word form in a specific context. So, again, interpretation precedes translation.

Now, above, I said that it was seemingly justifiable for one to understand interpretation as something added to translation. Why did I say that it appears to be justifiable since I disagree with it?

Because understanding how language works is hard. In the early years of computers (late 50's, early 60's) linguists believed that translation was an easy task and could be easily automated. People at that time believed that the U.N. translators' jobs would be eliminated by the late 60's and Star Trek handheld devices would be available as soon as we figured out how to miniaturize the big computers. As the study of computational linguistics took off, and people's understanding of how translation worked was built into software and applied to real world texts, it became embarrassingly obvious that we didn't know what we were talking about (there's an interesting, and complex, pun in that last clause). Even today, 40 years later, when our phones are more powerful than the computers of that age, Google's translation efforts, while helpful, make for a good laugh (and they have 10,000's of computers gridded together to work on the problem!).

What IS the whole problem? Well, translation is far from straight-forward because a person has to understand the text before it can be translated. Computers can't understand. And translation is far from mechanical. In other words, a person has to interpret the text before the person can translate. One of the funniest comments in this regard was in reply to Eugene Nida when he asked U.N. translators what was the most difficult thing they had to do. Their unanimous answer was the translating of political doublespeak. In other words, they had a very difficult time translating a text that was designed to have ambiguous (or even no) meaning even though it sounded good. Well, as soon as one understands that interpretation is prerequisite for translation the knee-jerk reply to the U.N. translators' statement is "Duh! Of course!" You can't accurately translate a text you can't get your mind around.

Also, as an aside, I've often thought that the process of translating a text from one socio-linguistic context to a significantly different socio-linguistic context forces the discovery of truth (all sorts of rough edges must be dealt with). I think this is true since accurate translation must be coherent--political doublespeak is purposely incoherent speech. I suppose if the translators were appointed to diplomatic positions in the U.N. we would either have peace, or all out war. Truth has a way of doing both; but, I digress.

Now, how can I illustrate that interpretation must happen before translating? The simplest case is with a given word. Clauses, sentences, and larger textual units are far, far more complex.

So, let's try a few simple exercises that will illustrate my point.

Translate this word into English: 'fraznutch'.

You done? You're not?!! Why not? Well, because you don't know what the word means. In other words, even with a single word, you must interpret it before you can translate it.

Now, translate this English word into any foreign language--you chose the language: 'rug'. How did you do? Maybe I should inform you that I was thinking of a hair piece commonly called a toupee. Here, you needed to know the context. In other words, there was interpretation involved in the process of translating. And there's even the issue of register that one must consider in order to generate accuracy in the resulting translation.

You can't escape interpretation in translating.

OK, so now let's make it hard: Translate this into any foreign language:

"A man, a plan, a canal: Panama."

You done? Good. Now, note that the intent of the author is to construct a palindrome. You didn't see that did you? Well, now that you know you need a palindrome, do you think you can come up with an accurate translation? When you get your mind around the implications of that problem, then you start to see the issues with accurate Bible translation. And how confounded difficult it is to answer the question that juxtaposes paraphrase and translation.

It feels simple to just say "paraphrase is loose translation" and "paraphrase adds interpretation to translation" and be done with it. Unfortunately, sometimes what looks like a paraphrase ends up being an accurate translation. I like to use 2 Cor. 6:12 where Paul accuses the Corinthians of having their bowels stuffed shut. It makes for a marvelously funny pun; but, it loses something in translation. Frankly, a paraphrase such as: "you have constipated affections and it's affected your mind" would probably be quite accurate (note Greek KARDIA is frequently more accurately translated as 'mind' in English, though I've thought a better meaning has to do with "the organ of intention").

May our bowels be moved for those in need of clear, accurate, and natural Bible translation.

Hmmmmmm...perhaps that last sentence needs translated.