Fragment 1501

The primary aim, for the sake of which all others were disregarded, was to preserve the literal meaning of the original…
T.J.Binyon
Pushkin

When I look at an article in Russian I say: “This is really written in English but it has been coded in some strange symbols. I will now proceed to decode.”
Warren Weaver
Translation

In the introduction to his highly controversial translation of Pushkin’s verse novel, Nabokov identifies three kinds of translation. (His remarks do not specify whether he means verse translation in particular, or whether he means literature in general.)

1. paraphrastic, a free version of the original, with omissions and additions prompted by the exigencies of the form
Good examples of this would be Christopher Logue’s renditions of the Iliad in War Music and Ezra Pound’s of Chinese poetry. The original is used as a springboard to launch the translator poet into a more personal flight of inspiration.

2. lexical, rendering the basic meaning of the words and their order
This is the kind of translation that one gets from a machine, such as Google translator, or Yahoo. We can call this a word-for-word translation.

3. literal, rendering, as closely as the associative and syntactical capacities of another language allow, the exact contextual meaning of the original
Personal examples of this would be Richmond Lattimore’s Homer, Stephen Mitchell’s Rilke, Auden’s Brecht, Terence Kilmartin’s Proust, Pasternak’s Shakespeare: monumental performances which stand the test of time and which set the bar.

1 and 2 may be seen as opposite ends of a spectrum. In 1, Out of Language is sacrificed in order to make a verse artifact in Into Language. In 2, Into Language (and often meaning) are sacrificed in order to stick as closely as possible to the syntax and lexis of Out of Language. Translations shift between these two poles, even in mid-sentence; they may be thought of as vibrating between them, like a finger on a taught string.

3 is the perfect (unobtainable?) balance of languages and sensibilities, in which nothing is compromised and no sacrifices are made to either language, neither in the linguistic demands they make of each other, nor to the meaning. It is a mythical beast; for Nabakov, the jinni of translation; for Eco, a unicorn.

The perfect literal translation persists as a kind of Ideal towards which all translators yearn but only seldom reach. Let’s look at how language communicates meaning in order to understand the difficulties involved in translation and in the process try to catch a glimpse of the unicorn.

Language communicates meaning in three ways: lexically, grammatically and semantically. In the first, key words have meaning in isolation, and the translator can look for exact synonyms across two languages: green in English becomes vert in French and lu in Chinese. It is this lexical correspondence (usually fortuitous, often suspiciously easy) which does more than anything to cause the perpetuation of the word-for-word translation myth. It is this level of meaning which Weaver talks about, and only this level which is capable of being transcoded.

The second way language communicates meaning, the grammatical, takes us to another level of complexity and difficulty. Here, grammar communicates meanings such as time, gender, number, position relative to other actors in the sentence, transitivity or non transitivity of actions, and so on. This kind of grammatical meaning is usually communicated through affixation, on the vertical axis. In addition, grammatical meaning is communicated through syntax on a horizontal axis: prepositional phrase order, subject-verb-object positions, definitional clauses, embedding, front- or back-loading devices such as the passive (I estimate- it is estimated) and cleft sentences (A man was walking –there was a man walking). Every language has its own balance of how grammatical meaning is distributed between the two axes, and translating from a language where the vertical axis predominates, into one where the horizontal one predominates increases the difficulty for the translator and causes the elusive unicorn to retreat a little further off.

The third way language communicates meaning is the semantic, a third level of complexity, in which the meaning of individual words is elevated or subsumed into some other, wholly conventionalised meaning, which cannot be grasped through the individual words. This area covers linguistic phenomena such as idioms, proverbs, compounds of more or less transparency such as phrasal verb structures, phatic phrases, colloquialisms, slang, euphemisms and so on. Here the unicorn becomes obscured by undergrowth, for an idiom in one language might use quite different words as the same idiom in another, making a word-for-word or even a literal translation impossible.

Now, so far our investigation has been based on three assumptions. First, that we are talking about surface (denotational) meaning only. We have not taken into account other more elusive types of meaning such as connotational: the way words taste and smell, their colour and weight, the way they cause us to react emotionally apart from their denotation; and referential: the way words call out to other words, the way texts echo other texts, either overtly or covertly, the way they refer to a shared knowledge of cultural and historical artifacts.

Second, we are leaving aside all such theoretical issues such as how meaning is created in the dialogue between the reader and the author, who may or may not be dead in Barthes’s sense of the word, how the reader’s response in Iser’s sense of the word may be quite different from the author’s intention in Wimsatt and Beardsley’s sense of the word; how meaning is temporary, highly individualised and created anew with each reading.

Third, our investigation so far assumes also that each language in the translation endeavour shares the same concepts, and that the difficulty of translation simply lies in finding equivalent terms for these shared concepts. However, once we take into account that some languages have concepts which do not exist in other languages; that the conceptual categories, not to mention the linguistic ones, may not even correspond, then we are in very great difficulties indeed.

Once we add these difficulties to the problems of translating purely denotational surface meaning, then we find ourselves in very deep forests, and the unicorn vanishes altogether.

Our translator works in the shadow of two muses: justice and loyalty, both of them blind. His aim is to do justice to both languages; but he is always pulled in either of two directions by his drive to stay faithful to the meaning. He cannot serve both muses all of the time, but must turn to each of them, meeting their insatiable demands, pacifying them, compromising between them, playing the diplomat.

In Pushkin’s Eugene Onegin, the narrator complains about the difficulty of translating Tatiana’s letter from French into Russian. He uses the simile of a young girl attempting to play the overture to Der Freischutz. For Pushkin, translation is like a musical performance, in which a work is created each time anew from marks on a written page, in which the perfect note may be struck. For our harassed translator, this might be a more helpful, a more consolatory metaphor than a decoding.