Literal TranslationINTRODUCTION The present excessive emphasis in linguistics on discourse analysis is resulting in the corresponding idea in translation theory that the only unit of translation is the text, and that almost any deviation from literal translation can be justified in any place by appealing to the text as an overriding authority. The prevailing orthodoxy is leading to the rejection of literal translation as a legitimate translation procedure. Thus Neubert (1983) states that one word of an SL text and a TL word in the translation rarely correspond semantically, and grammatically hardly ever. In the following three French sentences (about 75 words) and their English translation (68 words), every French content-word except taux has its English lexical counterpart, all with a corresponding grammatical function. Only about five function words have no one-to-one correspondents. Les autres pays ont augmente leurs defenses publiques relatives a I'enseignement supeneur plus que la Grande-Bretagne pendant les annees 1968-1970. (Le taux moyen d'accroisse- ment annuel des depenses relatives a I'enseignement supeneur est 24,71 en France, 18,07 an Japan, 28,09 en Suede, mais seulement 8,12 en Grande-Bretagne.) Mais notre pourcentage du PNB consacre aux depenses dans I'enseignement supeneur est quand meme plus grand que celui de presque tons nos voisins. The other countries have increased their public expenditure relative to higher educa- tion more than Great Britain in the years 1968-70. (The average annual increase in expenditure relative to higher education is 24.71 in France, 18.07 in Japan, 28.09 in Sweden, but only 8.12 in Great Britain.) But our percentage of GNP devoted to expenditure on higher education is nevertheless greater than that of almost all our neighbours. I do not think the French translation could be improved on, although one or two variants in the 'taste' area are always available. But about 90% of these three sentences are literally translated - which is perhaps exceptional, but not so surprising in this type of text. My thesis, however, is that literal translation is correct and must not be avoided, if it secures referential and pragmatic equivalence to the original. The meaning of many SL verbs is covered in English by a Romance and a rather less formal phrasal verb, which is likely to be Germanic. Your choice as translator will depend both on the object with which the verb is collocated and on the register of the passage. Thus in derriere lui un garqon distribuait pommes rissolees el petits pois, the verb distribuait is more likely to be 'was giving out' (fried potatoes and peas) than 'was distributing' which sounds, except in some idiolects, like a deliberate (formal) act, or 'was doling out', which betrays a translator's mania for colloquialisms. However, I would not use the most natural 'was serving', since this, by the back-translation test, would be servait. Note that other collocations also offer alternatives: for vivres, 'distribute' or 'share out'; counier, 'deliver' or 'hand out'; ordres, 'give' or 'deal out'; cartes, 'deal' or 'deal out'; argent, 'distribute' or 'hand round'; role, 'assign' or 'give out'. Whilst the second alternatives in the above examples are not literal translations in my definition, they all appear to be at variance with Neubert's proposition.
|
