You are on page 1of 7

SOURCE: BOOK TITLE: READINGS IN TRANSLATION THEORIES Below is one chapter of the book

7. Translation procedures J.-P. Vinay and J. Darbelnet Vinay and Darbelnet's book (1958) is a detailed and comprehensive contrastive analysis of English and French. The authors treat translation as an exact discipline, and seek to minimize the purely intuitive, subjective, artistic element. They are interested in the way different cultures think differently, along the lines of the Sapir-Whorf hypothesis. Their analysis of the translation process actually aims to 'follow the mind", as it were, from one language to another. The basic units of their analysis are units de pense `thought units', which correspond to the smallest part of an utterance that can be translated intact. They may be realized as morphemes, words, phrases, collocations, or even whole idioms. The major theoretical contribution of the book is the influential classification of translation procedures, given below. The present translation into English keeps to their technical terms, except that their use of quivalence (for procedure no. 6) has been rejected in favour of a more explanatory and less ambiguous term: equivalence now has quite a different sense in translation theory (see chapter 10 below). It should also be noted that their use of the term `literal translation' (procedure 3) is at variance with the usage of many other scholars: Vinay and Darbelnet mean approximately `literal and also grammatical', not `literal in the sense of word-for-word and thus (usually) ungrammatical'. The section in square brackets under Modulation contains examples of types discussed elsewhere in the book. All the procedures are further discussed at length in the book, in terms of lexis, grammar and discourse content. The procedure of transposition may be compared to Catford's class shift - see chapter 8, below. The general translation approach advocated by Vinay and Darbelnet leads them to conclude that the traditional controversy between "literal" and 'free" translation is a red herring. "The choice is not between literal and free translation, but between exact and inexact translation" (1958: 267).

***
1. Let us start with the general statement that there are two basic translation methods: direct or literal translation, and indirect or oblique translation. It may happen that a source language message can be transferred perfectly into a target language message, because it is based on parallel categories (structural parallelism) or on parallel concepts (metalinguistic parallelism). 61 But the translator may also be aware of gaps or "lacunae" in the TL which have to be filled by some equivalent means, so that the global impression is the same for the two messages. And there are also cases where, because of structural or metalinguistic differences, certain stylistic effects cannot be transferred to the TL without radical semantic or lexical changes. It is clear that in such cases the translator must have recourse to more roundabout procedures which may appear surprising at first sight but which nevertheless allow a rigorous analysis of equivalence: these are oblique translation procedures. The first three procedures outlined below are direct, and the others are oblique.

2. Procedure (1): borrowing In the case of a lacuna, usually a metalinguistic one (e.g. a new technique, an unknown concept),

borrowing is the simplest translation procedure of all. Indeed, it would scarcely be a procedure of any relevance here if the translator did not occasionally need to make use of it in order to create some particular stylistic effect. For instance, to introduce an element of local colour foreign terms are often retained, such as verst from Russia, sauna from Finland, tequila and tortilla from Mexico, etc. An English phrase such as "the coroner spoke" might be best translated into French, for instance, by retaining the English term coroner rather than by searching for some more or less felicitous equivalent title in the French legal system. There are many old loanwords which have become so much part of the lexis of the borrowing language that they no longer appear as loans; examples from English to French are alcool from alcohol, redingote from riding-coat, paquebot from packet-boat. Of more interest to the translator are the new loans, even personal loans. It is worth noting that loans often actually enter a language via translation as happens with semantic loans or false friends, against which one has to be on one's guard. Elements of local colour evoked by means of borrowings have an effect on the style, and consequently also on the message itself. 3. Procedure (2): calque A calque is a loan translation of a particular kind: a complete syntagma (syntactic unit) is borrowed, but its individual elements are translated literally. The result may be a calque of expression, which preserves the syntactic structure of the SL while introducing a new mode of expression (cf. French Compliments de la saison from Compliments of the season [on a Christmas card]); 62
or it may be a structural calque, which introduces a new construction into the TL (such as science fiction, used as such in French). As with borrowings, there are many old calques which have become fixed in the language; these too may undergo a semantic change which makes them false friends. The translator will be more interested in new calques which seek to fill a lacuna without recourse to an actual borrowing (such as French conomiquement faible `economically weak', originally from German; French l'homme dans la rue, straight from English the man in the street - more natural French would be l'homme de ['of'] la rue or even le Franais moyen `the average Frenchman'). It would seem that the best solution is usually to create a new lexical form from Greek or Latin roots, or to make use of conversion (hypostase - see Bally 1944: 257). This would avoid such awkward forms as thrapie occupationelle for occupational therapy, les quatre Grands for the Big Four, le Premier franais for the French premier [the French form, allowing for a small P, can also mean `the first Frenchman']; calques of this kind are often felt to be deplorable. 4. Procedure (3): literal translation Literal, word-for-word translation is defined here as one where the re sulting TL text is grammatically correct and idiomatic, but where the translator has not needed to make any changes other than those that are obviously required by the TL grammar itself [such as concord, inflectional endings]. Examples: I left my spectacles on the table downstairs -> J'ai laiss mes lunettes sur la table en bas; Where are you? ->

O tes-vous? In principle, literal translation is a unique solution, reversible and complete in itself. It is most commonly found in translations between closely related languages (e.g. French-Italian), and especially those having a similar culture. If literal translation is often possible between French and English, this is because shared metalinguistic concepts can equally well derive from a physical coexistence, periods of bilingualism, with the conscious or unconscious imitation that accompanies a certain intellectual or political prestige. Another reason is the general convergence of thought, and sometimes of structure, among the European languages (such as the creation of the definite article, the concepts of culture and civilization); interesting research has been done here in General Semantics. 5. Up to procedure (3) it has been possible to translate without recourse to special stylistic procedures. If this were always the case the present comparative study of two languages would be unnecessary, and translation, being 63 reduced to a univocal transfer from SL to TL, would be of no interest. The machine translation project at the Massachusetts Institute of Technology (see Locke and Booth 1955), working on scientific texts, is largely based on the existence of parallel segments of text, corresponding to parallel reasoning / processes, and as one might expect these are particularly frequent in scientific texts. However, if a literal translation (in the sense of procedure (3)) is felt to be unacceptable, the translator then has to turn to an oblique procedure. 'By unacceptable we mean that the literallytranslated message (a) has another meaning (b) has no meaning (c) is structurally impossible (d) does not correspond to anything in the metalinguistics of the TL (e) does correspond to something, but not at the same linguistic level. By way of illustration, consider these two examples: (i) He looked at the map; (ii) He looked the picture of health. The first can be translated literally: il regarda la carte; but the second cannot be translated Il paraissait l'image de la sant, unless for some particular expressive reason (e.g. an English person who speaks bad French, in a dialogue). If the translator ends up with a version such as Il se portait comme une charme, which is an idiomatic equivalent of the English, this is precisely because he recognizes this equivalence, at the level of the message; his ability to recognize this is due to his unique position "outside" both SL and TL. Equivalence of message, in the last resort, depends on identity of situation, and it is this alone that allows one to say that the TL retains certain features of reality that the SL does not recognize. Normally, if there were dictionaries of signifis, of meanings, all we would need to do would be to look up the situation identified by the SL message and translate accordingly. But since, practically speaking, such dictionaries do not exist, we have to start with words or units of translation, submitting these to certain procedures in order to arrive at the desired message. Since the meaning of a word is a function of its place in the utterance, it may well happen that the final solution is a choice of words so far from the point of departure that no dictionary would mention it. Given the infinite number of combinations of intervening signifiants [signifiers], it is clear that a translator cannot always find ready-made solutions in dictionaries. For he alone is in possession of the totality of the

message which guides his choice; in the last analysis it is in 64 terms of the message alone, as a reflection of the situation, that the parallelism between the two texts is judged. 6. Procedure (4): transposition Transposition means the replacing of one word-class by another, without changing the meaning of the message. The procedure can also be used within a language, as in rewording: thus He announced that he would return can be reworded, with the subordinate verb becoming a noun, as He announced his return. We call this second version the transposed form, and the original (verb) one the base form. In translation, two types of transposition can be distinguished: (i) obligatory, and (ii) optional transposition. For example, the French ds son lever requires more than a literal translation, because there must also be an obligatory transposition from the French noun lever `arising' to an English verb phrase: as soon as he gets / got up, since the English only has the base (verb) form here. Translating in the opposite direction, though, we have a choice between the calque and the transposition because French has both forms. By contrast, the two equivalent phrases aprs qu'il sera revenu and after he comes back can both be rendered as a transposition: aprs son retour, after his return. The base and transposed forms are not necessarily equivalent from the stylistic point of view. The translator must thus be prepared to carry out a transposition if the resulting version fits better in the sentence or allows a particular stylistic nuance to be retained. The transposed (nominal) form generally has a more literary character. [Transposition does not only occur between verbal and nominal expressions, but between any two word-classes.] Transposition often involves a reciprocal change (chass-crois): one change leads to another. [For instance French emport par le vent literally translates as carried away by the wind, but a transposed version would be simply blown away: here the meaning element in the French verb appears in the English preposition, and the sense of the French noun vent `wind' is now expressed in the English verb blown.] . 7. Procedure (5): modulation Modulation means a variation in the message due to a change in the point of view: seeing something in a different light. It is justified when a literal or transposed translation results in a form which is grammatically correct but not quite natural, going against the feeling of the TL. 65 As with transposition, we can distinguish free or optional modulations from fixed or obligatory ones. A classical example of an obligatory modulation is the phrase the time when ..., which must be rendered in French as le moment o [i.e. time becomes the moment, a unit of time; and when becomes where]. A common example of an optional modulation takes place when a negative expression in the SL becomes positive in the TL, although this is also closely linked to languagespecific stylistic features: it is not difficult to show ... -> il est facile de dmontrer `it is easy to show'. The difference between fixed and free modulation is one of degree. In the case of fixed modulation, a competent bilingual will not hesitate to have recourse to this procedure if it is supported by frequency of usage, total acceptance of usage, or a status established by the dictionary or a grammar book. With free modulation no fixation has taken place and the process must be underg6ne anew in each

case. However, this kind of modulation is not really optional in the strict sense; correctly carried out, it must result in the ideal TL solution corresponding to the SL situation. By way of comparison, one could say that a free modulation leads to a solution which makes the reader exclaim "Yes, that's just how it would be said." Free modulation thus nevertheless tends towards a unique solution; and this unique solution rests on a habitual mode of thought, which is imposed and not optional. Between fixed and free modulation, then, there is only a difference of degree; a free modulation may at any moment become a fixed one as soon as it becomes frequent, or as soon as it is felt to be the unique solution (this usually happens during the examination of bilingual texts, or discussions at a bilingual conference, or as a result of a famous translation which becomes established by virtue of its literary value). The evolution of a free modulation into a fixed one becomes complete when it is recorded in dictionaries and grammar books, becoming something to be taught. From that moment on, nonmodulation constitutes a mistake of usage and is condemned as such. [There are several types of modulation. Some examples are: concrete vs. abstract: give a pint of your blood -> donnez un peu de votre sang `give a little of your blood'; whole vs. part: he shut the door in my face -> il me claqua la porte au nez `he shut the door in my nose'; part vs. different part: he cleared his throat -> il s' claircit la voix `he cleared his voice'; converses: you can have it -> je vous le laisse `I leave it to you'; cause vs. effect: baffles analysis -> chappe d l'analyse `escapes analysis' means vs. result: firewood -> bois de chauffage `wood for heating'; 66 different sense: the rattle of a cab (sound) -> le roulement d' un fiacre (movement) `the rolling of a cab'.] 8. Procedure (6): total syntagmatic change We have repeatedly emphasized that two texts may account for the same situation by means of very different stylistic and structural devices. This is what procedure (6) (quivalence) is about. A classical example is the reaction of an amateur banging in a nail and hitting his finger by mistake: if he is French he will say Aie!, but if he is English he will say Ouch! This example, albeit rather oversimplified, illustrates one particular characteristic of this procedure: the change involved is usually syntagmatic, affecting the whole of the message. Most examples are thus fixed, to use this term again; they belong to the phraseological repertoire of idioms, clichs, proverbs, nominal or adjectival collocations, etc. Proverbs typically provide perfect illustrations of the procedure: like a bull in a china shop -> comme un chien dans un jeu de quilles `like a dog in a game of skittles'; too many cooks spoil the broth -> deux patrons font chavirer la barque `twoskippers will capsize the boat'. And the same is true of idioms: to talk through one's hat or as like as two peas must not be translated as calques on any account; yet this is what one notices among so-called bilingual populations which are in permanent contact with the two languages but end up mastering neither. Of course, some such calques may eventually become accepted by the other language, especially if the situation that they evoke is a new one and thus susceptible to acclimatization into the target language. But the responsibility for introducing calques into a language already perfectly organized should not be the translator's: only the author can allow himself such fantasies, and their succes or failure will reflect on him. In translating one must keep to more classical forms, for any attempt at innovation via calques will seem to carry the taint of an anglicism, a germanism, a

hispanism etc. 9. Procedure (7): adaptation This last procedure brings us to the extreme limit of translation; it is used in cases where the situation to which the message refers does not exist at all in ' the TL and must thus be created by reference to a new situation, which is judged to be equivalent. This is therefore a question of situational equivalence To take an example: it is culturally normal for an English father to kiss his daughter on the mouth, but a similar action would be culturally unacceptable in a French text. To translate he kissed his daughter on the mouth literally as ' il embrassa sa fille sur la bouche, in a context of a loving father returning home 67 after a long journey, would be to introduce into the TL message an element which is absent in the SL; this is a kind of overnanslation. A more appropriate translation would be il serra tendrement sa fille dans ses bras `he tenderly embraced his daughter in his arms', unless the translator wants to make a cheap attempt to introduce some local colour. The adaptation procedure is well known to simultaneous interpreters; there is a story that one interpreter into French, having adapted cricket into the Tour de France bicycle race, in the context of a sport that was particularly popular, was put into a difficult situation by the reply of the French delegate, who thanked the original speaker for referring to such a typically French sport. The interpreter had to reverse the adaptation and return to the English cricket... A refusal to make use of adaptations which are not only structural but also pertain to the presentation of ideas and their arrangement in the paragraph, leads to a text that is perfectly correct but nevertheless invariably betrays its status as a translation by something indefinable in its tone, something that does not quite ring true. This is unfortunately the impression all too frequently given by texts published by international organizations: either because of ignorance, or because of a misplaced concern with the literalness of translations, they demand versions which are as close to calques as possible. The result is a balderdash which has no name in any language, but which Etiemble has justly referred to as "North-Atlantic-ese" (sabir Nord-Atlantique). A whole text cannot be a calque, neither on the structural nor on the metalinguistic level. All the great literary translations have implicitly recognized the existence of the procedures outlined here, as Gide shows, for instance, in the preface to his translation of Hamlet. One wonders whether the reason why the Americans refused to take the League of Nations seriously was that many of its texts were non-modulated and non-adapted translations from a French original, just as North-Atlantic-ese can only derive from illdigested versions of Anglo-American originals. This is an extremely serious problem, but there is not the space here to discuss it further: intellectual, cultural and linguistic changes can, in the long term, give rise to important documents, scholarly manuals, scientific and technical articles, film dialogues etc, all produced by translators who cannot or dare not venture into oblique translations. At a time when excessive centralization and a lack of respect for other cultures are driving international organizations to adopt a single working language in which to draw up texts that are subsequently translated hastily by an unappreciated and insufficient staff, one may well worry that four-fifths of the 68 globe are nourished exclusively by translations, suffering a gradual intellectual decay under this diet of pulped catfood.

10. Application of the procedures The seven procedures outlined above apply equally to lexis, grammar, and the message itself. In addition to grammatical (structural) borrowings like science-fiction in French, one can for instance also speak of borrowings on the lexical level: bulldozer, stopover have entered French as loans; and borrowings on the message level are e.g. O.K., five o'clock tea in French and bon voyage in English. Finally, it is clear that within a single sentence several of these procedures may be used simultaneously, and that some translations depend on a highly complex technique that is difficult to define; for example, the translation of paper-weight by presse-papiers involves both transposition and modulation, both fixed of course. Similarly, the translation (on a door) of PRIVATE by DFENSE D'ENTRER `prohibition to enter' is at the same time a transposition, a modulation, and a total syntagmatic change. It is a transposition because there is a change from a statement ('This is private') to a warning (cf. wet paint -> prenez garde d la peinture `watch out for the paint'); and there is a total change of syntagma since the translation has been arrived at via the situation itself, not via the linguistic structure.

You might also like