0. The present article sets out to attempt to answer to following questions:
0.1. In what ways are the overall strategy, the concept and the ‘philosophy’ of the translator present in the resulting target language texts and how can they be detected?
0.2. If a given text in a source language is used to prepare several translations in a target language, by either the same or a different translator, is it possible to discern a fundamental strategic, conceptional or philosophical modification, or are there only relatively unimportant superficial differences (as concerns the language and the method); in other words, do language and other differences reflect some kind of change in the basic concept and, if so, what are the underlying reasons for this?
1.1. Before making an attempt to answer these questions, we must make some statements relating to theory and methodology.
1.1.1. The process of translating is considered to be an activity consisting of numerous linguistic and extralinguistic (psychological, mental and pragmatic) components, which cannot be dealt with by excluding the translator as was long the case.
1.1.2. A translator needs a certain strategy, idea, concept or ‘philosophy’ for all kinds of texts. There is a certain philosophy in all translations, even if the translator himself denies it. This philosophy has nothing to do with the genre of the text to be translated, whether it be a novel, an essay, a philosophical treatise, a poem, a technical text or the instructions for the use of an electric razor.
1.1.3. Although all texts have a given meaning, sense, style and semiotics, all texts have a different sense, meaning, style and semiotics, and, just as we cannot say that two texts are completely different, we cannot say that two texts are completely alike either.
1.2. If we adopt this policy, we can succesfully avoid many pointless disputes, and at the same time save ourselves trouble by not trying to define or in some way specify the notions of‘literariness’ and ‘literary text’. It seems to me that the defini
tions of these notions have not been fruitful enough, and are not convincing enough either, notwithstanding the fact that literary semiologists, aestheticians, literary critics and philosophers have devoted a great deal of energy to an analysis of this problem. We can also save ourselves trouble by not trying to define the difference between translations of literary works and technical texts, i.e. translations of liter
Sándor Albert
ary works/ philosophical texts/technical texts (cf. Ladmiral 1981: 23). Finally, if we adopt this policy, we can avoid the problems suggested by text typology which are as interesting for the scholars of text linguistics and text typology as the problem of trying to find ‘literariness’. I would merely like to refer to the fact - elaborated in detail elsewhere (cf. Albert 1988) - that I consider the existing text typologies too rigid, meaningless and lacking in delicate nuances, compared to the great variety of the actually existing texts.
1.3. If we start from the fact that the process of translating is a kind of practical activity, we have to accept the fact that its theory (the ‘theory of translation’) is not a rigid doctrine, a dogma or a mine of rules and regulations to be respected, but, perhaps in a somewhat paradoxical way, the theory of this activity, a kind of Hand- lungswissenschaft, théorie de l’application, which ‘is not judged by its a priori estab
lished scientific epistemological criteria, but by its a posteriori products, i.e. the ac
tual texts in the target language’ (Ladmiral 1979:189-190). In this approach the
‘theory’ of translation is nothing but philosophizing about problems encountered dur
ing the process of translating. Let us accept the point of view of Karl Jaspers, who says that philosophy is both an act and meditation on this act.
1.4. Let us consider again the above statement, that we cannot say that two texts are completely different or completely alike. Umberto Eco’s novel, The Name of the Rose is a good exemple to verify this statement. Undoubtedly, the structure of meaning and the semiotics in this novel are much more complicated, full of deli
cate nuances and contain more aspects than those of the instructions for the use of an electric razor for instance, but the fact that the latter also has a structure of meaning and a semiotics can only be doubted by those who have never translated such a text. A text typologist who attempted to put this novel in a certain category of texts would be in a fairly difficult situation, because everybody who knows this novel is well aware of its far-reaching, multiple structure of meanings, and of the trouble the translator has to go to to discover these hidden connotations and refer
ences (if this succeeds at all).
2.1. After this short theoretical introduction, let us now analyse the questions mentioned at the beginning of this paper.
2.2. A Danish author (Bjurström 1987) says that before the translator starts the translation of a text he should always devise (deliberately or intuitively) the overall strategy he will apply. This is certainly so, as the texts to be translated can, in an empirical way, be relatively easily and reliably categorized as a certain literary genre.
However, things are not so simple. The point is not that the translator will translate a novel in a novel—like way (i.e. with the help of the ‘overall’ strategy of novel trans
lation), a literary or philosophical essay in an essay-like way (i.e. with the help of the ‘overall’ strategy of essay translation) or a technical text in a ‘technical’ way, etc. The question is (here I would like to refer to an article recently published by a French linguist, philosopher and translator: Ladmiral 1986) whether he would do his translation with a ‘sourcier’ or ‘cibliste’ philosophy in mind. The former means that the translator tries to do his best to reveal the nuances of meanings and style of the source langage text, and it is the author (the sender) of the given text that would like to ‘transmit’ as accurately as possible when translating; the latter rather
considers the recipient of the targer langage text, i.e. the reader, and it is his needs that he respects; he tries ‘to be in his shoes’.
The two approaches (philosophies) are of course typical not of certain transla
tors (although there are definitely ‘sourcier’ and ‘cibliste’ translators, these terms not indicating any differences as regards value or quality); but of two fundamental positions (theories) on the translators’ part, the certain strategy mentioned by C. J.
Bjurström. Any text can be translated in both ‘sourcier’ and ‘cibliste’ ways, and even one translator is capable of doing either. Disregarding all other aspects, the two resulting target langage texts can be considered to be two versions of the given source language text, and the difference between them (and we would like to em
phasize this) is not in their value, but in their fundamental concept (interpreta
tion)!
2.3. Of course, we speak about an overall strategy or conception on the transla
tor’s part; obviously, it is very dangereous if the translator adheres rigidly to this overall strategy and is unable to depart from it even when he should. For instance, a translator who, in a novel, follows the same strategy when translating the dialogues and the narrative or descriptive parts, commits a grave mistake. This can have sim
ilar consequences to other problems, for instance, if he cannot recognize sarcastic tones, plays on words, hidden meanings, metalinguistic references or irony, or sim
ply confuses the functions of the language (e.g. he interprets the conative, phatic or metalinguistic functions as referential functions).
2.4. It is interesting that (although it happens very rarely) this difference in fun
damental concepts can be seen even in the case of a given translator. A Hungarian writer and translator, Marcell Benedek, for instance, twice translated a novel by Honoré de Balzac Illusions perdues. The first version was published in 1922, and the second in 1957. A comparison of the two texts in the target language (‘transla
tions’) is of interest, perhaps not mainly from a linguistic point of view, despite the fact that their titles are not exactly the same and more than 2500 linguistic differ
ences can be detected in them. It is more striking to observe the deliberate or intu
itive changes in the translator’s overall strategy or conception. It can be very well seen even from a quick comparison that the 1922 publication followed a ‘sourcier’
conception, and the 1957 one a ‘cibliste’ conception. Only suppositions can be made concerning the reasons for the change in conception: it could have been in
fluenced by the translator’s accumulated experience, or his better knowledge of the French language, but the fact that certain translation principles had changed significantly during the intervening 35 years cannot be disregarded either. (I will return to this point later).
2.5. It is even more interesting to compare texts made from the same source language, at more or less the same date, but by two different translators. It is very instructive, for instance, to analyse Hungarian translations of French ‘new novels’
(nouveau roman) of the 60s. A relatively long passage from the novel ‘Planetarium’
by Nathalie Sarraute appeared in a collection presenting the genre of the French new novel with the help of extracts in the early 60s. I could hardly recognise this part in the complete text (translated by someone else) published some years later.
The most interesting feature of the comparison of the two translations is the significant difference in the strategy or ‘philosophy’ of the two translators: the ear
lier translator, not being aware of the real innovations of the French new novel at
Sándor Albert
the time, translated the passage in the traditional way, following the ‘traditional’
patterns of novel translation, and a Hungarian reader had every right to ask: ‘What is so new about this new novel?’ Later, when the theory of literature, semiotics and aesthetics revealed the real innovations of the French ‘new novel’, subsequent trans
lators could form or modify their strategies or philosophies. This can well be seen from the text of the second (complete) edition. Thus, a knowledge of the back
ground, the structure of meanings and the semiotics of the given piece of work is indispensable if the translation is to be successful. If the translator ignores this back
ground information, his translation can well be linguistically accurate (at a level of simple linguistic ‘decoding’), but, from the aspect of equivalence, it is absolutely sure that his work will not be acceptable.
Although it is not closely related to the subject, it may be remembered that ‘the more a text is worked through by the problem of translation, the more untranslat
able it becomes’ (cf. Graham 1985, Johnson 1985: 146).
3.1. Next, let us examine the different methods, possibilities and means at the translator’s disposal to express his own philosophy in the text he creates in the tar
get language. The subject is a rather far-reaching one, I would mention only what I consider the most important part: the role of the significant features. Each text has certain features that a reader, a decoder, a translator, an analyser, etc. may con
sider more important than other features. In the translation of a poem, such signif
icant features can be, for instance, rhyme, rhythm, melodic line, different poetic forms, prosody based on stressed syllables or on quantity, versification, etc. Obvi
ously, in a translation of instructions for use all these play a negligible role. Instead, what is important here for the translator, is the quantity of information, the accu
racy of the contents and the proper application of terms in the target language. In a literary prosaic text, rhythm or stress can also be significant features, but the in
ternal structure of the sentences, their formulation, the relation of the sub-, co- and superordinations, the length of the sentences, the dialogues, or (as we have al
ready mentioned in the case of The Name of the Rose) the role of the different met
alinguistic connotations can also be important.
Clearly, the more complicated, the more multiple, or the more connotative a text, the greater the number of its components that may be significant. At the same time, a good translator will know very well, at the very beginning, that even if he recognizes all the significant features of the source language text (let us suppose so, though in practice this is out of the question), he will by no means be able to trans
mit them into the target language text. This is where and why he has to realize his own concept, his own philosophy: he himself has to decide which features of the text he considers to be more significant than the others, i.e. which features of the source language text he will definitely transmit into the target language text during the process of translating. No theory of translation will tell him which features are more significant than the others in a given text; he himself has to decide about this.
Thus, theoretically, a strange paradox can arise: the translator tries to consider all features significant and transmit all of them into the target language text. Let us suppose he succeeds in doing so. His translation will certainly not he good (will not be equivalent). Moreover, in extreme cases, his work cannot be considered to be a translation, but a simple exchange of codes, a simple linguistic Umkodierung, on
the level of a language or langue: though it would be stretching things to suggest that this could also be done by a machine. I think that this is the basic difference between a translation done by a man and a translation done by a machine. A man always has some kind of philosophy, strategy, background knowledge, experience and culture, whereas a machine has not. Thus, a man can only translate linguistic signs as a text (discours), and can only do so if he has understood and somehow in
terpreted them before. A machine decodes linguistic units on the level of a langue-, it will never interpret or philosophize, and will never add to its experiences or cul
ture (cf. Lederer 1987).
3.2. The role of these significant features can be illustrated very well with the help of some Balzac translations that appeared in Hungary towards the end of the 19th and the beginning of the 20th centuries. In these translations, the Hungarian translators ‘corrected’ the French author’s long-winded, complicated and multi- subordinated clauses: they broke them into 6-8 shorter units. They thought they would render the reader (and also Balzac) a service, as they made the novel some
what more understandable and easier to read, i.e. they polished the style of Balzac, who always seemed to be pressed for time. This principle (method) went to the opposite extreme in the ten-volume Balzac collection published in Hungarian in 1962-64. The new translators, probably following central instructions, took good care to adhere Balzac’s long-winded and complicated style even making an attempt to keep its structure in the Hungarian sentences. It is not easy to decide whether they did right. Obviously, the translators of the 1962-64 publication were more accurate than those at the turn of the century: they respected the principle of syn
tactic equivalence, considering the length and structure of the sentences to be signif
icant features. However, it seems that they disregarded the different functional rules of the two languages. Balzac’s sentences, in compliance with the peculiarities of the French language, are long and complicated. However, primarily owing to the accuracy and unambiguity of their linguistic referential relations, they are always clear and readily understood.
The same does not go for the Hungarian sentences of the same length. They are not always understandable at the first reading and seem to be ponderous and complicated. This is due primarily to linguistic features: in Hungarian, there are no genders, and thus there are no possibilities for referential (grammatical) sequences either; there is only one pronoun in the third person; etc. All this results in the fact that in Hungarian it is much too difficult to create complex, but still clear and un
ambiguous sentences: after a certain length the sentence simply ‘falls apart’.
3.3. The question can be formulated from an aesthetic aspect of course: which translation is better, more beautiful and more equivalent etc. for the reader, com
pared with the original? Which translation is the real ‘Hungarian Balzac’? Is it the one which, disregarding syntactic equivalence, consistently breaks down, one ofter another, the sentences with complicated structures (not considering their length to be a sinificant feature), or the one which, setting an example, respects the French author’s style (considering complicated sentences to be a linguistic expression of pompousness and circumstantiality) and attempts to transmit it by all means into the target language text? Naturally, this cannot be decided on the basis of such a single aspect. However, we could acquire a relatively exact idea about the ‘Balzac
Sándor Albert
perception’ of the different periods by analysing translations published one after another.
An analysis of translations made from the same source language text, but at dif
ferent times, can serve other purposes too. These translations can inform us about the relation of the translators (the readers and the society) to the language in a giv
en period. As far as I am concerned, for instance, the first Hungarian translation of a Balzac novel, Eugenie Grandet, published in 1843 (at a time when Balzac was still alive), is much more interesting (despite the fact that it is full of fundamental mis
translations, blunders and linguistic errors, the structure of its sentences is primi
tive, its dialogues are artificial, etc.) than the sixth translation published in 1957, which is in every respect a well-done, properly reviewed and almost perfect, pro
fessional piece of work.
The most important reason for this may be that the 1957 translation informs the reader not only about Balzac but also about the present condition of the Hun
garian language which, for me, is not particularly interesting. In contrast, the 1843 Eugenie Grandet translation informs me about a lot of other things: the condition of the Hungarian language 150 years ago, the relation between this condition and the public at the time and, last but not least, the philosophy and the efforts of the translator. For all these reasons, the translation made 150 years ago is much more interesting for me and has much more to say than the modern, in every respect better translation of the novel in contemporary Hungarian. (I must stress that this is only as far as I am concerned, because it would obviously be impossible to give this archaic text full of mistranslations to today’s Hungarian readers. They would not understand it and this would not be fair to Balzac either. This is one of the rea
sons why it is worth translating the works of a classic author several times: so that each period can have its own Balzac image. It would be very difficult to decide how often this ‘retranslation’ should be done: this depends on several factors.)
4.1
.
In summary, we can say that a given literary work can be translated into4.1