The main directions in the contemporary linguistic theory of translation.

The earliest linguistic theory of translation was developed by Soviet scholars Y. I. Retsker and A. V. Fedorov who pioneered in a linguistic analysis of translation problems. Their theory came to be known as the theory of regular correspondences. Translation, they argued, is inconceivable without a sound linguistic basis, and this basis can be provided by a contrastive study of linguistics phenomena and the establishment of certain correspondences between the language of the original and that of the translation. The authors of this theory were mainly concerned with a typology of relationships between linguistic units (equivalents - permanent correspondences, not sensitive to context, such as The League of Nations - Лига Наций, and context-sensitive variant correspondences, such as slander- клевета, навет, поклёп) but also investigated some of the translation techniques, such as antonymic translation, thus mapping out some ways of dealing with translation as a process.

In the '60 some linguists (V. Y. Rozentsveig in the USSR and E. Nida in the USA) proposed a theoretical model of translation, based on generative or transformational grammar. E. Nida subdivides the process of translation into three stages: a) analysis where an ambiguous surface structure is transformed into non-ambiguous kernel sentences to facilitate semantic interpretation (the foundation of a school - <- (somebody) founded a school or... - <- a school has a foundation); b) transfer where equivalents in the target language are found at a kernel or near-kernel level and c) restructuring where target-language kernel sentences are transformed into surface structures.

It is true that in some cases it is necessary to paraphrase the source-language structure to facilitate its translation. Such transformations come in handy especially when the source-language structure is ambiguous or when it has no parallel in the target language (e.g. He stood with his feet planted wide apart - <- He stood; his feet were planted wide apart -> Он стоял, его ноги были широко расставлены...; Он стоял, широко расставив ноги.

But transformations in terms of generative grammar are not the only type of paraphrases used in translation. What is more, in some cases, especially when close parallels exist between the source- and target-language structures, they are not even necessary.

The situational model of translationis based on situational analysis in linguistics, developed by V. G. Gak (USSR), J. Catford (UK) and others. It is based on the assumption that languages use somewhat different sets of semantic components (constituents of meaning) to describe identical extralinguistic situations. Russian verbs of motion contain the component of mode but not always the direction of movement while their English equivalents are often neutral with regard to the mode but always specify the direction (cf. Вот он идёт- Here he comes/Here he goes). The situational model provides some interesting insights into the mechanism of translation, especially when a situation is described in different semantic categories (cf. Проточный пруд and spring-fed pond) but does not seem to apply to sentences going beyond a mere description of a situation.

Different translation models complement each other and should therefore be combined in analyzing translation as a process.

The nature of translation.Translation may be viewed as an interlingual communicative act in which at least three participants are involved: the sender or source (the author of the source-language message), the translator who acts in dual capacity - as the receptor of the source-language message and as the sender of the equivalent target-language message), and the receptor of the target-language message (translation). If the original was not intended for a foreign-language receptor, there is one more participant: the source-language receptor for whom the message was originally produced. Translation as such consists in producing a text (message) in the target language, equivalent to the original text (message) in the source-language. Translation as an interlingual communicative act includes two phrases: communication between the sender and the translator and communication between the translator and the receptor of the newly produced target-language text. In the first phase the translator, acting as a source-language receptor, analyses the original message, extracting the information contained in it.

In the second stage, the translator acts as a target-language sender, producing an equivalent message in the target language and re­directing it to the target-language receptor.

In producing the target-language text the translator changes its plan of expression (linguistic form) while its plan of content (meaning) should remain unchanged. In fact, an equivalent (target-language) message should match the original in the plane of content. The message, produced by the translator, should evoke practically the same response in the target-language receptor as the original message in the source-language receptor. That means, above all, that whatever the text says and whatever it implies should be understood in the same way by both the source-language user for whom it was originally intended and by the target-language user. It is therefore the translator's duty to make available to the target-language receptor the maximum amount of information, carried by linguistic signs, including both their denotational (referential) meanings (i.e. information about the extralinguistic reality which they denote) and their emotive-stylistic connotations.

The problem of translatability.Conflicting views have been expressed by linguists concerning the problem of translatability ranging from an entirely negative stand, typical of Wilhelm von Humboldt who considered each language an embodiment of national spirit and the nation's world view and therefore regarded translation as an impossible task, to an unqualified positive attitude, found in many contemporary writings on translation. The very fact that translation makes interlingual communication possible is an argument in favor of translatability. Yet it is an oversimplification to claim that every meaningful element of the text is translatable.

In his preface to the Adventures of Huckleberry Finn Mark Twain says that he had reproduced in the book "painstakingly and with the support of personal familiarity" the shadings of a number of dialects (The Missouri Negro dialect, etc.). Naturally none of these fine distinctions can be reflected in the translation.

Yet by using low colloquial and substandard forms the translator can give an adequate impression of the character's social and educational status and will thus render the most essential functional characteristics of these dialect features. As compared to the determining semantic and functional properties of the text which are perfectly translatable, the untranslatable elements are marginal and relatively unimportant. Besides most of the losses can be to some extent compensated for. Therefore, we may speak of sufficient or adequate (though not necessary complete) translatability to permit effective interlingual communication and satisfactory rendering of communicatively relevant information.