Models based on componential analysis

Translation scholars (or to use H.J. Vermeer’s terms translationists or translatologists) have made endless attempts to come up with new ways of describing of what they understood by translation. In the opinion of A. Neubert, so far many theoretical studies enjoy a low reputation which is “directly linked to the insufficiency and hence non-acceptance of reductive models” that simplify and neglect quite important aspects of translation concentrating on some very peculiar problem areas instead of studying the overall field [Neubert 1989]. Nevertheless, it is obvious that most well-known models of translation have undoubtedly contributed to our understanding “of the pervasive heterogeneity of our elusive object of study” [ibid.] and require a most careful analysis.

Among the most well known approaches to translation are linguistic and pragmatic models some of which are considered below. The early theoretical models of translation are primarily linguistic in their character as their proponents linked the understanding of translation process with theoretical views on language and the interaction of languages in this process and they set themselves the aim of studying the relevant translation area from a selected point of view.

Componential method and practice of translation both relate to the analysis and interpretation of meaning hence the findings of this semasiological procedure were heavily relied upon in attempts to throw light upon semantics of correlated words in the two languages. The supporters of the componential analysis of meaning (J.J. Katz, J.A. Fodor, M. Bierwisch, E. Nida, O.N. Seliverstova, E. M. Mednikova, etc) proceed from the assumption that the smallest units of lexical meaning (semantic components, semes) of correlated words in any two languages seldom coincide in their quality, number, combination and order of arrangement which must be taken into account in translation. This model of translation singles out several stages in the process of translation including (a) establishing in SLT elementary units of content, (b) analyzing their componential structure which is then followed by (c) looking for such correspondences in a TL that are either analogous or closest in meaning to the units of translation. The equivalence of two texts is understood here as being measured on the basis of semantic equivalence of their elements. This model of translation provides a reliable tool for comparison and proper choice of units on various levels:

• correlated words may differ in the quality, number and arrangement of semantic components, cf. изба (деревянный крестьянский дом) – hut (a small, simple building with only one or two rooms); кроватка – (dimin. and affect. from кровать) – предмет домашней обстановки, служащий для спанья – длинная рама на ножках, с двумя спинками, на которую кладутся матрац и постельные принадлежности) – bed (a piece of furniture for sleeping on), cot (a small bed with high sides for a young child).
• correlated synonymic sets may differ in the number and semantic peculiarities of their members, cf. armistice (an agreement to stop fighting, usually for a short time), cease-fire (an agreement to stop fighting for a period of time, especially so that a more permanent agreement can be made), truce (an agreement between enemies to stop fighting or arguing for a short time, or the period for which this is arranged) [Апресян 1979] – перемирие; прекращение военных действий; прекращение огня; передышка, затишье ( НБАРС).

So, in view of the specified semantic differences between the synonyms in SL, as well as particular nuances in meanings between dictionary correspondnces in TL the choice of the proper way of translation is strictly delimited, e.g. ‘I began this book almost immediately after the Armistice …’ (R. Aldington ). – «Я начал писать эту книгу почти сразу же после перемирия …». One more example to prove it, ‘The first stage is a cease-fire. …The second is the signing of a truce agreement.’ – «Первая стадия переговоров должна привести к прекращению огня. На второй – подписывается соглашение о перемирии» (АРСС).

• Combinability of correlated words in the two languages where the difference often stems from the dissimilarities in their semantic structures, cf. the word issue contains the semantic component important, that’s why to translate the Russian word group важный вопрос it is necessary to resort to another adjective to avoid the repetition of the same semantic component, a vital / burning issue.
• Correlated semantic / conceptual fields may not coincide in SL and TL as different languages map differently respective spheres which accounts for the difference in the semantic volume of their members, cf. the semantic structure of the English adjective purple is rather broad and the role of markers in it is greater than that of distinguishers so in various contexts it may be translated by a number of Russian correspondences – фиолетовый, тёмно-лиловый, бордовый, пурпурный, пурпуровый, багровый, багряный, лиловый (БАРС).

Thus, translation models based on the principles of componential analysis of related units prove their value and use in the process of taking correct decisions and estimating the quality of translation. And yet, despite numerous merits, translation practice reveals limitations of such models since they cannot apply to and account for such correspondences in TL that result in adequate translation though there may be no complete coincidence in the componential structure of correlated units.