We have established thus far that stylistic approaches can look at different parts of the translation process, and that style itself can be seen to mean different things in relation to language, and therefore in its relation to translation. A third area about which there are very different views is the relationship of language, and specifically of style, to the universal and the particular. On the face of it, translation would seem to demand that at least something is universal, and for stylisticians concerned with the translation of style, an important question will always be this: which aspects of style are universal and which are inextricably linked with a particular language? In considering this question, it is worth looking at Jakobson's work. Though one of the criticisms commonly made of his work in stylistics by what Bradford (1994) refers to as the anti-Jakobson School of the 60s and 70s, typified by Fowler (e.g. 1975), is that it is overly formalist, and ignores the effects of situation and the reader, he makes rewarding reading for those working in translation, even today.
Jakobson was particularly interested in what made texts literary. In common with many structuralist and text-based writers (including American New Critics such as Wimsatt and Brooks and English proponents of close-reading such as Richards) he preferred to work with poetry, because he felt that its close and easily observable link between form and content embodied the essential nature of literature (1978b). However, according to Tabakowska (1993:11), all his work is based on the "assumption that 'normal' language is not qualitatively different from 'poetic language'", a view shared by Fowler. Nevertheless, while he pointed out that all language uses the same resources, Jakobson saw literary (and specifically poetic) language as being functionally different in that it used these resources differently in order to meet the "poetic function". In a 1960 article on poetics he states that this function "projects the principle of equivalence from the axis of selection onto the axis of combination " . In other words, if a non-poetic utterance might choose to say "the cat lay on the rug", a poetic one would select or insert words in the syntactic structure of the sentence in a way governed by their similarity to one another: the cat sat on the mat.
According to Bradford, who traces the life and views of Jakobson, the underlying theme in all his work (whether on phonology, film, aphasia, translation or poetry) was that the material substance of the sign is never fully distinguishable from its signifying properties" (1994:3). If this is so, then it does not bode well for translation, which by its nature needs to transport the sign into another language and situation, where its "signifying properties" will be different. So is that unity between form and meaning doomed to be lost in translation? Or worse, does it render translation impossible?
Jakobson's answer, in a 1959 article on translation, was that translation in a strict sense is not possible, at least for poetic texts, but "creative transposition" is possible because "cognitive experience" is universal, as are certain characteristics of poetry such as its concern with style and pattern (2000:118, 115). In focussing on the poetic function of language, Jakobson differed from the common structuralist view (e.g. Saussure 1959) that language is arbitrary. Poetic language was non-arbitrary, governed by its sound patterns, partially hermetic, and ambiguous, and thus functionally distinct not just from non-literary language but also from non-poetry.
Jakobson's concern, like that of Richards, went far beyond the formal aspects of poetry. He saw stylistic figures such as metaphor and metonymy as psychological processes (1987:100) and in this was the forerunner of later cognitive stylistics. However, the linguistics on which his stylistics was based was essentially a structuralist, inductive linguistics, and did not, therefore, place the human mind at the centre of linguistic explanation. Nevertheless, Jakobson's identification of common "cognitive values" (2000:115) of language, which, unlike the close bond between form and meaning in poetry, were translatable, emphasizes the interaction between universal and specific as the basis for translation.
This is an interaction noticed by many translators who distinguish the form-meaning relation fixed by the linguistics of a particular language from what Levine, for example, calls the secret bonds among all languages(1991:8). The latter make translation possible.