Stylistic shifts and markedness
The question of stylistic shifts in translation has received greater attention in more recent translation theory. This has to do with:
(1) interest in the intervention of the translator and his/her relationship to the ST author as exemplified through linguistic choices;
and (2) the development of more sophisticated computerized tools such as CAT and corpus-based methodsto assist analysis.
Schiavi (1996: 14) borrows a schema from narratology to discuss an inherent paradox of translation:
[A] reader of translation will receive a sort of split message coming from two different addressers, both original although in two different senses: one originating from the author which is elaborated and mediated by the translator, and one (the language of the translation itself) originating directly from the translator
The mix of authorial and translatorial message is the result of conscious and unconscious decision-making from the translator
For the analyst, the question is how far the style and intentions of the translator, rather than the ST author, are recoverable from analysis of the TT choices. Such analysis has been termed ‘translational stylistics’ by Kirsten Malmkjær (2003). It has also been advanced by the use of corpus-based methods. These have attempted to identify the ‘linguistic fingerprint’ of the translator by comparing ST and TT choices against large representative collections of electronic texts in the SL and TL.
So, for example, Baker (2000) compares the frequency of the lemma (forms of the verb) SAY in literary translations from Spanish and Portuguese (by Peter Bush) and Arabic (by Peter Clark), and uses the British National Corpus of texts as a reference to judge their relative importance. So, she finds that SAY occurs twice as often in the Clark TTs, and that the collocation SAY that is most common. But this could simply be because of the influence of the SL; the Arabic qaal is generally more frequent in the language than is English SAY because the repetition of the same reporting verb in English is frowned upon.
Most important, perhaps, is the analysis of the relative markedness of stylistic choices in TT and ST. Markedness relates to a choice or patterns of choices that stand out as unusual and may come to the reader’s attention. So, in English a sequence such as Challenging it is. Boring it isn’t is marked because of the unusual word order with the adjectives in first position. In Arabic the nominal sentence is marked because of the unusual word order compared with the verbal sentence which is unmarked or usual.
The key is to look for the reason behind the markedness. In this case, the wording is from a job advert (to recruit police in London), so the markedness functions to draw the reader’s attention to the advert and to illustrate that it is an unusual and challenging job. Similarly, the Arabic nominal sentence is used for emphasis.
In translation, it may usually be expected that a marked item in the ST would be translated by a similarly marked item in the TT but this is not always so.
Some work has investigated the possibility that translation may be less marked
Some work (e.g. Munday 2008) has also examined the distinctiveness of a specific translator’s work. So, comparing patterns in the work of the translator Harriet de Onís, Munday identifies:
-the manipulation of paratextual features (prefaces, footnotes, glossaries);
-a standardization of dialectal choices in dialogue (many different Latin
American dialects standardized into a less dynamic early twentieth-century American English);
-the choice of a rich literary lexicon (e.g. night was sifting through the jungle);
and certain syntactic patterns typical of condensed English style (e.g. the use of compound pre-modifiers such as the unusual tree-dense night and brancharched passage.
Other cognitive translation processes: Think Aloud Protocols (TAPs) :see the book