The notion of stylistic interaction also informs the article by Millán-Varela (2004) in the Special Issue just mentioned. This study of the style of a translation of Joyce into Galician aims to establish how the presence of the translator's voice interacts with all the other voices already present in the source text" (2004:38). She points to various examples of strangeness" in the translated text caused by keeping to the source text very closely and also traces of a previous translation. She found an overall"tendency to generalize", missing out details and linguistic features such as adverbials (2004:46: 52). Marco (2004), in the same Special Issue, besides looking at how differences in style between target text and its source text can change the reader's view of a particular character – in this case the governess in Catalan translations of Henry James' The Turn of the Screw (James 1984) - also considers, like Malmkjer,the style of the translator and how its presence is felt in the target text. He shows that a particular group of translators might have a particular approach to translating and that this,too,can be traced in the style of several target texts when compared. His interest is in an aspect of the study of style mentioned in its definition by Thornborrow & Wareing (1998) given at the start of this chapter: the style of a particular social group (see also Snell-Hornby 1995:24ff.). 

Besides looking for the visible presence of the translator in a target text, either more or less explicitly in interaction with the presence of the original author, stylistic studies of translated texts may provide evidence for the notion mentioned in 1.5 that translation can be seen as a separate literary type with a special language which has its own stylistic characteristics, a text-type distinguishable both from non-translated texts and other types of translation (cf. Venuti 2000:5).

That literary translation has its own stylistic characteristics is also suggested by Malmkjer's use of the term "translational stylistics", a particular type of stylistics which looks at the target text in its relation to its source text. If the target text is co-authored, the translator's voice will be additional to the voices in the source text, and Millán-Varela explicitly relates this additional voice to the Bakhtinian notion of "heteroglossia" (2004:38). There is in this sense more heteroglossia in a translated text than an original, although the normalizing tendency Millán-Varela notices may remove other aspects of heteroglossia, a point also observed by Thomson (2004). It is interesting that Millán-Varela's study suggests an enhancement of at least this literary quality of the original in the translation. I suggest that translated literature, and especially translated poetry, by requiring more work on the part of the reader to create a context, can also intensify poetic elements in the source text. In this sense, stylistic study of translated texts might be expected to find that such texts possess literariness to a high degree.

Baker (2000) suggests that studies of the style of individual translators can be made easier by the use of corpora, such as the Translational English Corpus. But she suggests that it is not necessarily easy to determine the influencing factors, such as the influence of the source language or the cultural and ideological positioning of the translator" (2000:258). This is a very important point, and it applies to some degree to all stylistics: it is one thing to abstract from the text to the choices it embodies and thus to a translator's (or any writer's) state of mind. But it is taking stylistic study one stage further to assess the influences on that state of mind, and at this stage the study becomes far more speculative. One of the difficulties of Critical Discourse Analysis and also of all translation studies which aim to suggest such influences on the translation (e.g. Boase-Beier 2003a) is their very speculative nature. Perhaps it is partly the precariousness of such undertakings which renders the appeal of the universal and its assumed cognitive (and, ultimately. corporeal) basis in studies such as Tabakowska (1993) so strong.

A further important issue connected with the relationship of the style of the target text to that of its source text is the question of domestication and foreignization. Venuti (2000) specifically links the stylistic nature of the target text to political questions, in the sense that for him fluency masks a domestication of the foreign text that is appropriative and potentially imperialistic. putting the foreign to domestic uses which, in British and American cultures, extend the global hegemony of English" (2000:341). He quotes Hilaire Belloc as saying that there should be no foreignizing or we shall lose our tradition of great writing (2000:14).

Venuti (2000:4) is careful to point out that foreignizing, which he traces from Schleiermacher and the German Romantic tradition, is not merely mi-metic of the source text. It does not exhaust itself in "direct" strategies such as Vinay & Darbelnet discuss, like "calque",a strategy "whereby a language borrows an expression form of another" or "oblique" strategies such as using the equivalent form of an onomatopoeic animal noise (1995:31-38). Instead, foreignizing goes beyond literalism to advocate .…. experimentalism: innovative translating that samples the dialects, registers and styles already available in the translating language" (Venuti 2000:341). The result is a style which is heterogeneous and "defamiliarizing" (ibid.). Examples of foreignizing in this sense have much in common, as Venuti's use of the word "defamiliarizing" suggests, with the sort of foregrounding practices discussed by the Russian Formalists and Prague Structuralists as well as many later stylisticians such as van Peer (1986) and Stockwell (2002a) as being characteristic of literary style in particular. Examples of stylistic foregrounding in this sense range from the mild (my "side" example in 3.1) to the striking (Scott's (2000) translations of Baudelaire). Venuti is often misinterpreted in that the perhaps unfortunate term foreignizing" is taken to suggest a translation that is stylistically close to the foreign text rather than a translation written in defamiliarizing language (see, e.g., Eco 2001:22ff.: Shamma 2005).

Gutt makes a similar point when he states that even in a direct translation it is close interpretive resemblance" that is aimed at, and not the reproduction of words, linguistic constructions or textual features" (2000:233); thus Gutt's direct translation is different from Vinay & Darbelnet's. But where Gutt's view is still essentially mimetic, albeit based on a mimeticism in which translation echoes the relationship between style and what it implies, implicates or is a clue to, Venuti's foreignizing has a virtually non-mimetic view of style: a translated text is foreignizing when it proclaims itself a translation by its unfamiliar use of language. The converse, a domesticating translation, aims not to tax the reader by unfaniliarity. In the sense of Relevance Theory, it requires minimal processing and the effects it can achieve will be only those of a text that might as well never have been foreign.

It is important to note, though, that it is as difficult to judge stylistic features in isolation as foreignizing or domesticating as it is to judge them documentary or instrumental. Because foreignizing is not measurable as an imitation of the foreignness of the original in any recognizable sense, but is a measure of foreignness per se, a word like side" in the example in 3.1 could be seen as foreignizing because unusual (to most speakers) or as domesticating (substituting a word used by a young Jewish woman in Lithuania by a woórd which might have been used by a non-Jewish woman from Cheshire because the latter fits a domestic system activated by the similarity of the two women in photographs). And as Venuti (1995:303) says, even a foreignizing translation domesticates insofar as what appears foreign to a target-language audience does so against the background of that particular language and culture.

The terms "foreignizing" and "domesticating" are thus best used to deseribe a particular translation strategy, or the potential effects on readers. However, this is not to say that foreignizing is never mimetic. Both the "side" example and the example of “coffins clapping shut from Dove in 3.1 shows that there is often a strong tendency, perhaps sometimes unconscious, on the part of the translator, to make stylistic choices which echo the original, or even, in the Holzman case, a potential original.

Foreignizing can also be seen, and is seen by Venuti, as a strategy which keeps alive an awareness of difference. Venuti sees in this a contrast to what he likes to call "linguistics-oriented approaches" (2004:148) which in his view favour universalism, though see the discussion of Venuti's criticisms in 1.3 and 1.4. Venuti's view of foreignizing takes up not only the Russian Formalist defamiliarizing and Lecercle's remainder (mentioned in 1.4), but also, more directly, de Man's term "rhetoric", as that which undermines meaning (1983:11), and Spivak's similar term "rhetoricity", the rhetorical nature (in de Man's sense) of a particular language, which works "in the silence between and around words" (2004:370-371).

This subversion of the grammar of a language by its rhetoricity" has, as MacKenzie (2002) points out, something in common with the views of writers such as Pilkington,considered, who work with cognitive pragmatic approaches, that is, the view that language carries meaning beyond what is explicitly stated.For Spivak, it is rhetoricity which subverts grammar much as for a relevance theorist it is implicature which goes beyond and can subvert (as in irony) utterance meaning. This subversion can often be in the attitude of the speaker (see Boase-Beier 2004b), and is one thing a translator takes into account by "attending to the author's stylistic experiments" (Spivak 2004:372) and by not neutralizing irony or other evidence of attitude. Such evidence can vary from the reference to auctions in the title of the scientific article mentioned in 2.2 to the presence of ambiguous structures and textual gaps in poems dealing with the Holocaust (Boase-Beier 2004c). But it remains the translator's decision, often influenced by her or his ethical views, whether and how to put the results of thus attending to stylistic experiments into the target text.

The factors will potentially influence the translator's choices. Corpus studies such as those mentioned by Winters (2004) or Baker (2000) will allow us to go some way towards reconstructing and comparing choices. But in order to explain the link between the translator's reading of source-text style and its realization in the target text, we need to have a clearer view of how style is processed cognitively. 

责任编辑：admin