In the previous section, we saw that weak implicatures are a way of formalizing the notion of meaning which goes beyond "primary" lexical or
syntactic meaning (Katz 1990). If such implicatures, with the openness that their description as weak entails, serve to provide clues to a state of mind, then the elements of style and in particular those aspects which are consistent enough to constitute a 'mind style' in the sense of Fowler (1977a) will be a starting point for creating a reading which captures something of what writers like Pope and Denham must have meant (see 1.2) by the "spirit" of the original text. This suggests that in many cases a translator will need to start with the style, not the content of a text. As an example, consider the following lines, from the poet Ernst Meister (1979:102):
Die meisten
the most
Sterne
stars
sind leblos.
are lifeless
These are translated (Boase-Beier 2003c:113) as
Most of
my stars
are lifeless.
In this translation, it is the cognitive state suggested by the semi-anagram of the poet's name in 'meisten Sterne'which has led to the inclusion of my" in the English translation. A full explanation of how this poem has been read and translated is in Chapter 5 (see also Boase-Beier 2004a), but it is enough here to note that the final stanza, which these three lines make up, is seen as introducing the way the poet might see himself, his usefulness, his afterlife. It would not be easy to introduce an anagram of the poet's name into an English version, but it is not necessary: what is needed is a reference to the subject, the voice in the poem. That 'my stars'echoes in its sound 'Meister' sufficiently for the reader to infer a connection (as I maintain in the article in question) may not be convincing to every reader, but it will not be disputed that the introduction of the my" in a poem which up to now has only addressed a "you" will be noticed. And it is this link to the cognitive state of the poet which is the decisive factor. This is what "starting with the style" is taken to mean. Meister's name is not actually there, but, for this translator, is implicated. The translator can exercise the freedom to include it or not, or include some other clue to the cognitive state the poem is seen to suggest. It will take a certain amount of reading and re-reading to see 'Meister'in 'my stars'. But this is true of the original, too, because it is the way poetry works: stylistic features demand engagement with the text. This is what I take direct translation to mean, but the distinction between direct and indirect translation does not correlate exactly with the literary / non-literary distinction; a translator will almost certainly want to do justice to the style of a critic such as Wilson Knight (e.g. 1961) or a medical historian such as Roy Porter (e.g. 1997). And the distinction is a question of degree, rather than an absolute opposition. One translation may be more direct (pay more attention to the style) than another, but still both may attempt to give more than just substance. Erkazanci, in an as yet unpublished thesis, explores how Turkish translators, bowing to target culture pressures, fail to do justice to heteroglossia in English texts as diverse as Lady Chatterley's Lover (Lawrence 1961) and Trainspotting (Welsh 1994), but do make some inconsistent attempts to indicate stylistic variety. Such translations are largely indirect, and could therefore be adjudged unsuccessful as examples of literary translation.
As Erkazanci (unpublished) points out, both domesticating and foreignizing translation practices could in fact be seen as instances of direct
translation, because both render the style of a text in one way (to fit target expectations) or another (to violate target expectations), but in neither case is style ignored.
As Sperber & Wilson (1995:38f.) make clear, not all interpretations of a text are available to all readers in all contexts. Assuming that the text will
work subject to relevance (not because the author obeys the relevance constraint consciously but simply because this is the way communication works) then the more literary the text is (in the sense given in 1.6, of being a text whose style carries much of the meaning),the more readers will look for maximal relevance, as opposed to the “optimal relevance"(Sperber & Wilson 1995:57ff.) of non-literary communication situations.