This view of the textual structures acting upon the reader's mind shares with Reader-Response Theory the insight that “the reader will be able to view things that would never have come into focus as long as [her or] his own habitual dispositions were determining (her or) his orientation" (Iser 1979:35). Mind stylk, in this approach, by revealing a new cognitive state, allows the reader to sec the world in new ways; in Relevance-Theory terms it produces "poetic effects" (Pilkington 2000).The cognitive stylistic view has similarities with thc view of discourse analysts such as Cook, who maintain that in order to explain "the cxtraordinary valuc and pleasure" people find in reading of "worlds and people who do not exist, of emotions and experiences which do not affect us... (of) banal facts which we know already in works which create patterns and play with expectations for no apparent reason at all", we must assume that “literary discourse has an effect on minds, refreshing or changing our mental representations of the world" (Cook 1994:4). For translation this suggests an approach in which the style of the text both conveys and creates a cognitive state. If the translation fails to capture such a cognitive state the target text will have less effect on its readers' minds. So a first step towards preserving the mind-altering effect of literature in translation will be to recognize it. The sort of cognitive content that might be conveyed by the text could uscfully be thought of as being organized into schemata, or "pre-existing knowledge structures" (Verdonk 1999:296), which individual readers have built up through a mixture of possibly innate knowledge such as how the body reacts in certain situations,as well as cultural knowledge and all manner of conventionalized patterns, beliefs and ideologies. The German speaker's knowledge that "nichts gewusst", in the example above, must be followed by haben or hätten is stored in such a (grammatical) schema and s/he also has schemata for more generalized aspects of knowledge such as how to read a poem.

Translated literature in particular could thus be seen to change the mind in three ways:

i) by acquainting us with thoughts or feelings we had not experienced, or reflected on, or known to exist;

ii) by showing us that other people experience them;

iii) by allowing us to experience those thoughts or feelings for ourselves.

In the von Törne example above, we learn the salutary lesson that, however much ambiguities might be seen as arising from the structure of the German language (which allows gaps such as in the example above), we, too, can experience a similarly ambiguous state of mind.

But there is something else that a literary translation must take into account: part of the way literary texts produce such enhancing effects on the reader's cognitive state is by focussing on the search for such effects. This explains why literary texts make that search difficult by stylistic means. Sperber & Wilson (1995:263ff.) express what might be considered a prototypical view when they say that in communication the search for cognitive effects stops when the reader (or hearer) assumes that the speaker's intention has been reached. But discussion underlined what translators, as readers of literature, have always – at least intuitively – known: that we cannot reach such an intention. Because the act of reading focuses on the search, rather than its (unattainable) end, our literary reading is geared towards maximal relevance (Sperber & Wilson 1995:261).If a literary author's aim is to make the reader search, the translation must keep whatever prolongs that search, not just because the enhancement of cognitive state will be greater, but because the search will be better. Translation which gives away too much too easily could be seen to have failed in this particular task. The tendency to make things easy for the reader, noted by Berman (2000). can be seen especially when the original text has several possible interpretations. For example "Birne der Betrachtung" (literally 'pear of observation') in a poem by Goll(Bullock & Schulte 1981:10-11) has, in Schulte's translation, become an observing pear", although the other possibility, that the pear is a focus of observation, would give a completely different reading. In this poem, entitled 'Crucifixion', autumn is represented as the object of crucifixion, so the pear could be an icon for us to look at or it could be a metaphor for observation itself. As each interpretation occurs to the reader, the poem is read differently, different contexts being built up first around the pear as an object to be looked at and secondly as an object symbolizing looking. The translation does not give us the opportunity to create these different contexts.

If "Birne der Betrachtung"contains two implicatures – that the pear represents observation and that it represents the observed - then the translation keeps neither, but uses a phrase which explicitly states the former. To translate both implicatures it is essential for a translator to notice them both, and then use an English phrase, such as 'pear of observation' which also has both.

Studies of the style of translated texts (e.g. Dahlgren 2005) generally see the style as the result of choice, and thus, ultimately, if not explicitly, of a cognitive state driving the choice. But, as Malmkjœr (2004:13) shows, such studies cannot ignore that a translation contains both the author's and the translator's choices. One way of approaching these different choices is to compare corpora of texts from source and translation (Baker 2000). Another way, examined in 5.1 (see also Boase-Beier 2003a), is to see the translator as taking on a particular translating persona, based on an interaction of her/his own view, or cognitive state, with that of an inferred author. Thus two translations of the poem "Two Donkeys'by Christian Morgensterm are seen as embodying two different views, held by two different translators, of the voice of the inferred author.

The notion of mind in the text can, then, be approached from the point of view of the mind constructed as inhering in the source text and affecting the reader in a particular way (as in the von Törne poem), or as being the cognitive state suggested by the interaction of inferred author mind style and translator mind style (as in the Morgenstern versions; see 5.1). But it is possible, and indeed sometimes necessary, to shift from the inferred author's, the reader's, or the translator's mind style to that of a whole group, people or culture (see Millán-Varela 2004). This will happen particularly when texts have multiple or unknown authors. A typical instance of this is the Bible. For Nida & Taber (1974), the style in the Bible was a tool for conveying God's message, and it could be adapted to suit its audience. The voice of God, one assumes, is not a stylistically marked voice. Indeed, it could not be, if we take the very varied provenance of the Bible into account. It serves the message and the sense of author comes from the content of the message rather than the way it is expressed. But Mary Phil Korsak takes a different approach to the voice of this authorless or multi-authored text: her translation "At the Start" (Korsak 1992) is, in the words of A.D. Moody in the Foreword, in "an English modified and refined by the characteristic qualities of the Hebrew original" and it is able to express the point of contact "between the mind of a modern reader and the primal Hebrew mentality"(Korsak 1992: xii-xiit).

The mind in the text can be seen as representing a cognitive state in all its aspects: it is influenced by idcology, it takes a particular attitude, or it embodies a particular feeling. The first, the effect of ideology on mind style, is examined in critical approaches such as that of Fowler who saw meaning as "inseparable from ideology" and dependent on "social structures" (Fowler et al 1997:2).Translation scholars, though not usually taking up Fowler's approach explicitly, have also focussed on the ideological aspects of style (e.g. Prasad 1999). One possible difference between literary and non-literary texts is the greater presence of an informing mind in the former. In the von Törne poem with the missing auxiliary, though a gap cannot tell us what to think, it tells us that we must think, in itself an ideological position. It should not be forgotten,though, that non-literary texts are never fully objective, as analyses caried out by proponents of Critical Discourse Analysis (such as Fairclough 1992), and mentioned in 1.4, have shown. Nevertheless, if the cognitive state informing a text is seen to be more central to literature than to other texts, then a literary translation would need to pay particular attention to this aspect of the text.

责任编辑：admin