The notion of mind as that which mediates between the world and the text has always been important both for stylistics and translation studies, and has been one of the aspects considered to a greater or lesser extent whenever style has been taken into account in translation. This is particularly the case for literary translation. If literary texts are indeed "a product of the mind" (Graham 1992:xiv), then we would expect to find traces of the mind of the originator in the product, that is, in the style of the source text. Schleiermacher, in the early part of the nineteenth century, saw language as both a system of

common knowledge and a reflection of a speaker's mind (Schleiermacher 1938; Mueller-Vollmer 1985:11; MacKenzie 2002:6). But there is much in the mind of a writer which remains implicit and the translator needs to pay attention to such implicit meanings. Schleiermacher's view is that translation involves either a movement of the reader towards the writer, or of the writer towards the reader (Schulte & Biguenet 1992:42). It is significant that the meeting is of the author,and not the text, with the reader; the concern is with the author's "peculiar way of thinking and feeling" (Schulte & Biguenet 1992:39), a view also held by Spitzer, who said that style "reveals the soul of the author" (1948:15).

But there is no consensus about how the mind inhabits the text, just as there is no consensus about how the mind inhabits the body (Mithen 2003:31ff.: Sheldrake 2003:1-16). Where the notion of "mind style" has been used in approaching translation, it is often Fowler who is cited as a source.

Thus mind style is usually seen as "any distinctive linguistic presentation of an individual mental self" (1977a:103). But Fowler was not entirely consistent in his use of the term, at one point calling it "an impression of a world-view" (1977a:73). This formulation suggests a mind style is not something located in the linguistic structures of the text but something more tenuous: an "impression", presumably in the mind of the reader. Semino & Swindlehurst distinguish mind style from point of view; the latter is the perspective from which the fictional world is presented"but mind style appears to be an attribute of mind: "the way in which the fictional world is perceived" (1996:145). In a recent article (Boase-Beier 2003a) I distinguish mind style as a textual feature from the corresponding cognitive state which can be attributed to it.

It is important to remember at this point that, at least since Wimsatt & Beardsley (in Wimsatt 1954) and other New Critics rejected the emphasis on authorial intention, there has been a realization that we cannot identify the mind in (or behind) the text unequivocally with that of the author (see also Fowler 1977:36). As point (vi) above suggests, the translator might reconstruct an authorial voice, but still be aware that it is a reconstruction and that it is not identical with the real author's voice. Because Tennyson wrote in the voice of a suicidal murderer in 'Maud' (1894), it does not mean this was Tennyson's own situation, or even inclination, in spite of the view of some of his contemporary critics (Thwaite 1996:322). The concept of "authorial voice" is useful for the translator because it allows for the perception and creation of a coherent style, without having to claim unequivocally that this is the voice of the original author; the notion of inferred author (or character. or narrator) suggests that the burden of discovery lies with the reader or translator. But Reader-Response Theory in particular has been careful to point to the reciprocal nature of inferring through reading and thereby attributing implication to the writer: Wolf (1971) notes that an author might have a particular readership in mind, and so mind style in the text might be seen as what Iser (1979:34) calls "a textual structure anticipating the presence of such an recipient", the implied reader" (Fowler 1977b:33), as seen from the reader's own perspective. Stockwell (2002b), who does not actually use the term mind style", speaks of the cognitive stance which the reader takes up, guided by the changing deictic position reflected in the text.

In a recent article (2004b),I suggest that one way a reader is thus guided is by the presence of actual gaps in the text that have to be filled. The example I use is this sentence from a poem by von Törne:

Ich lese in der Zcitung, dass dic Mörder /von Mord und Totschlag nichts gewusst--.

                                                                                           (von Törne 1981:56)

The sentence means something like 'T read (present tense) in the paper, that the murderers .... known nothing of murder and manslaughter*. The dashes indicate a missing auxiliary verb in the German. What the translator needs to be aware of is that the German reader has a cognitive schema for structures such as "Ich lese, dass sie nichts gewusst (aux)" which clearly suggests there should be an auxiliary at the end of the sentence. This auxiliary could be understood as cither indicative haben ("have") or subjunctive hätten ("are said to have/might have").The immediate linguistic context implies both options, and this causes the reader to imaginc both options, with their attendant consequences. For the translator, the presence of the gap in the original is likely to focus the attention on an attempt to keep open these two very different possible ways of thinking: translating that which goes beyond the words on the page involves finding a way to reproduce the gap in English, so that the same options (as far as possible) are open to the English reader.