As translators we can study the style of a text and on this basis make a good case for what a text means, but we cannot guarantee, even if we wanted to, that we know what the author meant.

Understanding any act of communication involves a certain degree of pretence about having achieved such certainty. When we respond to a colleague's statement that it is too hot by opening a window, or a friend's remark that it is late by going home, we are making an assumption that we know what is meant. The colleague could be suggesting that he has a high temperature and needs to see a doctor, or merely making an observation. The friend could be suggesting we decide an important issue quickly, or remarking that the evenings are getting longer. We arrive at the meaning for which we have most evidence and act upon it, while being aware, if we think about it, that there are other possible interpretations. We are used to the fact that in some situations statements are more difficult to interpret than in others; absence of body language in anemail can make us especially careful to question the writer as to her or his intentions.

Translation is no different from any of these other communicative situations in that we have to arrive at an interpretation for which there appears to be a reasonable amount of evidence. The answer to the sixth question about balancing ignorance with a need to act is thus quite simple: if we have enough evidence for interpreting a text, we can act upon it, even while being aware that any such interpretation is not final, absolute or exclusive. But in literary translation there are two complicating factors. The first is that literary texts by their very nature allow multiple interpretations, a point made by critics as diverse as Empson (1930), Iser (1979) and Downes (1993), and mentioned many times in this book. The second is that translation cannot end with reading. While the critic might be content to note the indeterminacy of literary meaning (to a greater or lesser degree, depending upon her or his point of view) a translator has to go on to produce a target text which will be seen by its readers to be a reasonable rendering of a reasonable number of aspects of the source text.

Much of what gives a literary text its special literary character lies in its style. It is the style which allows the text to be open-ended and ambiguous (as poststructuralist and postmodernist views suggest it is), to invite the involvement of the reader (as in Reader-Response Theory), to contain implicatures which the reader can base inferences upon (as in Relevance Theory), to give expression to feelings and attitudes and to do all this by virtue of its relationship with meaning. A cognitive stylistic view of translation suggests that as readers we see style as a reflection of mind, and attempt to grasp that mind in reading and to recreate it in translation.