This article, based on a survey of scholarship and world-wide-web 'metatexts' (reviews, translators' guidelines, etc.), discusses the options open to translators when translating older literary and religious texts. It is argued that different decisions along a scale ranging from 'hyperarchaisation' to 'violent modernisation' give different deictic signals, which point the reader to different aspects of the temporal and cultural relationship between source and target text. Translators' decisions and readers' interpretations are mediated by cognitive factors: translators may be more or less skilled in producing certain target-text styles or conveying certain signals, and readers more or less able or willing to process certain styles. They are also mediated by translational and literary norms, though these may vary across time, between cultures and between interest-groups. In the recent English-reading world, the interaction of pre-modern, modernist and post-modern norms can give different attitudes towards the use of modernising and archaising techniques: archaisation in poetry translation, for example, tends to be seen as hackneyed 'Victorian' translationese rather than as signalling the source text's specific historicity, whereas archaisation in religious translation can be seen as integrating the text into a liturgical tradition.