An Overview of the Volume

The volume is divided into three sections, according to the methodology and the type of contrastive analysis carried out (cf. Aijmer 2008).

Contrastive Analysis with Parallel Corpora

The first section comprises four papers based on parallel (translated-text) corpora.

Karin Aijmer’s contribution discusses obligation across languages and genres. The starting-point is the observation that in both English and Swedish the meaning of obligation can be expressed by a modal auxiliary. However must and its Swedish cognate maste are not always each other’s translation equivalents reflecting the fact that there are several grammatical and lexical ways to express obligation. The translations of Swedish maste into English showed that maste was frequently translated by the semi-modal have to especially in fiction. If have to and had to were conflated the frequency would be even higher than for must as a translation choice. Need to, should and ought to, on the other hand, were all more frequent in non-fiction than in fiction. In the Swedish translations from English maste was most frequent both in fiction and non-fiction. Other alternatives are fa (‘may’), ska(ll), skulle (‘shall’, ‘should’), behova (‘need’), bor (‘ought to’). Must and have to express different meanings as a translation of maste. Have to, especially when qualified by will, is downtoning and polite, it can have a general or generic meaning and it can indicate negative evaluation. In Swedish ska(ll)/skulle can be used to express power and fa indicates that an action is unwelcome to the hearer. In non-fiction must, have to and need (and their Swedish correspondences) are associated with interactional goals and how these are evaluated as good or bad. By using impersonal structures with a collective we as the grammatical subject or an agentless passive the speaker can get the message across to the hearer with maximum hedging.

The aim of Lieven Buysse’s contribution is to examine the mutual translat- ability of Dutch dus and English so. The corpus used is the Dutch-English component of the Dutch Parallel Corpus (1997-2009). The texts belong to five text types: fictional and non-fictional literature, journalistic texts, instructive texts, administrative texts, and external communication. The corpus has been balanced for text type and for translation direction and amounts in all to five million words. Since it is a bidirectional corpus all the examples of so translated to dus (and dus from Dutch into English) were included as well as ‘back-translations’. The functional ranges of these two discourse markers were shown to be remarkably similar - the polysemies of dus and so overlapped almost completely - but there were significant differences in frequency and distribution, with dus being both more frequent overall and more associated with inference than so, which occurred more typically in resultative contexts. Significant genre effects were found: as well as being unevenly distributed across genres, dus and so also tended to occur with different functions in different genres and were differently distributed according to language. Thus the study not only demonstrates how semantic equivalence does not result in translation equivalence, with only a sixth of source-text dus being translated by so in the corpus, but also how genre constrains the markers differently in the two languages.

Michaela Martinkova and Marketa Janebova wanted to investigate the evidential and epistemic senses of the Czech particle pry by studying the functions of pry reflected in the correspondences in another language. The authors used the English-Czech and Czech-English sections of the Czech National Corpus- InterCorp, which is a multilingual parallel corpus of texts written in 39 different languages with their Czech counterparts. Their study focused on three registers which were represented in the corpus: fiction, journalistic texts and spoken language. The j ournalistic texts are represented by the PressEurope database (2009-2014). The spoken language in InterCorp comes from Proceedings from the European

Parliament and a corpus of Subtitles. Czech as source and as target language were not differentiated in order to obtain a sensible amount of text from the European Parliament. The sub-corpora vary in size and come from different periods. Moreover there were more translations than target texts in the corpus.

The authors found different patterns of pry usage and different frequencies according to genre as well as different patterns of translation. From these translation patterns, and with the great majority of correspondences being evidential, the authors concluded that the epistemic uses of pry are context-bound: the interpretation of the particle as conveying doubt may arise in the context as an inference from the context.

Magdalena Szczyrbak discusses the correspondences between English modal adverbs of certainty and their Polish correspondences in argumentative legal writing. The material used for the study consists of 30 Opinions of Advocated General at the European Court of Justice, issued between 2011 and 2013 comprising about 576,000 words. The data has been drawn from source texts in English and their Polish translations. The English texts were written by a native speaker of English, whereas the translations were made by professionals having Polish as their native language. At the outset the most frequent modal adverbs were identified in the English sub-corpus and then the equivalents in the Polish sub-corpus were determined.

The genre of Opinion was chosen because it was assumed that it would be rich in persuasive devices. Modal adverbs of certainty have been shown in previous studies to be useful rhetorical devices inextricably linked to stance and argumentation. They are for instance used both to foreground and background legal arguments and to demonstrate power and authority. The modal adverbs studied were indeed, necessarily, of course, clearly and obviously and their Polish correspondences. The translations were used to study both the conventional meanings of the adverbs and the ad hoc meanings associated with the particular genre. It is shown that there were noticeable differences between the English adverbs and their Polish correspondences with regard to the degree of persuasiveness and that the author’s presence was less visible in the Polish translations. Omission of the modal adverb in the translation was shown to lessen the rhetorical force of the translated text and its ability to influence the reader.