dc.description.abstract | The present study investigates the passive and related constructions in English and Swedish. It is a bi-directional study that uses empirical fiction and non-fiction material in the form of original texts and their translations from the English-Swedish Parallel Corpus. Using a contrastive research methodology, the aims are to characterise the passive and its use, and to chart the relationships between the passive and related constructions in a network.
The first part of the investigation is based on original texts and shows that the passive is used with varying frequencies in English and Swedish in the text types investigated. It is most frequent in Swedish non-fiction data, followed by English non-fiction, English fiction and, lastly, least frequent in Swedish fiction. The be¬-passive and the s-passive are the dominant and unmarked forms, whilst the get-passive, the bli-passive and the vara-passive display marginal use.
In the second part, the contrastive analysis focuses on non-congruent translation correspondences, i.e. those that were translated into a non-passive construction. The results show intricate cross-linguistic relationships, where semantic properties such as animacy, participant role, and degree of agentivity and transitivity are central features. The linear organisation of information in terms of theme-rheme structure was found to be a determining factor and is extensively discussed. The networks display a wide variety of clause-structural changes along with changes in lexis. Some strategies, such as the Intransitive construction and the Noun construction, turn out to be strong in both languages, whereas other strategies play a more marginal role. Several strategies are characterised by reduction in transitivity in comparison with the passive, which raises questions about the nature of the passive and about the way it is used in the languages. Another main finding is the overall tendency to retain the thematic structure of the source text in translation, which suggests that textual concerns commonly override syntactic ones when they do not correlate. | sv |