Clefts in English and Swedish: A contrastive study of IT-clefts and WH-clefts in original texts and translations

Abstract: This study investigates the use of cleft constructions in English and Swedish on the basis of a bidirectional translation corpus consisting of original English and Swedish texts and their translations into the other language. This design minimizes the problems inherent in corpora of original texts alone, viz. that no pair of texts is perfectly matched for content, as well as translation corpora, viz. possible interference from the source language on the target language. With the current corpus design, results obtained from the comparison of original texts can be corroborated in the translation data. The study compares three types of clefts in the two languages: IT-clefts, (basic) WH-clefts and reversed WH-clefts. The main observation is that clefts of different types vary in frequency in English and Swedish. Thus, IT-clefts are more common in Swedish, whereas basic and reversed WH-clefts are more common in English. This tendency is evident in the original texts in the corpus and is corroborated in the translations by the proportions of ‘congruent’ (same-type) translations used by professional translators. A second result is that clefts of different types behave differently with respect to the types of non-congruent translations used by the translators. For example, Swedish IT-clefts are typically rendered by non-congruent translations that are ‘unmarked’ syntactically, i.e. ordinary SVO declaratives. English reversed WH-clefts, on the other hand, often receive Swedish IT-clefts translations or translations where the cleft constitutent in the original occurs in a ‘fronted’ position. Clefts introduced in the course of translation show the same tendencies, sometimes more clearly. Different types of explanations for the tendencies are proposed. A fundamental difference between English and Swedish seems to be that Swedish tolerates syntactic operations like clefting and fronting to a higher extent than English in order to create ‘coherence’ in texts. Thus, English IT-clefts appear to require ‘contrastive’ contexts to a higher extent than their Swedish counterparts. The difference is also evident in that Swedish permits anaphoric and pronominal elements as cleft constituents in IT-clefts to a higher extent than English. A second, functional, explanation of the distributional differences is to be found in the way different syntactic constructions are conventionally used to perform different discourse functions. Thus, Swedish IT-clefts and fronting structures can be understood to perform functions similar to those performed by English reversed WH-clefts. Building on previous work on the relation between information structure and discourse function, the hypothesis is advanced that the different distributions of clefts in the two languages follow in part from differences in the mapping between types of information structure and discourse functions.