This paper presents corpus-elicited contrastive data relevant to the inter- and intralanguage study of cleft constructions. It is shown that even in closely related languages like English and Swedish, the frequency and distribution of different types of cleft constructions vary greatly. Previous approaches to the study of clefts in English have attempted to derive distributional differences between types of clefts from their syntactic, semantic, or informational properties, or from processing constraints. However, these analyses do not easily carry over to the contrastive data presented here. In particular, attempts to account for the distribution of different types of cleft in English in terms of independent factors cannot account for the correspondences found in translated texts. The aims of this paper are (1) to present a methodology for establishing differences between languages based on the study of authentic texts, (2) to show that closely related languages like English and Swedish differ in the frequency and distribution of different types of cleft, and (3) to propose a parameter along which the function of marked word-order constructions may vary across languages.