摘要
The motivation for this study is two-pronged. A number of studies on modality or conditionals have put forward the claim that conditionality, and conditionals, have a special relationship with modality. However, this claim has not been empirically investigated, or established quantitatively, nor has the nature of this relationship been examined. Furthermore, although existing classifications of conditionals take account of modal marking, they do not do so consistently, nor do they take account of all modality types. This study, therefore, examines the relation of conditionals to modality using the written BNC, and develops two complementary metrics for the extent of modal marking (modal load): modal density and modalisation spread. It establishes the modal load in a random sample of if-conditionals, and compares it to that of written English, as well as other conditionals, and a number of non-conditional bi-partite constructions. The examination also compares the modal load in different types of if-conditionals, and their two parts (protasis and apodosis), and motivates the development of a typology of ifconditionals. Finally, the study examines the modal nature of if-conditionals in light of two current theories, Lexical Grammar and Construction Grammar. The analysis confirms, and provides a quantitative measure of, the connection between conditionals and modality. It also supports a two-dimensional classification of if-conditionals which takes into account the interaction of the modal function of the conditional and the nature of link holding between its two parts, while providing frequency information on the types established. The analysis also indicates that there is a correlation between types of conditionals and modal load patterns in their respective protases and apodoses. In light of the above, the study proposes a conception of conditional constructions as environments of indeterminacy, drawing on mental space theory and quantum mechanics, and proposes an expansion of the notion of construction family.