Abstract
The present paper deals with the constructions WILL V and SHALL V from the perspective of distinctive collexeme analysis (DCA), a quantitative corpus-linguistics method developed by Gries and Stefanowitsch (2004a). The data were extracted from the academic subgenre of the Corpus of Contemporary American English (COCA). The primary aim of the paper is to identify the semantic categories of verbs that are statistically attracted to each construction and to examine how they relate to the modal meanings of will and shall. The results show that the two constructions mainly attract the verbs from distinct semantic categories. The WILL V construction is predominantly associated with the verbs expressing cause-effect relationships, enablement and necessity, change and development, and research-related actions. Across these categories, will conveys predictive meaning, presenting developments, outcomes, and research consequences as expected. On the other hand, SHALL V mainly attracts verbs denoting academic commitment, institutional and procedural acts, text structuring and discourse navigation, and acts of definition and delimitation. Within the construction shall typically expresses intention, and, in combination with institutional and procedural verbs, it conveys meanings related to norms, obligations, and formal procedures.