Weather domain in Croatian: a corpus–based overview of precipitation and non–precipitation expressions

Abstract

Th is study investigates weather expressions in Croatian, focusing on precipitation and non–precipitation weather phenomena. It highlights the complexity of the weather domain as described by extant syntactic, semantic and lexical studies showing variation both inter– and intralinguistically, and as the focus of future investigations. Croatian is examined with respect to proposed typologies of weather expressions, built around the notion of coding weather occurrences with the predicate or the argument of the sentence, or both, as well as investigations of lexicalization patterns of weather verbs with regards to Figure, Path and Manner. Though such classifications are based on cross–linguistic comparisons, their applications for systematizing intralinguistic variability are investigated on Croatian data. Since descriptions of weather expressions in Croatian have mostly focused on a subset of the most common weather verbs, verba meteorologica, a corpus–based approach is used to expand the dataset with different types of syntactic structures and predicates used in describing weather events, complemented by lexicographic sources. Weather expressions in Croatian are analyzed according to the semantic properties of dynamic and static weather phenomena, contextual grounding of a weather expression, manner of occurrence and the notions of primary and secondary weather expressions. Based on the overview of lexicographic and corpus data, this paper sketches an outline for future systematic study of the weather domain in Croatian.