| Summary: | Under a linguistic categorization hypothesis causal connectives are taken as categorization devices. Indeed, corpus studies suggest that connectives strongly specialize in one specific causality category, but also that their use is not restricted to the causality categories they are prototypically associated with. If we assume that the meaning of causal connectives can adequately be described with reference to well-defined conceptual categories – how can we explain that variation in the actual usage of connectives occurs? We focus on volitional causal coherence relations, which count as the prototypical usage context of the Dutch causal connective daarom “that’s why”. Volitional causal relations can alternatively be marked with the connective dus “so” which is prototypically used in epistemic causal relations. Our hypothesis is that volitional causal relations marked with daarom vs. dus systematically differ in terms of subjectivity. We discuss a model of analysis that contains multiple operationalizations of subjectivity and distinguishes between different levels of complexity (sub-clause, clause, and discourse). We find that volitional causal relations with dus contain subjective elements more often than volitional causal relations with daarom. We interpret this patterning within a usage-based theoretical framework, and propose to analyze cases of volitional dus as non-prototypical instantiations of dus’s inherent subjective, prototypically epistemic meaning.
|
|---|
