The Building Blocks of Grunts: An Application of the Compositional Model on Brabantian Dutch Conversational Grunts

Abstract

Despite their ostensible ubiquity, conversational grunts have largely defied effective categorisation, which is why Ward (2006) developed his Compositional Model (CM) for American English (AmE). Although Ward (2006) presents suggestive evidence for the applicability and veracity of his model, further research is still needed to determine whether his evidence is robust. In this study, Ward’s (2006) CM is applied to Brabantian Dutch to examine whether its applicability extends beyond AmE. The Corpus Gesproken Nederlands (CGN) was used for this study. Data was gathered by listening to approximately 82 minutes of spontaneous conversations between at least two Brabantian Dutch speakers and subsequently annotating the various (non-)lexical conversational items through a bottom-up process. The annotated items were then compartmentalised into hypothetical grunt components and at least one hypothetical meaning was formulated for each component through a bottom-up process. The meanings were then evaluated by testing whether the proposed meanings for each component were upheld in each annotated instance of the components. The results suggest that the CM is applicable to Brabantian Dutch as well, since 7 out of the 9 tested components seemed to have adequate explanatory power. Furthermore, evidence is found to suggest that AmE and Brabantian Dutch share non-lexical components, because their accuracy scores are comparable (AmE: [m] = 95%, /o/ = 94%, dental click = 77%; Brabantian Dutch: [m] = 94%, /o/ = 94%, [|] = 76%). Notably, the existence and meaning of the word-final [ʔ] in Brabantian Dutch is described in this study, despite a surprising lack of research on this consonant in this morphological context in Dutch.