Abstract: In this talk I address three well-known typological generalizations concerning word and morpheme order. In each case, a number of competing hypotheses have been proposed in the literature to explain them, and I present experimental and corpus-based evidence adjudicating amongst them. First, in the domain of noun phrase word order, previous research has argued that innate syntactic representations constrain possible ordering patterns (Cinque 2005, Steedman 2020). I show instead that learned conceptual knowledge and a preference for transparent mappings between levels of representation can explain why certain orders are more frequent than others. Second, in the domain of morpheme order, distributional properties like frequency and parsability (related to the co-occurrence frequency of stems and morphemes) have been argued to be primarily responsible for how multiple morphemes stack onto a stem (e.g., Hay 2001, Hahn et al. 2021). I show that differences in meaning play a crucial role in how nominal morphemes like number and case are ordered, in some cases overriding any effect of frequency. Third, the suffixing preference has long been argued to be driven by universal features of our perceptual system (e.g., Hawkins & Cutler 1988, Hupp et al. 2009). I show instead that a preference for prefixes or suffixes depends on previous language experience, providing indirect support for alternative theories of suffixation as a result of grammaticalization processes.
Zoom Link to Talk: https://yale.zoom.us/j/93677552383