Phonosemantic conditioning in Kirundi loanword noun class allocation

This paper examines the phonological and semantic factors in the noun class allocation of loanwords in Kirundi. I provide a formal account of the criteria used in determining the noun class of loanwords during the lexicalization process. The data indicates an interaction between phonological and semantic factors. I formalize the relevant factors as Optimality Theory constraints and describe a ranking capable of predicting the noun class allocation of novel loanwords in Kirundi. The analysis taken here calls for greater unity between the disciplines of semantics and phonology, traditionally taken to be distinct save for certain phenomena, such as prosody.