This paper provides the first detailed analysis of the models for classing nouns found in the early Pali exegetical texts traditionally assigned to the fifth-century monk-scholar Buddhaghosa. It identifies three fourfold models and argues that: (a) each model has a distinctive distribution; (b) the original occurrence in each model serves a specific purpose, which can be identified through studying the relation between these original occurrences and their respective context; (c) Buddhaghosa did not invent these models but repurposed pre-existing models to fit specific exegetical needs; (d) Buddhaghosa's reuse of these inherited models is consistent with his view expressed elsewhere that Pali is a sacred, perfect language; and (e) in developing these models, the unknown authors combined influences from the Pali canon and Sanskrit grammatical sources. Finally, I briefly consider whether studying the models sheds new light on the early history of Pali grammar.