Abstract
Pedro de Cárceres, a Spanish friar from New Spain (Mexico), wrote the first grammar of Otomi (Oto-Pamean, Oto-Manguean) in 1580. The stage of the Otomi language reflected in Cárceres is called 'Colonial Otomi' (sixteenth and seventeenth centuries). Later, dialectal diversification gave rise to a family of different languages. Cárceres describes Colonial Otomi as having two verb conjugations. However, modern languages displaying great morphological innovation (e.g. Mezquital Otomi) show no trace of such conjugations. This led some authors to discredit Cárceres' work by portraying it as confusing and inaccurate. In this article, I show that the analysis by Cárceres is fundamentally right, and I propose that many of the inconsistencies in his work can be solved by positing three conjugations instead of two, especially in the light of the evidence from modern languages like Tilapa Otomi, which has preserved three such conjugations. © The Philological Society 2011.