Payne’s rule and the late *o > a shift in Mojeño Ignaciano

Abstract

Ramirez & França’s (2019) claim that a change *a > o took place multiple times and without any discernible conditioning factor - which the authors present as an explicit counterproposal to a series of claims made in Carvalho & Rose (2018) - is methodologically and empirically flawed. We show here that the supposed evidence from loanwords is artifactual, that the comparison between 18th Guaná and modern Terena rests on an arbitrarily selective and, ultimately, misleading treatment of the relevant sources, and that the claimed ancestor-descendant relationship between Old Mojeño and Trinitario is at odds with other well-established claims about the history of Mojeño dialects, in addition to being an unnecessary and extraneous assumption. Moreover, a lexical stratification of the Mojeño lexicon in terms of basic and less-basic strata shows that the main correspondence favored by Ramirez & França (2019) as the reflex of Proto-Arawakan *a in Mojeño is essentially restricted to nonbasic vocabulary, a finding that vindicates Carvalho & Rose’s (2018) interpretation of this pattern as reflecting dialect borrowing or diffusion.