Promiscuous Paradigms and the Morphologically Conditioned “Ergative Split” in Texistepec Popoluca (Zoquean)

Abstract

0. Introduction Many authors have struggled to capture the relationship between the various phenomena to which the label “ergativity” is applied. Languages can conflate transitive objects with intransitive subjects, to the exclusion of transitive subjects, at several different levels: syntactic structure, morphological case marking, and verbal agreement systems (Dixon 1994). While some of these patterns may partially overlap in a single language, the overlap is never complete—no language seems to be 100 percent ergative, by any definition (Dixon 1977, 1994). The diversity of these patterns both within and across languages has challenged efforts to define ergativity in a way that is both informative and restrictive. I argue against the assumption that ergative patterns share some underlying syntactic commonality, based on evidence that, in verbal agreement systems, the source of “ergativity” or “split ergativity” may originate in the morpho-phonology, rather than the assignment of Case in the syntax. This paper advocates a position first adopted by Woolford (1999), that there are two distinct types of ergative agreement. One type is parasitic on Case, typically involving agreement only with Nominative (a.k.a. “Absolutive”) arguments, as in Hindi. A second type occurs in languages with no ergative case morphology on nominals, and crucially does not depend on the assignment of Ergative Case in the syntax (Woolford 1999). I argue that the second type is just one of many examples of phonology and morphology “intrusively” affecting the choice between syntactically distinct agreement paradigms. In support of the distinction between ergative agreement systems that are based on Case and those based on morphological paradigm selection, I present