Heidi Harley, Mercedes Tubino, Jason D. Haugen
To appear in The morphosyntax-phonology connection: Locality and directionality at the interface, edited by Vera Gribanova and Stephanie Shih. Oxford: OUP.
Publication year: 2016

In this paper, we present further evidence from the Uto-Aztecan language Hiaki that, although it may appear at first to be a counterexample to the very notion of locality of suppletion, it actually conforms to the hypothesis that suppletion is locally triggered, as suggested by both Embick (2010) and Bobaljik (2012). We show that number suppletion in Hiaki is consistently local, since it invariably occurs within the same (Root) phase that contains both the suppletion target (the verbal Root) and the trigger (its complement). The configuration between trigger and target of suppletion which we argue for demonstrates that locality is relevant to the Hiaki case. Our argument is robustly supported by the syntactic incompatibility between intransitive suppletive verbs and high applicatives found in Hiaki. This phenomenon, consistent with Pylkkänen (2002, 2008) and first discussed in Harley, Tubino-Blanco & Haugen (2009), is highly relevant to our argument inasmuch as it suggests that intransitive suppletive verbs must be unaccusative. Structurally speaking, this entails that the suppletive verb and its argument (the suppletion trigger) are sisters, thus occurring within the same phase (relevant for Embick’s notion of locality) and, indeed within the same XP (relevant for Bobaljik’s notion of locality). Further evidence that the incompatibility is syntactic rather than pragmatic in nature is the fact that the restriction applies only in the applicative construction, not with the functionally equivalent postpositional benefactive adjuncts, since postpositional benefactees are perfectly possible in identical applicative-less constructions involving intransitive suppletive verbs.