Transparadigmatic Output-Output Correspondence: A Case Study from Ese Ejja

Abstract

One prominent theory capturing unexpected phonological similarity across morphologically-related forms is Output-Output Correspondence (OO-Corr – Benua 1997; Burzio 1998; a.o.). Classic OO-Corr involves correspondence between an output [x-y] and a subconstituent base [x], e.g. a derived stem with a bare root [√-deriv]x↔ [√]x (párent-hood ↔ párent). Further, Paradigmatic OO-Corr (Pa-OO-C) involves correspondence between an output [x-a] and a base [x-b], where the affixes [a] and [b] share a morphosyntactic feature [+f] which places them together in a morphological paradigm (McCarthy 2005, Hall & Scott 2007), e.g. [√-infl[1sg past]]x↔ [√-infl[1sg pres]]x. In contrast, this paper proposes a novel type of OO-Corr termed Transparadigmatic Output-Output Correspondence (Tr-OO-C), involving correspondence between an output [x-y-z] and a base [x-z]. Under Tr-OO-C, morphologically-related forms share the same root as well as share the same outer morphology, but differ in terms of an inner morphology, e.g. [√-deriv-infl]x↔ [√-infl]x. I argue that Ese Ejja (Takanan: Bolivia) is an example of Tr-OO-C, wherein there is unexpected stress uniformity between inflectional and derivational verbal forms. I argue that derivational forms are in correspondence with an inflectional base enforced by highly ranked constraints Id-BO(stress), Corr-BO(root), and Corr-BO(infl), modeled after Agreement-By-Correspondence (Rose & Walker 2004). In total, Ese Ejja prosody is a crucial case study for OO-C and adds to the body of evidence showing how the phonology of words can be shaped by morphologically-related forms.