Pseudo-ABA patterns in pronominal morphology

Abstract

In this paper, I present an analysis of pseudo-ABA patterns of morphology found in pronominal forms. I argue that an analysis that assumes unrestricted phonologically null allomorphy or unrestricted impoverishment overgenerates, allowing all the logically possible patterns of syncretism to appear. An analysis that includes spanned portmanteau exponents generates all and only the attested patterns of syncretism. Pseudo-ABA patterns arise when the complete pronominal tree (the structure for the anaphor) is exponed by a spanned exponent for [a [d]] and an exponent for p. Spanned portmanteau exponents are compatible with an analysis in which a and d are cyclic nodes, and one in which they are not. However, the analysis in which a and d are cyclic nodes is incompatible with another morphological behaviour of pronominal forms, namely variable exponence. To provide a unified analysis of pseudo-ABA patterns and variable exponence, a and d cannot be cyclic nodes.

*ABA pronominal stem allomorphy without containment

“…In the decade and change following Bobaljik's (2012) masterful analysis of comparative suppletion, there has been an efflorescence of work on so-called *ABA environments: pockets of natural language morphology in which non-adjacent categories within a paradigm fail to co-supplete or co-syncretize to the exclusion of intervening categories. *ABA phenomena have been spotted in a panoply of domains, to include pronouns (Ganenkov 2018, Smith et al 2019, Middleton 2021, case (Caha 2008(Caha , 2017McFadden 2018;Zompì 2019;Davis 2021), number (Mare 2021), gender (Gray & Gregor 2019), tense and temporality (Franco 2013, Adamson 2019, possessives (Van Baal & Don 2018), complementizers (Wiland 2018)-and most recently, honorifics (Truong 2022) and fractionals (Sudo & Nevins 2022) In the AAA case, root constancy is maintained throughout the paradigm. In the ABB case, co-suppletion obtains across the two non-positive grades.…”