Assimilation, dissimilation, and surface correspondence in Sundanese

Abstract

Much recent work has approached long-distance agreement effects using the notion of correspondence between surface segments, driven by relative phonological similarity. This theory of correspondence also has consequences for dissimilatory interactions. Sundanese exhibits a complex [r]∼[l] alternation, which may arise by assimilation or dissimilation. This alternation is analyzed as the result of constraints on surface correspondence, which give rise to both assimilation and dissimilation in complementary distribution, with the choice between them determined by the structural configuration of the interacting liquids. Harmony occurs where surface correspondence between liquids is permitted; dissimilation occurs where such correspondence is prohibited. Dissimilation is argued to emerge from the interaction of constraints that require similar consonants to correspond, and constraints that limit this correspondence, rather than being an effect of anti-similarity constraints.

Long-Distance Phonotactics as Tier-Based Strictly 2-Local Languages

“…While there exist many proposals for constraintbased analyses of consonant harmony as well as dissimilation, we focus here on the predictions of the Agreement by Correspondence framework (ABC; Rose & Walker 2004/2010a. This approach has seen relative success in accounting for the typology of consonant harmony, and has more recently been extended to analyses of long-distance consonant dissimilation (Bennett 2013(Bennett , 2015; see §3.2 below). The ABC framework posits constraints that require pairs of segments to enter into a surface correspondence relation if they surpass some similarity threshold, due to sharing a certain set of feature values (e.g.…”

“…The only case that exhibits anything resembling a strictly beyond-transvocalic pattern, Sundanese rhotic dissimilation (Bennett 2015), is replete with other complications which make it far less persuasive as a test case (infixing morphology, co-existence with lateral harmony, sensitivity to stem-initial vs. non-stem-initial position, root vs. affix affiliation and onset vs. coda status). Furthermore, artificial language learning experiments show that when exposed to a strictly beyondtransvocalic dependency, English adults fail to learn such a pattern and tend to instead interpret it as unbounded (Hansson & McMullin 2014); importantly, this is true for harmony and dissimilation alike, even though the latter is predicted to be possible by the ABC model.…”

Kashaya [asp] assimilation and dissimilation by correspondence

“…This pattern is "robustly attested" crosslinguistically (Bennett 2013), including the famous example of Grassmann's Law in Greek and Sanskrit.…”

“…An aspirated stop in a Kashaya prefix becomes a plain voiceless stop by dissimilation when the root begins with an aspirate; but it also assimilates to an unaspirated stop in the same position. The OCP cannot help with assimilation, so I propose a unified analysis of the two phenomena within Agreement by Correspondence (Rose & Walker 2004, Bennett 2013. Unlike somewhat similar cases in the literature with both assimilation and dissimilation, here there is no structural difference.…”

“…A reason that the ABC theory is especially promising for Kashaya is that it can also be used to generate dissimilation (Bennett 2013, 2015, Bennett & Rose 2016, Ozburn & Kochetov 2013. In analyzing dissimilation, correspondence across some boundary is penalized by a CC·EDGE constraint; in the present case, it is the prefix-root boundary, but for other languages a prosodic edge such as the syllable can be used for morpheme-internal dissimilation.…”