Past Event

One Repair Strategy or Two?   The Implications of Phonological Persistence 

Languages vary in how disfavoured phonological strings are repaired. Hiatus, for example, may be resolved by processes such as deletion, epenthesis, glide formation, or coalescence. Within a single language it is also not uncommon to countenance multiple resolution strategies for a disfavoured phonological string. Sometimes these different strategies are motivated purely phonologically, as are the three hiatus resolution strategies of coalescence, glide formation and elision in Xhosa (Casali 2011). Other times these differences are morpho-syntactically motivated.

Newell (2008) and Newell and Piggott (2014) argue that the different hiatus resolution strategies in Ojibwe are due to the persistence of phonological structure built on inner derivational cycles. Within a cycle hiatus is resolved by deletion, while hiatus may be resolved by epenthesis if the vowels are interpreted in separate cycles (if not left unresolved). In (2a) the verb and all of its suffixes are interpreted in the same cycle (vP, a cycle within phase theory (Chomsky 1999)), while in (2b) the tense morpheme and verbal root are interpreted in separate cycles (CP and vP, respectively).

This presentation examines what motivates different phonological repair strategies within a single language. It proposes that inter-cyclic repair is cross-linguistically more persistent than intra-cyclic repair. ‘Persistence’ here is the apparent dispreference within a system to delete structural relations that have been built up during a derivation. Phonological rules are sensitive in general to whether a morpheme is undergoing, or has already undergone,

interpretation in the phonological module of the grammar. This sensitivity manifests itself as persistence, defined as follows:

(3) Phonological Persistence (PP): Phonological operations that apply after structure building has occurred cannot be more destructive that operations that apply during structure building.

This preservation of structure in the phonology is not absolute, however, but gradient. I propose the following constraint on Phonological Persistence:

(4) Phonological Subset Generalization (PhSG): If a phonological rule X in a language applies both early (during) and late (after structure building operations), the late application of the rule will target a subset of phonological structure targeted in its early application.

This pattern is seen cross-linguistically. It is seen in (1) and (2) above, where one segment is completely removed from the output within the first phono-syntactic cycle, but the resolution strategies in a second cycle preserve (at least some aspect of) the segments in question. I will demonstrate that this pattern also occurs in hiatus resolution in Berber, Ilokano, and European Portuguese, and NC-cluster resolution in Malagasy, and Acholi among others.

The existence of multiple resolution strategies for a single phenomenon within a single language raises the question of whether the language contains a single repair strategy that applies differently depending on its timing of application, or whether the language countenances two operations to accomplish the same repair. For example, in Ojibwa, is the hiatus resolution rule as in (6) (see data in (2)), or are (6a) and (6b) two separate rules in the language?

(6) Hiatus Resolution: *VV
Sub-clauses: (a) deletion (b) epenthesis

It will be argued that patterns like those seen herein indicate that resolution is accomplished by a single rule that is governed by PP and the PhSG. If this were not the case, we would predict any possible pattern of language internal resolution, including those that contravene (5).

References: 

Casali, R. 2011. Hiatus resolution. In The Blackwell Companion to Phonology. M. van Oostendorp, C. Ewen, E. Hume, K. Rice eds.

Chomsky, N. 1999. Derivation by Phase. MIT Occasional Papers in Linguistics, no. 18, MIT Working Papers in Linguistics. Department of Linguistics and Philosophy, Cambridge, Mass.

Newell, H. 2008. Aspects of the morphology and phonology of phases. PhD dissertation. McGill University

Newell, H. & G. Piggott. 2014. Interactions at the syntax-phonology interface: evidence from Ojibwe. Lingua 150, 332-362.

About the Presenter

Dr. Heather Newell received her PhD from McGill University in 2008, and now teaches at UQAM. Her work focuses on the interface of morphosyntax with phonology, especially on the phonological consequences of phase theory, and on Ojibwe.