I view the study of suprasegmental phonology as essential to understanding the interrelation between modules of grammar, by virtue of their long-distance nature. In the case of nasality, the focus of my dissertation and much of my research to date, its role in the phonology is mediated by articulatory factors, which often result in mismatches between phonetic, phonological and morphological boundaries.
Nasalization at the phonetics-phonology interface
Atchan: In my dissertation, I examine the typologically unusual system of nasal contrast in Atchan: vowels contrast for nasality, but there are no underlying nasal consonants. Surface nasal consonants are predictable allophones of sonorants -- a natural class which comprises implosives, approximants, and /h/. Though implosives have sometimes been considered obstruents, I show in collaborative work with Alexandra Pfiffner and Lindsay Hatch that the Atchan 'implosive' is characterized by a lack of heightened pressure in the oral cavity, rather than true rarefaction, and that it patterns phonologically with sonorants, not obstruents. Nasality in Atchan is a syllable-level, rather than segmental, feature, and its surface realization is contingent on the sonority of segments within the syllable. I also contend that a distinct rightward nasalization process in the language is driven by articulatory pressures on coarticulation resulting from a nasal 'appendix' which extends after nasal vowels. My work contributes novel empirical data to the discussion of the typology of nasal contrast, as I demonstrate that the system of nasal contrast in Atchan is best treated as one in which nasality is a property of a prosodic, not segmental, unit, and I suggest that similar analyses may be appropriate for other, perhaps all, languages which have been described as having a contrast for nasality in vowels but not consonants.
Tupian: In collaborative work with Myriam Lapierre at the University of Washington, I have conducted a typological survey of nasal harmony systems across the Tupian family in South America. We propose an Agreement by Correspondence analysis incorporating the use of Q- Theoretic representations, which decompose the segment into temporally ordered subsegments, allowing us to model assimilation which targets edges of segments. Specifically, we introduce a phonotactic constraint which ensures surface correspondence between adjacent subsegments across segment boundaries, requiring agreement in the feature [±nasal]. Our analysis accurately predicts the typology of nasal and oral assimilation processes across Tupian.
Nasalization at the phonology-morphosyntax interface
Atchan: Atchan exhibits what appears on the surface to be nasal harmony in which the affected domain differs based on the identity of the trigger. After some nasal vowel triggers, just the following segment is nasalized, whereas after other triggers, a much larger amount of structure is affected. This pattern poses a challenge for traditional phonological theories, which assume that all harmony triggers should behave identically, and that nasalization is always driven by local coarticulation. In my dissertation, I examine distinct places in the grammar of Atchan which show morpheme-specific nasalization patterns, and argue that the attested differences in surface phonological output in Atchan are actually tied to differences in underlying \textit{morphosyntactic} structure, rather than differences in phonological representation. Nasalization represents the exponence of singular pronominal subject features, expressed on every verbal head within a particular morphosyntactic domain. I connect my analysis of the grammatical role of nasalization in Atchan to larger questions about possible domains of phonological processes, proposing a historical pathway from automatic processes driven by coarticulation, which are phonologically delimited, to phonologized processes operating over morphosyntactic constituents.
Guaraní: Informed by initial primary data collection with two native speakers of Paraguayan Guaraní, I conducted a corpus study of sociolinguistic interviews in order to investigate factors which influence the actual application rate of nasal harmony. I found that several factors significantly affect application rate, including one morphosyntactic factor in particular: proclitics harmonize at a significantly lower rate than prefixes, despite the two types of morphemes competing for the same slot in the morphology. This demonstrates that morphosyntax can be a significant factor conditioning variation in harmony.
Tone at the phonology-morphosyntax interface
Atchan: In a paper submitted to the proceedings of ACAL 55, I show that an unexpected High tone in Atchan appears if and only if a verb (i) is not marked for perfective aspect, and (ii) immediately precedes a phonologically overt complement. This generalization applies to all multisyllabic verbs in the language, and to a subset of verbs with initial voiced obstruents. I argue that this 'floating' High tone is part of the lexical tone melody of the verb, and that its surface realization is contingent on the mapping of syntactic structure to prosodic phrasing. I show that referencing the mapping from syntax to prosody is crucial in order to account for asymmetries in lexical High-tone association.
Eastern Kru: In a paper submitted to the proceedings of ACAL 56, I survey the lexical tone inventories of 11 Eastern Kru language varieties, finding that some languages are in the process of reducing the complexity of tone melodies through the collapse of distinct melodies into inflectional tone classes. In a manuscript in prep, I survey grammatical tone marking across 24 Eastern Kru varieties, finding that tonal imperfective and negative marking have developed over time from the phonological erosion of segmental morphs which have been reduced to just tone, albeit with different results in different varieties. This work contributes to the small but growing literature on the diachrony of tone, and additionally represents a piece of my larger research program on the typology and historical evolution of suprasegmental phonological features.