I am primarily interested in the reconstruction of Proto-Indo-European from a morphological perspective (both nominal and verbal), though I also investigate how sound changes may occur on the basis of acoustic and perceptual phonetics.

Case collapse across Indo-European: In joint work with Tamisha L. Tan, I investigate patterns in the type of case used to mark participial Absolute Constructions over time across the Indo-European family, observing a diachronic tendency for oblique/inherent cases (e.g. the ablative or instrumental) to be replaced by structural cases (e.g. the accusative) or overt prepositions.  From a formal and comparative perspective, we connect these developments to phonological erosion and syntactic integration/restructuring, drawing on diagnostics such as co-reference, negation, and cliticization to present an analysis in support of disjunctive and configurational approaches to case assignment (Baker 2015).

•  Diachronic patterns of case-marking across Indo-European absolute constructions. 2025. (contact for ms.)

Diphthongization in Greek: Several Ancient Greek dialects, Lesbian, Cyrenean, and crucially Elean, show diphthongal outcomes for the second compensatory lengthening, each with different distributions. I argue that these diphthongs arise due to perceptual factors such as formant transitions that lead listeners to misanalyze the nasal as a nasalized glide. This glide then diphthongizes before losing its nasalization. I also account for the odd distribution of this sequence in Elean based off an analogical restoration.

• A 'Forma(n)tive' Misperception: The rise of diphthongization in Elean. 2021.  East Coast Indo-European Conference XL (ECIEC). 

The outcome of *-os in Slavic: Slavic famously has two outcomes for the sequence of word-final *-os, -ъ and -o, with Old Novgorod showing an additional -e outcome.  Recent accounts (Olander 2012) have argued in favor of a sound change involving an intermediate stage where pre-PS *a became *ə before (R)s#. However, I propose that this phonological change must be further constrained in order to properly account for the attested distributions.

Many languages show declension or conjugation class phenomena where allomorphy is conditioned by seemingly arbitrary (neither syntactic nor phonological) features. Some models of morphology (e.g. Distributed Morphology) explain such aberrant class types via memorized features known as diacritics, whereas other refer to such idiosyncratic distributions as morphomic. A key strand of my research is in doing critical framework comparison through investigating the empirical evidence for morphology as an independent level of representation, drawing on both diachronic and synchronic data.

Morphomes and Paradigms: In exploring several case studies of the diachronic innovation and loss of putative morphomic distributions across a range of Indo-European languages, this dissertation investigates the empirical and theoretical factors influencing the actuation and directionality of analogy from the perspective of paradigm leveling. In addition to providing a novel typology of non-root alternations, it argues for the importance of natural morphosyntactic classes (as defined by features) and frequency in accounting for biases in analogical change. 

• A Morphological Typology of Non-Root Alternations: Invasion, Suppletion, and Allomorphy. 2023. Doctoral Dissertation, Harvard University.

Nanosyntax and Allomorphy:  Recent work in Nanosyntax (Caha et al. 2019, Caha 2020) introduces an account for diacritic phenomena based off of the assumption that lexical entries contain syntactic structure. Different allomorphs are conditioned by differently sized roots, removing the need for diacritics. I argue that this theory is too restrictive and fails to account for the full breadth of cross-linguistic variation, primarily using data from plurals in German, Welsh, and Albanian.

• The Limits of Root-Controlled Allomorphy. 2021. North East Linguistics Society (NELS 52).

• On Non-Adjacent Allomorphy: Evaluating Predictions from Nanosyntax. 2021. Syntax, Phonology and Language Analysis 14 (SinFonIJA 14).

Based off of elicitation work, I have also worked on impersonal passive constructions in Sakha (Yakut), a Siberian Turkic language spoken in Yakutia, Russia. 

Impersonal Passives in Sakha : Building off of Legate & Akkuş (2017), Tamisha L. Tan and I argue that the passive morpheme in Sakha -IlIn- is the fossilized outcome of a doubled 'passive of passive' construction that is still found in other Turkic languages. We show that this morpheme serves not only as a passive morpheme in the traditional sense but also as an impersonal, using several syntactic diagnostics. In addition, we present an account that has important implications for the reconstruction of the Turkic voice system.

•  The Double Duty of the Sakha "Passive". 2021. Tu+5 Workshop. [handout].

• An Impersonal Look at Sakha Passives. LSA 94th Annual Meeting.