How do we learn patterns of alternations like that of the English plural ending seen with cans [khænz] and caps [khæps]? Major interests for phonologists include understanding how speakers internalize the structural relationship between phonological representations. I investigate this question with a primary focus on the acquisition of morphophonology.


My research approaches individual phenomena in the acquisition of phonology from three directions: (a) formal analysis within the framework of generative grammar, (b) experimentation accessing learners’ phonological knowledge, and (c) computational modeling incorporating components of phonological theory.

Research area 1: phonological learning biases

One of my core areas of interest is the study of universal learning biases. In my Dissertation, I propose that child production of inflected forms is modulated by various learning biases in the intermediate stages of morphophonological acquisition. I am currently revising a manuscript reporting an experimental testing of the uniformity bias, and another manuscript on the description of a computational simulation of biased learning process.

My work with Adam Albright of MIT uses an artificial language learning paradigm to explore adults’ learning of phonological alternations. We propose that a bias against alternations, a bias in favor of alternations that target broader classes of segments, and a substantive bias against perceptually salient alternations play a role in learning patterns of alternations. We then raised a question about the source of phonological learning biases- we recently presented the results of experimental and computational test  reporting a source of paradigm uniformity preference. We argue that uniformity bias is innate phonological bias, as opposed to learned preference from learners’ L1 or experimental environments.

The goal of these investigations of the specific nature of learning biases is to understand (a) whether typologically unmarked patterns reflect patterns of biases in phonological acquisition, (b) why learners acquire certain patterns more readily than others, and (c) why language change shows systematic directionalities.

Research area 2: Tone and accent 

My interest in segmental alternations also extends to the investigation of accent and tonal alternations and their interactions. In my articles published in Lingua(Do, Ito, and Kenstowicz, 2014) and Studies in Phonetics, Phonology, and Morphology (Do and Kenstowicz, 2011), We argue that native speakers assign accent based on the action of UG markedness constraints, whose weights are determined by the frequency of the accent patterns in the native lexicon. Another article published in Korean Linguistics(Do, Ito, and Kenstowicz, 2014) proposes that while both segmental and tonal analogical changes are two separate developments, both are motivated by the same grammatical mechanism: selecting the alternant with which learners have the greatest confidence of producing a correct utterance, as predicted by the forms’ morphological base. In a current collaborative project on this topic, I am tracking the development of the segmental variants and their tonal correlates in first language acquisition. This line of research has already produced interesting results, and I believe it will allow us to better understand the role played by segmental and suprasegmental components in children’s acquisition of phonological alternations.

Research area 3: L2 morphological acquisition and processing 

My interest in L1 acquisition extends to L2 acquisition and processing, as well. In a submitted article co-authored with a Georgetown student Yoonsang Song, I investigate late L2 learners’ morphological processing of English derived words to better understand the mental representation of morphologically complex structures in the L2 grammar. I also work with a Georgetown student Kate Riestenberg onL2 learning of tones in Zapotec language.

Research area 4: Sound change 

I also have great interest in sound change. With Elizabeth Zsiga and Jonathan Havenhill at Georgetown,  we questioned how patterns and directionality of sound change are guided by phonetic naturalness. We are currently preparing a manuscript based on our experiment simulating this language contact situation. Related projects on sound change include a collaborative study with Jonathan Havenhill exploring the role of visual cues in sound change. This project is motivated by the fact that the acoustic characteristics of sound change are well documented, while research on articulatory components is comparatively limited. Using video and ultrasound analysis, we are investigating the contribution of lip configuration and tongue position to the Northern Cities Vowel Shift in American English.

In sum, while my research is diverse and wide-ranging, it is always motivated by my desire to understand how child and adult learners internalize morphological relationships between phonological representations, both at segmental and suprasegmental levels, and from both diachronic and synchronic perspectives. By approaching questions on phonology from three angles – generative, experimental, and computational – my work unveils the specific nature of universal learning biases, the role of language-specific experience in language acquisition, and the effects of resources from other linguistic layers on the shaping of our phonological grammars.