Research

L1 and L2 processing of empty subjects in control constructions

When processing English control sentences such as 'John promised Mary __ to wash,' comprehenders immediately utilize control information from the matrix verb (promise) to associate the antecedent NP John (controller) with the infinitival empty subject (controllee). However, in verb-final languages like Korean, verb information only becomes available at the end of the sentence, making immediate use of verb information impossible.

In this project, under the supervision of my advisor Dr. William O'Grady and Dr. Amy J. Schafer, I investigated (i) whether L1-Korean comprehenders draw on morphosyntactic cues early in the sentence to interpret infinitival empty subjects in Korean complement control constructions, and (ii) how L1-English comprehenders and L1-Korean L2ers of English interpret adjunct infinitive control constructions in English, with a particular focus on purpose clauses (e.g., Sarah hired Tim __ to finish the work quickly).

Findings revealed that (1) complementizer cues facilitate the interpretation of empty subjects in Korean control constructions even without control verbs, and (2) L2 participants, even highly advanced learners, exhibited greater variability than L1-English participants in interpreting empty subjects. Moreover, the L2 group was also more sensitive to task type. Their comprehension was less consistent in online tasks than offline tasks, whereas the L1 group performed consistently across tasks.

(Top) At the 22nd East-West Center International Graduate Student Conference in Honolulu, HI, USA in February 2023

(Bottom) At the 30th Japanese/Korean Linguistics Conference in Vancouver, Canada in March 2023

Project outputs:

Lee, Youngin. (November 2025). L1 and L2 processing of English adjunct control. The 50th Boston University Conference on Language Development (BUCLD50), Boston, USA. [Poster]
Lee, Youngin. (October 2025). Complementizers as Cues to Control: Evidence from Korean. The 50th Anniversary International Conference of The Linguistic Society of Korea, Seoul, Korea. [Talk]
Lee, Youngin. (May 2025). Processing of Empty Subjects: Evidence from Control Constructions (Dissertation)
Lee, Youngin. (May 2024). The processing of filler-gap dependency in Korean control constructions. The 37th Annual Conference on Human Sentence Processing (HSP 2024), Ann Arbor, USA. [Poster]
Lee, Youngin. (December 2023). Integration of morphosyntactic information in incremental processing: evidence from Korean control sentences. The Architectures and Mechanisms for Language Processing - Asia (AMLaP Asia) 2023, Hong Kong, China. [Poster]
Lee, Youngin. (August 2023). Examining incremental sentence processing in Korean: the case of control constructions. The 4th International Conference on Theoretical East Asian Psycholinguistics (ICTEAP-4), Seoul, Korea. [Talk Slides]
Lee, Youngin. (2023). The role of complementizers in Korean subject and object control Constructions. In S. Williamson, A. Aminat Babayode-Lawal, L. Bosman, N. Chan, S. Cho, I. Fong, K. Holubowsky (Eds.), Proceedings of the 30th Japanese/Korean Linguistics (pp. 619–629). Vancouber, Canada: CSLI Publications.
Lee, Youngin. (February 2023). The use of complementizer for comprehension of Korean control constructions. The 22nd East-West Center International Graduate Student Conference (IGSC22), Honolulu, USA. [Talk]
Journal manuscript (Under review)

L1 and L2 acquisition of negated disjunction

Another strand of my research focuses on L2 acquisition of negated disjunction. I have been working on this project under the supervision of Dr. Bonnie D. Schwartz from the Department of Second Language Studies at UHM. The interpretation of negated disjunction (e.g., The pig didn't eat the carrot or the pepper) varies cross-linguistically (Szabolcsi, 2002). For instance, negation scopes over disjunction (NOT>OR) in English ("The pig ate neither"), whereas disjunction scopes over negation (OR>NOT) in Japanese ("The pig ate either, but not both"). O'Grady, Lee, and Lee (2011) found that Korean native speakers appear to allow both interpretations, although NEG>OR predominated. Addressing methodological concerns in that research (e.g., pragmatic infelicity), this L2 study on Korean negated disjunction employs a modified TVJT with two critical conditions (NOT>OR story vs. OR>NOT story), presenting stimuli bimodally (aural-written) in future tense.

Project outputs:

Lee, Youngin. (Accepted). How adults interpret disjunction under negation in native and nonnative Korean. To appear in Language Acquisition and Language Disorders, Special Issue: New Frontiers and Connections in Second Language Acquisition: Selected Proceedings of the 17th Generative Approaches to Second Language Acquisition (GASLA-17). [Peer-reviewed]
Lee, Youngin & Schwartz, B., D. (2024). The acquisition of negated disjunction in (native and) nonnative Korean. In H. Abdullah Ali AlThagafi & J. Ray (Eds.), Proceedings of the 48th annual Boston University Conference on Language Development (pp. 299-312). Somerville, MA: Cascadilla Press.
Lee, Youngin. (May 2024). How adults interpret disjunction under negation in native and nonnative Korean. The 17th meeting of Generative Approaches to Second Language Acquisition (GASLA17), Urbana Champaign, USA. [Poster]
Lee, Youngin & Schwartz, B, D. (November 2023). Negated disjunction in (native and) nonnative Korean. The 48th Boston University Conference on Language Development (BUCLD48), Boston, USA. [Talk] [Slides]

(Top) At the 48th Boston University Conference on Language Development in Boston, MA, USA in November 2023

(Bottom) At the 17th meeting of the Generative Approaches to Second Language Acquisition in Urbana-Champaign, IL, USA in May 2024

Scope assignment in L1 adult Korean

Previous research has shown that Korean native speakers fall into two groups when interpreting sentences with the universal quantifier motun 'every' and negation. For example, in the sentence 쿠키몬스터가 모든 쿠키를 안 먹었다 ('Cookie Monster didn't eat every cookie'), some speakers interpret it to mean that Cookie Monster did not eat any cookies at all (Every>NOT reading), while others understand it as meaning that it is not the case that Cookie Monster ate every cookie (= He may have eaten some) (NOT > Every reading). Han, Lidz, and Mosulino (2007, 2016) attributed this variation to V(erb)-raising, whereas Zeijlstra (2024) explained it in terms of Korean speakers' sensitivity to the universal quantifier motun. In this study, we aim to tease apart these two proposals by conducting a Truth-Value Judgment Task, using ellipsis to cancel polarity sensitivity (Sag 1976). Our findings suggest that the observed interspeaker variation among Korean native speakers is not due to V-raising, but rather arises from the lexical variation of in the interpretation of motun.

Project outputs:

Ohba, A., Lee, Youngin, Hwang, H., & Shimada, H. (June 2025). Testing the Split Grammar Hypothesis: ±V-raising or PPI/polarity insensitive? The 32nd Japanese/Korean Linguistics Conference, Ithaca, USA. [Talk]
Journal manuscript (In prep)

L1 acquisition of pre-subject 'only'

At the child center, where we conducted an experiment (under the supervision of Dr. Kamil Deen)

Children are known to misinterpret pre-subject only (e.g., "Only the cat got a car”) as if it were preverbal only (e.g., "The cat only got a car”). Though numerous theoretical explanations have been put forward (Crain, 1992;1994, Patterson et.al., 2006, Gualmini et al., 2003, Hohle et al., 2009, Kim, 2012, Hackl et.al., 2015, a.o.), we show that the reported challenge of pre-subject-only is in fact methodological in origin - previous studies failed to satisfy a crucial felicity requirement of only: to extract a subset from an established superset of referents. We use a modified TVJT (Crain & Thornton, 1998) to show that when this felicity condition is properly met, children are able to interpret pre-subject-only correctly.

Project outputs:

Deen, K., Brennan, P., Chang, Y-T., Daniels, R., Lee, Youngin., Lin, K., Ohba, A., Reddy, A., Tang, A., Tonoike, S., Topelian, A., Wang, J., Yusa, M. & Zubiri, L, A. (November 2023). Subsets from supersets: How children correctly interpret pre-subject only. The 48th Boston University Conference on Language Development (BUCLD48), Boston, USA. [Poster]
Paper in prep

Other projects

Processing of evidentiality in Korean and Turkish

Evidentiality is a grammar form that marks information sources from which a speaker knows events in his/her statement. Korean obligatorily marks this form in their grammars via inflectional forms that are suffixed to finite verbs. In this project, we investigate whether evidentiality processing is affected in healthy aging, and if so, how far declining working memory and source memory capacities contribute to evidentiality processing difficulty in aging. For this purpose, we conducted a word-by-word reading experiment in order to investigate how evidentiality processing is modulated in healthy aging in a life-sample group of Korean adults in the 18-80 age range.

Project output:

Arslan, S., Kim, S. Y., Lee, S., Lee, M., Lee, S., Lee, Youngin, Cho, C., Yeom, H., & Cho, S. W. (2022). Processing of evidentiality in Korean. In Proceedings of the Discourse and Cognitive Linguistics Society of Korea
Arslan, S., Kim, S. Y., Lee, S., Lee, M., Lee, S., Lee, Youngin, Cho, C., Yeom, H., & Cho, S. W. (November 2022). Processing of evidentiality in Korean. The Discourse and Cognitive Linguistics Society of Korea Conference, Seoul, Korea. [Slides]

Development of speech acts and prosody

It has been reported in the literature that children with autistism spectrum disorder (ASD) are largely impaired in their communicative use of language. Certain skills such as commenting, acknowledging the listener, and requesting information were completely absent while other speech acts including responding to questions and requesting objects or actions and protesting were present (Cho, 2003; Hong et al., 2010). This project seeks to determine the extent to which direct pragmatic intent is properly delivered in the utterances produced in a structured setting by Korean young children with and without autistic spectrum disorder (ASD).

Project outputs:

Lee, Youngin, Jung, S. J., Lee, K., Lee, S., Jang, H., & Cho, S. W. (2021). Direct pragmatic language skills in Korean-speaking children with autism spectrum disorder. Korean Journal for Infant Mental Health (KJIMH), 14(2), 1-27.
Lee, Youngin, Lee, S., Jung, S., Lee, K., Chung, S., & Cho, S. W. (August 2021). The selective impairment of direct speech acts in Korean young children with autistic spectrum disorder. The Korean Association of Language Sciences, Seoul, Korea.
Lee, K., Jung, S. J., Cho, S. W., Lee, Youngin, Lee, S., Yoo, H., Shin, Y. J. (June 2021). A Pitch analysis of the spontaneous speech in Korean toddlers with autism spectrum disorder. The 17th World Association for Infant Mental Health (WAIMH) World Congress, Brisbane, Australia.

Acquisition of a story grammar and null subject in child Korean

Stories are told everywhere across various culture and communities in the world for thousand of years. Previous studies indicate that stories have an underlying structure called a story grammar (Strong, 1998), and that it begins to develop at around four years of age (Peterson and McCabe, 1983). In this project, I examined how coherence develops in Korean-speaking young children’s storytelling skills. 27 Korean children aged 5 to 7 were told a story and asked to retell it. The data were analyzed in terms of both global and local coherence, which involved the story structure with a focus on purposiveness on one hand, and subject ellipsis on the other.

Project outputs:

Lee, Youngin. (2014). Coherence in Young Children’s Storytelling. In Proceedings of Discourse and Cognitive Linguistics Society of Korea (pp. 67-87).
Lee, Youngin. (2012). Coherence in Young Children’s Storytelling. M.A. Thesis at the Sogang University, Korea.

Development of the human-robot dialog interface

In this interdisplinary research project to develop of the human-robot dialog for patients dementia, we argue that, in determining semantic similarity, Korean words should be recategorized with a focus on the semantic relation to ontology in light of cross-linguistic morphological variations. Our results indicate that languages must be analyzed by varying methods so that semantic components across languages may allow varying semantic distance in the vector space models.

Project outputs:

Lee, Youngin, Lee, H., Koo, M., & Cho, S. W. (2015). Korean Semantic Similarity Measures for the Vector Space Models. Journal of the Korean Society of Speech Sciences, 7(4), 49-55.
Yang, H., Lee, Youngin, Lee, H. J., Cho, S. W., & Koo, M. W. (2015). A Study on Word Vector Models for Representing Korean Semantic Information. Journal of the Korean Society of Speech Sciences, 7(4), 41-47.

Page updated

Google Sites

Report abuse