My main research interest is in the cognitive science of reading and language. This research spans the nature of reading processes and second language (L2) learning. My research aims to understand the cognitive and neural mechanisms involved in learning and processing L1 and L2 using multiple research methods, specifically behavioral, functional magnetic resonance (fMRI), and event-related potentials (ERPs). Recently, I have used a megastudy approach, which is in line with the emphasis on Big Data trends, to depict a complete picture of human cognition. I see my research as interdisciplinary, and I am still enthusiastic about developing more integrated research perspectives and collaborations.
Due to the nature of morphemes and the vast number of morphologically complex words in many languages, morphological sensitivity has been thought to be a crucial part of skilled reading. In order to better understand its psychological reality and underlying processing mechanisms, I have conducted several research projects regarding the processing of three types of morphological structures with different language groups, such as derived words in Korean L1 (Kim, Wang, & Taft, 2015) and Korean-English bilinguals (Kim & Wang, 2014; Kim et al., 2011), compound words in Korean-English bilinguals (Ko et al., 2011), and inflected words in Korean L1 and English L1 (Kim, 2012). These studies also extended into an ERP study of Korean learners’ English (Chung, Park, & Kim, 2019), showing that the amplitudes of the brain potentials to morphologically complex words were significantly modulated by asymmetric lexical representation between L1 and L2.
I have conducted several research projects on neural correlates of L1 and L2 reading processes during my postdoc period (Kim et al., 2016; Kim et al., 2017; Kim et al., 2020). We first examined the factors that influence the balance between assimilation (similar reading network for L1 and L2) and accommodation (additional networks for L2 than L1). I hypothesized that the nature of mapping between orthography and phonology is one of the critical factors. To test the hypothesis, a group of native Korean speakers who learned English and Chinese were recruited. The results of this study, which utilizes a unique population (Korean-Chinese-English trilinguals), provide the clearest support for the assimilation and accommodation account and indicate the nature of mapping between orthography and phonology in L1 and L2 (Kim et al., 2016).
In addition, I was particularly interested in whether differences in brain activation during L1 reading are reflected in L2 reading. The results suggest that the brain network for reading English as an L2 is shaped by the L1s (Korean vs. Chinese). We could conclude that L2 acquisition is constrained by the existing L1 system in late bilinguals (Kim et al., 2017). Finally, another study analyzed the fMRI data to reveal whether language modality (visual vs. auditory) plays a role in determining the brain mechanisms of L1 and L2 (Kim et al., 2020). In our trilingual group, we examined the neural representation similarity in neural networks between L1 and L2 in spoken and written word processing. The findings of this study provided important insights about spoken and written language processing in the bilingual brain based on the evidence supporting more differentiated networks for written word processing in the three languages than spoken word processing.
I adapted a large-scale data approach (the so-called megastudy approach) to provide a complete picture of compound word processing (Kim, Yap, & Goh, 2019). This study is very meaningful not only because we provided the database of more than 2800 English compound words’ semantic transparency ratings at the constituent level but also because the results were compared to those using existing studies based on either human-judged ratings at the whole-word level or distance scores derived from the prediction-based distributional semantic algorithm at the constituent levels. This study has beed extended to Korean compounds' semantic transparency database construction (Kim & Kim, under review).
In addition, I participated as one of the researchers responsible for the Korean data in the Multilingual Picture (MultiPic) database, which contains naming norms and familiarity ratings for 500 coloured pictures across 32 languages or language varieties from around the world (Dunabeitia et al., 2022).
Collaborators
Seckin Arslan (CNRS)
Fan Cao (University of Hong Kong)
Krisda Chaemsaithong (HYU)
Taehong Cho (HYU)
Hyeongjeong Jeong (Tohoku University)
Yoonjung Kang (University of Toronto)
Miseon Lee (HYU)
Philip Monahan (University of Toronto)
Sanghoun Song (Korea University)
Hyungwook Yim (HYU)