‘Vehicle Change’ Project

In 2017, I proposed an online processing study through Northwestern Undergraduate Research Grant (URG) Program. In this study, I was trying to test the process called ‘vehicle change’ (Fiengo and May 1994, Merchant 2001). It has been observed that in the ellipsis construction, a form of omission, (e.g., VP-ellipsis or sluicing), the content of the ellipsis site and the antecedent can be mismatched. ‘Vehicle change’ is one of such mismatch. More specifically, it has been shown that within the ellipsis site, a name like John or Mary is understood as a pronoun like he or she. There have been various studies of the processing of ellipsis constructions, but I could not find any study that tested how such mismatched ellipsis site is processed. I got interested in how readers achieve the interpretation of the ellipsis site when such mismatch is tolerated. How can readers recognize such mismatch? How easy is it to interpret such mismatched ellipsis constructions? So, I proposed an experiment to reveal if readers employ complex mental mechanism to convert proper nouns to pronouns when they process sentences with ellipsis. Revealing what process is involved in the interpretation of the ellipsis site can lead to a better understanding of the mechanism of sentence comprehension in general, and help better understand the abstract mental structures that exist beyond the visual form of the sentence. I am currently collaborating on this project with other researchers from the Department of Linguistics at Northwestern University.

Russian’s “Command” of Cataphoric Pronouns

I conducted another research project in the area of syntax. This project is related to one of my primary interests which is the structure of Russian since it is the language I know tacitly, and the project results in my Linguistics Honors thesis on syntax of Russian adverbial clauses. In my thesis, I am analyzing the syntactic structure of Russian sentences that are introduced by the subordinating connective poka and investigating how this connective is different from its English counterpart while. English and Russian seem to follow the same restriction on coreference. However, after detailed investigation into the adverbial clauses in Russian, it becomes clear that in the cases when the subordinating connective is replaced by while (poka in Russian), the coreference is still possible in English, but the coreference is not possible in the same configuration in Russian (Kazanina 2005). If Russian has the same structure as English, as Kazanina (2005) and Kazanina and Phillips (2010) contend, then the impossibility of coreference is problematic for BPC (Chomsky 1981); although the pronoun does not c-command the name, the coreference is not possible. Thus, it has been suggested that this ‘exceptional’ anti-coreference effect should be attributed to some idiosyncratic property of the poka-construction and it is not a configurational constraint (Kazanina & Phillips, 2001, 2003, 2010). This study aims to offer an alternative analysis of the poka-construction, and to show that the anti-coreference effect is instead predicted from the analysis of the poka-construction that I defend. Specifically, I argue that the poka-construction has the structure of clausal coordination where TPs are coordinated. Thus, I argue that the clause introduced by poka is not an adjunct clause. Furthermore, I point out that in clausal-coordination structures, the pronoun in the subject position in the first conjunct-clause cannot be coreferential with the name in the subject position in the second conjunct-clause (Langacker 1969). I show that the BPC that crucially refers to Phase Command (Bruening 2014) can correctly capture the anti-coreference effect seen in the poka-construction.

Implicit Agents in Ukrainian: Evidence from Retrieval Interference in Sentence Processing

In one of my ongoing projects, I am investigating sentences including passive constructions in Ukrainian, and testing whether implicit agents represented syntactically give rise to retrieval interference. This study aims to reveal certain aspects of the parsing mechanisms that support language processing.