Case-mismatches in Korean left-node-raising: An experimental study

The goal of this article is to investigate Case-mismatches in Korean Left-Node-Raising (LNR) via the utility of experimental syntax techniques. To achieve this goal, we use the 2 × 2 factorial design of two Case-mismatch types (accusative vs. dative Case) of Korean LNR with respect to the locus of Case-licensing/mismatches (the first vs. second conjunct). The result of the experiment suggests that the first conjunct Case-licensing of the fronted NP in LNR is crucial across Case types, indicating that symmetric approaches (cf. Nakao’s (2009, 2010) across-the-board (ATB) Scrambling and Chung’s (2010) Multidominance) to LNR is not the right avenue to pursue. It also suggests that there is a distinction between structural Case-licensing and inherent Case-licensing in LNR (cf. Chomsky 1986, 1995; Bošković 2008). We review three syntactic analyses of Case-mismatch effects in LNR (ATB Scrambling, Multidominance, and Scrambling + pro) and defend the Scrambling + pro analysis.