초록
영어
This study examines the processing of Korean how many-phrase questions that include filler-gap dependencies. A corpus analysis and a reading time analysis were conducted. The reading time data were collected in an experiment employing a self-paced reading task with 43 native Korean-speaking adults. Korean question words can appear in situ or be scrambled to appear in various positions. The materials were sentences with how many-phrases and a variety of word orders. The reading time results indicate that the readers began to search for the gap as soon as they had identified the fronted how many-phrase in scrambled constructions. This finding tells us that the readers’ sentence processing mechanism is not guided by probabilistic expectations; that is, it is not shaped by the distributional patterns in their language experience. However, the finding is consistent with the prediction that the parser engages in an Active Filler Strategy, indicating that the processing mechanism is guided by an active process of dependency formation.
목차
1. Introduction
2. Background
2.1. Processing Filler-Gap Dependencies in English
2.2. Processing Filler-Gap Dependencies in Japanese and Korean
2.3. Language experience and processing filler-gap dependencies
2.4. Distributional analyses in the corpus data: myech-phrases
3. Experiment
3.1. Participants
3.2. Materials and Procedure
3.3. Results
4. Discussion
5. Concluding Remarks
References