Mitchell (1987) conducted a self-paced reading experiment that showed that people experienced difficulty reading a noun phrase when it immediately followed an intransitive verb. From this, he argued for a two-stage theory of parsing, in which verb subcategorization information is initially ignored. In response, Adams, Clifton, and Mitchell (1998) found no evidence to support this claim in an eye-tracking experiment and argued that Mitchell's segmentation procedure distorted the parsing process. We report an eye-tracking experiment, in which materials similar to those in Adams et al., but with longer noun phrases, were used, that showed a pattern of difficulty similar to Mitchell's findings. Hence, Mitchells results did not depend on the use of an artificial method of presentation. Me results cast further doubt on the adequacy of constraint-based accounts of parsing.
|Number of pages||7|
|Journal||Psychonomic Bulletin & Review|
|Publication status||Published - Dec 2001|
- SYNTACTIC AMBIGUITY RESOLUTION
- FIXATION TIMES