No abstract

No abstract

This article provides an experiment-based study on scrambling involved in binding. There have been a large number of attempts to analyze scrambling in terms of A/A’-distinctions. These analyses are dependent upon the canonical Binding Theory(BT). Binding relations, however, do not always determine which copy of a scrambled phrase is interpreted. Besides, different acceptability judgments on the same sentence among researchers have been one of the impediments to proposing a far-reaching theory regarding scrambling. For this reason, this paper conducted an acceptability judgment testing to scrutinize the mixed nature of scrambling and binding. In the experiment, ‘scrambling effects’ are observed in all but one of the conditions; it suggests that scrambling-involved data can be analyzed by means of the inactiveness of the base copy. The parallels of the pronoun ku and the anaphor caki observed in scrambled contexts support the idea that the dependence on the BT for understanding the nature of scrambling is problematic.

This study investigates why some quantifiers incur intervention effects while others do not at LF. Based on the understanding that controversial acceptability judgments of the relevant sentences make it difficult to precisely distinguish interveners from non-interveners, this study tests Korean intervention constructions via a formal experimentation. With reliable data from the experiment, I argue that an intervening factor is linked to the epistemic "non-specific" property and there is no absolute set of intervening quantifiers. The intervening status of a quantifier can be "off" when the addressee perceives it as a particular individual by context or real-world knowledge. This novel perspective provides a more convincing and unified account of the intervention phenomenon and its gradience in acceptability than previous studies in which the effect of "context" is ignored or underestimated.