1 |
Managing Privileged Knowledge about Identity in Language Comprehension
|
|
|
|
Abstract:
A central claim in research on language comprehension is that assumptions about other’s knowledge are essential to guide language users’ understanding of utterances in both interactive conversations and fictional narratives. Past studies have suggested that listeners or readers are able to suppress privileged knowledge about the existence of entities from the earliest moments of processing. In the present dissertation, it is investigated whether this ability extends to cases where comprehenders possess privileged knowledge about the identity of an entity. This question is motivated by the central role played by identity in various semantic phenomena, as well as children’s well-documented difficulties in managing “identity” discrepancies in theory of mind tasks. Using a visual world eye-tracking paradigm, Experiment 1 shows that knowledge that is in conflict with an object's outward features (e.g., knowing that the object that looks like a lightbulb is actually a candle) is readily accessible to comprehension systems as listeners hear object descriptions unfold in time. Experiment 2 then verifies that, without the foreknowledge, the visually-misleading objects are identified in terms of their outward appearance. The validated stimulus objects are then used in Experiments 3 and 4, which evaluate the extent to which listeners suppress their privileged knowledge about object identity during real-time conversation. Similarly, Experiments 5 and 6 are conducted to explore whether listeners suppress privileged knowledge about the identity of entities during the comprehension of fictional narratives. In this case, interpretation patterns are examined as listeners hear the discourse “spoken” by fictional characters whose knowledge states are again varied. The results from both kinds of tasks reconfirm previous findings showing that language users rapidly take into account a speaker’s awareness of the existence of a referential entity. In contrast, however, the experiments show that comprehenders do not suppress private knowledge about the identity of an entity during referential processing—an effect that occurs even when the perspective of another individual is emphasized and regardless of whether the context involves conversational interaction or fictional narratives. Together, the results reveal important differences in the way in which distinct types of perspective-based cues are integrated in real- time language comprehension. ; Ph.D.
|
|
Keyword:
0621; Appearance-reality distinction; Common ground; Conversation; Discourse; Perspective-taking; Real-time comprehension
|
|
URL: http://hdl.handle.net/1807/71580
|
|
BASE
|
|
Hide details
|
|
|
|