Efficient language processing:the role of memory and visuo-spatial context

language processing;language comprehension;memory;eye-tracking

Ryskin, Rachel A

关键词: language processing;    language comprehension;    memory;    eye-tracking;   

Others  :  https://www.ideals.illinois.edu/bitstream/handle/2142/92997/RYSKIN-DISSERTATION-2016.pdf?sequence=1&isAllowed=y

美国|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

A hallmark of adult language processing is the efficiency with which it unfolds despite the massive amount of ambiguity present in the linguistic signal. Prior work has shown that visuo-spatial (e.g., Heller et al., 2008; Ryskin et al., 2014) and pragmatic (Sedivy et al., 1999) contextual information can be rapidly integrated during the comprehension process. In this dissertation, I explored the use of this information during on-line language processing and how it is supported by learning and memory mechanisms. In Chapter 2, I examined whether individual differences in working memory, basic recall, and memory cue generation predict the ability to make use of information about visual perspective differences during comprehension. A targeted re-analysis of these data (Experiment 2), examined how perspective-taking evolves within a conversational pair, over the course of the conversation. In Chapter 3, I explored the use of spatial perspective-taking during comprehension. In two experiments (Experiments 3 and 4), I examine how listeners encode spatial perspectives in a speaker-specific fashion and use this information during interpretation of spatial language. In Chapter 4, I examined whether listeners can learn high-level, pragmatic information about a speaker through exposure and integrate it with the visual context on-line, during sentence processing. I examined the ability of listeners to learn and transfer this information across contexts (Experiments 5a and 5b), as well as the ability to use this information within a context (Experiment 6). The results of the work reported in this dissertation point to an important role for context-specific learning in the ability of the language processing system to rapidly and efficiently integrate visual, spatial, and pragmatic information during on-line comprehension.