Title: CRM: An Integrated Model of Word Processing and Eye-Movement Control During Chinese Reading
Speaker: 李兴珊 (Xingshan Li )
Time: 15:00, Wed, 9 September 2020
(Hong Kong time)
Venue: https://cuhk.zoom.us/j/779556638
https://cuhk.zoom.cn/j/779556638 (Chinese Mainland)
About the Speaker
Xingshan Li is a Professor of Psychology at the Institute of Psychology, Chinese Academy of Sciences. He acquired his Phd degree in Psychology at University of Massachusetts Amherst in 2007. His research interest focusses on Chinese reading and language processing. In the last few years, he has tried to understand how Chinese readers deal with some unique properties of Chinese text during reading. In one line of his research, he used eye tracking and computational modelling to examine how Chinese readers parse characters into words and how Chinese readers guide their eye movements. These questions are interesting given that there are no inter-word spaces to mark word boundaries in Chinese. He has published many articles at journals such as Psychological Review, JEP: General, Cognitive Psychology, JEP: HPP and JEP: LMC. He is an Associate Editor of Quarterly Journal of Experimental Psychology, and the Director of the CAS Key Laboratory of Behavioral Sciences.
CRM:An Integrated Model of Word Processing and Eye-Movement Control During ChineseReading
Xingshan Li
Instituteof Psychology, Chinese Academy of Sciences
In the talk, I will introduce a new model of Chinese reading (CRM). In the Chinese writing system, there are no inter-word spaces to mark word boundaries. To understand how Chinese readers conquer this challenge, we constructed an integrated model of word processing and eye-movement control during Chinese reading. The model contains a word-processing module and an eye-movement control module. The word-processing module perceives new information within the perceptual span around a fixation. The model uses the interactive activation framework (McClelland & Rumelhart, 1981) to simulate word processing, but some new assumptions were made to address the word segmentation problem in Chinese reading. All the words supported by characters in the perceptual span are activated and they compete for a winner. When one word wins the competition, it is identified and it is simultaneously segmented from text. The eye-movement control module makes the decision regarding when and where to move the eyes using the activation information of word units and character units provided by the word-processing module. The model estimates how many characters can be processed during a fixation, and then makes a saccade to somewhere beyond this point. The model successfully simulated important findings on the relation between word processing and eye-movement control: how Chinese readers choose saccade targets, how Chinese readers segment words with ambiguous boundaries, and how Chinese readers process information with parafoveal vision during Chinese sentence reading. The current model thus provides insights on how Chinese readers address some important challenges, such as word segmentation and saccade-target selection. Future directions will be briefly introduced at the end of the talk.
Virtual Psycholinguistics Forum:
(https://cuhklpl.github.io/forum.html)
