Chinese character recognition in native and second language readers

Yao, Yun

Permalink

https://hdl.handle.net/2142/78767 Copy

Description

Title

Chinese character recognition in native and second language readers

Author(s)

Yao, Yun

Issue Date

2015-04-21

Director of Research (if dissertation) or Advisor (if thesis)

Packard, Jerome L

Doctoral Committee Chair(s)

Packard, Jerome L

Committee Member(s)

• Christianson, Kiel
• Sadler, Misumi
• Golato, Peter S

Department of Study

E. Asian Languages & Cultures

Discipline

E Asian Languages & Cultures

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Chinese
• Second Language Acquisition
• Orthographic Processing
• Radical
• Context Effect
• L1 Transfer

Abstract

Unlike phonographic writing systems such as that of English, Chinese has a logographic writing system, in which each grapheme encodes both sound and meaning (Sun, 2006). Thus, Chinese provides a unique perspective for the development and examination of reading theories, which are primarily based on the results of alphabetic scripts. Explaining the success of Chinese native readers is also a critical step to understand L2 character recognition and help nonnative speakers become effective Chinese readers. Although there is a growing body of research on both L1 and L2 character recognition, the literature on L1 processing has focused on context-independent character identification, and the majority of the previous studies on L2 character recognition used offline tasks. Therefore, more experimental data are needed to understand how context effects affect character recognition and the nature of L2 character processing in real time. This study examined how characters were processed by native speakers (N = 104) and L2 learners (N = 105) under context-independent and context-embedded conditions. By using two primed-character decision tasks, Experiments 1 and 2 explored the orthographic, phonological, and semantic activation of radicals in compound character recognition. Experiment 3 tested context effects on character recognition by comparing radical activation in highly constraining and non-constraining sentence contexts using a self-paced/masked priming paradigm (Luke & Christianson, 2012, 2013). Finally, the influence of L1 orthography on L2 character recognition was examined by comparing the processing patterns of Chinese learners from four typologically different L1 writing systems including abjad, abugida, alphabet, and syllabary/logography. The results suggest that both native speakers and L2 learners activated the orthographic, phonological, and semantic properties of radicals during character recognition under the context-independent condition. Moreover, native speakers activated radical-level and character-level representations in parallel when reading compound characters in non-constraining sentence contexts and made specific predictions about the identity of the upcoming character in highly constraining contexts. L2 learners did not activate radical-level information when reading sentences. Nevertheless, they were sensitive to contextual cues and adopted different reading strategies as their Chinese proficiency increased. Although native speakers processed phonetic radicals and semantic radicals in qualitatively and quantitatively similar ways, L2 learners displayed different processing patterns for different types of radicals. Semantic radicals were activated earlier than phonetic radicals; L2 learners who had acquired a limited number of characters were more sensitive to phonetic radicals, whereas those who had acquired a larger number of characters were more sensitive to semantic radicals. Lastly, the results indicate that L1 orthographies may influence L2 character recognition. L2 learners from alphabetic L1 backgrounds had faster responses than those from abugida L1 backgrounds, and learners from shallow L1 orthographies had higher accuracy rates than those from deeper L1 orthographies. These findings are consistent with previous research on context-independent character recognition and can be accommodated in the Interactive Activation Model (e.g., Taft, 2006), which posits that simple characters and complex characters are represented at different levels. The observed context effect demonstrates the methodological importance of examining character recognition in experimental settings that approximate natural reading. As for pedagogical implications, the results suggest that L2 learners may benefit from explicit instruction of phonetic radicals and teaching approaches that are sensitive to their L1 backgrounds. The study also implies that instructors should draw L2 learners’ attention to character decomposition and radical properties when teaching sentence-level comprehension, so that L2 learners will be able to develop effective character processing strategies that are applicable to natural reading.

Graduation Semester

2015-5

Type of Resource

text

Permalink

http://hdl.handle.net/2142/78767

Copyright and License Information

Copyright 2015 Yun Yao

Owning Collections