Orthographic variation and brain processes: A non-alphabetic perspective.

Saturday, October 01st, 2011 [08:30 - 10:30]

SY_13. Orthographic variation and brain processes: A non-alphabetic perspective

Lee, C. ^1, ³ , Lee, J. ^2, ³ & Tzeng, O. J. ^1, ³

1 Brain and Language Laboratory, Institute of Linguistics, Academia Sinica
2 Department of Educational Psychology, National Taiwan Normal University
3 Laboratories for Cognitive Neuroscience, National Yang-Ming University, Taiwan

Languages vary significantly in how graphemic symbols are mapped onto spoken languages. Yet, studies have shown very similar brain circuits are activated in readers across different writing systems. It is imperative to know how the orthographic variation determines cross-language differences in the grain size of lexical representations and the neural mechanisms for reading. Chinese is classified as a logographic writing system. Chinese characters represent morphsyllabic rather than phonemic information. The first talk provides evidence for the early sublexical phonological priming and discusses how the orthographic and phonological processes are weighted differently in reading Chinese. In the second talk, a series of ERP experiments demonstrates the interplay between orthographic density and phonetic consistency in the different stages of lexical processing and suggests that Chinese readers pick up subtle statistical regularities about orthographic as well as phonological patterns. Meanwhile, the pervasive homophony among Chinese characters implies the importance of the graphic form in selecting meaning and avoiding homophonic confusion. The third talk will demonstrate the homophone density effect in the brain to highlight the feedback processing in visual word recognition. The fourth talk will demonstrate that the characteristics of the Chinese writing system affect not only the reading process of Chinese characters but also the mechanisms that support verbal short-term memory. A series of behavioral and neuroimaging experiments reveal clear contribution from orthographic processing to retention of Chinese characters. The fifth talk provides eye-movement findings on how Chinese readers utilize the intrinsic properties (i.e., orthographic and phonological codes) and the extrinsic property (i.e., word predictability) of characters and words in reading sentences. In the last talk, a conspiracy theory of reading, based upon the affordance theory and the orthographic equilibrium hypothesis, is proposed to map the cerebral reading circuit based upon data generated from the educated eyes to the educated brain.

TALKS

SY_13.1 - Two-Way Mapping between Phonology and Orthography in Reading Chinese?

Lee, J. ^1, ³ , Chang, T. T. ^2, ³ & Shen, W. Y. ^1, ³

1 Department of Educational Psychology, National Taiwan Normal University
2 Institute of Neuroscience, National Yang Ming University
3 Laboratories for Cognitive Neuroscience, National Yang-Ming University, Taiwan

The role of phonology in reading has long been emphasized in the study of alphabetical languages. In contrast, the role of phonology in reading Chinese, a logographic writing system, has been debated. The issues of whether phonology is necessary for Chinese reading and whether the contribution of phonology to Chinese is as great as that to alphabetical languages remain open. Although previous studies have revealed the role of phonology in Chinese reading, the emphasis was placed on the high demand of visual-spatial processing in orthographic analysis and on the addressed phonology of syllabic (whole-character) units. In this talk, I will present evidence of early sublexical priming of phonological processing from an ERP experiment. I will also present the brain activation revealed in an fMRI experiment that is associated with orthographical analysis in reading Chinese when both semantic and phonological contribution is minimized. An attempt will be made to integrate the roles of early phonological and late orthographic information in reading Chinese by proposing the feed-forward and feed-backward processes. I will argue that, in comparison with reading alphabetic languages, reading Chinese does not involve different processing. However, reading Chinese places different weights on various stages from reading alphabetic languages due to the characteristics of the Chinese writing system.

SY_13.2 - The neighborhood effects in Chinese visual word recognition

Lee, C. ^1, ²

1 Brain and Language Laboratory, Institute of Linguistics, Academia Sinica
2 Laboratories for Cognitive Neuroscience, National Yang-Ming University, Taiwan

Researches on word recognition have clearly shown that a word’s identification is affected by its neighborhood properties and suggest the visual word recognition relies on the co-activation among orthographic, phonological and semantic units. Languages vary in orthographic depth which may determine cross language differences in the grain size of lexical representations, the reading strategies developed, and the rate at which children acquire reading skills. Chinese is often classified as a logographic writing system. The reading unit, character, represents morphosyllablic rather than phonemic information. It is an intriguing question of how Chinese readers establish the efficient mapping between orthography and phonology in learning to read Chinese. In this talk, a series of neuroimaging and behavioral studies will demonstrate how Chinese readers capture the statistical mapping consistency between character and sound and the neural correlates responsible for the Chinese orthography-to-phonology transformation. Meanwhile, the ERP data reveals the consistency effect in three different time windows, N170, P200 and N400, especially for reading phonograms with large orthographic neighborhood (phonetic combinability) and suggests the interplay between orthographic density and the mapping from orthography to phonology in the different stages of lexical processing. This is further supported by the developmental data which shows the vocabulary size determines when and how the knowledge of phonetic consistency is acquired in learning to read. Based on these cross-linguistic evidences, it appears that the statistical learning approach holds that Chinese readers pick up subtle statistical regularities about orthographic as well as phonological patterns reflected in the early language inputs. Although the functional units underlying the reading mechanism may vary due to the orthographic depth, similar functional operation is assumed.

SY_13.3 - Homophone Effect in Chinese Reading and its Neural Representation

Kuo, W. ^1, ² , Lee, C. ^1, ³ , Lee, J. ^1, ⁴ & Tzeng, O. J. ^1, ³

1 Laboratories for Cognitive Neuroscience, National Yang-Ming University, Taiwan
2 Institute of Neuroscience, National Yang-Ming University, Taiwan
3 Brain and Language Laboratory, Institute of Linguistics, Academia Sinica
4 Department of Educational Psychology, National Taiwan Normal University

In recent years, the role of phonology to orthography feedback consistency in visual word recognition has been a major challenge to the current models for reading. To tackle this issue is sometime problematic for the alphabetic writing system due to its nature of the high degree of coupling between grapheme and phoneme. Chinese, as an ideographic writing system, has approximately 1300 syllables which correspond to around 5000 characters. The pervasive homophony of Chinese implies the importance of a graphic form for selecting meaning and escaping homophony in reading Chinese. Most importantly, many Chinese homophone mates can be completely different in their orthographic patterns which provide a ground for a better understating of the feedback processing in visual word recognition. The present fMRI study manipulates the homophone density and character frequency in a lexical decision task. The reaction time data reveal the homophone density effect in reading low-frequency characters, and this psychological effect was mirrored in the left inferior temporal-occipital junction which a brain area emphasized for word from processing. The implication of how an orthographic form could benefit from the feedback connection from its homophonic mates will be discussed.

SY_13.4 - Orthographic contribution to verbal short-term memory of Chinese characters: Evidence for the close relationship between linguistic knowledge and retention

Wu, D. ^1, ² & Lin, E. Y. ^1, ³

1 Laboratories for Cognitive Neuroscience, National Yang-Ming University, Taiwan
2 Institute of Cognitive Neuroscience, National Central University, Taiwan
3 Institute of Neuroscience, National Yang-Ming University

Previous literature of short-term memory (STM) indicates that verbal materials are dominantly retained in phonological representations while other linguistic information (e.g., orthography, semantics) only contributes to verbal STM minimally if not negligibly. However, accumulating evidence suggests that retention of verbal materials is jointly supported by multiple linguistic components correspondent to orthographic, phonological, and semantic representations in long-term memory. That is, the underlying mechanisms of verbal STM are closely related to those of long-term linguistic knowledge. Similarly, the organization and processing of verbal STM should reflect the characteristics of a specific language to be retained. Because logographic characters of Chinese are visually more complex than alphabetic letters, the correspondence between orthography and phonology is less transparent, and there are abundant homophones, reading and remembering Chinese should rely more on information additional to phonology than reading and remembering alphabetic languages. In behavioral experiments with a probed recognition task, participants’ STM of Chinese characters was clearly affected by orthographic similarity of to-be-remembered stimuli, irrespective of phonological similarity of the materials. Experiments with functional magnetic resonance imaging further revealed that remembering homophonic or orthographically similar characters was associated with higher activation in a left-lateralized network including the brain regions that are sensitive to orthography, consistency, and homophone computation of Chinese characters (i.e., the middle and inferior frontal gyri, the insula, the middle temporal gyrus, and the junction of the fusiform gyrus). Remembering orthographically similar characters also involved bilateral superior and inferior parietal lobules, which might be associated with visuo-spatial processing of STM. The findings from our investigations provide converging evidence for the indispensible contribution of orthographic representations to verbal STM of Chinese characters. These results also suggest that linguistic characteristics of a specific language not only determine the processing mechanisms of that language but also delineate the organization of verbal STM for that language.

SY_13.5 - Eye Movement Guidance in Reading Chinese Sentence

Tsai, J. ^1, ²

1 Department of Psychology, National Chengchi University
2 Laboratories for Cognitive Neuroscience, National Yang-Ming University, Taiwan

In contrast to alphabetic scripts, the structural and functional units of Chinese written system are less transparent for word recognition in reading sentence. First, Chinese character is the basic written unit of the spoken language and usually maps onto morphemes and syllables. At the sub-lexical level, approximately 80% of the characters are phonetic compounds that are made up of a semantic radical and a phonetic radical. Comparing to letters of alphabetic languages, Chinese characters encompass richer information of orthography, phonology, and semantic in a packed region. Second, over 76% of Chinese words are compounds of characters and many characters can stand alone as individual words or be the constituent of compound words. The relationship of character meaning to the meanings of words containing them is often not apparent. Extracting word in sentence is even more difficult since there is no visual space separating words. Therefore, Chinese readers need to make use of the rich lexical information and contextual constraint, in order to correctly recognize characters and extract words from the character string in a sentence. We report eye movement experiments of sentence reading to address the special features of Chinese written system. One is the parafoveal preview benefits of orthographic and phonological codes. The other is the word predictability and word frequency effects. The eye movement data provide the evidence that Chinese readers utilize the intrinsic properties (i.e. orthographic and phonological codes) and the extrinsic property (i.e. word predictability) of characters and words in the process of reading. These factors not only contribute to Chinese reading but also show their influence in early lexical processing.

SY_13.6 - Mapping the Literate Eyes to the Educated Brain in an Non-Alphabetic Script: What Have We Learned over the last three Decades’ Research on Reading Chinese

Tzeng, O. J. ^1, ²

1 Brain and Language Laboratory, Institute of Linguistics, Academia Sinica
2 Laboratories for Cognitive Neuroscience, National Yang-Ming University, Taiwan

One important aspect of learning to read a printed text, in which a series of graphic symbols are arranged to represent the key features of the corresponding spoken language, is learning to move the eyes to search for relevant information from the text. The question is what information is available in the prints that would help the readers to accomplish the act of successful reading. Physically, there is the graphic information and its spatial layout. Linguistically, there is phonetic information embedded in the script/speech mapping relationship and morphological information which characterizes the meaning components in the prints. More importantly, there is orthographic information which prescribes the transitional probability from graphic component to the next within a character (letter) and/or transitional probability from one character (letter) to the next within a word. Results in our laboratories clearly demonstrate that an educated reader picks up graphic, phonological, morphological, and semantic information, supported by a language-based short-term memory, in early processing. Since no word boundary is provided in a Chinese text, because all printed characters are spaced equally, examining how to gather “word”-related information parafoveally during eye-fixation in order to facilitate subsequent reading in a Chinese text would shed light for how the reading circuit is organized within the brain. Important progress has been made over the past three decades. Three sets of experimental data related to the neurophysiological processes have been generated under different kinds of experimental paradigms and with different types of brain imaging techniques. Are they consistent, compatible, confirmatory, complimentary, or in conflict among one another, with respect to their implied underlying cognitive neuropsychological processes? A conspiracy theory of reading, based upon both the affordance theory and the orthographic equilibrium hypothesis, is proposed to map the cerebral-reading circuit based upon data generated from the educated eyes to the educated brain.