Phonology-to-orthography consistency at sublexical level in Chinese writing

Abstract

Background The process of writing can be achieved by either a lexical route, based on direct lexical item retrieval, or a sub-lexical route, based on phonology-to-orthography conversion (Ellis & Young, 1994). Since the two routes operate together among normal individuals, analysing the nature of phonology-to-orthography conversion is difficult if not impossible. The current study achieved this by observing the dictation performance on characters with various phonology-to-orthography relations of a Chinese dysgraphic patient with storage problem in the orthographic output lexicon (OOL). Method WCY, 57-year-old male with 7-month post-onset left hemisphere stroke, was self-referred to our clinic. Initial assessment using the Cantonese Aphasia Battery (Yiu, 1992) yielded a diagnosis of mild anomic aphasia with an AQ score of 93.5. Follow up assessment using the logogen model (Ellis & Young, 1994) as theoretical framework with the stimuli set used in Law, Yeung, Wong & Chiu (2005), the Pyramid Palm Tree Test (PPT; Howard & Patterson, 1992) and the Birmingham Object Recognition Battery (BORB; Riddoch & Humphreys, 1993) was conducted. Furthermore, a writing-to-dictation task was administered using a total of 120 Chinese characters selected in a 2 (Frequency) X 3 (Feedback consistency) construct. Feedback consistency of stimuli were categorised into three levels: a) characters with no homophones, b) characters with homophones that are associated with more than three phonetic radicals, and c) characters with at least two homophones, all sharing the same phonetic radicals. Results & Discussion Results of PPT (25/28) and BORB (24/25 in test 7, 25/25 in test 8, and 23/23 in test 12) revealed an intact semantic processing. WCY’s below average performance in oral picture naming (174/217) and written picture naming (94/217), but good reading aloud performance (120/120) confirm that he has anomia and dysgraphia without dyslexia. In the writing-to-dictation task, Chi-square test results showed that WCY’s performances in the six categories of characters are not identical (p <.01). Frequency effect was observed (38/60 for high frequency characters vs 6/60 for low frequency characters) indicating a potential storage problem in the OOL. Among the high frequency character trials, he demonstrated relatively better performance (15/20) in writing characters in category (c) compared to his performance (12/20 and 11/20 respectively) in writing characters in (a) and (b). The better performance in (c) indicated facilitation from the possible usage of syllable-to-radical conversion in writing-to-dictation by WCY. Error analysis results further confirm our claim. Among the 23 errors he produced in (c), seven of them contain the target phonetic radicals. On the other hand, only three out of 27 in (b) contain the target phonetic radicals. Finally, none of the 26 errors in (a) contain the target phonetic radicals. The difference between (c) and (b) is potentially attributed to the advantage of having one-to-one associations between syllables and orthographic forms at the radical level in (c) compared to the one-to-many condition in (b). Finally, in the case of (a), although there may also exists a one-to-one association between syllables and radicals, the absence of other neighbours makes similar facilitation as in (c) impossible. Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed.