Whole-word phonological representations in the Chinese lexicon S.-P. Law a, * , W. Wong a , Karen M.Y. Chiu b a University of Hong Kong, Hong Kong b Tung Wah Hospital, Hong Kong Accepted 8 July 2005 Available online 15 September 2005 The recognition and production of multimorphemic words in in- flected languages have been a focus of psycholinguistic research. Similarly, questions have been raised about how morphologically complex words are represented in languages with little inflectional morphology, such as Chinese, where the great majority of multimor- phemic words are compounds composed of stem morphemes. Recent models of the Chinese lexicon have converged on an architecture with two basic characteristics: (i) there are three types of interconnected units, semantic features, single syllables, and single characters; (ii) whole-word and morphemic representations are purely semantic in nature, and their relationship is not hierarchical (Taft, Liu, & Zhu, 1999; Zhou & Marslen-Wilson, 2000). The former purports to reduce the redundancy in representation at the phonological and orthographic levels, as the majority of multisyllabic or multicharacter words is simply the concatenation of their constituents. The interconnection among the three levels of representation have found support from brain-injured individuals with preserved reading aloud despite se- mantic deficits ( Law & Or, 2001; Law,Wong, & Chiu, in press; Weekes & Chen, 1999; Weekes, Chen, & Yin, 1997; ). Furthermore, the assumption of single syllable units is compatible with the occur- rence of two types of reading errors reported of dyslexic individuals— reading a pronounceable constituent in the target character and pro- ducing a contextually inappropriate pronunciation for a character with different phonological forms in different word contexts, i.e., homo- graphic heterophone (Law, 2004). However, the issue of the presence of whole-word representations remains, especially for multisyllabic words containing homographic heterophones. If the phonological lexicon contains only single syllable units, correct reading aloud of the characters in question must be supported by semantic information. In this study, we assessed a Cantonese speaker with lexical non-semantic dyslexia, YKM, on a reading task with items containing homographic heterophones and a comprehension task involving synonymy judgments. His good per- formance on reading these words in contrast with his poor compre- hension of them argues for phonological representations corresponding to whole words. YKM was a 61-year-old right-handed male speaker of Cantonese with a university degree. At the age of 50, he suffered sub-arachnoid haemorrhage at the anterior communicating artery with arteriovenous clipping done. In January 2002, he was admitted to the hospital with left basal ganglion haemorrhage. CT scans showed left intracerebral hae- matomas with small left frontal effusion and a small right cerebral infarct while no arteriovenous malformation aneurysm was detected. He had right-sided hemiparesis of the limbs and was diagnosed to have hyper- tension and cataract. His speech was non-fluent with mild dysarthria of the lips. YKM was previously assessed on word reading, pictured object naming, and comprehension of verbal and non-verbal materials (Law et al., in press). He exhibited a dissociation between preserved reading aloud and severely impaired picture naming (see Table 1). He performed at below-normal levels on all verbal and non-verbal comprehension tests. Hence, his performance pattern was hypothesized to be due to semantic deficits vis-a ` -vis the largely intact direct reading route. A total of 42 homographic heterophones were selected to generate 42 word pairs for the reading task. As for the synonymy judgment task, each of the 84 words was paired with a synonym and an unrelated word, respectively. The subject would score one point if he correctly accepted the synonym and rejected the unrelated word. YKM correctly read aloud 72/84 (85.7%) of test items (or both members in 31/42 word pairs). Of the 12 errors he made on homo- graphic heterophonic characters, nine were contextually inappropri- ate errors (e.g., pin4 ji4 ‘cheap’, a ` bin6 ji4 where bin6 lei6 ‘convenient’). In contrast, he only scored 44/84 (52.4%) in the com- prehension task. Of the 40 errors, 30 were acceptance of the synonym and the unrelated distractor; nine involved rejections of both words, and one case with incorrect decisions on both trials. More impor- tantly, there did not seem to be a relationship between YKM’s abilities to read aloud and to comprehend a word. He could not comprehend 35/72 words he correctly read aloud, and for the 12 stimuli he misread, he showed some degree of understanding of the words in seven cases. The present findings are incompatible with the view that recognizes phonological representations of single syllables only. On the other hand, YKM’s performance can be accounted for by models that as- sume the existence of whole-word phonological representations. In reading aloud, characters of a disyllabic or multisyllabic word inde- pendently access their syllable unit; at the same time, the characters converge on the whole-word unit. As long as the direct reading route is Brain and Language 95 (2005) 215–216 www.elsevier.com/locate/b&l * Corresponding author. E-mail address: splaw@hkucc.hku.hk (S.-P. Law). 0093-934X/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.bandl.2005.07.111