Regular-, irregular-, and pseudo-character processing in Chinese : the regularity effect in normal adult readers

Abstract

Background Unlike alphabetic languages, Chinese uses a logographic script. However, the pronunciation of many character’s phonetic radical has the same pronunciation as the character as a whole. These are considered regular characters and can be read through a lexical non-semantic route (Weekes & Chen, 1999). Pseudocharacters are another way to study this non-semantic route. A pseudocharacter is the combination of existing semantic and phonetic radicals in their legal positions resulting in a non-existing character (Ho, Chan, Chung, Lee, & Tsang, 2007). Pseudocharacters can be pronounced by direct derivation from the sound of its phonetic radical. Conversely, if the pronunciation of a character does not follow that of the phonetic radical, it is considered as irregular and can only be correctly read through the lexical-semantic route. The aim of the current investigation was to examine reading aloud in normal adults. We hypothesized that the regularity effect, previously described for alphabetical scripts and acquired dyslexic patients of Chinese (Weekes & Chen, 1999; Wu, Liu, Sun, Chromik, & Zhang, 2014), would also be present in normal adult Chinese readers. Method Participants. Thirty (50% female) native Hong Kong Cantonese speakers with a mean age of 19.6 years and a mean education of 12.9 years. Stimuli. Sixty regular-, 60 irregular-, and 60 pseudo-characters (with at least 75% of name agreement) in Chinese were matched by initial phoneme, number of strokes and family size. Additionally, regular- and irregular-characters were matched by frequency (low) and consistency. Procedure. Each participant was asked to read aloud the stimuli presented on a laptop using the DMDX software. The order of stimuli presentation was randomized. Data analysis. ANOVAs were carried out by participants and items with RTs and errors as dependent variables and type of stimuli (regular-, irregular- and pseudo-character) as repeated measures (F1) or between subject’s factor (F2) using SPSS 19. Results The results for RTs and errors showed a main effect of type of character both by participants and items. Simple effects showed that irregular characters were treated significantly slower than regular and pseudocharacters. For errors, regular and pseudocharacters were significantly less error prone than irregular characters. Discussion The regularity effect found here suggests that irregular characters are read via the lexical semantic route while regular and pseudocharacters are read through the lexical non-semantic route. These results are in line with -and extend- the literature on surface acquired dyslexia in Chinese (Weekes & Chen, 1999; Wu et al., 2014) since, to the best of our knowledge, this is the first study that addressed regular, irregular and pseudocharacter reading in Cantonese Chinese healthy adults. Figure 1, please see publisher web.