Strategies in the writing of Chinese characters by intermediate and advanced students of Chinese as a foreign language

The dual-route model of spelling has occupied a central position in discussions of theories of spelling alphabetic languages. Is such a model positing a direct route and an indirect or assembled route mediated by phonology adequate to explain the writing of Chinese characters by students of Chinese as a foreign language? More particularly, is there evidence of more than one assembled strategy, for example, graphic and semantic, as well as phonological? In addition, is there evidence that whatever strategies are employed by CFL learners change as skill increases? This study looks at the writing of characters by five levels of low intermediate to advanced students (i.e., end of second year college-level Mandarin to fourth year) to answer this question. Subjects completed one or more of five tasks: (1) A comprehensive spelling test of 100 characters selected from the 1000 most frequent characters in a specific character list. Subjects wrote the character for one romanized syllable presented in a short phrase with accompanying English gloss. (2) Essay writing. (3) Follow-up rewriting of incorrectly written characters from the essays. (4) Phonologically plausible pseudo-character generation. (5) Semantically plausible pseudo-character generation.;An analysis of the knowledge errors made by subjects (as opposed to performance errors or slips of the pen) is consistent with a view that there is more than one assembled strategy. Examination of the use of the graphic and phonological strategies over the five levels suggests that higher level students use more graphic strategies than lower level students. In contrast, the phonological strategy appears to be equally available to all levels of students in the writing of real characters suggesting that this skill is not affected by studying Chinese. However, in the creation of pseudo-characters higher level students use more appropriate phonetic components in the creations than lower level students.;A technique for distinguishing between performance-based slips of the pen and competence-based knowledge errors is given. Implications for pedagogy and areas of future research are also discussed.