探索語音轉碼在英文或日文為母語的中文學習者閱讀中文短文時所扮演的角色

Phonological recoding (PR) refers to the processes for converting printed words into sounds. Many researchers believe that PR is a natural and necessary process in first language (L1) reading. However, the role of PR in foreign language (FL) reading is unclear because the processes of the acquisition of L1 and FL are different. In this study, I conducted several eye-tracking experiments to investigate whether English or Japanese learners of Chinese would rely on PR to read Chinese texts, whether the properties of those learners’ L1 would affect their reliance on PR when reading Chinese texts, and whether text complexity would influence their propensity for relying on PR as well. A 5 by 2 by 3 mixed factorial design will be used for a series of eye-tracking experiments, with reading conditions and text complexity being the “within-subjects” factors, and L1 background (European language/Japanese/Chinese) being the “between-subjects” factor. Each subject was required to read two Chinese short passages under five reading conditions (articulatory suppression (AS)/ reading while listening to irrelevant speech (IRS)/ reading aloud (RA)/ reading while listening to the same text (RWL) / silent reading (SR)) and two text complexities (easy/ more difficult) (i.e., 20 passages in total). The dependent variables will include subjects’ reading times for each passage, the accuracy rates of the immediate posttests, and the global and eye-movement measurements. The results showed that phonological information did not play an important role in Chinese participants’ reading comprehension. However, it had significant effect on European students’ reading comprehension, while it only weakly affected the reading comprehension of Japanese students. In AS-condition, both European and Japanese students performed significantly worse in the immediate posttests. Nevertheless, RA and RWL did not enhance reading comprehension, and they even impaired reading comprehension. These results indicated that European and Japanese students did not rely on phonological recoding for comprehending Chinese texts. The eye-movement analyses provided further explanations for the results just mentioned: Subjects’ mean saccade lengths became significantly longer and the regression ratios were much higher in the AS-condition, which suggests that subjects reduced the use of phonological recoding and adopted skimming as their reading strategy under articulatory suppression. This led to fragmented and inaccurate reading comprehension. Besides, AS may interfere with the function of articulatory rehearsal and reduces subjects’ ability in integrating meanings in the texts. IRS did not show any effect on reading. RA prolonged reading time, reduced saccade length and regression ratio. However, it did not enhance reading comprehension at all. When the texts were more difficult, RA even significantly reduced European and Japanese students’ reading comprehension. RWL interfered with the reading of Japanese and Chinese students, but it could accelerate the reading speed of European students.