Saccade target selection in L2 reading: Evidence from Chinese-English bilinguals

Recent evidence shows that saccade target selection in second-language (L2) English reading is similar to that in native-language (L1) English reading. However, only bilinguals with languages belonging to the same writing system were examined in these studies. The present study investigated saccade target selection and whether it changes with L2 proficiency in cross-writing-system bilinguals, specifically Chinese-English bilinguals. In Study 1, we analyzed an L2 reading data set of Chinese-English bilinguals. For both high- and low-L2-proficiency bilinguals, we observed an inverted U-shaped preferred viewing location curve and found a word length effect on the initial landing position, suggesting that bilinguals selected the word central position as the saccade target. However, when the launch site was five letters or farther from the target word, the PVL curve peaked at the word beginning. In Study 2, we manipulated target word frequency, parafoveal information validity, and L2 proficiency in a factorial experiment. A significant three-way interaction was found in outgoing saccade length. Low-proficiency bilinguals exhibited longer outgoing saccades from high-frequency than low-frequency target words, but only when parafoveal information was valid. With the increase of L2 proficiency, this interaction ceased to exist. These results suggest decreased reliance on ongoing processing for saccade target selection as L2 proficiency increases. Combining the two studies, we found for Chinese-English bilinguals' L2 reading a specific processing-based strategy that varied with L2 proficiency and a universal word-based strategy. These findings provide new insights into reading strategy and eye movement control in cross-writing-system bilinguals' L2 reading. (PsycInfo Database Record (c) 2026 APA, all rights reserved).