Differential Effects of Initial and Final Character Positional Probabilities During Chinese Novel Word Learning in Reading: Evidence From Children’s Eye Movements

Abstract

Chinese children aged 8 to 10 years can utilize character positional probability—the likelihood of a character appearing at initial or final word positions—as a cue to facilitate word segmentation and identification during novel word learning. Given the distinct roles of initial and final characters in Chinese word recognition, we conducted two parallel experiments to investigate how positional probability influences children’s processing during reading. Sixteen two-character pseudowords were constructed as novel target words. In Experiment 1, we manipulated the positional probability of the initial character while keeping the final character constant. In Experiment 2, we reversed this manipulation for the final characters. Each novel word was embedded in six semantically constrained sentence frames. Forty-five children’s eye movements were recorded as they read these sentences. Results revealed that novel words containing characters with high positional probability in their preferred positions elicited shorter fixation durations compared to those with low positional probability characters. This effect emerged earlier for initial than for final characters, suggesting that Chinese children are sensitive to positional probability cues associated with both initial and final characters when learning new words during reading, but with a processing advantage for initial characters. Moreover, the positional probability effects remained stable across multiple exposures, indicating that children consistently rely sub-lexical cues such as positional probability before fully forming new lexical representations. Our findings highlight the protracted nature of lexical learning in children and their extended reliance on statistical positional information during early stage of lexical acquisition.