Abstract:

In alphabetic languages, such as English, the spaces between words are one kind of cue of visual word segmentation that guides the reader in selecting the word center as the saccadic target. Previous studies have shown that in English reading, such visual word boundaries can be extracted from parafoveal vision. However, no visual word boundary serves as a saccadic target for selection in Chinese reading. Most recently, several researchers have claimed that Chinese readers adjust their saccade length to accommodate lexical processing, rather than moving their eyes to specific saccadic targets. The properties of fixated words affect the subsequent saccadic target selection. For example, the saccade length is longer for high- than for low-frequency words. Moreover, parafoveal processing also has an important influence on the modulation of saccade length. Some researchers have nonetheless found that the load of fixated word processing modulates the processing of parafoveal words, where readers then adjust the selection of the next saccade target, though several studies have failed to find an interaction between the fixated word and foveal and parafoveal processing.

Given the evidence that the frequency of fixated words and the strokes of parafoveal words significantly influence the upcoming saccade length, in the present study, the frequency of fixated words in foveal processing and the strokes of parafoveal words were manipulated to explore the relationship between foveal processing and parafoveal processing on saccade target selection. If the load of processing of fixated words modulates the processing of the word in parafoveal vision, then we predicted that the effect of parafoveal word strokes from adjusting the current saccade length would be more pronounced when the fixated word is processed in high- rather than in low-frequency foveal processing.

The results showed that participants fixated on low-frequency words for a significantly longer time than for high-frequency words; they fixated on parafoveal words with a low number of strokes for a significantly shorter time than those with a high number of strokes; the saccade length of high-frequency words was longer than that of low-frequency words; the upcoming saccade length in parafoveal vision was significantly longer for words with a low than a high number of strokes; and the location of the initial fixation on words with a low number of strokes was closer to the word centre than with words with a high number of strokes. Unfortunately, we failed to observe significant interactions between these two factors for any eye movement measures used in the present study.

The results indicate that the frequency of fixated word and the number of strokes of parafoveal words independently affect the selection of saccadic targeting in Chinese reading.

Key words: Chinese reading, saccade-target selection, foveal processing, parafoveal processing