Do chinese readers obtain preview benefit from word n + 2? Evidence from eye movements.

Abstract

The boundary paradigm (K. Rayner, 1975) was used to determine the extent to which Chinese readers obtain information from the right of fixation during reading. As characters are the basic visual unit in written Chinese, they were used as targets in Experiment 1 to examine whether readers obtain preview information from character n + 1 and character n + 2. The results from Experiment 1 suggest they do. In Experiment 2, 2-character target words were used to determine whether readers obtain preview information from word n + 2 as well as word n + 1. Robust preview effects were obtained for word n + 1. There was also evidence from gaze duration (but not first fixation duration), suggesting preview effects for word n + 2. Moreover, there was evidence for parafoveal-on-foveal effects in Chinese reading in both experiments. Implications of these results for models of eye movement control are discussed.

Wordn + 2 preview effects in three-character Chinese idioms and phrases

“…Experiment 1 was an attempt to replicate findings from previous studies on n+2 preview benefit in Chinese reading, using the same stimuli as Yang et al (2009) and Yang et al (2012). The reason for including this replication of "benchmark" n+2 preview effects was to be certain that parafoveal processing was occurring in the way that we would anticipate based on previous findings.…”

“…For example, if we obtained null n+2 preview effects in both the idioms and matched phrases we wanted to be sure that it was due to the particular way the idioms were being processed rather than our participants simply not extracting information from word n+2 regardless of how favourable the circumstances were. Replicating Yang et al's (2009) effect would allow us to reject the latter explanation. Conversely, if we observed n+2 preview effects in both the idioms and phrases we wanted to ensure that it was due to some characteristic of our stimuli that differed from those used by Yang et al (2012), rather than our participants extracting a greater deal of parafoveal information than those tested in prior studies.…”

“…Finally, we hypothesised that the idioms should be read faster than the matched phrases, even with the predictability of word n+2 being controlled. In addition to investigating whether idioms are processed as single lexical units during reading, we also attempted to replicate the findings of Yang et al (2009) and Yang et al (2012) (i.e. word n+2 preview effects are only observed in Chinese text when word n+1 is highly frequent) in the same testing session as our main study.…”

“…Yang, Wang, Xu, and Rayner (2009) manipulated the previews of the second and third character after a boundary, which together formed a single word. This word was preceded by a single character high-frequency word, and significant preview benefits were observed.…”

Parafoveal processing across different lexical constituents in Chinese reading

“…This means that parafoveal linguistic information is closer to the point of fixation in more than less dense languages. It is not surprising, therefore, that many studies have shown stronger and earlier preview effects in Chinese reading than in English reading (Wang, Tong, Yang, & Leng, 2009;Yan, Richter, Shu, & Kliegl, 2009;Yang, Wang, Xu, & Rayner, 2009;Yen, Radach, Tzeng, Hung, & Tsai, 2009). Thus, Chinese has three important properties that make it an extremely interesting language in which to consider how readers carry out morphological processing of words when they fall both in foveal and parafoveal vision.…”