中文摘要：

本研究采用边界范式,控制单字词N＋1的预视加工负荷和词N＋2首字的预视信息类型,探讨了汉语阅读中词N＋2的预视对高频单字词N＋1加工产生的影响。结果发现,当词N＋1的预视为低加工负荷时,词N＋2首字的预视对词N＋1的加工产生了显著的影响;当词N＋2首字的预视字与词N＋1正字法相似,并且只呈现在副中央凹中时,显著的阻碍了词N＋1随后的加工。结果表明,副中央凹中词N＋2能够获得预视加工,可能获得正字法预视信息并对随后词N＋1的加工产生了影响,并且这种影响受词N＋1的预视加工负荷的调节。实验结果符合注意梯度指引模型的假设。

英文摘要：

Whether readers can obtain preview information from word N＋2 in the parafovea is a key problem for solving the controversy of whether words in the perceptual span are processed in serial or in parallel. Several studies using the boundary paradigm have demonstrated that the preview of word N＋2 influences fixations on word N＋1（Kliegl, 2007; Yan, 2010; Yang, 2009; Risse, 2012）, but it has not been investigated what kind of preview information drives these effects. The present study used the boundary paradigm to explore the influence of the preview of word N＋2 on the processing of word N＋1 during Chinese reading. Three experiments are reported. In experiment 1, we manipulated the previews of a single character high-frequency word N＋1 and the first character of word N＋2. The results showed that fixation durations on word N＋1 given an identity preview of word N＋1 were significantly longer given an identity compared to non-word preview of word N＋2. In experiment 2, we manipulated the previews of word N＋1 and the first character of word N＋2, and controlled the identity previews of both characters to be orthographically similar. We hypothesized that the orthographic information gained from the first character of word N＋2 in the parafovea would affect the processing of word N＋1. Our results showed no effect of the N＋2 preview. However, due to the orthographically similar character N＋2 being available during fixations on both words N and N＋1 we were unable to determine if our effect was driven by it being processed far into the parafovea during fixations on word N, or closer to the fovea during fixations on word N＋1. To address this possibility we conducted experiment 3. We used the same manipulation as in experiment 2, but ensured that the orthographic similarity between the two characters was only present during fixations on word N. In this experiment there was orthographic overlap between the identity preview of word N＋1 and the false preview of word N＋2. As such, the orthograp