中文摘要：

采用眼动记录技术，考察不同背景音对中文篇章阅读和词汇加工的影响。结果发现：（1）与安静和白噪音背景相比，无关言语背景下的总阅读时间更长、阅读速度更慢、注视次数更多、平均眼跳幅度更小。（2）不同背景音下的首次注视时间和凝视时间无显著差异，而无关言语背景下的回视路径时间和总注视时间较白噪音条件下的更长。结果表明，无关言语干扰了读者的阅读过程，其语义成分和声学特征变化是干扰产生的重要原因；这种干扰体现在词汇加工中的语义整合阶段。

英文摘要：

Reading is an important way to obtain information, and also a complicated cognitive activity. Most readers have to read in quiet environment, while background noises are unavoidable. what the influences of these noises on our reading, and what is the underlying cause？ A lot of researchers have dedicated to this issue, and found that the irrelevant speech seriously interfere with the process of reading. Some of them suggested that the semantic information in ISE disrupts our reading （Tremblay, Nicholls, Alford, ＆ Jones 2000; Oswald, 2000）, while the role of acoustical variation in sound on reading has been ignored. Moreover, the previous researches only concerned the overall influence on reading, tell us little about the way the semantic processes involved in the reading task are disrupted by background speech. So the present study was to investigate the effects of background noise on passage reading and word recognition, then conform what is the actual mechanism of ISE on reading. The experiment design was a single factor 3 levels （background noise： irrelevant speech, white noise and silence） within-subjects design. For the passage materials, there are 9 Chinese expositions, 3 passages for practicing and 6 for experiment. For the background noise, irrelevant speech was the dialogue chosen from the TV plays in the past 3 years. For the target words, 144 two character words were chosen for further analysis. During the experiment, 27 participants were asked to read passages under three background noises （irrelevant speech, white noise, silence） which were displayed according to the order of Latin square. The audio materials were played by Window Media Player on Acer 5750G computer, and participants listened to the sound with SONY MDR-E9LP stereo headphone. The eye movements were recorded with a SR Research EyeLink 2000 eye tracker. The sampling rate was 1000Hz. The distance between the subject and the screen was 65 cm and each Chinese character was 28×28 pixels and subtended 0.96°. The results we