Abstract: There are more characters than the syllables in Chinese. That means there will be a lot of homophonic characters and phonologically similar characters. It is an interesting question that how the phonologically similar or homophonic characters influence the Chinese reading. Tongue twisters are sentences that contain a number of words with the same initial consonants and/or the same vowels. Previous studies have found the tongue twister effect (TTE). The tongue twisters took more time than the normal sentences both in silent reading and oral reading. Few studies used eye tracking technique to explore the TTE. The current study employed two experiments to investigate the TTE in Chinese reading and effect of the tongue twister on the perceptual span in Chinese reading. In Experiment 1, the experiment design was one-way (sentence types: tongue twister, normal sentence) within-subjects design. There are 56 tongue twisters and 56 normal sentences. Normal sentences were matched with tongue twisters in terms of the syntax structure, number of character strokes and character frequency. A total of 20 undergraduate students (they did not participate the following experiment) rated the materials on difficulty, naturalness and plausibility. There is a significant difference between the tongue twister and the normal sentence in terms of the sentence difficulty while there are no difference between the tongue twister and the normal sentence in terms of the naturalness and plausibility. 19 undergraduate students participated in the experiment. Their eye movements were recorded with SR Research Eyelink 2000 eye tracker. The results showed an effect of tongue twister，that is there were longer reading time, slower reading rate, more fixations and shorter mean saccade length for tongue twisters than for normal sentences. In Experiment 2, the experiment design was 2 (sentence types: tongue twisters, normal sentences) × 6 (window size: 1characters, 3characters, 5characters, 7characters, 9characters and full line) mixed design. The sentence type was the between-subjects factor, and the window size was the within-subjects factor. The sentences used in the experiment 2 are the same as the Experiment 1 and they were rotated in Latin square. A total of 50 undergraduate students participated in the experiment. Their eye movements were recorded with SR Research Eyelink2000 eye tracker. The result shows that the perceptual span was one to two characters to the right of the fixated word the tongue twister and the perceptual span was two-three character to the right of the fixated word for the normal sentence. The overall results show that the perceptual span of tongue twisters became smaller compared to the normal sentences. In conclusion, the present study provides further evidence to demonstrate the TTE in Chinese reading and this effect influences the perceptual span in reading. Why do the tongue twisters decrease the reading rate and perceptual span? This is related to the inner speech. Due to the overlap of the phonemic representations automatically activated during reading, tongue twisters slow down the reading speed and interfere with comprehension and memory (Rayner et al., 2012). The tongue twister effect increases the difficulty of sentence reading, so the perceptual span was smaller.

Key words: tongue twister effect, Chinese reading, perceptual span

摘要： 绕口令效应指无论朗读还是默读，绕口令的阅读时间都长于正常句子的阅读时间。本研究采用Eyelink 2000型眼动仪，通过两个实验考察汉语阅读中的绕口令效应及其对知觉广度的影响。实验1结果表明：与阅读非绕口令相比，绕口令的阅读时间和注视次数均显著增加而阅读速度和平均眼跳幅度均显著下降。实验2结果表明：绕口令的知觉广度为注视点右侧1~2个汉字，控制句的知觉广度为注视点右侧2~3个汉字。结果说明汉语中存在绕口令效应，使读者知觉广度变小，阅读效率下降。

关键词: 绕口令效应, 汉语阅读, 知觉广度