| On the basis of previous researches, we conducted three experiments, five sub-experiments, to explore the processing manner of compound word and provide some evidences for the debate of serial model and parallel model. The critical question about the processing manner of compound word was whether the two characters were processed one by one or at the same time. Meanwhile, whether words in a line are processed one by one, or two or more words could be processed in parallel, is a major dispute between serial attention shift (SAS) and distributed lexical processing models of eye movement control. The present study aimed to test the two types of models by examining their different predictions on preview processing during reading:(1) the level of information from parafoveal processing; (2) whether information extracted from parafovea could exert influences on the processing of foveal.Research 1, for the purpose of studying the time course of when the parafoveal word and foveal word were or not a whole word, we discussed the parafoveal processing of monomorphemic word, compound word and phrase in normal Chinese reading. Using the boundary technique, the preview of a character was either correct or was visually dissimilar and meaningless but resembling a Chinese character. An invisible boundary was set between the two constituents; when the eyes moved across the boundary, the previewed character was changed to its intended form. We observed higher preview effect for monomorphemic word meanwhile the equal preview effect for compound word and phrase. Moreover, we found parafoveal-on-foveal effect for monomorphemic word. The results showed the higher predictability for monomorphemic word induced the larger papafoveal effect. The uncertainty for one word and the complexity of structure for compound word leaded to similar parafoveal processing for compound word and phrase.Research 2, using the boundary technique, we examined the effect of structure and characteristics of compound word for the parafoveal processing. An invisible boundary was set between the two constituents; when the eyes moved across the boundary, the previewed character was changed to its intended form. Experiment 1, we studied the parafoveal processing of adjective-noun and noun-noun compound word. We also adopted two preview conditions:identical and inconsistent. We found larger preview effect for adjective-noun compound word compared to noun-noun compound word. And more importantly, we have found marginal significant parafoveal-on-foveal effect for both kinds of words. Experiment 2, readers'eye movements were monitored as they read Chinese two-constituent compound words in sentence contexts. The first compound word constituent was either an infrequent character with a highly predictable second constituent or a frequent character with an unpredictable second constituent. The parafoveal preview of the second constituent was manipulated, with four preview conditions:identical to the correct form; a semantically related character to the second constituent; a semantically unrelated character to the second constituent; and a pseudo-character. An invisible boundary was set between the two constituents; when the eyes moved across the boundary, the previewed character was changed to its intended form. The major findings were:(1) A visuo-orthographic parafoveal-on-foveal preview effect when the second constituent was highly predictable from the first constituent; (2) a semantic parafoveal effect for the high-constraint but not for the low-constraint condition.Research 3, in an eye-tracking experiment we examined the influence of adding word demarcation between the words on preview effects during Chinese sentence reading. Using the boundary technique, the preview of a character was either correct or was visually dissimilar and meaningless but resembling a Chinese character. An invisible boundary was set before the target word; when the eyes moved across the boundary, the previewed character was changed to its intended form. Experiment 1, we examined the influence of adding word spaces between the words on preview effects. Besides the preview manipulation (identical or inconsistent), two further manipulations were implemented:(a) the target word was one-character or two-characters (compound word) in length and (b) the sentence was either unspaced, or spaced with the boundary location before or after the blank space preceding the target word. Experiment 2, we examined the influence of adding word demarcation on preview effects by using a highlighting manipulation. Besides the preview manipulation (identical or inconsistent), two further manipulations were implemented: (a) the target word was one-character or two-characters (compound word) in length and (b) the sentence was either normal or highlighted. We observed a larger preview effect in the word spaced condition compared to the unspaced and highlighted conditions for one-character, but not two-character targets. Meanwhile, we did not found significant parafoveal-on-foveal effect. The lack of effect for the two-character words was likely caused by subjects mis-segmenting the target in the unspaced condition. In all likelihood, the larger preview effect for one-character targets in spaced compared to unspaced text occurs due to a combination of reduced lateral masking and a veridical cue to the short length of the word.Overall, our results indicate preview can obtain visuo-orthographic, lexical and even semantical information. More importantly, readers can obtain faint and instability parafoveal-on-foveal effect. Preview effect is affected by the structure and characteristics of the word. The results support the compound words may be processed compositionally. In all, we conclude that readers can process more than one word at the same time during Chinese reading. The data from these experiments fit better with the theoretical assumptions of parallel processing models of eye movement control in reading.
|
PDF Full Text Request