采用意义适合性判断任务，考察了语境位置对不同认知方式个体歧义句理解过程中抑制内部干扰的影响。结果发现：语境位置影响歧义句的歧义消解，对歧义句不适当意义的抑制作用方面，前语境优于后语境；在自然阅读的条件下，在句子加工的初始阶段，场独立个体能够更好的利用前语境信息抑制内部无关信息的干扰，而场依存个体在句子加工的初始阶段对前、后语境的阅读时间上不存在差异，而在句子加工的后期需用更多时间才能完成对内部干扰的抑制。

In the process of cognition, inhibition is defined as a mechanism that either prevents the irrelevant information from entering the working memory or eliminates the irrelevant information after it has entered working memory. The function of the inhibition is preventing access, deletion and restraint. With the development of related studies in inhibition, more attention had been paid to the special groups. At present, researches on individual differences in inhibition mainly focused on different groups with different ages, different comprehension capacities and different working memory capacities. Many studies demonstrated that cognitive style (field-dependence/ independence) was associated with the cognitive reconstruction, the span of verbal working memory and the spatial selective attention. Shouxin Li et al (2010) found that field independence (FD) individuals have higher reading efficiency than the field dependence (FI) individuals. This difference takes place in the latter period of working memory process. The purpose of present study is to investigate the inhibition mechanism differences between FD and FI individuals in inhibitory processes using different position of context. The Embedded Figure Test (EFT) was adopted to select FI and FD individuals. Totally 120 undergraduates were tested by the EFT. 16 typical FD participants and 16 typical FI participants were selected in the experiment 1. The same numbers of FD and FI participants were selected in the experiment 2. The experimental materials are 24 ambiguous sentences, 24 unambiguous sentences and 6 practice sentences. 24 ambiguous sentences are designed for four conditions: 6 pre-context ambiguous sentences, 6 pre-context unambiguous sentences, 6 post-context ambiguous sentences and 6 post-context unambiguous sentences. E-prime 1.0 was used to write procedure, and the computer (1.7 GHz CPU Pentium4.0, 17 inch display, resolution 1024 x 768) was used to present stimulation in experiment 1. Eyelink 2000 eye tracking system produced by SR Research Company in Canadian was used, and the pupil + corneal model were designed for experiment 2. The fixed speed verbatim rendering paradigm was used in Experiment 1. The reaction time and accuracy rate were recorded and analyzed. In Experiment 2 participants controls the speed and the sentences were presented as a whole. The first run dwell time and the second run dwell time were recorded and analyzed. 2 (cognitive styles: FD, FI) × 2 (positions of context: pre-context, post-context) × 2 (types of sentence: ambiguous sentence, unambiguous sentence) mixed design was used, and cognitive style was between group variable, position of context and type of sentence were within group variable. The materials were presented visually to the subjects. Eye movements were recorded by using a video-based eye tracker. Reaction time, accurate rate, the first run dwell time and the second run dwell time were analyzed with a three-factor mixed design ANOVA. The results indicate that: (1) Inhibitory efficiency is affected by the position of context. The inhibitory effect of pre-context is superior to post-context. (2) Under natural reading conditions, FI individuals are better able to take advantage of the pre-context information to suppress the internal interference in the early stage of sentence processing, while there is no difference in the reading time on the condition of pre-context and post-context to FD individuals. In the latter stage of sentence processing, FD individuals require more time to complete the internal interference suppression than FI individuals do.