中文摘要：

采用自定步速阅读方法与图片核证范式（pictureverificationparadigm）探讨状态不确定独立否定句（如“裙子不是红色的”）的动态表征过程，包括3个实验。结果发现：（1）在否定加工的初期，即250ms间隔时间条件下，读者表征事件的否定状态；肯定句已经通达事件的实际状态。（2）在否定加工的中期，即750ms间隔时间条件下，读者仍然表征事件的否定状态，但已开始对事件的备择选项进行搜索；肯定句仍然保持对事件实际状态的表征。（3）在否定加工的后期，即1500ms间隔时间条件下，读者表征否定标记与事件的否定状态，同时完成对备择选项的搜索：肯定句依旧保持对事件实际状态的表征。据此，本文提出的锚激活与限制满足模型（anchor—basedactivationandsatisfactionconstrainedmodel）得到初步证明。

英文摘要：

Affirmation and negation are semantic and grammatical categories present in virtually all languages. Many different theories have been proposed to explain the processing mechanism of negation. The schema-plus-tag model and the two-step simulation hypothesis have been accepted by many researchers. However, neither of these hypotheses cannot explain the processing mechanism of isolated negative sentences with uncertain states,such as the skirt is not red. In this study, we proposed anchor-based activation and satisfaction constrained model, to solve this problem. The sentence-verification paradigm was adopted in three experiments to explore the processing mechanism of negated sentences with uncertain states. In these experiments, participants were presented with affirmative or negative sentences with target objects, such as the skirt is/isn ＇t red. After reading the sentences with a delay condition of either 250ms, 750ms, or 1500ms, pictures were presented instantly. The task of participants was to verify whether the objects in the pictures appeared in the previous sentences. In these experiments, we designed three kinds of probing pictures. The first kind was ＂N＂, which depicted the negated state of affairs of negated sentences, such as a red skirt for the sentence The skirt is not red; the second kind was ＂x＋N＂, which depicted the actual state of affairs of negated sentences, such as a red skirtplus negated marker （x）; the last kind was ＂A＂, which depicted one of the alternative states of affairs of negated sentences, such as a blue skirt. Results of Experiment 1 indicated that both of the data patterns of the affirmative and negated sentences were ＂N = x＋ N, x＋ N〈A and N〈A＂, which suggested that, in the 250ms delay condition, the actual state of affairs were simulated in affirmative sentences and the negated state of affairs were simulated in negated sentences. Results of Experiment 2 were similar to those of Experiment 1, the data pattern of both kinds of sentence were also ＂