话题结构对语篇情绪更新的影响

The Influence of Topic Structure on Emotional Updating in Discourse Context

读者在阅读理解的过程中,能够推断出故事主角的情绪状态,并能随着语篇中的情绪变化做出情绪更新。在加工情绪信息的同时,读者也加工语篇结构方面的信息,二者皆有可能随着语篇的展开而发生变化。本研究采用自定步速阅读的方法,分别在外显和内隐两种情绪加工任务下,考察话题结构对语篇情绪更新的影响。结果发现,在实验1的外显情绪判断任务下,话题结构未显示出对语篇情绪更新的作用;在实验2和实验3的内隐情绪理解任务下,话题延续时,情绪转换句的阅读时间长于情绪延续句,此时情绪更新需要额外的加工时间;而话题转换时,二者没有显著差异,说明此时读者在新结构下建立当前句的情绪表征,并不受到先前情绪的影响。

It is shown that readers can make inferences of the protagonist＇s emotions during reading comprehension, and they can update their emotional representations when the emotions in the discourse shift （e.g., from positive to negative）, reflected by increased reading time on the emotion-shifted sentences as compared with the emotion-continuation ones. However, during reading comprehension, not only the emotional information is processed, but also the information about discourse structure is processed by readers, and these two factors may both vary with the development of the story. So the present study was interested in whether topic structure could make influence on emotional updating in discourse context. Besides, through explicit and implicit emotional tasks, the present study investigated whether the influence of topic structure on emotional updating was modulated by the processing of emotional information. In our three experiments, two factors were manipulated： topic structure （topic-continuation, topic-shift） and emotion （emotion-continuation, emotion-shift）. Through the explicit task in experiment 1, participants were instructed to read 48 critical discourses and 72 fillers （each discourse consisted of two sentences） using a sentence-by-sentence self-paced reading procedure, and were asked to choose a number from 1-9 to make judgment of the discourse emotion （1 represented for the most negative, 9 represented for the most positive）. Then the rating grade and the rating time were recorded. Also, the reading time of the second sentence was recorded. In experiment 2 ＆ 3, through the implicit task, participants were instructed to answer a comprehension question after reading 1/3 of the materials. In experiment 2, the reading time of the second sentence was recorded. In experiment 3, besides the second sentence, there was a third sentence which included an emotional word. Therefore, the reading time of the emotional word and the word following it were recorded. The results from Experiment 1 showed that for the explicit task where subjects were asked to make emotional judgment, topic structure did not affect emotional updating. However, the results from Experiment 2 and 3 showed that for the implicit task where subjects were asked to read for comprehension, topic structure significantly affected emotional updating in such as way that emotion-shifted sentences were read longer than emotion-continuation sentences in the topic-continuation conditions, but not in the topic-shifted condition. This suggested that the process of emotional updating was disrupted in the topic-shifted condition. To conclude, through the explicit and implicit emotional processing tasks, the present study conducted three self-paced reading experiments to investigate how topic structure influenced emotional updating in discourse context. It was shown that topic structure did not affect emotional updating under the explicit task. However, under the implicit task, emotional updating needed extra processing time for the topic-continuation conditions, but not for the topic-shifted conditions. These results provide evidence for a strong effect of discourse structure on emotional processing in discourse context.