学习方式如何影响新词语整合到语义网络的过程

PSYCH OpenIR > 认知与发展心理学研究室

	学习方式如何影响新词语整合到语义网络的过程
其他题名	How learning types influence the integration of novel words into semantic network
	陈双
	2014-05
摘要	在遇到一个新词语后，我们是如何学习新词语的意义，建立新词语的概念表征，并将其整合到已有的语义网络中的？不同的学习方式对这一过程有什么影响？本研究采用ERP 技术和语义启动范式，通过5 个实验，考察了上述问题。实验一考察了直接学习和推断学习两种学习方式对新词语建立的概念-特征表征的影响。新学习的词语在语义启动范式中作为启动词，目标词包括对应概念和特征相关词。结果发现在两种学习条件下，新词语都能启动概念词，而只有在推断学习条件下，新学习的词语可以启动特征相关词。这些结果表明学习方式影响了新词语的概念-特征表征，只有在推断学习条件下，新词语可以建立和长时记忆中相关特征之间的联系。接下来的实验进一步考察了通过推断学习方式获得的新词语如何建立和长时记忆中已知词语之间的语义联系。我们探讨了两种重要的语义关系：类别关系和主题关系。被试在推断词义时可能利用了学习语篇中提供的语义特征或者学习语篇中描述的情境。实验二的结果表明在特征推断学习条件下，新学习的词语不能启动主题相关词；实验四的结果则发现在情境推断学习条件下，新学习的词语只能启动情境相关的主题相关词，而不能启动无关情境中的主题相关词。通过这两个实验的比较，我们发现了学习新词语意义的过程中涉及的不同方面的信息影响了新词语和长时记忆中已知词汇之间的主题关系的建立。另外，实验五在同一个实验中比较了特征推断学习和情境推断学习两种条件对新词语的类别关系建立的影响。结果发现在视觉语义启动任务中，学习过程影响了新词语的类别关系，只有在特征学习条件下，新词语建立了和类别相关词之间的语义联系。然而在听觉语义启动任务中，两种条件下学习的新词语都启动了类别相关词。这表明学习过程在一定程度上影响了新词语的类别关系表征，在特征推断学习条件下，新词语和类别相关词之间的语义联系更强一些。另外，实验三表明已知词汇可以同时启动类别相关词和主题相关词。这表明新学习的词语受限于学习过程，只能获得部分的词汇知识。本论文通过这一系列实验发现了学习方式和学习过程对新词语的语义表征和语义关系的影响。我们发现新词语的概念可以通过直接学习或推断学习获得。而利用语义特征推断词义的过程中会激活长时记忆中的相关特征，导致新词语可以和这些相关特征之间建立联系。另外，推断词义过程中利用的信息影响了新词语和长时记忆中的已知词语之间建立的语义关系。利用描述的情境来推断新词语意义的时候，新学习的词语可以建立和学习情境相关的主题相关词之间的联系；而利用语义特征来推断新词语意义的时候，新学习的词语可以建立和类别相关词之间的联系。
其他摘要	Word meaning should be grounded in conceptual representation. When we encounter a novel word, how do we construct the representation of the novel word? How is the novel word integrated into the preexisting semantic network? And how do the different learning conditions affect the processes? In five experiments, we used ERPs in semantic priming paradigm to explore these questions. We explored how the learning conditions affected the concept-feature representation of the learned novel word in Experiment 1. We compared the direct mapping and inferred learning conditions. In a semantic priming paradigm, the learned novel word served as the prime, while the corresponding concept and feature related word served as the targets.We found that the novel word learned in both condition primed the corresponding concept. Nevertheless, the processing of feature related word was only facilitated by the novel word only in the inferred learning condition. These results suggest that the learning conditions affect the concept-feature representation of novel word. We further explored the integration of the learned novel word into the semantic network via inference in the following experiments. We focused on two kinds of semantic relationships: the taxonomic relation and the thematic relation. Participants inferred the word meaning from descriptive features or episodes. We found that the novel word learned from features could not prime the thematically related word in Experiment 2. However, in Experiment 4, the novel word learned from descriptive episodes could prime the thematically related word. However, the effect was constrained by the learning episodes. The results of these two experiments revealed that the learning of the thematic relation of novel word was affacted by the learning experience. In Experiment 5, we examined the taxonomic relation of novel word learned from descriptive features or episodes. In the visual semantic priming task, only the novel word leanrd from features primed the taxonomically related word, whereas in the auditory semantic priming task the newly learned word in both learning conditions primed the taxonomically related word. These results indicated that the learning processes influenced the taxonomic relation of novel word to some extent. The association between the learned novel word and the taxonomically related word was stronger in the feature learning condition than in the episodic learning condition. In addition, we found the known word could prime both taxonomically related word and the thematically related word. By comparing the known word and the newly learning word, we suggest that learner only acquried partial word knowledge of novel word in the constraint of the learning experience. In summary, we uncovered the role of learning conditions and learning experience on the representation and semantic relations of newly learned word. We found that the concept can be learned via direct mapping as well as inferred learning. Moreover, the inference of word meaning from features could activate some related features in long-term memory, leading to the connections between the novel word and the feature related word. In addition, the information involved in the inference processes affected the relations between the learned novel word and other known words stored in the long-term memory. The novel word learned from descriptive episodes could connect to the thematically related word in the same episode, while the novel word learned from features could connect to the taxonomically related word.
学科领域	基础心理学
关键词	词汇学习语义表征概念特征类别关系主题关系学习方式
学位类型	博士
语种	英语
学位专业	心理学
学位授予单位	中国科学院研究生院
学位授予地点	北京
文献类型	学位论文
条目标识符	http://ir.psych.ac.cn/handle/311026/19491
专题	认知与发展心理学研究室
作者单位	中国科学院心理研究所
推荐引用方式 GB/T 7714	陈双. 学习方式如何影响新词语整合到语义网络的过程[D]. 北京. 中国科学院研究生院,2014.

条目包含的文件
文件名称/大小	文献类型	版本类型	开放类型	使用许可
陈双－博士学位论文_.pdf（4453KB）	学位论文		限制开放	CC BY-NC-SA	请求全文