时间顺序标码的层次网络表征研究

【作者】 李伯约；

【导师】 黄希庭；

【作者基本信息】 西南师范大学 ， 基础心理学， 2001， 博士

【摘要】 任何事件都发生在某个特定的历史时间，因此在对事件编码时，就附着上了时间信息作为其时间标记。这种标记可以是一种直接表示事件的顺序的符号，本研究称之为顺序标码。时间顺序标码是按照层次网络的结构组织起来的。通过对这种层次网络组织中的时间顺序标码进行检索这种特殊的推理方式，就可确定该事件在过去经历过的事件中的顺序。因为这种时间信息的顺序标码的多层次网络组织有什么特点，以及人们怎样在这种组织中进行进行检索加工，是人们的时间认知的一个很重要的方面，所以本研究力图找到新的思路与视野来对之进行研究。通过本研究的结果，可以对时间顺序标码的层次网络的组织得出如下认识： 首先，不按序呈现的时间信息的顺序标码同按序呈现的时间顺序标码以及按序呈现的关系标码之间的效应是相同的，是人们对各个项目的先后顺序认知的主要依据。因此，以不按序呈现的时间顺序标码为刺激材料，进行实验，既可以保持同过去的研究之间的继承性，又可以避开语义顺序的干扰。但更重要的是，因不按序呈现顺序标码的分布范围较广，不受工作记忆的限制，方能有效地探讨时间顺序标码的多层次的网络组织。再者，时间顺序标码的表征是言语表征而不是空间表象表征，这说明被试没有在心理上重新排列各个项目，而是对其标码直接进行推理判断。总而言之，这种范型是通过推理来探讨时间顺序标码层次网络组织的有效的手段。 人们为了提高时间认知的效率，只要项目的数目达到几个时就要开始对这些时间信息进行分组，形成组块即小组，然后再把小组结合为大组，进而组成更大的组，以至于形成一个具有层次网络结构的组织形式。当若干的项目被结合为没有明确的边界的小组时，若两个项目之间间隔达二个项目时，就会出现距离效应；当若干的项目被结合为有明确的边界的小组时，两个组的项目之间就会出现越界效应；当层次网络形成 时，即对不同层次的项目之间就会出现层次效应。这些效应表明，时间顺序标码是按 照层次网络的结构组织起来的，若干项目组成一个小组，形成了高一级的层次；若干 小组组成更大的组，形成更高一级的层次。在这个层次网络结构中，高层的信息控制 着低层的信息。高层次的顺序信息也可通过连线而为低层次的项目自动继承。而在时 间顺序标码的表征中蕴涵着这种层次网络信息。因此，一旦知道一个项目的时间顺序 标码的层次网络表征，就可知道该项目同其它项目的先后关系。本研究的结果，同 Coffins和Quillian首倡的语义层次网络模型，以及Lee和Estes等人的主观时间信息的 层次网络的模型、mttenlocher等人的客观时间信息的层次网络的模型基本上是一致。叭 的。 对两个项目的先后顺序的推理，是以对其标码的表征按层次进行直接比较的加工方 式为主，必要时也采用运算的加工方式。山于对时间顺序标码的层次网络表征的加工 是从顶层向底层逐步进行的，因此两个项目之间的距离越大，越在高层次上分开，就 只需要在越高的层次上进行加工，涉及到的层次就越少，反应时越短，反应的正确率 越高。反之，同一组块内的两个相邻的项目，因需要一直加工到最底层，涉及到的层 次最多，反应时必然最长，正确率也最低。这一结果，回答了用距离效应来对层次网 络结构的质疑。 对两个项目的先后顺序的椎理，是以对其标码的表征按层次进行直接比较的加工方 式为主，必要时也采用运算的加工方式。 在线性的时间的基础上形成的时间顺序标码层次网络组织是人为的，而且由于各个 项目的标码之间没有明确的特征上的差异，组成层次网络的依据则是各个项目的标码 在顺序上的接近性，因而就不可避免地具有自己的独特之处，这表现在对时间顺序标 码层次网络表征的加工具有自动性、内隐性和模糊性上。被试对于没有明确的层次、 组界的连续的时间顺序标码进行分组、形成层次网络的加工，都是自动进行的，而且 这种小组乃至层次网络的存在是内隐的，是没有为被试所意识到的。个体对时间顺序 标码的层次网络表征可以进行模糊的加工。这种模糊的加工导致了具有连缕性的时间 顺序标码的层次网络的模糊边界的特殊问题。当然，对时间顺序标码层次网络表征的 加工也可具有控制性、外显性和糟确性的特点，且其加工的效果分别优于自动的、内 隐的和樱糊的加工的效果。 由于对项目之间的挂近性的把扛是因人而异的，甚至于就同一个体而言，也会因其 经验或其它因景的不同而产生相当的差异，时间顾序标码的层次冈络表征的最根本的 特点就是它的主观性。即使是依据日历标码而形成的“客观的”时间顺序标码的层次 网络表征，也要受到这种主观性的制约。如果一个客观更多还原

【Abstract】 Any event happens at a certain time, so the information of this time is attached to the code of the event as its temporal marker. If this marker is a symbol to denote the order of the event directly, it can be called temporal sequential marker in this research. Each temporal sequential marker has a hierarchical network organization information. Through searching temporal sequential markers in such hierarchical network, a special form of reasoning, the orders of events can be decided. The natures of hierarchical network structure of such temporal sequential markers, and the process of searching under such structure are very important for human cognition. Therefore, in this research, we strove to find a new paradigm and perspective to the studies of them. On the basis of the findings of this research, the following comprehension of temporal sequential markers can be gotten:At first, temporal sequential markers with unordered display have the same effect of temporal sequential markers and relational markers with ordered display, therefore, they can be the main basis for human cognition of the order of events. Therefore, temporal sequential markers with unordered display as stimulus are most suitable for conducting successive experiments with the previous researchers, and are exempt from the disturbance of semantic orders. Temporal sequential markers with unordered display can be distributed more broadly to go beyond the limitation of working memory, so they are uniquely useful to explore structure of hierarchical network of t~emporal sequential markers. The representation of temporal sequential ~~iarkers is verbalSrepresentation, and this means that the items are not rearranged in the mind by the ~ubjccts, so they axe judged directly through their markers. In general, the paradigm used in this research is a useful method to explore temporal sequential markers undei?hierarchical network structure.Hr raising the efficiency ot the cognition of timc, the people should combine items into one chunk, this means, into a group when there are several Lens, and then, combine several smaller groups into one bigger group, and so on and so forth. This combination will lead to an organization with the hierarchical network structure. When several items are combined into groups?without a clear boundary, the distance effect will exist between two items with an interval of other two items. When several items were combined into groups with a clear boundary, the boundary effect will exist between two items in different groups. And when a hierarchical network is formed, the hierarchical effect will exist between two items in different layers. All these effects shows that temporal sequential markers are organized in accordance with the hierarchical network structure, that means, when several items are combined into a group, a higher hierarchy is formed: when several groups are combined into a bigger group, an even higher hierarchy is formed. In this hierarchical network organization, the information of upper layers is in the control over the information of lower layers. And the items of lower layers can obtain the sequential information of upper layers. In the representation of temporal sequential markers, such hierarchical network information is comprised. Therefore, once an item抯 hierarchical network representation of temporal sequential marker is available, the order of this item with other items can be judged. At last, the findings of this research are consistent with the semantic hierarchical network model initiated by Collins and Quillian, and consistent with the hierarchical network models of subjective temporal information by Lee and Estes, and of subjective temporal information by Hut tenlocher.The method process of direct judgment is taken mainly for the reasoning of the order of two items, but at necessary time, the process of calculation is taken also. If two items are separated by longer distance, they should be9divided at upper layer. Because the process of the hiera更多还原