叙事是人类信息传递的重要载体，而口口相传则是社会信息传播的重要途径。然而，故事是如何随着口口相传而发生改变的我们仍不清楚，其背后的神经基础也仍有待探讨。本论文采用“链式传递”范式，结合功能磁共振成像技术，在高生态效度的故事自然传递场景下，通过三个研究对这一问题进行回答。
在研究一中，我们结合自然语言处理和网络分析的研究方法，探讨故事的不同成分是如何随着代际传递而不断发生改变的。结果表明，故事的结构相比于故事内容来说具有更高的保真度，即故事的内容更容易随着传递而发生形变，其中，高中心化的核心情节维持稳定，而低中心化的次要情节更容易发生形变。研究二对其背后的神经基础进行了进一步探索，结果表明左侧角回参与了故事结构的传递，而故事内容传递的表征与右侧中央前回和左侧视觉皮层相关。此外，功能连接模式分析表明，海马前部与默认网络子网络的连接模式相似性可以跨代际预测结构和内容的高保真传递。通过模型比较，我们还发现这种功能连接模式相似性对故事结构与内容的高保真传递的预测发生在“听故事”和“讲故事”的不同阶段。研究三进一步进行探索性分析，探讨了形变的故事的表征构建如何影响了对于原始故事表征的重构，通过组内-组间一致性差异分析，表明右侧背内侧额叶受到先验形变故事的影响，进而影响了后续表征重构，且故事的形变程度可以预测神经改变程度。
综上，本研究对叙事传递的认知模式和神经基础进行了系统性探索，结果表明故事的结构和内容传递很有可能涉及了情景建构、社会学习等不同的认知加工过程，且故事结构与内容在大脑上表征模式存在分离，默认网络的不同子网络也在其中扮演着不同的角色。最后，背内侧额叶区域会受到先验形变表征的影响，进一步影响后续原始故事的理解。本研究对于信息传播、文化演化与谣言澄清等研究与应用有着重要的启发性意义。

Narrative, the transportation of culture, plays a key role in cultural accumulation and evolution. However, the cognitive and neural mechanisms of multi-generation narrative propagation (i.e., in a linear transmission chain) are not well understood.
In the first study, by combining the transmission chain design with the computational language model and network analysis method, we tried to explore the dynamic changes of different narrative components (structure and content) throughout the transmission chains. Results showed the structure of the story sustained higher fidelity than the content, which means the content might be distorted much more easily than the structure. Moreover, high centrality events showed higher fidelity in all 4 generations and in both chains than low centrality events
Further, using fMRI technique and the method of Inter-Subject Representation Similarity Analysis (IS-RSA), we expected to answer how our brains represent different components of narratives and cope with the narrative transmission. Results showed that the left angular gyrus was involved in the transmission of story structure, and the transmission of content was involved with the right precentral gyrus and left visual cortex. Functional connectivity pattern analysis showed the connectivity pattern similarity between the anterior hippocampus and subnetworks of default mode network could predict the high-fidelity transmission of structure and content across generations. Through model comparison, we found that narrative comprehension and production played distinctive roles in the prediction of narrative transmission.
Study 3 explored how the representational construction of the deformed story affected the reconstruction of the representation of the original story. The intra-group consistency difference analysis showed that the right dorsal medial prefrontal cortex was affected by the prior deformed story, which in turn affected the subsequent representations reconstruction, and the deformation of the story can predict neural changes.
In conclusion, this study systematically explored the cognitive model and neural basis of narrative transmission. This research provides evidence for the studies and application of information propagation, cultural evolution, and rumor clarification, and fills the gap of previous research.