580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How do neuropsychological and cognitive processes underpin narrative comprehension and production?
This theme explores the neural substrates and cognitive mechanisms underlying the human abilities to understand and generate narratives, highlighting the overlap and interrelation between comprehension and production. Understanding these processes is crucial as narrative functions are fundamental for communication, memory, social interaction, and clinical interventions.
Key finding: This review integrates neuroimaging and patient research identifying a consistent network—frontal, temporal, and cingulate regions—involved in both narrative comprehension and production. It establishes that working memory... Read more
Key finding: The paper introduces the concept of the 'deictic center'—a mental spatio-temporal-personal reference point constructed by readers—to explain how narrative comprehension is achieved. This cognitive construct allows readers to... Read more
Key finding: This study formalizes narrative comprehension as the integration of explicit textual information with implicit background world knowledge via preference-based argumentation. It demonstrates that successful comprehension... Read more
Key finding: Empirical results show that discourse cue placement patterns, observed in expert human tutors' speech, significantly influence narrative text comprehension. In particular, core-contributor ordering facilitated processing... Read more
2. How can computational methods model and generate narratives to enhance AI narrative intelligence?
This theme focuses on computational approaches to automatically understand, manipulate, and generate narrative text and structures. It addresses challenges such as preserving coherence, temporal and causal ordering, and narrative richness through AI and cognitive architectures, with applications including storytelling, educational technology, and video summarization.
Key finding: Introduces the Narrative Reordering task where an existing narrative is rewritten into a different temporal order while preserving plot coherence. The paper presents a human-annotated dataset and demonstrates that narrative... Read more
Key finding: Proposes narrative as a fundamental cognitive model for knowledge representation beyond traditional scripts and cases, emphasizing three narrative features—transformation, conflict, and unactualized events—that enable richer,... Read more
Key finding: Presents an integrated architectural approach combining micro and macro narrative generation mechanisms, supported by AI and neural network technologies including deep learning. The study develops narrative simulation systems... Read more
Key finding: Defines a goal-driven narrative generation framework that produces coherent narratives from long-form video data, emphasizing comprehensive user goal understanding and open-domain input handling. The paper bridges social... Read more
Key finding: Surveying AI systems and tools developed since the 1970s for narrative understanding and generation, the paper highlights conceptual dependency frameworks and story grammars that enable goal-driven and knowledge-based... Read more
3. What roles do narrative competence and embodied narrative understanding play in literacy, education, and human-AI collaboration?
This theme examines narrative competence as a foundational skill influencing literacy development, learning, and human understanding. It also investigates human-AI co-creative writing experiences and embodied cognition perspectives that view narrative as an emotional and motivational scaffold, contributing to educational outcomes and reconfiguring human interactions with AI systems.
Key finding: Argues that narrative competence—defined as identifying, interpreting, creating, and communicating multi-layered narratives—is an essential but overlooked foundation underpinning broader literacy skills. The paper highlights... Read more
Key finding: Demonstrates that narrative understanding emerges from embodied, affective interactions beginning in fetal development, wherein action-based 'stories' build sensorimotor intelligence. These affectively driven, self-generated... Read more
Key finding: Through autoethnographic studies utilizing GPT-3, the paper reveals the challenges and potentials of human-AI co-writing in fiction. It finds that while AI lacks semantic coherence for independent creativity, human-AI... Read more
Key finding: Provides an historical overview of narrative intelligence in AI, emphasizing the foundational role of narrative in human cognition and learning from childhood onwards. It documents early AI efforts in story understanding and... Read more
Key finding: Utilizing behavior analytic frameworks, this paper elucidates the functional classes of verbal behavior underlying narrative activities, including extended sequences like storytelling and paraphrasing. It emphasizes the... Read more