Narratives provide a powerful means of making sense of our world. They are cognitive tools that we use to examine and shape ourselves, as well as the environment around us. Every day, we seek out and tell stories to reflect on our past, navigate our present, and direct our future.

Because narratives embed the complexity of the universe as we know it, it is challenging to develop a complete representation of narrative meaning. Some proposed theories focus on the perceptions and interactions of narrative agents. Others emphasize the communicative impact that a narrative has on its audience. Many more interpretations have been explored, but relatively little work exists that seeks to reconcile these multiple definitions and applications of narrative.

Narrative models may reflect more about their creators and their unique scenarios than the elusive tangles of meaning they are meant to capture. They are, after all, stories we are telling about stories. Each presents a lens in which to view, and subsequently define, narrative for a particular context. Regardless, perhaps we can learn from these diverse perspectives. Their congruence illuminates key features of story, while their discrepancies progressively carve out the rich, explorable space of all possible narratives. Although we may never reach the boundaries of that space, our explorations could give rise to new insights in narrative understanding.

This thesis seeks to determine whether we can devise a framework that can usefully encode multiple narrative representations in a human-readable format. To do so would promote annotation, comparison, and sharing of narrative data among researchers. Further, a framework that can express distinct narrative representations should also reveal methods for automatically transforming their fundamental units. I will provide examples for how this type of adaptation may be useful for personalizing human-computer interaction scenarios which rely on narrative structures.