Possible Worlds, Artificial Intelligence, and Narrative Theory
Possible Worlds, Artificial Intelligence, and Narrative Theory
Growing up with Nell: a narrative interface for literacy
Proceedings of the 11th International Conference on Interaction Design and Children
| Hi-index | 0.00 |
Curveship, a Python framework for developing interactive fiction (IF) with narrative style, is described. The system simulates a world with locations, characters, and objects, providing the typical facilities of an IF development system. To these it adds the ability to generate text and to change the telling of events and description of items using high-level narrative parameters, so that, for instance, different actors can be focalized and events can be told out of order. By assigning a character to be narrator or moving the narrator in time, the system can determine grammatical specifics and render the text in a new narrative style. Curveship offers those interested in narrative systems a way to experiment with changes in the narrative discourse; for interactive fiction authors and those who wish to use of the system as a component of their own, it is a way to create powerful new types of narrative experiences. The templates used for language generation in Curveship, the string-with-slots representation, shows that there is a compromise between highly flexible but extremely difficult-to-author abstract syntax representations and simple strings, which are easy to write but extremely inflexible. The development of the system has suggested ways to refine narrative theory, offering new understandings of how narrative distance can be understood as being composed of lower-level changes in narrative and how the order of events is better represented as an ordered tree than a simple sequence.