Our goal is to investigate the structure of narratives that produce phenomena such as suspense, curiosity, surprise. We will be focussing on the first of these: suspense. The fundamental question guiding this research could be phrased thus:
What are the minimal requirements on formal descriptions of narratives such that we can capture suspense and thus generate novel narratives which produce it ?
A successful answer to this question might suggest a series of models conforming to these minimal requirements which could be used to generate novel suspenseful narratives. One might, in addition, expect such models to be able to:
- track the level of suspense at each step of a narrative,
- generate different narrative orderings and suggest the most suspenseful versions.
Whilst the broader generative goal guides our research, our first step will be to identify the mechanisms which produce suspenseful narratives. We will examine the notions of suspense which occur in the psychological, literary and computational literature together with some more intuitive notions, and attempt to create a formal model of suspense which takes some of these notions into account. This model will then be used to create some short narratives whose suspensefulness could be tested on a sample of readers.