ABSTRACT
SAM (Script Applier Mechanism) is a system of computer programs written to investigate how knowledge of context can be used to aid in understanding stories. The basic knowledge source SAM applies is the script. Using scripts of varying degrees of complexity, SAM can read (by the process of script application) not only simple stories, but also newspaper articles referring to domains as diverse as car accidents and state visits. Each of these types of texts involve certain invariant components, such as what can happen, the order in which things happen, and who is involved. This consistency of form and content enables the script-based model of reading to be used.
