Induced Pictorial Representations
Abstract
Researchers agree that mental representations of discourse are established at many levels, including a model of the situation described by the discourse. I describe two sets of studies investigating spatial properties of mental models induced by text. In the first set, Holly Taylor and I have found that descriptions written from different perspectives, route and survey, seem to induce the same perspective-free spatial models termed spatial frameworks. In the second set of studies, Nancy Franklin and later David Bryant and I have gathered detailed data on a spatial framework capturing a common situation, of an observer surrounded by objects. That spatial framework is not perception- like, but rather reflects conceptions of space. Extensions of both paradigms are discussed. This early research indicates that situation models constructed from text contain spatial properties, such as relative locations and directions, but are not perception-like or image-like. They are more general than a particular view, allow different perspectives, and have differential access to different parts. Keywords: Mental images; Learning.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 05, 1989
- Accession Number
- ADA216472
Entities
People
- Barbara Tversky
Organizations
- Stanford University