ABSTRACT
GIS has been widely proclaimed as offering an efficient, user-friendly environment for the examination and presentation of digital data. While this may be true in some cases, a notable example being the way geographically-referenced census data can be examined, treatment of more qualitative data categories has proved far less satisfactory . The capa bilities of GIS in the quantitative arena has long masked the considerable inadequacies which the technology exhibits when considering the generation and of less deterministic data. An example of this would be qualitative descriptions of landscape. In addition. it has been for some time that GIS has application shortcomings in another field � that of visual determination. With respect to functions such as viewshed generation and visible area analysis, commentators (Fisher. De 1 994) have illustrated inadequacies in the way in which GIS identifies viewshed characteristics and how they are implemented.
