ABSTRACT
The electronic and · spatial data revolution of the last two decades has resulted in the development of a variety of data-management tools for handling the vast amounts of temporal and georeferenced data that are being created. The advent of GIS marks a major step forward in integrating multiple layers of spatial information from multiple sources. However, this technology is very deficient when it comes to coping with the subsequent need for spatial analysis and modelling, and there is a noticeable absence of suitable and relevant technologies. One reason is the slow recognition that human systems modelling is of an order of magnitude more complex than classical statistical and conventional mathematical modelling methods can readily handle. The lack of adequate, prior, theor etical knowledge has resulted in the growing failure of quantitative geography to do much, or indeed anything, with most of the data being created by GIS other than to map it . Additionally, rcal doubts may be expressed about whether the prevailing Aristotelian logic-based paradigm of science provides an adequate framework for spatial science (see, for example, Burrough and Frank, 1 995) . The existing crisp, logic-based science para digm may well be too limited to represent real-world spati al phenomena adequately, and alternative texts, languages and representations may need to be considered. One route is to dabble with alternative non-scientific philosophies and ways of thinking (e.g. post Modernism), but there are likely to be major practical problems with the incorporation of these ideas into GIS . Another approach, and one which would fit more easHy in a GIS framework, is to investigate whether there are more scientific alternatives that can cope with the complexity of geographical phenomenon without appealing to the paranormal for inspiration.
