ABSTRACT
It is axiomatic to say that all maps are an abstraction of reality (Robinson et al. 1978). They are not exhaustive in their coverage (not all phenomena are shown on all maps), nor are they precise in the locational information they portray. Thus, in both location and in attribute they are no more than acceptable representations of the mapped phenomenon, be it physiography, topography, soil, population or some other phenomenon (Chrisman 1987; Hsu and Robinson 1970; Jenks and Caspall 1971; MacDougall 1975; Morrison 1971; Monmonier 1977; Muller 1977). This being so there is inevitably a significant level of unreliability in the resulting map product. This unreliability is naturally incorporated in any geographic database for which the map may be an input.
