Abstract

Effective resource management requires an intimate knowledge of the resource base and an equally thorough understanding of the types of impacts and their effects. Current resource managers are hampered because they, and the rest of the archaeological community, have a less than desirable knowledge of both. As a result, typical decision making matrices are relatively simple, employing discontinuous variables (high, medium, low), and binary assumptions (there is an impact/thrre is no impact). This is not the best level of decision making that can be achieved, and is particularly ineffective for managing regional scale impacts such as forestry and agriculture. It is argued here that we should start looking at the resource management process in a much broader context. Ideally, we would look at probability of site occurrence, type of site, probability of detection, type of impact, equipment used, soils, weather, vegetation cover, etc. In doing so, we are forced to move from a two dimensional to an 'n' dimensional decision matrix. Without ingesting harmful substances, people don't like to think in 'n' dimensions. Computers are not as inhibited – what's a dimension? We can create expert systems to determine the best possible management solutions and then apply these solutions, as rules, to the landscape using GIS. This paper discusses a proposal for an appropriate pilot study and outlines some of the potential benefits of such an approach.