Internet Archaeol. 16. Rajala. Subjectivity and fuzziness of data

3.6 Subjectivity and fuzziness of data

All data and system models put limitations on the way research units, their spatial relationships and meanings can be presented. The biases of digital concepts, Boolean logic and standard set theory have been sources of critique. However, the 'fuzziness' of archaeological categories (Zubrow 1990a, 68) has been acknowledged.

Harris and Lock (1995) suggested that increasing the accuracy of the collection of archaeological data on one hand and the integration of fuzzy categories and fuzzy logic on the other could remedy problems with archaeological data. The use of fuzzy logic has often been discussed in a preliminary fashion (e.g. Kirkinen 1999), but finished projects with workable fuzzy categories have rarely been published (cf. Reeler 1999; Crescioli et al. 2000). The problem with fuzzy logic is that it tries to quantify, often arbitrary, subjective decisions. It looks scientific but is a product of qualitative evaluation.

Fuzzy logic may not be the ultimate solution but it has produced valid theoretical discussions. Subjectivity per se is not negative; GIS is not perceived as value neutral or objective. The 'loss of innocence' and the end of an uncritical application of deterministic models happened during the 1990s (cf. Wheatley 1993). The selective nature of data was discussed thoroughly by, for example, Harris and Lock (1995) and this subjectivity of data can be handled according to a realistic view of philosophy. When all research is seen as subjective, controlled by agreed disciplinary processes, the maintenance of standards of digital creation of data assures reasonable data accuracy. However, the inherited archaeological weaknesses of data must be regarded as fundamental to GIS approaches.