Internet Archaeol. 16. Rajala. Spatial patterns

5.2 Spatial patterns

Although predictive modelling was never dominant in Europe (cf. Harris and Lock 1995), the evaluation of the significance of different co-occurrences was based on statistics and research strategies were nevertheless geared towards this type of modelling. Gaffney and Stančić (1991) stressed the ability of GIS to handle large datasets in model building and testing. Likewise, van Leusen (1993) was looking for environmental, socio-cultural and statistical explanations using Boolean overlays and mathematical decision rules. Similarly, Harris and Lock (1995, 354) considered that knowledge was created through spatial exploratory data analysis. In this way, GIS contributed to the creation of systematic mapping instead of allowing intuitive exploration and interpretation of spatial patterning (Wheatley and Gillings 2002, 3-5).

Wheatley and Gillings (2002, 6-8) have suggested that patterns are quantifications of neutral and meaningful spaces. The significance can be shown using explanatory factors that are based on observed or analogical behaviour and anthropological theories. Over recent years, past social meanings have been approached using viewshed and multiple viewshed analyses, cost surfaces and their combinations (cf. Harris and Lock 1995; Gaffney et al. 1995b; 1996; Llobera 2000; Wheatley and Gillings 2000).