The long-standing debate on the position of (GIS-based) spatial analysis in archaeology has not yet resulted in a clear consensus on whether these methods are actually helping to arrive at more sophisticated interpretations of the past. Characterizations of GIS as being a ‘non-novel’ technology, heavily favouring deterministic visions of the past have appeared next to claims that GIS-based spatial analysis can actually tell us more about such diverse issues as the individual perception of the landscape or long-term socio-economic dynamics. In this paper we want to make the case that, more than anything else, archaeology lacks practice in framing research questions in such a way that they can be fruitfully analyzed in GIS.
We want to illustrate this through a case study that would seem to be a relatively clear-cut example: the reconstruction of a 5 km stretch of a Roman road in the Dutch region of South Limburg. Reconstructing this section through least-cost path analysis would seem to be a sensible approach to determining the most probable route. However, despite all the technological means and data sets at our disposal, the archaeological part of the problem is still one of extremely scattered knowledge and information, and of widely diverging interpretations. We hope to illustrate that reframing the research questions in the language of spatial technology will not only result in possible routes that can be tested on the ground; it also helps us to better understand where and why the archaeological interpretative process is turning facts into faith.