Umeå universitets logga

This article discusses the multiscale nature of modelling in archaeology and its relationship with higher-level spatial analysis. The application and purpose of modelling in archaeology is as varied as the multidisciplinary field itself. With the increasing integration of geographical information systems (GIS) and other digital methods into the archaeological workflow, both new opportunities and potential pitfalls present themselves. The struggle of balancing informal inferences of human behaviour in a formal system, such as GIS, has been the subject of much discussion, as well as the questioning of whether some modelling tasks would be better suited for implementation outside the GIS environment. Higher-level spatial analysis is dependent on a number of lower-level models, each building on the other, inheriting both information and uncertainties. These nuances can be difficult to demonstrate clearly once they have been incorporated into another model, potentially obscured further when restricted by the “geographical space” that is central to GIS. Rather than forcing informal models into a formal environment, an alternative would be to opt instead for the visualization of these within the more flexible “variable space,” where the data are front and centre, and spatial and temporal concepts can function as a means of explaining patterns in the model. This article discusses aspects of the challenges and opportunities involved in these types of analysis and provides examples of alternate approaches that could be considered non-spatial.

Northern Fennoscandia is a geologically complex region affected by both glacial and postglacial processes. Quartzite was a key material type utilized by hunter-gatherers in Northern Sweden around the period 4 000 – 2 000 BP, and is thus critical to the understanding of raw material procurement and material flow within the region. However, there is a severe lack of methodological development in the characterization of these materials, and provenance of locally available geological material is complex and fraught with uncertainty. 126 quartz/quartzite points and preforms were sampled from 47 archaeological sites along the upper Ångerman river valley in Västerbotten, Sweden. The material has been analysed non-destructively using three separate portable spectroscopic instrumentations (Near-infrared, Raman, X-Ray Fluorescence). Evaluation of the spectra and exploratory data analysis using Principal Component Analysis demonstrates detectable differences in the material that likely stem from diagenetic/paragenetic origin. The presence of graphite, muscovite and biotite could likewise provide information on the material’s metamorphic grade. In addition to reaffirming the potential of field-based screening instrumentation, these results will benefit future surveys of geological sources in the region. They also indicate potential for the construction of a predictive model that could classify the quartzite based on its chemical characteristic. Such a model would prove useful in future spatial analysis and testing of models of raw material management.