This paper presents a GIS procedure to use (legacy) site-based survey datasets in settlement pattern analysis. Over the past few decades, an impressive quantity of archaeological data has been registered by field-survey projects. Large datasets have been produced, most notably in Mediterranean countries where field-survey has been widely applied for the study of regional landscapes. A majority of these datasets is site-based, which means that only well-defined surface concentrations of artifacts were recorded by surveyors, who usually registered them as dots on topographic maps. To realize the potential of these legacy site-based survey data, methodological procedures that cope with their limits, and more importantly with the distortions on data patterns caused by biasing factors, are necessary. By using these procedures, archaeologists can aim for correct interpretations of the past settlement behavior underlying the survey data. In this paper a GIS method is proposed consisting of a set of interrelated quantitative and qualitative approaches to infer significant settlement patterns from site-based survey datasets. Specifically the GIS procedure consists of two parts. One part regards the assessment of the effect of biases on the spatial pattern exhibited by legacy data. The other part aims to extract historical meaning from data patterning. To show how the proposed GIS research procedure works in practice, a case-study is employed. As part of the research by the Landscapes of Early Roman Colonization project (NWO, Leiden University, Royal Netherlands Institute in Rome), site-based datasets collected by previous survey projects in central-southern Italy are examined to investigate settlement patterns in the early Roman colonial period (3rd century BC).
(Anita Casarotto)
