Survey, ceramics and statistics The potential for technological traits as chronological markers

In contrast to excavated datasets, the lack of stratigraphic context for ceramic survey material often means that a considerable part of the collected sample is essentially non-diagnostic and can offer little beyond a presence/absence for determining the spatial extent of human activity. Whilst detailed regional comparison between excavated ceramic sequences and survey assemblages of feature sherds presents the best opportunity to identify specific periods of activity across the landscape, this paper explores the potential to characterise broader chronological periods into which less diagnostic sherd material might be reliably placed. We discuss the results of a preliminary statistical analysis on ceramic material from the Keros Island Survey to establish the degree to which values of specific technological variables are spread throughout the assemblage, since parameters such as wall thickness, fabric coarseness, firing characteristics and level of preservation may be considered indicative or ‘diagnostic’ for broad chronological periods. Increasing the diagnostic potential of material collected during a pedestrian field survey can offer a more nuanced and also robust interpretation of the nature of human activity in a particular region and how it changed through time. In the case of the Keros Island Survey, this approach has contributed to the interpretation of diachronic land use strategies that extend from the establishment of the Early Bronze Age maritime sanctuary at Dhaskalio-Kavos up until the recent abandonment of the island in the 1950s.