Pottery analysis and the archaeological context. Chosing samples, counting, weighing and interpreting materials in context

Most dig houses in the world and certainly all those in the Near East are characterised by thousands of crates with millions of pottery sherds. At the nearly 60-year-campaign-long Arslantepe dig house crates are wooden boxes lined up along the edge of the mound, that overshadow our work area and that get regularly spread out on a more than 100m long line of tables: our “sherd yard” and the “mega puzzle area” for the restorers (Bollati and Ghedin present volume, fig. 1). The pottery specialists - each of us dedicated to one of the periods of occupation of Arslantepe - spend hours, days, weeks, months on these tables, again, like at many other excavations. Differently to other excavations though, we have Marcella Frangipane: forty-three years at the site and it is nearly impossible that she hasn’t noticed something before you and rare that you have an intuition before she does. The work of a potter analyst may be repetitive and tedious, but with Marcella there is little automatism in this activity. She has showed us why not to embrace an a priori methodology; every choice should be weighted and evaluated on the basis of the character of the assemblage and of the context of retrieval. Every data collection should spring from specific questions. How should we choose the contexts to analyse in detail? Should sherds be counted, weighted? Should vessel num- bers be estimated? Should diagnostic sherds and amorphous ones be treated differently? The choice of one or the other depends on the aim with which the count is done and on the type of context analysed. The present paper shall explain how the processing of pottery material at Arslantepe is carried out as to aid its analysis, and then use the case of Late Chalcolithic 3-4 (Arslantepe period VII, 3900-3400 BCE) to discuss why specific methodological choices in pottery analysis have been adopted.