Islands in a stream? Reconstructing site formation processes in the late Middle Paleolithic site of 'Ein Qashish, northern Israel

Abstract

Information retrieved from open-air sites is necessary for a more complete reconstruction of Middle Paleolithic behaviors, which is currently often based mainly on data from caves and rockshelters. However, open-air sites present methodological and analytical challenges that differ from those encountered in cave sites. Being essentially an integral part of their paleo-landscape, open-air sites are affected by localized as well as landscape-scale processes that need to be distinguished from the anthropogenic signatures. 'Ein Qashish is an open-air late Middle Paleolithic occurrence located in the Yizra'el Valley just east of Mt. Carmel, situated on the Pleistocene floodplain of the Qishon stream. The site is found at the interface of the sediments deposited by the Qishon stream, which drains the Yizra'el Valley, and paleo-Wadi Qashish, which flowed off Mt. Carmel. Artifacts and bones are dispersed over a vertical distance of some 90 cm. Observations in geological trenches located at variable distances from the excavated area indicate that this cluster of finds occurred at a circumscribed locality on the landscape rather than as a part of a continuous distribution on the paleo-surface. We 'reverse-engineer' site formation processes, starting with the latest (post-depositional), moving to earlier (syn-depositional) processes, and relate them to their most likely causative agents through the examination of a series of explicit models. GIS techniques were applied in order to parse the variable lines of information, which include lithic and faunal taphonomy, size distributions and spatial and stratigraphic dispersal of objects, fabric analysis and OSL results. We discuss the implications of the results for distinguishing anthropogenic from non-anthropogenic effects and for reconstructing the timeline of depositional events that led to the current distribution of artifacts. Insights into these questions inform our understanding of human activities at the site, suggesting that the excavation samples the margins of an occupation horizon where activities focused on knapping and resource processing rather than task-specific behaviors such as hunting.