Journal of Archaeological Science (2001) 28, 597–612 doi:10.1006/jasc.2000.0591, available online at http://www.idealibrary.com on A Taphonomic Perspective on Neolithic Beginnings: Theory, Interpretation, and Empirical Data in the Western Mediterranean Joan Bernabeu Auban Departament de Prehisto `ria i Arqueologia, Universitat de Vale `ncia, Blasco Iba ´n ˜ez, 28, 46010—Vale `ncia, Spain C. Michael Barton Department of Anthropology, Arizona State University, PO Box 872402, Tempe, AZ 85287-2402, U.S.A. Manuel Perez Ripoll Departament de Prehisto `ria i Arqueologia, Universitat de Vale `ncia, Blasco Iba ´n ˜ez, 28, 46010—Vale `ncia, Spain (Received 10 February 2000, revised manuscript accepted 8 May 2000) The fills of caves and rockshelters generally comprise complex depositional palimpsests, making fine scale chronological resolution extremely difficult. Nevertheless, these settings remain very important in archaeology because they often preserve long records of cultural change. This is true for the initial appearance of food producing economies in the western Mediterranean. The chronologically ambiguous nature of cave and shelter deposits is one of the reasons for the continued debate over the processes responsible for the beginning of the Neolithic in this region. We employ taphonomic studies of the archeofaunal record from Mesolithic and early Neolithic cave and shelter sites in Mediterranean Spain to disentangle some of the formation processes affecting relevant deposits in order to better understand the processes of cultural change that led to the spread of agricultural communities.  2001 Academic Press Keywords: NEOLITHIC, EUROPE, TAPHONOMY, ZOOARCHAEOLOLGY, CAVES/ROCKSHELTERS, GEOARCHAEOLOGY. Introduction B ecause they are sediment traps in which artifact- bearing deposits can accumulate over long periods of time, caves and rockshelters have been the focus of archaeological investigation since the inception of the discipline (Barton & Clark, 1993; Straus, 1990). This characteristic has made them in- valuable for recording long-term patterns of prehis- toric social change. Nonetheless, as geoarchaeological and taphonomic studies have accumulated for cave and shelter deposits, and the artifacts they incorporate, it has become increasingly apparent that the interpret- ation of assemblages from these contexts is often problematic. Although artifact assemblages from discrete strati- graphic units have long been interpreted as accumulat- ing during the occupation of a locale by a discrete social group for a single season or for a short sequence of closely spaced occupations, there is a growing recognition that this is the case rarely if at all (Barton & Clark, 1993; Barton & Neeley, 1996; Colcutt, 1979; Dibble et al., 1997; Jelinek, 1976). Rather, assemblages are likely palimpsests of repeated occupations of diverse length, separated by varying time spans, and representing intervals of varying but often multi- generational length—even in carefully controlled excavations. ‘‘Living floors’’ in such contexts are more a product of archaeological practice than prehistoric behaviour patterns. Given the frequent inability of cave and rockshelter deposits to provide archaeologists with a series of superimposed snapshots of past society, they are of less use for detailing the process of social change than they are for preserving a record of the long-term results of change. In spite of these potential uncertainties about the depositional integrity of cave and shelter deposits, they still commonly serve as the basis for inferences about the processes of social change. In many areas of the world, these locales remain by far the best known 597 0305–4403/01/060597+16 $35.00/0  2001 Academic Press