The complexity of production spaces at the Neolithic site of La Draga

Abstract

La Draga (Banyoles, Girona) is an early Neolithic pile-dwelling settlement that was occupied uninterruptedly between ca. 5300-4800 BC. This archaeological site stands out for being one of the first evidences of the Neolithisation process in the NE of the Iberian Peninsula and, for having provided an exceptional sample of organic remains thanks to its archaeological layers preserved in an anoxic environment. The research performed during last years at La Draga has been aimed at acquiring knowledge about the strategies developed, by this first farming society, on subsistence and technology. Recently, with the available information an intra-site spatial analysis has been developed with the purpose of approaching the organizational strategies of the same community. In this paper, the results obtained from the spatial analysis of sector B of the settlement will be presented. The focus is on evidence of social space organisation and inferring socio-economic strategies and social organization. The goal of the spatial analysis performed was to predict the nature of artefact dispersal and to account for the proper place were each particular social activity took place. From the observed degree of spatialization of subsistence and technological activities the nature of social relations can be discerned. The approach applied relies first, on an ontological categorization of the archaeological observables in terms of social activities, and second on the nature of spatial dispersion, distinguishing the social reason for randomness or its alternative, aggregation. On this regard, a null-hypothesis of spatial random distribution is formulated to check whether there was some intentionality in the proper location of activities that generated the observed remains. The detection of concentrated patterns is interpreted in terms of the conscious selection of a particular place for a particular activity. Those hypotheses are evaluated geostatistically, measuring spatial autocorrelation and anisotropy. 2D Kernel density methods are used to test the spatial pattern of individual findings, whereas surface interpolators (IDW and Kriging) are used to test the frequency patterns at the spatial level. Results for all the archaeological categories treated are integrated using a spatially constrained Multivariable Correspondence Analysis to check whether different categories are spatially correlated. All those analysis, at the univariate and multivariate level allows to create probability maps for the different social activities that generated the observed archaeological record.