Excavations at a hunter-gatherer site known as 'Grootrif G' shell midden, Lamberts Bay, Western Cape Province

INTRODUCTION Since excavations at Elands Bay Cave thirty years ago, a small number of nearby sites were excavated in the late 1970s and early 1980s in order to provide complementary observations to those made from this large shelter (Horwitz 1979; Klein & Cruz-Uribe 1987; Manhire 1987; Noli 1988; Parkington & Poggenpoel 1987; Robey 1987). In the midst of these studies, Parkington et al. (1986) explored the nature of the impact of pastoralism on local hunter-gatherer populations by attempting to integrate various lines of evidence, a topic revived later with a rock art perspective and additional quantified data by Yates et al. (1994). A synthesis of the local cultural sequence was proposed by Parkington et al. (1988), in which various issues requiring further research were also highlighted. Some of these issues were taken up by further studies focusing on sites located to the north (Lamberts Bay) and south of Elands Bay and by refining the chronology of already excavated sites (Wahl 1994; Jerardino 1995a, 1996, 1998; Jerardino & Yates 1996, 1997; Jerardino et al., in press). While much of the published observations of this latter research have helped to reconstruct and re-model the local cultural sequence between 8000 and 2000 BP, the same measure of understanding is not yet available for the pottery period (last 2000 years). Instead, separate and sitespecific observations remain either unpublished or not fully integrated into one cultural sequence (Horwitz 1979; Klein & Cruz-Uribe 1987; Robey 1987; Noli 1988; Wahl 1994; Jerardino 1998). Moreover, emphasis on the analysis of spatial patterns and ceramic stylistic changes at Dunefield Midden has, unwittingly, channelled research efforts away from establishing a cultural sequence for the pottery period (Parkington et al. 1992; Kent 1998; Orton 2002; Stewart 2006). Clearly, quantified data pertaining to the last two millennia in the study area are in need of being updated and synthesized. Although this is a task beyond the purpose of this paper, a productive approach for now is to build a good empirical basis for the pottery period. Consequently, the purpose of this paper is to report on the findings at ‘Grootrif G’ (GFG), a site with late-pottery period occupation in the Lamberts Bay area, and to discuss these in the light of available contemporary observations. Excavations were initially undertaken at GFG with the aim of rescuing material eroding from exposed sections. In later years it also became evident that excavations at GFG could also increase the number of observations for the pottery period in the northern-most part of the study area, as what is known about the pottery period has been derived almost entirely from sites located to the south around Elands Bay.