~.THE DIGGING STICK Volume 23, No 2 ISSN 1013-7521 August 2006 SURVIVAL AND CULTURE IN THE COASTAL DESERT OFNAMAQUALAND What people ate and where they sat to eat it Genevieve Dewar The time has come to finish the lab work and The cave (30.18°S 17.216°E) is the only well­ write my PhD thesis. I have been excavating, documented site on the Namaqualand coast and cataloguing and analysing 11 sites for the past yet it provided early dates for the introduction of three years, focusing on the faunal remains from sheep into South Africa (Sealy & Yates 1994; the Later Stone Age (LSA) in Namaqualand. This Webley 1992, 1998, 2002). The paucity of is a time of introspection, of sifting through data research in the region is based on two factors: and theory and integrating years of hard work. Now is the time to piece together the picture of life in the +V5, Buffels River coastal desert of Namaqua­ ,'fSK40 land. Why the archaeology of SwartIinji.es Namaqualand? Ri,rer The Namaqualand coastal ) strip of the Northern Cape ~,,L~){2100,4-~1l '\ " ~oeg:rnrer Province (Fig 1) is an arid Atlantic ~,~, marginal landscape with cap, e" T,own Ocean ' ~­ less than 25 mm of rain a / +tc Groen Riv#-' year, occupied by large burial sand dunes, low bushes and animals adapted to aridity, .. Scale 1: 100.000 such as steenbok and tor- toise. However, the region Fig 1: Map of Namaqualand, South Africa. The map on the left shows the also has a beautiful rocky Namaqualand Coast, with the study region in the box. The map on the right coastline teeming with shell- shows the location of sites discussed in the text. fish, crayfish, marine birds and mammals such first, Spoegriver Cave is the only cave for 150 km as fat-rich seals, which has brought people to its and previous research focused on the excavat­ coast since the Early Stone Age (-250000 BP). ion of caves; second, De Beers Namaqualand Namaqualand has received very little attention Mining Co. owns the greatest part of the coastal from the South African archaeology community, strip and to obtain access to this coastline is yet what has been published, namely Uta Web­ difficult. ley's work at Spoegriver Cave (Webley 1992, Fortunately, for the past two decades the Arch­ 1998, 2002), has yielded important information. aeology Contracts Office of the University of Cape Town has been compiling an extensive Genevieve Dewar, BSc, MA, is a doctoral student in the database of sites on the Namaqualand coast Department of Archaeology at the University of Cape Town, while conducting archaeological assessments gendewar@mweb,co.za ahead of open-cast diamond mining. Today, this South African Archaeological Society database includes over 1 300 open sites, a rich is evident from the many hand axes and negative resource indeed, from quarries to shell middens. scar boulder cores found at silcrete quarries So while some may think that a desert coastline along the marine terrace. The region also has may be quite uninhabitable without large caves evidence of occupation right through the Middle to provide shelter, the vast quantity of shell Stone Age, including both the Still Bay and middens along this coast proves otherwise and Howieson's Poort industries. Thus, we can say will hopefully reveal evidence of prehistoric that Homo erectus, archaic Homo sapiens and patterns of mobility and systems of exchange (cf. Homo sapiens inhabited the region. Sealy & van der Merwe 1986; Wadley 1989). Evidence from the LSA (20 000 to 400 BP) is An unexpected advantage of studying the marg­ fragmented. The earliest evidence for LSA occu­ inal Namaqualand desert is the close proximity pation is the large scrapers found in the base and similar environment to the Kalahari desert layers of Spoegriver Cave (Webley 2002), sug­ (Fig 1). This means that analogies derived from gestive of an early Holocene date (-8 000 BP, extensive ethnographic research on the Kalahari Albany industry), but without radiocarbon dates a (cf. Lee & Devore 1976; Beisele 1993) are more definitive answer is elusive. As in other areas of relevant when applied to archaeological evi­ the country (Deacon 1974 for the interior and dence for human behaviour from Namaqualand Parkington et al 1988 for the west coast), there deposits. In addition, Bleek and Lloyds' /Xam may have been a hiatus of human occupation informants (1911) were from the Northern Cape, -along the Namaqualand coast until just after the introducing an historical ethnographic approach arid mid-Holocene period, since the oldest to interpreting archaeological remains from Nam­ radiocarbon dates cluster around 4 500 to 4 300 aqualand. Also, it is important to study human BP. Caution is needed, however, since the dates adaptations to arid environments as we face the come from open sites located in deflation hollows onset of global warming in our own time. of sand dunes that may only have formed during the mid-Holocene (Chase 2005). Thus, any sites The beauty of the region, the importance of that were deposited before the mid-Holocene Spoegriver Cave, the vast number of identified would be hidden under many metres of sand. sites, the potential of integrating archaeological After the mid-Holocene there was continuous data with ethnographies and histories, and the occupation along the coast right up to the arrival potential to study arid adaptations makes this a of the Trek Boers in the 18th century. truly exciting region to study. For many thousands of years, hunting and The study region gathering was the sole mode of production at the coast. Eventually, herding and agriculture The research area is a strip of coastline within the appeared in the region, but not until the second winter rainfall zone of the north-west coast of the th millennium AD and the 18 century respectively. northern Cape. All of the identified sites are More specifically, Spoegriver Cave has early located within De Beers property, from just below dates for the introduction of sheep at 2 105 ± 65 Port Nolloth to just below Spoegriver (Fig. 1). My BP (Sealy & Yates 1994), implying some form of research covers the entire length of the property, herding subsistence strategy, but there is no including the shoreline, the coastal dunes and other clear evidence yet for pastoralist culture in the plateau up to 10 km inland. There are three Namaqualand. While pottery is present in the major rivers, the Buffels, the Swartlinjies and the region, there are few intact pots and so far the Spoeg. Major landforms include rocky platforms earliest radiocarbon date is 1 930 ± 50 BP (Web­ and headlands, dune corridors and large dune ley 2002), suggesting that pottery did not arrive seas along the shoreline and coastal plane, and as a package with sheep (Webley 2002). In an inland plateau that terminates at a marine addition, I have analysed well over 60000 bones terrace 7 km inland. However, coverage of these from seven open sites and not a single sheep landforms has been patchy and focused primar­ bone has been identified. Possibly herder sites ily on the coastal plain, as access to the mining have not yet been identified or herders did not area was dependent on De Beers' scheduled keep/slaughter sheep atopen sites, only at cave mining activities. sites. Archaeological evidence for pastoralists remains elusive, even in a region named for a Shell middens: who and when Khoikhoi pastoralist tribe: the Namaqua (Elphick Habitation of this coast from the Early Stone Age 1977; Penn 2005). The Digging Stick 2 Vo123(2) August 2006 The research project taboos against the presence of women during a hunt (Beisele 1993). However, there were also My interest lies in the survival strategies and taboos relating to expected behaviour of women culture of people who lived along the Namaqua­ and children around dead springbok (Bleek & land coast, with the focus on arid adaptations, Lloyd 1911) that do not apply to other species, analysis of what people ate (subsistence strat­ suggesting that they may have been present egies) and where they sat to eat it (settlement when springbok were killed. This could indicate strategies). While I have not yet compiled the that trapping was not in fact regarded as 'hunting' data to answer these questions for the broader - a sacred act cond ucted by men - but was some region, I will discuss the results from a few indiv­ 'other' type of activity. The third social implication idual sites. I chose 11 LSA sites with good org­ of mass trapping springbok relates to sharing anic preservation to determine subsistence strat­ and the vast bounty of available meat. SK400 egies, using a behavioural ecology approach. To has the remains of over 120 animals, indicating date, the sites with interesting faunal samples the need for processing and distribution. An are SK400 (Dewar et al 2005), the micromammal event of this magnitude would have been very site KV502 and the penguin midden LK2004-11. significant in solidifying social bonds and SK400 (29.682°S 17.063°E) is the first archaeo­ relationships. There would have been so much logically recorded mass-kill site where people meat that women and the elderly of the trapping trapped herds of springbok during a summer group would have received a large enough share drought. The radiocarbon age of the site is 420 ± of the meat to in turn permit them to strengthen 45 BP (Pta-91 05) and the best fit calibrated date other bonds by sharing down the line. is 1 478 AD (ranging from 1 459 to 1 612 AD). The second site, KV502 (29.500oS 17.055°E) Summer drought migrations of springbok are has an interesting story to tell involving evidence well documented and involve upwards of 20 000 of people eating mouse and rat bodies (Dewar & individual animals driven by thirst madness.