ChapterPDF Available

Provisioning Responses to Environmental Change in South Africa’s Winter Rainfall Zone: MIS 5-2


Abstract and Figures

The later Middle Stone Age of southern Africa witnesses a number of important changes in lithic technology including the early appearance of bifacial point and microlithic systems. Though radiometric ages for these changes remain contested, they can in places be reconciled with elements of climatic variation. This paper examines the organization of provisioning systems through four successive industrial phases – the Still Bay, early Howiesons Poort, later Howiesons Poort, and post-Howiesons Poort – using data from rock-shelters and open sites. We show that the dominant means of technological delivery shifted from individual to place provisioning and back to individual provisioning through these industries, likely reflecting variation in humidity through the last glacial. We also show that the apparent abundance of sites and richness of assemblages in industries such as the Howiesons Poort may be in part a consequence of a research focus on rock-shelters; we found limited evidence for Howiesons Poort sites in the open. This is in contrast to the other industries, most clearly the Still Bay, for which we identified several large open-air gearing-up locations. Finally, we suggest that while interrogation of the archaeological record at the industry level is viable, it also masks behaviorally meaningful variation that should serve to encourage finer-scaled analyses.
Content may be subject to copyright.
13© Springer International Publishing AG 2018
E. Robinson, F. Sellet (eds.), Lithic Technological Organization and
Paleoenvironmental Change, Studies in Human Ecology and Adaptation 9,
Chapter 2
Provisioning Responses toEnvironmental
Change inSouth Africa’s Winter Rainfall
Zone: MIS 5-2
AlexMackay, EmilyHallinan, andTeresaE.Steele
The later Middle Stone Age (MSA) archaeological record of southern Africa is
known for the diversity and antiquity of its complex lithic technologies, including
early examples of the manufacture of blades and bladelets, bifacial points, and
backed artifacts (or microliths) (Volman 1980; Mitchell 1988; Soriano etal. 2007;
Wadley 2007; Villa et al. 2009, 2010; Högberg and Larsson 2011; Porraz etal.
2013a, b; Wurz 2013). Associated production techniques include the use of pressure
aking, marginal percussion with soft stone, and heat treatment (Brown etal. 2009;
Soriano etal. 2009; Mourre etal. 2010; Schmidt etal. 2012; Porraz etal. 2013b).
These lithic systems underwent turnover on the order of 2–10 kyr throughout the
Late Pleistocene (Jacobs etal. 2008a, Tribolo etal. 2012; Porraz etal. 2013a).
There has been considerable discussion of the drivers underlying technological
variability through this period (Ambrose and Lorenz 1990; Wurz 1997; Ambrose
2002; McCall 2007; Mackay 2009; Powell etal. 2009; McCall and Thomas 2012;
Ziegler etal. 2013; Mackay etal. 2014a, Soriano etal. 2015). While the focus of
these arguments has largely been on the form of typical technological items, we
consider evidence for changes in provisioning– effectively, the ways in which lithic
technologies were delivered to their point of need– through the period from ~75 to
~50 ka. We focus on southern Africa’s winter rainfall zone and the relationship
between provisioning systems and environmental variation, as it is known.
A. Mackay (*)
Centre for Archaeological Science, University of Wollongong, Wollongong, NSW, Australia
E. Hallinan
Division of Archaeology, University of Cambridge, Cambridge, UK
T.E. Steele
Department of Anthropology, University of California, Davis, CA, USA
The Archaeology andEnvironments ofSouthern Africa’s
Winter Rainfall Zone
Southern Africa is divisible into three broad climatic regions– the winter rainfall
zone (WRZ), summer rainfall zone (SRZ), and year-round rainfall zone (YRZ)
(Chase and Meadows 2007) (Fig.2.1)– based on factors controlling precipitation.
The WRZ is situated in the southwest of the subcontinent and receives most of its
annual precipitation from the northward migration of westerly winds during the
austral winter. Controls on the delivery of winter rain in the region are complex, but
the bulk of the available theory and data suggest an expansion of the zone of west-
erly inuence under cooler conditions associated with expansions of Antarctic sea
ice (Nicholson and Flohn 1980; Chase and Meadows 2007; Toggweiler and Russell
2008; Chase 2010; Mills etal. 2012; Stager etal. 2012; Truc etal. 2013). In conse-
quence, increases in the extent and duration of the rainy season are expected to
accompany glacial conditions, resulting in greater regional humidity. This is most
evident in marine isotope stage (MIS) 4 and MIS 2, with evidence for decreases in
humidity into MIS 3 (Cowling etal. 1999; Klein etal. 1999; Shi etal. 2000; Stuut
etal. 2002; Avery etal. 2008; Bruch etal. 2012).
The major vegetation community in the WRZ is fynbos, characterized by low
shrublands and high rates of endemism (Mucina and Rutherford 2006). Fynbos taxa
show considerable resilience to climatic change (Meadows and Sugden 1993;
Valsecchi etal. 2013), possibly a consequence of edaphic controls arising from the
Fig. 2.1 Study area showing the major shelter sites (KKH Klein Kliphuis, DRS Diepkloof) and
minor sites of relevance to the paper (VR003 Varsche Rivier 003, KFR Klipfonteinrand), as well as
the rainfall zones of southern Africa
A. Mackay et al.
low-nutrient soil related to underlying sandstone geology (Valsecchi etal. 2013;
Carr etal. 2016). Increased humidity in the WRZ has been associated, in some cases
at least, with increases in Afromontane forest taxa, though in higher-nutrient areas,
it may be associated with expansions of grasslands (Klein 1976; Cowling etal.
1999; Cartwright 2013). Aridity clines to the north and northeast result in increased
prevalence of arid-adapted plants, most notably in the form of succulents (Mucina
and Rutherford 2006).
The later MSA archaeological sequence in the WRZ can be broken down into
several components or industries based on the characteristics of aked stone artifact
assemblages. The earliest of these may be referred to as the earlier Middle Stone
Age (MSA) which some researchers subdivide into multiple phases (Volman 1980;
Wurz 2002). Generally dating to greater than 80 ka, the earlier MSA is outside the
scope of interest of this paper and will receive little further discussion.
Chronologically, the earlier MSA is followed by the Still Bay industry in which
bifacial points are common and the prevalence of ne-grained rocks often increases
(Henshilwoodetal. 2001; Högberg and Larsson 2011; Porraz etal. 2013b) (Fig.2.2).
Researchers have noted an underrepresentation of cores in Still Bay sites across
southern Africa (Henshilwoodetal. 2001; Wadley 2007; Porraz etal. 2013b), and
some have suggested that this may have been a period in which residential mobility
was emphasized, based largely on inferences drawn from technological systems
design theory (Mackay 2009; McCall and Thomas 2012). Most ages place the Still
Bay around the transition from MIS 5 to MIS 4 between 75 and 70 ka, in the
context of rapidly cooling conditions and increasing humidity (Stuut etal. 2002).
These ages, however, are contested (Jacobs etal. 2008a, Tribolo etal. 2009; Högberg
and Larsson 2011; Tribolo etal. 2012; Jacobs etal. 2013). Conicting OSL ages at
the WRZ site of Diepkloof place the Still Bay either 75–71 ka or >100 ka. It should
be noted that while bifacial points occur throughout MSA sequences at sites in the
SRZ (e.g., Kaplan 1990; Wadley 2012; de la Pena etal. 2013; Porraz etal. 2015),
this pattern has not been documented in the WRZ or YRZ (Mackay etal. 2014a).
Bifacial points recovered from excavated contexts in these regions invariably occur
as a single discrete temporal band within sequences (Henshilwood et al. 2001;
Vogelsang etal. 2010; Porraz etal. 2013b, Will etal. 2015; Steele etal. 2016) and
where dated exceed 70 ka.
The termination of the Still Bay at rock-shelter sites across southern Africa is com-
monly marked by an occupational hiatus (Jacobs et al. 2008a, b; Vogelsang et al.
2010; Högberg and Larsson 2011). Diepkloof in the WRZ is an exception, with appar-
ently continuous occupation after the end of the Still Bay, into a period characterized
by the production of small blades, backed artifacts, and pièces esquillées (Igreja and
Porraz 2013; Porraz etal. 2013b). Porraz and colleagues refer to this industry as “early
Howiesons Poort” as it shares many of the characteristics of the widespread later
industry. An apparently similar early Howiesons Poort industry is also known from
Pinnacle Point 5/6in the YRZ (Brown etal. 2012). Both sites have OSL ages for the
start of this industry at around 70 ka, though Diepkloof also has ages >100 ka for the
same layers produced by a different team of analysts (Jacobs etal. 2008a; Tribolo
etal. 2012). From 65 ka, “classic” Howiesons Poort assemblages are in evidence at
2 Provisioning Responses toEnvironmental Change inSouth Africa’s
numerous sites across southern Africa. These assemblages are marked by the
production of blades, notched akes, and backed artifacts (Fig.2.2) and also, in many
instances, by extremely dense accumulations of aked stone (Volman 1980; Singer
and Wymer 1982; Jacobs etal. 2008a; Mackay 2010; Vogelsang etal. 2010; Porraz
etal. 2013a, b). In the WRZ, the classic or later Howiesons Poort is notable for the
preferential selection of ne-grained rocks and specically silcrete. Various research-
ers have argued for the Howiesons Poort as a time of, alternatively, increased residen-
tial mobility (Ambrose and Lorenz 1990; Porraz etal. 2013a) and decreased residential
mobility (Mackay 2009; McCall and Thomas 2012) in response to the cool, humid
conditions suggested by environmental archives (Chase 2010).
Fig. 2.2 Distinctive artifacts from the study period. Panel (a) Still Bay bifacial points (all from
Hollow Rock Shelter); (b) cores, blades, backed artifacts, and notched akes from the Howiesons
Poort (from Klein Kliphuis, Putslaagte 8, and Varsche Rivier 003); (c) cores, retouched akes, and
unifacial points from the post-Howiesons Poort (all from Klein Kliphuis). Scale bars are 10mm
A. Mackay et al.
The Howiesons Poort terminates around the MIS 4/3 transition ~58–60 ka at
most sites (Jacobs etal. 2008b; Piennar etal. 2008; Guerin et al. 2013). Various
proxies suggest increased temperatures and decreased humidity at this time (Chase
2010). Technologically, blade production fades, and backed artifacts are replaced by
unifacial points (morphologically equivalent to convergent scrapers) as the domi-
nant implement type (Fig.2.2) (Volman 1980; Soriano etal. 2007; Villa etal. 2010;
Mackay 2011; Conard et al. 2012; Porraz etal. 2013b). In the WRZ, these “post-
Howiesons Poort” assemblages also witness decreases both in artifact discard and
in the prevalence of silcrete.
The archaeological signal in the WRZ becomes increasingly muted after 50 ka
(Mackay 2010; Faith 2013). Few sites have ages in this range, and those that do are
often associated with only ephemeral accumulations of aked stone. The causes of
this quietude are presently not well understood. Notably, it is not matched by a simi-
lar occupational absence in the SRZ where signals are relatively robust (Mackay
etal. 2014a). After ~25 ka, the occupational signal in the region strengthens once
again, but by this time, the MSA has ended and Later Stone Age (LSA) technologies
dominate (Deacon 1978; Deacon etal. 1984; Orton 2006; Mackay 2010).
Provisioning, Mobility, andPopulation
Provisioning can be dened as the means by which hunter-gatherers overcame the
spatial and temporal mismatch between opportunities to make stone tools and loca-
tions at which stone tools were deployed. Kuhn (1995) differentiates two idealized
forms of provisioning: provisioning of individuals and provisioning of places.
Provisioning of individuals involves foragers equipping themselves with tools suf-
cient to underwrite tasks encountered before the next opportunity to retool.
Provisioning of places involves foragers equipping landscape nodes with tool stone,
turning them into sources, and allowing them to function as gearing-up locations.
The inuences on these provisioning systems, and their archaeological manifes-
tations, are expected to be different. Individual provisioning can be deployed when
impending tasks are difcult to anticipate, providing maintainable gear of general
utility (Nelson 1991). In contrast to individual provisioning, place provisioning can
function where extended or repeated occupation of a given location can be antici-
pated (Kuhn 1995) and therefore presumes some foreknowledge of the nature and
duration of impending activities. Necessarily then, place provisioning is predicated
on predictable (though not necessarily abundant) subsistence conditions. Where
individual provisioning might emphasize the transport, use, and conservation of a
small number of largely utilitarian items in maintainable toolkits (Shott 1986; Kuhn
1994), place provisioning is expected to result in on-site reduction of transported
rocks – most likely cores – and the consequent accumulation of manufacturing
debris at residential bases (Binford 1980; Nelson 1991; Kuhn 1995; Riel-Salvatore
and Barton 2004; Mackay 2005).
2 Provisioning Responses toEnvironmental Change inSouth Africa’s
A key difference between these provisioning systems is in the way in which gearing
up occurs. Place provisioning allows groups to gear up at selected locations– presum-
ably central places in a foraging range. Provisioned places can then function as nodes
for logistical forays into the surrounding landscape. The viability of such an approach,
as noted above, depends on the predictability of subsistence resources, and particularly
water, at that occupational node (Kelly 1983, 1995; Read 2008). In the absence of place
provisioning, however, gearing up needs to have occurred on the landscape at points
where the foraging round intersected appropriate sources of usable stone.
While there has so far been little discussion of provisioning systems in the Late
Pleistocene of southern Africa, there has been some discussion of changing patterns
of mobility (Ambrose and Lorenz 1990; McCall 2007; McCall and Thomas 2012).
Perhaps the most inuential of these is the Ambrose and Lorenz (1990) model of
mobility during the Howiesons Poort. In that model, cool and dry conditions pre-
sumed to be associated with the glacial MIS 4 were inferred to have resulted in
decreased resource abundance. As a consequence, this increased the distance and
frequency of movements required to fulll minimum subsistence needs. In alterna-
tive arguments, Mackay (2009) and McCall and Thomas (2012) have argued that
the Howiesons Poort was a time in which logistical movements from residential
bases were emphasized.
The relative abundance of artifacts in Howiesons Poort assemblages, coupled
with the apparent humidity of that time (Chase 2010), seem to argue against the
Ambrose and Lorenz (1990) model of constant long-range population movement;
however, it is still viable if there was signicant change in a third factor– popula-
tion. Recent arguments have attempted to link the perceived increases in technologi-
cal and behavioral complexity during the Still Bay and Howiesons Poort periods to
population increases across southern Africa (Powell etal. 2009). The absence of
direct evidence for population increase in the Howiesons Poort notwithstanding
(Mackay 2011), and allowing that only the Howiesons Poort seems to be distinc-
tively abundant, it is possible to reconcile increases in mobility and relative abun-
dance during the Howiesons Poort if there were dramatic increases in the number of
individuals on the landscape at this time. Under these conditions, while people may
have been moving further and more often, the increased population could conceiv-
ably have resulted in greater accumulations of artifacts given a sufcient number of
visits to sites, however brief. (It is unclear why population increase should align
with a period of resource duress in this scenario, but this is not important with
respect to testing the model).
Taken together, these various principles and models allow us to develop some
expectations for changes in provisioning systems through the Late Pleistocene in
southern Africa. First, if humid conditions were those most favorable to place provi-
sioning, then this is most likely to have occurred during the mid to late MIS 4 (follow-
ing Stuut etal. 2002; Chase 2010). We would expect to see these periods associated
with increases in numbers of artifacts and specically cores, as tool- making potential
was transported to specic sites. Assuming direct-to-site transport of cores, rather than
their transportation and maintenance as part of a mobile toolkit, we would expect to
see multiple stages of reduction on-site, potentially including initial or early decortica-
A. Mackay et al.
tion. Gearing up would also occur at these locations, and this may have resulted in the
production and discard of complex tools if there were sufcient levels of subsistence
risk (Bousman 1993; Collard etal. 2005; Read 2008). The other side of this inference
is that beyond the selected landscape “nodes” which were provisioned, we would see
little evidence of gearing up and consequently fewer sites around the landscape gener-
ally. Sites beyond the nodes in these periods would be represented by isolated artifacts
from episodic discard events.
Second, if less humid conditions were less suited to place provisioning and
instead encouraged a greater reliance on individual provisioning, then we would
expect to see this during late MIS 5/early MIS 4 and potentially during MIS 3.
Gearing up would be more likely to occur at landscape locations where tool stone
was directly available. Otherwise, sites would show relatively small accumulations
of artifacts with fewer cores and potentially more tools.
Third, if human populations were abundant, and regardless of the provisioning
systems used, we might reasonably expect an abundance of archaeological material
at the landscape scale. In effect, the entire signal of such periods should be particu-
larly “loud” relative to others.
Study Area
Our study area for this paper is centered on the northern Cederberg mountains situ-
ated in the heart of the WRZ in the Western Cape Province of South Africa (Fig.2.1).
Rainfall arrives predominantly from northwesterly winds during winter storms, and
there is little summer rain. In the middle of the study area is the Olifants River valley
located between the low coastal mountain range to the west and the Cederberg
mountains which rise to >2000 masl to the east. The Cederberg and coastal moun-
tains are composed mainly of deeply incised Table Mountain Sandstone (Visser and
Theron 1973). The Olifants River is the only permanently owing watercourse in
the study area. The river is presently dammed in the area of Clanwilliam, creating a
10-km-long body of standing water, the volume of which exhibits marked seasonal
uctuations. Further to the west is the Sandveld, a thick coastal sandsheet falling
between the coastal range and the sea. The Sandveld is punctuated by sandstone
ridges which are often widely spaced. Drainages in this area tend to be relatively
ephemeral with short catchments rising to the west of the coastal range.
Rocks suitable for the manufacture of aked stone artifacts show a variable
distribution across the landscape. Quartzite and quartz are relatively abundant
and widely distributed. Quartzite occurs as massive geological bands and asso-
ciated scree and as river cobbles in the Olifants River. Quartz occurs as small
pebbles eroding from conglomerate beds and is dispersed across land surfaces.
The ne-grained sedimentary rock silcrete is less common than the other major
rock types and tends to occur as discrete, infrequent outcrops (Roberts 2003;
Porraz etal. 2013b). Cobbles of silcrete have not been observed in the Olifants
River during surveys.
2 Provisioning Responses toEnvironmental Change inSouth Africa’s
The archaeological sites for this study include both excavated rock-shelter
deposits and numerous open artifact scatters. The two main rock-shelters are Klein
Kliphuis in the Olifants River valley and Diepkloof in the Sandveld. Both sites have
later Howiesons Poort and post-Howiesons Poort components, while Diepkloof
additionally has an earlier Howiesons Poort and a Still Bay industry (Mackay 2009,
2010; Porraz etal. 2013b). The Diepkloof data presented here derive from the L6
column sequence (Mackay 2009).
Surveys for open sites have generally not been systematic with the exception of
those conducted by one of us (EH) in the area surrounding Clanwilliam Dam
(Hallinan 2013). In those surveys, temporally diagnostic artifact markers– bifacial
points for the Still Bay, backed artifacts for the Howiesons Poort, and unifacial
points for the post-Howiesons Poort– were used to characterize distributions for
various industries rather than “sites” per se (see discussion in Hallinan 2013). In
addition to these surveys, we will present some artifact distribution data from loca-
tions around other excavated contexts in the region, specically around Varsche
Rivier 003 immediately to the north of the present study area and along the Doring
River to the east (Steele etal. 2012).
Our analytic methods for shelter sites largely follow Riel-Salvatore and Barton
(2004) in using artifact density and proportions of retouched akes to assess place
and individual provisioning. Density of artifacts per unit volume of deposit is not
presently available for Diepkloof, and artifact discard per unit time is complicated
by the contested OSL ages. Consequently, we use numbers of artifacts per industry
as a basic measure of discard and supplement this with discard per unit time within
the available age constraints.
Caveats aside, we expect that where place provisioning has occurred, large
numbers of akes and cores will have been discarded with relatively low propor-
tions of retouched akes. The opposite should hold where individual provisioning
dominates. The proportion of cores to retouched akes should thus provide an
approximation of shifts between place and individual provisioning.
In addition to these data, we also consider evidence for cortex proportions, par-
ticularly in the silcrete that is not locally available at either of the shelters under
consideration. If cores were transported and reduced as part of toolkits, then we
would expect to see evidence for reduction consistent with spatial distance to source.
That is, there should be little cortex on silcrete akes. Alternatively, if place provi-
sioning occurred and cores were transported more or less directly from source to
site, then distance to source should less effectively predict cortex prevalence. In that
sense, it is pertinent to remember that spatial distance to source is relevant only
insofar as it is a proxy for time. Where spatial distance to source is held constant,
direct movements from source to discard location are expected to produce less
reduction than where movements in the intervening space are tortuous.
A. Mackay et al.
For the open sites, our interest is in the distribution and abundance of implements
relative to stone sources. We use the term “implement” in this sense to refer to
morphologically regular retouched akes and more specically those that are dis-
tinctive of different industries, here, bifacial points, backed artifacts, and unifacial
points. If individual provisioning dominates and assuming that, in any given
industry, multiple implements are being made at procurement opportunities, we
would expect to see clustering of these implements at or near sources of stone.
Where place provisioning dominates, implements should cluster at selected and
presumably central foraging nodes that are not predicted by the distribution of
stone sources.
We start our results by looking at the data from rock-shelter sequences and move on
to open site distribution data.
Discard Rates, Cores, andRetouched Flakes
At Diepkloof and Klein Kliphuis, artifact numbers are exceptionally large in the later
Howiesons Poort, with weaker discard during the Still Bay and post- Howiesons
Poort and moderate discard during the earlier Howiesons Poort (Table2.1). As noted,
these data are based on discard per industry as opposed to discard per unit time.
We can factor for time if we allow for multiple potential outcomes given the
Table 2.1 Numbers of artifacts and discard rates per industry for Diepkloof and Klein Kliphuis
calculated using the Jacobs etal. (2008a, b) and Tribolo etal. (2012) chronologies separately
Site Industry
Duration (kyr)
(central age range) Discard rate (artifacts / kyr)
(2008a, b)
DRS Still Bay 1519 7–3 29–9 217 506 52 169
Early HP 2852 12–8 39–20 238 357 73 143
Later HP 9604 7–3 47–33 1372 3201 204 291
Post-HP 614 11–8 56 77
KKH Later HP 26139 12–6 2178 4357
Post-HP 5060 7–3 723 1687
Note that no discard rates are calculated for the post-Howiesons Poort at Diepkloof using the
Tribolo etal. (2012) chronology because only a single central age is given and the duration of the
industry cannot be ascertained
2 Provisioning Responses toEnvironmental Change inSouth Africa’s
discrepancies between the Jacobs et al. and Tribolo etal. chronologies and their
inherent uncertainties (Table2.1). The dates suggest that, regardless of the chronol-
ogy used or whether discard rates are calculated for the maximum spread at one
sigma error or just for the central ages, the later Howiesons Poort always returns the
highest values. At Klein Kliphuis, discard rates are typically three times higher in
the Howiesons Poort than the post-Howiesons Poort. At Diepkloof, later Howiesons
Poort discard rates are roughly double those seen in the earlier Howiesons Poort
and Still Bay and roughly 25–40 times higher than in the post-Howiesons Poort,
albeit the post-Howiesons Poort in this part of the site may be truncated (Porraz
etal. 2013b).
Core and retouch numbers show similar patterns in terms of absolute quantities
to total artifact numbers; however, their relative proportions are quite variable
(Table2.2). Retouched akes are proportionally most common in the Still Bay at
Diepkloof and in the post-Howiesons Poort at Klein Kliphuis, while cores are most
common in the early Howiesons Poort at Diepkloof. A χ2 test for variance across
both sites is signicant at 0.05 (p< 0.001, df= 5, Cramer’s V=0.145). Adjusted
residuals suggest that the variance is being driven principally by the oversupply of
cores in the early Howiesons Poort and the oversupply of retouched akes in the
Still Bay (at Diepkloof) and post-Howiesons Poort (at Klein Kliphuis). With respect
to the post-Howiesons Poort at Diepkloof, it should be noted that this has the small-
est sample of any of the units considered.
It should alsobe noted that the Howiesons Poort industries (both early and later)
include numerous backed artifacts which were probably not maintained tools; thus,
even though retouched ake proportions are relatively low in these layers, they
probably still overestimate the proportion of maintained tools in the assemblage.
Raw Material Proportions andCortex
Considerable changes occur in the proportions of different rock types through the
sequences (Table2.3); only the major rock types- quartz, quartzite and silcrete- are
included, as these combined account for more than 90% of artifacts in all units.
Table 2.2 Proportions of cores to retouched akes and adjusted residuals, for all industries at
Diepkloof and Klein Kliphuis
Site Industry
Cores Retouched akes
Cores/Retouched akesnAdj. res. nAdj. res.
DRS SB 35 3.4 77 3.4 0.45
Early HP 54 2.1 41 2.1 1.21
Later HP 132 0.9 139 0.9 0.95
Post-HP 13 0.4 13 0.4 1
KKH Later HP 267 1.9 274 1.9 0.97
Post-HP 52 3.0 97 3.0 0.53
A. Mackay et al.
Both sites have high proportions of silcrete in the later Howiesons Poort and
post-Howiesons Poort. At Klein Kliphuis, silcrete peaks in the former; at Diepkloof,
it peaks in the latter. The abundance of silcrete at both sites in the later Howiesons
Poort industry, which also has the largest numbers of artifacts and cores, suggests
that large quantities of that material, though likely not available in the immediate
surrounds of the site, were being transported to sites in these periods.
Silcrete cores are also particularly common in these industries. There are 133
complete silcrete cores in the later Howiesons Poort at Klein Kliphuis, compared
with 20in the post-Howiesons Poort. At Diepkloof, there are 57 silcrete cores in the
later Howiesons Poort; the remaining industries combined have only 16. Thus, there
is a lot of silcrete, particularly cores, being transported to sites in the later Howiesons
Poort. The remaining question is whether silcrete cores were being reduced in a way
that suggests extensive transport and maintenance prior to arrival on-site.
Cortex data suggest that silcrete akes with >50% cortical coverage on their
dorsal surface are more common in Howiesons Poort than other units at both sites,
though the patterns are statistically weak ( χ2Diepkloof: p = 0.87, df = 3, Cramer’s
V=0.113; χ2KleinKiphuis: p=0.81, df=1, Cramer’s V=0.041)1. Variance at Diepkloof
is driven chiey by the absence of cortical silcrete akes in the Still Bay and their
comparative abundance in the later Howiesons Poort. While typical “Howiesons
Poort” cores (cf., Villa etal. 2010) retain a cortical lower surface throughout their
reduction, akes with 50% cortex likely relate to initial core setup and ongoing
maintenance (Porraz etal. 2013b). Thus, while large numbers of silcrete cores were
being transported to both sites in the Howiesons Poort, they appear often to have
been in sufciently early stages of reduction to have produced highly cortical akes.
This observation is not easily reconciled with consistently extensive reduction of
cores prior to arrival on-site (Table2.4).
While the later Howiesons Poort and Still Bay seem to have relatively clear
signals with respect to provisioning systems, the post-Howiesons Poort and early
Howiesons Poort are perhaps more enigmatic. The post-Howiesons Poort has very
high percentages of silcrete in the context of systems that seem otherwise consistent
1 These data are generated by site rather than for the whole sample given differences in the raw
material availability in the Sandveld and in the Olifants River discussed earlier, where all data are
considered together p=0.001, df=5, and Cramer’s V=0.093.
Table 2.3 Relative proportions of the major raw material types per industry for Diepkloof
and Klein Kliphuis, artifacts >15mm only
Site Industry
Quartz Quartzite Silcrete
% n % n % n
DRS SB 18.3 178 54.9 533 26.8 260
Early HP 49.6 555 38.2 427 12.2 137
Later HP 32.5 745 16.4 377 51.1 1173
Post-HP 18.6 41 10.0 22 71.5 158
KKH Later HP 8.4 411 5.9 288 85.7 4200
Post-HP 8.7 154 28.6 505 62.7 1109
2 Provisioning Responses toEnvironmental Change inSouth Africa’s
with individual provisioning. In the early Howiesons Poort, cores are relatively
common but artifact discard rates are moderate. Proportions of retouched akes
indicate low frequencies of ake maintenance, but silcrete artifacts are uncommon
and silcrete cores – which might be taken to indicate place provisioning with
high- quality rocks – are rare. The most abundant rocks in this industry are locally
available quartz and quartzite. Based on this, the provisioning possibilities that arise
are either episodic place provisioning with local rocks or individual provisioning
principally with maintained cores. We are unable to resolve these possibilities with
the available data.
Open Site Data
So far, the data presented suggest that the later Howiesons Poort has a particularly
strong archaeological signal in the study area as measured by the abundance and
discard rates of artifacts and that this may be a result of place provisioning. A further
test of this proposition comes from the distribution of open sites. If the later
Howiesons Poort is a consequence of place provisioning to allow gearing up at spe-
cic locations, then we may or may not nd open site expressions of that industry.
On the other hand, if the abundance of the later Howiesons Poort is simply a product
of massive increases in the number of people using the landscape, then we would
expect to nd the signal of the later Howiesons Poort widely distributed in open site
contexts. Furthermore, if the Still Bay and post-Howiesons Poort industries reect
individual provisioning, we expect to nd evidence of that in gearing-up sites near
sources of selected tool stone.
Clanwilliam Dam Surveys
Extensive surveys in the area surrounding Clanwilliam Dam were conducted by one
of us (EH) with assistance from students at the University of Cape Town. These
surveys resulted in the location of artifacts dating to the Earlier, Middle, and Later
Table 2.4 Proportions of silcrete akes with >50% and <50% cortex and adjusted residuals, for
all industries at Diepkloof and Klein Kliphuis
Site Industry
Cortex <=50 Cortex>50
% n Adj. res. % n Adj. res.
DRS SB 100 79 2.1 0 0 2.1
Early HP 95.7 45 0.1 4.3 2 0.1
Later HP 93.8 316 2.3 6.2 21 2.3
Post-HP 98.0 50 1.0 2.0 1 1.0
KKH Later HP 90.2 1329 1.7 9.8 145 1.7
Post-HP 93.2 316 1.7 6.8 23 1.7
A. Mackay et al.
Stone Ages. The MSA component included bifacial points likely indicative of the
Still Bay, backed artifacts that may relate to the Howiesons Poort2, and unifacial
points likely related to the post-Howiesons Poort.
Bifacial points were identied at ten localities, distributed throughout the land-
scape (Fig.2.3a). Most locations had between one and three points, but one site,
known as Clanwilliam Dam East, has at least 50. Clanwilliam Dam East occurs in a
small embayment on the edge of the dam where seasonal uctuations in the water
level have eroded a shallow body of sediment, which is exposed during the dry sum-
mer months and submerged through most of winter. The 50 points so far located
represent a minimum for the site. Of the 37 analyzed points, ~95% are made of
silcrete and ~46% of these retain cortex. The assemblage includes both heavily
reduced and recycled points and those in the earliest stages of manufacture (Fig.2.4).
While no silcrete source has been identied at the site, a proximate source cannot be
precluded due to the dam, which forms the western boundary of the site.
A single backed artifact was located during the surveys (Fig.2.3b). The artifact
was made of gray-green silcrete and measured 33mm, making it comparable in
size to the later Howiesons Poort backed artifacts from Klein Kliphuis (Mackay
2011). Interestingly, the artifact was found in the same embayment as the bifacial
points (Fig.2.4). Given the overlap of bifacial points and backed artifacts in the
latest Still Bay phase at Diepkloof (cf., Porraz etal. 2013b), it is possible that this
backed piece is not in fact Howiesons Poort. This aside, the surveys revealed no
evidence for the concerted production of backed artifacts at any open site location
around Clanwilliam Dam.
Unifacial points, like bifacial points, were widely distributed in the survey area
(Fig.2.3c). A total of 26 unifacial points was identied, spread over 9 localities.
Three sites had ve, three, and two points, respectively, and ve were found as
single points within larger assemblages. At Clanwilliam Dam East, 11 unifacial
points were recorded, but given their context alongside a substantial number of bifa-
cial points, these may represent a stage in bifacial point manufacture (Högberg and
Larsson 2011). As with backed artifacts, there was no evidence for the concerted
production of these implements at any location.
The Clanwilliam Dam survey data reveal some interesting patterns but they have
limitations. The presence of the dam precludes survey of the margins of the Olifants
River in this area, meaning that certain, specic locations remain unexplored. We
can to some extent assess the validity of the patterns found there by considering
partial survey data from other areas.
2 Two points are madehere. First, it is not presently possible to differentiate the earlier and later
phases of the Howiesons Poort solely on the basis of implement form, and the two are therefore
considered together here. Second, backed artifacts do recur in the LSA, but these are commonly
smaller than their MSA counterparts. We return to this point later in this section.
2 Provisioning Responses toEnvironmental Change inSouth Africa’s
Fig. 2.3 Triangles showing the distribution of (a) bifacial points, (b) backed artifacts, and (c)
unifacial points around Clanwilliam Dam
Fig. 2.4 Clanwilliam Dam East, looking east. Black oval shows the location of the concentration
of bifacial points. Points on the right-hand side include early-stage (bottom half) and late-stage/
reworked points (top half). The uppermost point image shows both faces of a recycled point which
had been snapped at one end and subsequently aked to remove blades from the margins
Doring River Surveys
To the east of the Olifants River valley is the Doring River. The two rivers run paral-
lel for much of their course before merging and owing to the sea together as the
Olifants. The Doring River has remnant Late Pleistocene terraces along more than
50km of its course, and these are often covered in MSA artifacts (Mackay etal.
2014b). Surveys so far have resulted in the identication of bifacial points at eight
localities (Fig.2.5a). These include two instances of isolated points, two instances
of two points together, three instances of multiple points (six, six, and eight points),
and one instance of >140 bifacial points of which 136 have been analyzed. This
large assemblage of points occurs as a spatially discrete component of a much larger
MSA scatter known as Kleinhoek 1, situated on a large terrace (Fig.2.6). More than
95% of these points are made of relatively ne-grained blue-gray quartzite which
outcrops on the ridge and scree slope that forms the western boundary of the site.
The assemblage includes early-stage and late-stage points.
A. Mackay et al.
So far, we have only identied two backed artifacts, both at a single locality–
Uitspankraal 7 (Fig.2.5b). These artifacts were both identied as isolated nds dur-
ing repeated nonsystematic surface sampling of the site.
Unifacial points have been identied at seven localities, ve of which also have
bifacial points (Fig.2.5c). Four included single, isolated points. Of the remainder,
one included two unifacial points, and another included four points though these
were in the same cluster as the bifacial points at Kleinhoek 1 and may therefore
relate to bifacial point manufacture (Högberg and Larsson 2011). The nal site,
Uitspankraal 7, like Kleinhoek 1, is a large terrace with multiple archaeological
components including MSA and LSA artifacts. In the center of the scatter is a dense
concentration of heated silcrete cores and akes and 18 unifacial points which we
Fig. 2.5 Triangles showing the distribution of (a) bifacial point sites, (b) backed artifact sites, and
(c) unifacial point sites along the Doring River. Rock-shelters (circles) are PL8 Putslaagte 8, KFR
Klipfonteinrand, and MRS Mertenhof
Fig. 2.6 Kleinhoek 1, looking northeast. The scatter covers the entire erosion feature; however, all
of the ~140 points are constrained to the area covered by the black oval. The points on the right-
hand side include examples of early-stage points (bottom half) and late-stage/reworked points (top
half). Scale bars are 10mm. Note that the scree slope at the bottom of the site photo includes the
same raw material of which the points are made
2 Provisioning Responses toEnvironmental Change inSouth Africa’s
believe to be associated with the post-Howiesons Poort (Will etal. 2015). While
there are also bifacial points at this site, they are located in a different area of the
extensive scatter, and their distributions do not appear to overlap. Another similarly
dense post-Howiesons Poort scatter has also been located on the Tankwa River– a
tributary of the Doring River– a further 40km southeast of Uitspankraal 7 under a
different survey program (Hallinan and Shaw 2015).
The 16 artifact-bearing localities on terraces so far identied along the Doring
River have yielded more than 160 bifacial points and 28 unifacial points, with
only two backed artifacts. In spite of this, three of the four excavated rock-shelter
sites in the Doring River catchment have Howiesons Poort components
(Klipfonteinrand, Mertenhof, Putslaagte 8), and only two each have the Still Bay
and post-Howiesons Poort (Högberg and Larsson 2011; Mackay etal. 2015; Will
etal. 2015). Consistent with the patterns elsewhere, in the only site to contain all
three of these industries (Mertenhof), the density of artifacts is greatest in the
Howiesons Poort (Will etal. 2015).
Knersvlakte Surveys
The nal open site data come from around the excavated site of Varsche River 003
(Steele et al. 2012; Steele et al. 2016), located in the Knersvlakte region of
Namaqualand. Varsche River 003 includes Still Bay and Howiesons Poort industries
but as yet no post-Howiesons Poort. Limited surveys were undertaken around the
site during excavation, and full details will be published elsewhere. We note here
only that we recorded bifacial points at ve sites. Three of these were isolated, and
one site included three points. At the fth site, we identied 60 bifacial points dur-
ing our initial analysis (Mackay etal. 2010) and have subsequently expanded this
sample to 142, with three unifacial points. The site– Soutfontein– occurs on the
banks of the Varsche River immediately below a large outcrop of quartz. Quartz
accounts for ~67% of points. No other unifacial points have been identied in the
surrounding area and no backed artifacts.
The data presented here depict fairly clear patterns with respect to provisioning in
some industries and less clear patterns in others. It seems probable that the com-
monly noted density of artifacts in the later Howiesons Poort is a consequence of
place provisioning in the context of cool and humid conditions of mid to late MIS
4. Large numbers of cores from spatially restricted sources were transported to shel-
ter sites in this industry and subsequently reduced from an initial state in which they
still retained considerable cortical coverage. The fact that many of these retouched
pieces were not maintainable suggests that they would have been suited to logistical
and possibly task-specic trips (McCall 2007; Mackay 2009; McCall and Thomas
A. Mackay et al.
2012). In support of the artifact data, geoarchaeological evidence for site use
suggests extended periods of fairly intensive occupation during the later Howiesons
Poort (Goldberg etal. 2009; Miller et al. 2013; Karkanas et al. 2015), consistent
with the use of sites as residential bases.
The near-complete absence of evidence for Howiesons Poort sites on the land-
scape provides important additional information. Site occupation was clearly very
specic at this time, with gearing up occurring mainly if not exclusively at selected
sites away from stone sources. Even though backed artifacts are small and perhaps
more easily missed than the other implement types on which we focused, had there
been instances of substantial backed artifact production sites in the open compara-
ble to those in shelters, we expect we would have seen them. While other research-
ers have found Howiesons Poort backed artifacts in duneeld settings under perhaps
more ideal survey conditions, it is notable that they too have failed to nd clear
open-air Howiesons Poort manufacturing sites (e.g., Dietl etal. 2005; Kandel and
Conard 2012) (though note Carrion etal. 2000). Overall, it seems to us improbable
that the signal we are witnessing is the result of very large numbers of highly mobile
people moving across the landscape.
The Still Bay inverts many of these outcomes. Discard rates are generally low,
cores are rare, and the proportion of retouched akes is high. The dominant imple-
ment form– bifacial points– could be characterized as maintainable. While there
was a reasonable amount of silcrete transported to Diepkloof in this period, the
paucity of cortical akes combined with the shortage of cores argues against place
provisioning. We also found clear evidence for dedicated gearing-up locations out
on the landscape. At least two such instances were in close proximity to sources of
stone that comprised the majority of assemblages, and this could not be precluded
at the third (Clanwilliam Dam East). Given that the Clanwilliam Dam East assem-
blage was dominated by a single stone type, included a mix of heavily reduced and
early-stage points, and had a reasonably high proportion of points on which cortex
was still visible, it is likely that a source of stone was located fairly close to the site.
Overall, we believe that the Still Bay signal is consistent with gearing up at sources
of stone to provision individuals with maintainable implements. This system of
organization occurred in the context of rapidly cooling and perhaps increasingly
humid conditions from MIS 5 to MIS 4 (accepting the Jacobs chronology) or under
variable conditions of earlier MIS 5 (accepting the Tribolo chronology with its pro-
portionally large uncertainties).
The earlier Howiesons Poort presents a less clear signal with the use of local
rocks, moderate rates of discard, reasonable numbers of cores, and low proportions
of retouched akes, among which the dominant implement type is not considered
maintainable. Either limited/episodic place provisioning with local rock or indi-
vidual provisioning with a combination of cores and tools– or perhaps variable use
of both– may explain this pattern. This issue requires further consideration. The
Jacobs chronology places this industry in the context of cool and humid conditions
of early to mid MIS 4. In the Epica Dome C southern hemisphere temperature
record, this is the coldest period of that stage (Jouzel etal. 2007).
2 Provisioning Responses toEnvironmental Change inSouth Africa’s
The post-Howiesons Poort, like the earlier Howiesons Poort, has a somewhat mixed
signal. Most of its characteristics in shelter sites– low discard rates, few cores, high
proportions of retouched akes, and maintainable implements– suggest individual pro-
visioning. However, this needs to be balanced against the abundance of silcrete which
is not local to either of the sites studied. As Binford (1979) notes, where individuals are
provisioned with a limited amount of transported gear, we expect them to make supple-
mentary use of locally available rocks; yet the evidence for this at Diepkloof and Klein
Kliphuis is weak. Furthermore, the clearest post- Howiesons Poort site identied on the
landscape appears in many respects to reect place provisioning. At Uitspankraal 7, we
see preferential transportation and reduction of cores of nonlocal rock at a specic
landscape location where multiple implements were made.
We suspect that the solution to this puzzle lies in the coarse nature of industries
and their complex relationship with behavior. An industry is generally identied
based on a limited number ofartifact characteristics. In the case of our open site
data, it was one variable alone that we used to determine industry afnity. Yet tech-
nologies are complex mixtures of mobility, provisioning, material selection, core
reduction and discard, and implement production. These variables may change at
different rates and in relation to different controls. The Howiesons Poort to post-
Howiesons Poort transition, for example, has been shown to involve gradual
decreases in blade production and increases in ake size (Soriano etal. 2007; Villa
etal. 2010; Mackay 2011). In contrast, the change from backed artifacts to unifacial
points seems to have been relatively rapid. At Klein Kliphuis, artifact densities
remain quite high in the earliest post-Howiesons Poort (Mackay 2010). These are
also the layers in which silcrete is abundant; thereafter, quartzite and eventually
quartz come to dominate. Cores are also common in the earliest post-Howiesons
Poort, but what is notable is that their size at discard increases through time from the
later Howiesons Poort into the earlier post-Howiesons Poort before cores become
altogether infrequent. It may be that while other elements of technology were
changing across the Howiesons Poort to post-Howiesons Poort transition, the nature
of provisioning initially remained much the same– that of large quantities of sil-
crete transported to site, sometimes as cores. However, the return on this provision-
ing investment, measured here as the reduction of cores prior to discard, may have
begun to decrease as increased aridity resulted in shorter and perhaps less predict-
able durations of site use. Thus, the change from place to individual provisioning
may have occurred some time after the transition from the Howiesons Poort to post-
Howiesons Poort, resulting in the mixed provisioning signal for the later industry–
something suggested by the core to retouched ake ratios in the earliest
post-Howiesons Poort at Diepkloof. Viewed in this light, the open site post-
Howiesons Poort occurrence at Uitspankraal 7 may be seen to document the con-
tinuation of place provisioning in the earliest post-Howiesons Poort but a spatial
shift from selected and presumably preferred landscape locations in rock-shelters to
a better-watered location on the Doring River. The value of the post-Howiesons
Poort here may be in cautioning us against making an oversimplistic association
between industry and behavior (Mackay 2016).
A. Mackay et al.
The objective of this paper was to explore changes in provisioning through the later
MSA in southern Africa’s WRZ and to compare these changes to environmental
variability. The latter goal is complicated by chronological uncertainties and the
paucity of terrestrial paleoenvironmental archives. Nevertheless, we feel it has been
possible to identify some clear changes in provisioning and, where chronologies
allow, to suggest that place provisioning occurred most notably in the context of
cool humid conditions during the later Howiesons Poort. During the Still Bay,
though environmental conditions are poorly resolved, we feel condent that indi-
vidual provisioning dominated, with gearing up occurring at raw material sources.
Provisioning in the other industries was less easily resolved, but we posit that at
least part of our difculty here arose from the concept of industries and their capac-
ity to mask behaviorally meaningful variation.
Beyond provisioning, our paper also suggests that the noted abundance of
archaeology during the later Howiesons Poort is biased by an occupational system
focused on rock-shelters and an approach to archaeology which often does very
much the same. Consideration of the rock-shelter record alone can thus be mislead-
ing. It seems quite likely that rock-shelters had specic conditions of use and that
these conditions were only met periodically. An extension of this observation is that
the lack of an observable archaeological record in the WRZ from 50 to 25ka, on
which others and we have commented (Mitchell 2008; Mackay 2010; Faith 2013),
cannot necessarily be taken at face value. If occupation can focus on shelters to the
exclusion of open sites under some conditions, it is plausible that the inverse could
hold under others (Mackay etal. 2014b).
Acknowledgments AM’s research has been supported by grants from the Australian Research
Council (DP1092445, DE130100068). Fieldwork conducted by EH was funded by an NRF African
Origins Platform grant awarded to John Parkington.
Ambrose, S. H. (2002). Small things remembered: Origins of early microlithic industries in sub-
Saharan Africa. In R. G. Elston & S. L. Kuhn (Eds.), Thinking small: Global perspectives on
Microlithization, Archaeological Papers of the American Anthropological Association, no 12
(pp. 9–30). Arlington, VA: American Anthropological Association.
Ambrose, S.H., & Lorenz, K.G. (1990). Social and ecological models for the Middle Stone Age in
Southern Africa. In P.Mellars (Ed.), The emergence of modern humans (pp.3–33). Edinburgh:
Edinburgh University Press.
Avery, G.D., Halkett, D., Orton, J., Steele, T.E., & Klein, R.G. (2008). The Ysterfontein 1 Middle
Stone Age Rockshelter and the evolution of coastal foraging. South African Archaeological
Society Goodwin Series, 10, 66–89.
Binford, L.R. (1979). Organization and formation processes: Looking at curated technologies.
Journal of Anthropological Research, 35(3), 255–273.
2 Provisioning Responses toEnvironmental Change inSouth Africa’s
Binford, L. R. (1980). Willow smoke and dogs tails: Hunter-gatherer settlements systems and
archaeological site formation. American Antiquity, 45(1), 4–20.
Bousman, C. B. (1993). Hunter-gatherer adaptations, economic risk and tool design. Lithic
Technology, 18(1&2), 59–86.
Brown, K.S., Marean, C.W., Herries, A.I. R., Jacobs, Z., Tribolo, C., Braun, D., Roberts, D.L.,
Meyer, M.C., & Bernatchez, J.(2009). Fire as an engineering tool of early modern humans.
Science, 325(5942), 859–862.
Brown, K.S., Marean, C.W., Jacobs, Z., Schoville, B.J., Oestmo, S., Fisher, E.C., Bernatchez, J.,
Karkanas, P., & Matthews, T. (2012). An early and enduring advanced technology originating
71,000 years ago in South Africa. Nature, 491(7425), 590–593.
Bruch, A.A., Sievers, C., & Wadley, L. (2012). Quantication of climate and vegetation from
Southern African Middle Stone Age sites: An application using late Pleistocene plant material
from Sibudu, South Africa. Quaternary Science Reviews, 45, 7–17.
Carr, A.S., Chase, B.M., & Mackay, A. (2016). Mid to late quaternary landscape and environmen-
tal dynamics in the Middle Stone Age of Southern South Africa. In B.A. Stewart & S.Jones
(Eds.), Africa from MIS 6–2: Population dynamics and Paleoenvironments (pp. 23–47).
Dordrecht: Springer.
Carrion, J.S., Brink, J. S., Scott, L., & Binneman, J.N. F. (2000). Palynology and palaeoenvi-
ronment of Pleistocene hyaena coprolites from an open-air site at Oyster Bay. South African
Journal of Science, 96, 449–453.
Cartwright, C. R. (2013). Identifying the woody resources of Diepkloof Rock Shelter (South
Africa) using scanning electron microscopy of the MSA wood charcoal assemblages. Journal
of Archaeological Science, 40(9), 3463–3474.
Chase, B.M. (2010). South African palaeoenvironments during marine oxygen isotope stage 4:
A context for the Howiesons Poort and Still Bay industries. Journal of Archaeological Science,
37(6), 1359–1366.
Chase, B.M., & Meadows, M.E. (2007). Late quaternary dynamics of southern Africa’s winter
rainfall zone. Earth-Science Reviews, 84(3–4), 103–138.
Collard, M., Kemery, M., Banks, S., & S. (2005). Causes of toolkit variation among hunter-
gatherers: A test of four competing hypotheses. Canadian Journal of Archaeology, 29, 1–19.
Conard, N.J., Porraz, G., & Wadley, L. (2012). What’s in a name? Characterising the ‘post- Howiesons
Poort’ at Sibudu. South African Archaeological Bulletin, 67(196), 180–199.
Cowling, R.M., Cartwright, C.R., Parkington, J.E., & Allsopp, J.C. (1999). Fossil wood charcoal
assemblages from Elands Bay Cave, South Africa: Implications for late quaternary vegetation
and climates in the winter-rainfall fynbos biome. Journal of Biogeography, 26, 367–378.
de la Pena, P., Wadley, L., & Lombard, M. (2013). Quartz bifacial points in the Howiesons Poort
of Sibudu. South African Archaeological Bulletin, 69(198), 119–136.
Deacon, H. J., Deacon, J., Scholtz, A., Thackeray, J. F., Brink, J.S., & Vogel, J. C. (1984).
Correlation of palaeoenvironmental data from the late Pleistocene and Holocene deposits at
Boomplaas cave, southern cape. In J.C. Vogel (Ed.), Late Cainozoic Palaeoclimates of the
southern hemisphere (pp.339–352). Rotterdam: Balkema.
Deacon, J.(1978). Changing patterns in late Pleistocene/early Holocene prehistory in southern Africa
as seen from the Nelson Bay Cave stone artifact sequence. Quaternary Research, 10, 84–111.
Dietl, H., Kandel, A.W., & Conard, N.J. (2005). Middle Stone Age settlement and land use at the
open-air sites of Geelbek and Anyskop, South Africa. Journal of African Archaeology, 3(2),
Faith, J.T. (2013). Taphonomic and paleoecological change in the large mammal sequence from
Boomplaas Cave, Western Cape, South Africa. Journal of Human Evolution, 65(6), 715–730.
Goldberg, P., Miller, C.E., Schiegl, S., Ligouis, B., Berna, F., Conard, N.J., & Wadley, L. (2009).
Bedding, hearths, and site maintenance in the Middle Stone Age of Sibudu Cave, KwaZulu-
Natal, South Africa. Archaeological and Anthropological Sciences, 1(2), 95–122.
Guerin, G., Murray, A.S., Jain, M., Thomsen, K.J., & Mercier, N. (2013). How condent are we
in the chronology of the transition between Howieson’s Poort and Still Bay? Journal of Human
Evolution, 64(4), 314–317.
A. Mackay et al.
Hallinan, E. (2013). Stone age landscape use in the Olifants River Valley, Western Cape. MPhil,
University of Cape Town.
Hallinan, E., & Shaw, M. (2015). A new Middle Stone Age industry in the Tankwa Karoo, Northern
Cape Province, South Africa. Antiquity Project Gallery, 89, 344.
Henshilwood, C.S., Sealy, J.C., Yates, R., Cruz-Uribe, K., Goldberg, P., Grine, F.E., Klein, R.G.,
Poggenpoel, C., van Niekerk, K., & Watts, I. (2001). Blombos Cave, Southern Cape, South
Africa: Preliminary report on the 1992–1999 excavations of the Middle Stone Age levels.
Journal of Archaeological Science, 28(4), 421–448.
Högberg, A., & Larsson, L. (2011). Lithic technology and behavioural modernity: New results
from the Still Bay site, Hollow Rock Shelter, Western Cape Province, South Africa. Journal of
Human Evolution, 61(2), 133–155.
Igreja, M., & Porraz, G. (2013). Functional insights into the innovative early Howiesons Poort
technology at Diepkloof Rock Shelter (Western Cape, South Africa). Journal of Archaeological
Science, 40(9), 3475–3491.
Jacobs, Z., Hayes, E. H., Roberts, R. G., Galbraith, R. F., & Henshilwood, C.S. (2013). An
improved OSL chronology for the Still Bay layers at Blombos Cave, South Africa: Further tests
of single-grain dating procedures and a re-evaluation of the timing of the Still Bay industry
across southern Africa. Journal of Archaeological Science, 40(1), 579–594.
Jacobs, Z., Roberts, R.G., Galbraith, R.F., Deacon, H.J., Grun, R., Mackay, A., Mitchell, P.,
Vogelsang, R., & Wadley, L. (2008a). Ages for the Middle Stone Age of southern Africa:
Implications for human behavior and dispersal. Science, 322(5902), 733–735.
Jacobs, Z., Wintle, A.G., Duller, G.A. T., Roberts, R.G., & Wadley, L. (2008b). New ages for
the post-Howiesons Poort, late and nal Middle Stone Age at Sibudu, South Africa. Journal of
Archaeological Science, 35(7), 1790–1807.
Jouzel, J., Masson-Delmotte, V., Cattani, O., Dreyfus, G., Falourd, S., Hoffmann, G., Minster, B.,
Nouet, J., Barnola, J.M., Chappellaz, J., Fischer, H., Gallet, J.C., Johnsen, S., Leuenberger,
M., Loulergue, L., Luethi, D., Oerter, H., Parrenin, F., Raisbeck, G., Raynaud, D., Schilt, A.,
Schwander, J., Selmo, E., Souchez, R., Spahni, R., Stauffer, B., Steffensen, J.P., Stenni, B.,
Stocker, T.F., Tison, J.L., Werner, M., & Wolff, E.W. (2007). Orbital and millennial Antarctic
climate variability over the past 800,000 years. Science, 317(5839), 793–796.
Kandel, A.W., & Conard, N.J. (2012). Settlement patterns during the earlier and Middle Stone
Age around Langebaan Lagoon, Western Cape (South Africa). Quaternary International, 270,
Kaplan, J.(1990). The Umhlatuzana rock shelter sequence: 100 000 years of stone age history.
Natal Museum Journal of Humanities, 2, 1–94.
Karkanas, P., Brown, K.S., Fisher, E.C., Jacobs, Z., & Marean, C.W. (2015). Interpreting human
behavior from depositional rates and combustion features through the study of sedimentary
microfacies at site Pinnacle Point 5-6, South Africa. Journal of Human Evolution, 85, 1–21.
Kelly, R. L. (1983). Hunter-gatherer mobility strategies. Journal of Archaeological Research,
39(3), 277–306.
Kelly, R.L. (1995). The foraging spectrum: Diversity in hunter-gatherer Lifeways. Washington
D.C.: Smithsonian Institution Press.
Klein, R. G. (1976). The mammalian fauna of the Klasies River mouth sites, southern Cape
Province, South Africa. South African Archaeological Bulletin, 31, 75–99.
Klein, R.G., Cruz-Uribe, K., Halkett, D., Hart, T., & Parkington, J.E. (1999). Paleoenvironmental
and human behavioral implications of the Boegoeberg 1 late Pleistocene hyena den, Northern
Cape Province, South Africa. Quaternary Research, 52, 393–403.
Kuhn, S.L. (1994). A formal approach to the design and assembly of mobile toolkits. American
Antiquity, 59(3), 426–442.
Kuhn, S.L. (1995). Mousterian lithic technology. Princeton: Princeton University Press.
Mackay, A. (2005). Informal movements: Changing mobility patterns at Ngarrabullgan, Cape York
Australia. In C.Clarkson & L.Lamb (Eds.), Lithics down under: Recent Australian approaches
to lithic reduction, use and classication (pp.95–108). Oxford: Archaeopress.
2 Provisioning Responses toEnvironmental Change inSouth Africa’s
Mackay, A. (2009). History and selection in the Late Pleistocene archaeology of the Western Cape,
South Africa. PhD, Australian National University.
Mackay, A. (2010). The late Pleistocene archaeology of Klein Kliphuis rockshelter, Western Cape,
South Africa: 2006 excavations. South African Archaeological Bulletin, 65(192), 132–147.
Mackay, A. (2011). Nature and signicance of the Howiesons Poort to post-Howiesons Poort
transition at Klein Kliphuis rockshelter, South Africa. Journal of Archaeological Science,
38(7), 1430–1440.
Mackay, A. (2016). Technological change and the importance of variability: The Western Cape of
South Africa from MIS 5 to MIS 2. In S.Jones & B.A. Stewart (Eds.), Africa from MIS 6–2:
Population dynamics and Paleoenvironments (pp.49–63). Dordrecht: Springer.
Mackay, A., Jacobs, Z., & Steele, T. E. (2015). Pleistocene archaeology and chronology of
Putslaagte 8 (PL8) rockshelter, Western Cape, South Africa. Journal of African Archaeology,
13(1), 71–98.
Mackay, A., Orton, J., Schwortz, S., & Steele, T.E. (2010). Soutfontein (SFT)-001: Preliminary
report on an open-air bifacial point-rich site in southern Namaqualand, South Africa. South
African Archaeological Bulletin, 65, 84–95.
Mackay, A., Stewart, B.A., & Chase, B.M. (2014a). Coalescence and fragmentation in the late
Pleistocene archaeology of southernmost Africa. Journal of Human Evolution, 72, 26–51.
Mackay, A., Sumner, A., Jacobs, Z., Marwick, B., Bluff, K., & Shaw, M. (2014b). Putslaagte 1
(PL1), the Doring River, and the later Middle Stone Age in southern Africa’s Winter Rainfall
Zone. Quaternary International, 350, 43–58.
McCall, G.S. (2007). Behavioral ecological models of lithic technological change during the later
Middle Stone Age of South Africa. Journal of Archaeological Science, 34(10), 1738–1751.
McCall, G.S., & Thomas, J.T. (2012). Still Bay and Howiesons Poort foraging strategies: Recent
research and models of culture change. African Archaeological Review, 29(1), 7–50.
Meadows, M.E., & Sugden, J.M. (1993). The late quaternary palaeoecology of a oristic kingdom:
The southwestern Cape South Africa. Palaeogeography, Palaeoclimatology, Palaeoecology,
101, 271–281.
Miller, C.E., Goldberg, P., & Berna, F. (2013). Geoarchaeological investigations at Diepkloof
Rock Shelter, Western Cape, South Africa. Journal of Archaeological Science, 40(9),
Mills, S.C., Grab, S.W., Rea, B.R., Carr, S. J., & Farrow, A. (2012). Shifting westerlies and
precipitation patterns during the late Pleistocene in southern Africa determined using glacier
reconstruction and mass balance modelling. Quaternary Science Reviews, 55, 145–159.
Mitchell, P.J. (1988). Early microlithic assemblages of southern Africa. World Archaeology, 20(1),
Mitchell, P. J. (2008). Developing the archaeology of marine isotope stage 3. South African
Archaeological Society Goodwin Series, 10, 52–65.
Mourre, V., Villa, P., & Henshilwood, C.S. (2010). Early use of pressure aking on lithic artifacts
at Blombos Cave, South Africa. Science, 330(6004), 659–662.
Mucina, L., & Rutherford, M.C. (2006). The vegetation of South Africa, Lesotho and Swaziland.
Pretoria: South African National Biodiversity Institute.
Nelson, M. (1991). The study of technological organization. Archaeological Method and Theory,
3, 57–100.
Nicholson, S.E., & Flohn, H. (1980). African environmental and climatic changes and the general
atmospheric circulation in the late Pleistocene and Holocene. Climatic Change, 2, 313–348.
Orton, J.(2006). The Later Stone Age lithic sequence at Elands Bay, Western Cape, South Africa:
Raw materials, artefacts and sporadic change. Southern African Humanities, 18(2), 1–28.
Piennar, M., Woodborne, S., & Wadley, L. (2008). Optically stimulated luminescence dating at
rose cottage cave. South African Journal of Science, 104, 65–70.
Porraz, G., Parkington, J. E., Rigaud, J.-P., Miller, C. E., Poggenpoel, C., Tribolo, C., Archer,
W., Cartwright, C. R., Charrié-Duhaut, A., Dayet, L., Igreja, M., Mercier, N., Schmidt, P.,
Verna, C., & Texier, P.-J. (2013a). The MSA sequence of Diepkloof and the history of southern
African Late Pleistocene populations. Journal of Archaeological Science, 40(9), 3542–3552.
A. Mackay et al.
Porraz, G., Texier, P.-J., Archer, W., Piboule, M., Rigaud, J.-P., & Tribolo, C. (2013b). Technological
successions in the Middle Stone Age sequence of Diepkloof Rock Shelter, Western Cape, South
Africa. Journal of Archaeological Science, 40(9), 3376–3400.
Porraz, G., Val, A., Dayet, L., De la Pena, P., Douze, K., Miller, C., Murungi, M.L., Tribolo, C.,
Schmid, V., & Sievers, C. (2015). Bushman Rock Shelter (Limpopo, South Africa): A perspec-
tive from the edge of the highveld. South African Archaeological Bulletin, 70(202), 166–179.
Powell, A., Shennan, S., & Thomas, M.G. (2009). Late Pleistocene demography and the appearance
of modern human behavior. Science, 324(5932), 1298–1301.
Read, D.W. (2008). An interaction model for resource implement complexity based on risk and
number of annual moves. American Antiquity, 73, 599–625.
Riel-Salvatore, J., & Barton, C.M. (2004). Late Pleistocene technology, economic behavior, and
land-use dynamics in Southern Italy. American Antiquity, 69(2), 257–274.
Roberts, D. L. (2003). Age, genesis and signicance of South African coastal belt silcretes.
Pretoria: Council for Geoscience.
Schmidt, P., Porraz, G., Slodczyk, A., Bellot-Gurlet, L., Archer, W., & Miller, C.E. (2012). Heat
treatment in the South African Middle Stone Age: Temperature induced transformations of
silcrete and their technological implications. Journal of Archaeological Science, 40(9),
Shi, N., Dupont, L.M., Beug, H.-J., & Schneider, R. (2000). Correlation between vegetation in
Southwestern Africa and oceanic upwelling in the past 21,000 years. Quaternary Research,
54(1), 72–80.
Shott, M.J. (1986). Technological organisation and settlement mobility. Journal of Anthropological
Research, 42(1), 15–51.
Singer, R., & Wymer, J.(1982). The Middle Stone Age at Klasies River Mouth in South Africa.
Chicago: The University of Chicago Press.
Soriano, S., Villa, P., Delagnes, A., Degano, I., Pollarolo, L., Lucejko, J.J., Henshilwood, C., &
Wadley, L. (2015). The Still Bay and Howiesons Poort at Sibudu and Blombos: Understanding
Middle Stone Age technologies. PloS One, 10(7), e0131127.
Soriano, S., Villa, P., & Wadley, L. (2007). Blade technology and tool forms in the Middle Stone
Age of South Africa: The Howiesons Poort and post-Howiesons Poort at Rose Cottage Cave.
Journal of Archaeological Science, 34(5), 681–703.
Soriano, S., Villa, P., & Wadley, L. (2009). Ochre for the toolmaker: Shaping the Still Bay points at
Sibudu (KwaZulu-Natal, South Africa). Journal of African Archaeology, 7(1), 45–54.
Stager, J.C., Mayewski, P.A., White, J., Chase, B.M., Neumann, F.H., Meadows, M.E., King,
C.D., & Dixon, D.A. (2012). Precipitation variability in the winter rainfall zone of South Africa
during the last 1400 yr linked to the austral westerlies. Climate of the Past, 8(3), 877–887.
Steele, T.E., Mackay, A., Fitzsimmons, K., Igreja, M., Marwick, B., Orton, J., Schwortz, S., &
Stahlschmidt, M. (2016). Varsche Rivier 003: A Middle Stone Age site with Still Bay and
Howiesons Poort assemblages in southern Namaqualand, Western Cape, South Africa.
PaleoAnthropology, 2016, 100–163.
Steele, T.E., Mackay, A., Orton, J., & Schwortz, S. (2012). Varsche Rivier 003, a new Middle
Stone Age site in southern Namaqualand, South Africa. South African Archaeological Bulletin,
67(195), 108–119.
Stuut, J.-B., Prins, M.A., Schnieder, R.R., Weltje, G.J., Jansen, J.H. F., & Postma, G. (2002).
A 300-kyr record of aridity and wind strength in southwestern Africa: Inferences from grain-size
distributions of sediments on Walvis Ridge, SE Atlantic. Marine Geology, 180, 221–233.
Toggweiler, J.R., & Russell, J.(2008). Ocean circulation in a warming climate. Nature, 451,
Tribolo, C., Mercier, N., Douville, E., Joron, J.L., Reyss, J.L., Rufer, D., Cantin, N., Lefrais, Y.,
Miller, C.E., Porraz, G., Parkington, J., Rigaud, J.P., & Texier, P.J. (2012). OSL and TL dating
of the Middle Stone Age sequence at Diepkloof Rock Shelter (South Africa): A clarication.
Journal of Archaeological Science, 30, 3401–3411.
Tribolo, C., Mercier, N., Valladas, H., Joron, J.L., Guibert, P., Lefrais, Y., Selo, M., Texier, P.J.,
Rigaud, J.P., Porraz, G., Poggenpoel, C., Parkington, J., Texier, J.P., & Lenoble, A. (2009).
2 Provisioning Responses toEnvironmental Change inSouth Africa’s
Thermoluminescence dating of a Stillbay–Howiesons Poort sequence at Diepkloof Rock
Shelter (Western Cape, South Africa). Journal of Archaeological Science, 36(3), 730–739.
Truc, L., Chevalier, M., Favier, C., Cheddadi, R., Meadows, M.E., Scott, L., Carr, A.S., Smith,
G.F., & Chase, B.M. (2013). Quantication of climate change for the last 20,000 years from
Wonderkrater, South Africa: Implications for the long-term dynamics of the Intertropical
Convergence Zone. Palaeogeography, Palaeoclimatology, Palaeoecology, 386, 575–587.
Valsecchi, V., Chase, B.M., Slingsby, J.A., Carr, A.S., Quick, L.J., Meadows, M.E., Cheddadi,
R., & Reimer, P.J. (2013). A high resolution 15,600-year pollen and microcharcoal record from
the Cederberg Mountains, South Africa. Palaeogeography, Palaeoclimatology, Palaeoecology,
387, 6–16.
Villa, P., Soressi, M., Henshilwood, C.S., & Mourre, V. (2009). The Still Bay points of Blombos
Cave (South Africa). Journal of Archaeological Science, 36(2), 441–460.
Villa, P., Soriano, S., Teyssandier, N., & Wurz, S. (2010). The Howiesons Poort and MSA III at
Klasies River main site, Cave 1A. Journal of Archaeological Science, 37(3), 630–655.
Visser, H.N., & Theron, J.N. (1973). Clanwilliam. 3218, Geological series. Pretoria: Government
Vogelsang, R., Richter, J., Jacobs, Z., Eichhorn, B., Linseele, V., & Roberts, R.G. (2010). New
excavations of Middle Stone Age deposits at Apollo 11 Rockshelter, Namibia: Stratigraphy,
archaeology, chronology and past environments. Journal of African Archaeology, 8(2),
Volman, T.P. (1980). The Middle Stone Age in the Southern Cape. PhD, University of Chicago.
Wadley, L. (2007). Announcing a Still Bay industry at Sibudu Cave, South Africa. Journal of
Human Evolution, 52(6), 681–689.
Wadley, L. (2012). Two ‘moments in time’ during Middle Stone Age occupations of Sibudu, South
Africa. Southern African Humanities, 24, 79–97.
Will, M., Mackay, A., & Phillips, N. (2015). Implications of Nubian-like Core reduction systems
in Southern Africa for the identication of early modern human dispersals. PloS One, 10(6),
Wurz, S. (1997). The Howiesons Poort at Klasies River: from artefacts to cognition. MA,
University of Stellenbosch.
Wurz, S. (2002). Variability in the Middle Stone Age lithic sequence, 115,000–60,000 years ago at
Klasies River, South Africa. Journal of Archaeological Science, 29(9), 1001–1015.
Wurz, S. (2013). Technological trends in the Middle Stone Age of South Africa between MIS 7 and
MIS 3. Current Anthropology, 54(S8), S305–S319.
Ziegler, M., Simon, M.H., Hall, I.R., Barker, S., Stringer, C., & Zahn, R. (2013). Development of
Middle Stone Age innovation linked to rapid climate change. Nature Communications, 4, 1905.
A. Mackay et al.
... The Tankwa Karoo study area can be divided into two zones: Cape (characterised by Cape Supergroup geology and Fynbos vegetation) and Karoo (Karoo Supergroup geology and Succulent Karoo vegetation). Since each zone shares physical and ecological similarities with the archaeologically well-studied Western Cape (Mackay 2016;Hallinan and Parkington 2017;Mackay et al. 2018) and Upper Karoo (Sampson 1968(Sampson , 1985Sampson et al. 2015) (Figure 1; Tables 1 and 2), it is expected that aspects of ESA and MSA behaviour will resemble the patterns observed for these regions in Cape and Karoo contexts respectively. These expectations are set out in Table 3 with behavioural differences assessed through lithic artefacts in terms of landscape distribution, raw material preferences and technological organisation. ...
... Cape expectations based on Western Cape observations by Hallinan and Parkington (2017) and Mackay et al. (2018) Karoo expectations based on Upper Karoo observations by Sampson (1968Sampson ( , 1985 ESA The present-day vegetation of the area is composed of plant communities of the Fynbos and Succulent Karoo Biomes, broadly coinciding with Cape and Karoo bedrock geology respectively. Both biomes are renowned for their high levels of biodiversity and endemism, with common biotic features including abundant geophytes and a lack of trees (Procheş et al. 2006;Rebelo et al. 2006). ...
... A particular regional feature is the occurrence of large-scale open-air bifacial point production sites (Table 2). These are located on river terraces, associated with raw material outcrops: Clanwilliam Dam East on the Olifants River (silcrete), Kleinhoek 1 on the Doring River (fine-grained quartzite) and Soutfontein on the Sout River (quartz) (Mackay et al. 2018). More widely, no bifacial points or Still Bay deposits are reported inland of the Cape Fold Belt or sandstones of the Natal Group in eastern South Africa, although Soutfontein, Varsche Rivier and Apollo 11 (in southern Namibia) are exceptions. ...
Full-text available
Southern Africa is an ecologically highly varied region, yet many generalisations about past human behaviour are drawn from rock shelter sites in coastal and montane Fynbos Biome environments. The Tankwa Karoo region offers the opportunity to extend our archaeological knowledge from the well-researched Western Cape into the arid interior Karoo in order to better capture behavioural variability and identify specific adaptations to more marginal conditions. This research presents the results of off-site surveys in the Tankwa Karoo, which spans the Cape to Karoo transition, mapping surface stone artefacts from the Earlier and Middle Stone Ages. The observed patterns in landscape use and lithic technology for each time-period were tested against a set of expectations based on previous research in the Western Cape and the Upper Karoo. The results indicate that in the Earlier Stone Age the most arid parts of the Tankwa Karoo saw only ephemeral use, with the better-watered mountain fringes preferred. In contrast, various strategies in the Middle Stone Age allowed groups to occupy these marginal parts of the landscape, including new kinds of technological behaviour suggestive of specific adaptations to this environment.
... These differences can also be interpreted in the context of provisioning systems (e.g. Kuhn, 1995;Mackay et al., 2018). Two distinct provisioning systems occur, provisioning of individuals and provisioning of place (Kuhn, 1995). ...
... Individual provisioning systems prepare mobile individuals for unreliable resource availability. Such toolkits for individual provisioning mostly include cores, from which tools can be manufactured as needed, and formal multi-purpose tools, which can be resharpened after use (Kuhn, 1995;Mackay et al., 2018). ...
... On-site manufacture is visible by the presence of multiple cores and debitage recording different stages of the reduction sequence. Place provisioning is also characterized by prolonged or re-occurring occupations of a site and therefore, predictable resources close by (Kuhn, 1995;Mackay et al., 2018). In the KRM assemblages discussed here, local raw materials are used and reduced on-site, as indicated by the presence of cobbles, cortical flakes, cores and small debitage. ...
Klasies River contains an extensive MIS 5 MSA sequence, mostly from what has been described as the MSA ll, MSA 2a or the Mossel Bay techno-complex. The current Witness Baulk excavations undertaken between 2015 and 2016 allows for a renewed and detailed investigation of the variability of the lithic technology from the lowermost part of the MSA ll. This assemblage is equivalent to Deacon's SASL sub-member and Layers 17a and b of Singer and Wymer. Two phases are recognized: An MIS 5c phase in layer SMONE (Singer & Wymer layer 17a) and an MIS 5d phase in layers BOS One and BOS Two (Singer & Wymer layer 17 b). The two phases are characterized by commonalities such as a focus on quartzite utilisation, the presence of a main unidirectional reduction system, similar end products which comprise of points, blades and flakes and a low amount of formal tools. However, this high-resolution investigation of the layers reveals temporal variability. In the Shell Midden ONE (SMONE) layer cores and products are relatively lighter and small debitage is more frequent. There are more flake end products and compared to the lower layer, there are fewer tool types. In the Black Occupational Soil (BOS) layer points and blades are more numerous, products are heavier, core types are more variable, and a higher frequency and variety of tool types occur. Such detailed differences within the MIS 5 assemblages from Klasies River, not described before, shows that the MSA ll is not a homogeneous entity, and that subtle patterning occurs that may link to different technological strategies. Compared to other MIS 5 sites on the southern Cape, namely Blombos Cave, Pinnacle Point and Cape St. Blaize, a common pattern in place provisioning is seen, although the technology shows differences between the sites. This study indicates the value of more detailed studies of MIS 5 assemblages as a tool to understand variability from a more refined perspective.
... Nevertheless, the principle behind the approach is supported in research elsewhere with larger samples and greater chronological control. Mackay et al. (2018) identify varied provisioning strategies employed in the MSA based on changes in technological organization and artifact density as seen in dated cave assemblages and temporally equivalent surface scatters. A provisioning model was also tested in the Tankwa Karoo (Hallinan, 2018(Hallinan, , 2021, with the case study presented below demonstrating how varied sampling strategies-both in standardized sample squares and as a continuous and truly off-site survey-can be fruitfully combined. ...
... Bower also notes that demographic issues were not prominent research goals in the 1980s. Demography is a major theme in research today and is often interlinked with landscape-scale studies that seek to tackle broad-scale trends in settlement and behavioral change (e.g., Mackay et al., 2018;Stewart & Mitchell, 2018). Bower's final point of critique is that CRM surveys and landscape research were largely separate strands that rarely combined results. ...
Full-text available
Surveys are an important reconnaissance tool in African archaeology, but surface-oriented research is still relatively limited. Thirty years on from John Bower’s “survey of surveys,” this article revisits the state of surface archaeology in Southern Africa and reviews its role in landscape archaeology more broadly. Drawing on examples of Middle Stone Age research in particular, the article considers how archaeologists have addressed Bower’s methodological concerns of site definition, data collection, survey logistics, and interpretation. Recent research in the Tankwa Karoo region of South Africa is presented as a case study to demonstrate the value that surface research holds for understanding past behavioral variability at a landscape scale.
... This is particularly problematic in the Still Bay for which, until recently, relatively few sites had been studied in detail. Those few studies that have incorporated multiple Still Bay sites likely to be within a single land-use system provide results that contrast with the expected high mobility of Still Bay foragers (Hallinan 2013;Hallinan and Parkington 2017;Mackay et al. 2018Mackay et al. , 2010Shaw et al. 2019). ...
... In each of these catchments, moreover, there is at least one open-air site containing a large number of bifacial Fig. 1 Location of prominent Still Bay sites against proportion of winter rain (data from WorldCLIM 2.0 (Fick and Hijmans 2017)). Location of study area shown by white box points, between 38-251 points per site and dominated by a single raw material, as well as numerous smaller openair sites with 10 or fewer bifacial points (Hallinan and Parkington 2017;Mackay et al. 2018). Consistent with the expectations of McCall and Thomas (2012), the pattern implies structured variation, comprising a primary gearing-up location and a mix of multiple small occupational nodes and isolated discard events throughout the catchment. ...
Full-text available
South Africa’s Still Bay technocomplex (77–70 ka) is an early example of a technological system organised around the production of bifacial points. Noting the diversity of raw materials used and the frequency of non-local raw materials found among excavated bifacial point assemblages, numerous researchers have argued that Still Bay foragers were highly mobile. This pattern, however, is in contrast to that observed in some open-air surface Still Bay assemblages, where raw material diversity among bifacial points is low and local rocks dominate. We resolve this apparent discrepancy by combining information on raw material distribution, least-cost path analysis, and artefact data from two rock shelters and numerous open-air sites located along the Doring and Olifants Rivers in South Africa. The results demonstrate that raw material selection for bifacial point production was responsive to geological resources within river catchments but that bifacial points were transported regularly between catchments over minimum distances of 30–60 km. Our data appears to support the inference that Still Bay foragers were wide-ranging.
... Numerous cave and rock shelter sites in the coastal-montane belt have provided key evidence for complex and innovative behaviour in a succession of distinctive technocomplexes, particularly during late Marine Isotope Stage (MIS) 5 and MIS 4 [1][2][3]. Specifically, the Still Bay and Howiesons Poort have received special attention due to the early evidence of art and symbolism alongside high levels of technological investment in producing characteristic artefacts, bifacial points (Still Bay) and backed artefacts (Howiesons Poort) [4,5]. The Fynbos vegetation biome, where these tend to occur, provides a dense and predictable food supply in its juxtaposition of terrestrial and marine resources, the latter often linked with increasing technological, social and cognitive developments in the MSA [6][7][8]. ...
... Until recently, interest in the MIS 4 Still Bay and Howiesons Poort technocomplexes of the South African MSA has eclipsed the study of the subsequent post-Howiesons Poort and final stages of the MSA in MIS 3 [11,98,[153][154][155][156]. Initial suggestions that human behaviour experienced a devolution, regression or behavioural reversal following the innovative bursts seen in MIS 4 are no longer upheld [157][158][159] and the period is now generally viewed as reflecting shifts in technological organisation and adaptive strategies [4,115,160,161]. The climate of MIS 3 was not uniformly characterised by hyper-aridity as is sometimes stated [162][163][164], instead seeing rapid fluctuations and considerable variability in South Africa's different biomes [11]. ...
Full-text available
The Middle Stone Age record in southern Africa is recognising increasing diversity in lithic technologies as research expands beyond the coastal-montane zone. New research in the arid Tankwa Karoo region of the South African interior has revealed a rich surface artefact record including a novel method of point production, recognised as Nubian Levallois technology in Late Pleistocene North Africa, Arabia and the Levant. We analyse 121 Nubian cores and associated points from the surface site Tweefontein against the strict criteria which are used to define Nubian technology elsewhere. The co-occurrence of typically post-Howiesons Poort unifacial points suggests an MIS 3 age. We propose that the occurrence of this distinctive technology at numerous localities in the Tankwa Karoo region reflects an environment-specific adaptation in line with technological regionalisation seen more widely in MIS 3. The arid setting of these assemblages in the Tankwa Karoo compares with the desert context of Nubian technology globally, consistent with convergent evolution in our case. The South African evidence contributes an alternative perspective on Nubian technology removed from the ‘dispersal’ or ‘diffusion’ scenarios of the debate surrounding its origin and spread within and out of Africa.
... After that, we discuss how the assemblages from KRM analyzed here compare to the two relevant phases at PP13B and the M3 phase at BBC in relation to shellfish exploitation and lithic technology. The latter provides information on provisioning systems, which Mackay, Stewart, and Chase (2014) and Mackay, Hallinan, and Steele (2018) propose to be an essential piece of the puzzle to characterize MSA occupations. Kuhn (1995) defines provisioning of place and provisioning of individuals as the two extremes on a continuous range. ...
... Kuhn (1995) defines provisioning of place and provisioning of individuals as the two extremes on a continuous range. Place provisioning infers long-term or repeated occupation of a site with predictable resource availability (Kuhn 1995;Mackay, Hallinan, and Steele 2018). On-site manufacture of implements with cores that are made from local raw material or have been transported to the site characterize place provisioning. ...
Coastal adaptation in the southern Cape can be seen around 100,000 years ago in sites such as Klasies River, Blombos Cave, and Pinnacle Point, representing the occupation of a new niche by early Homo sapiens in this region. However, there is relatively little information available on the details involved in fully entering this niche from a regional perspective. At Klasies River main site (KRM), evidence for coastal adaptation occurs in early MIS 5. Here we explore the variability in shellfish exploitation and how it links to lithic technology, in deposits dating to ca. 93,000–110,000 years ago. We compare this to broadly contemporaneous assemblages from Pinnacle Point 13B and Blombos Cave. The lithics in all the layers from KRM investigated here have been produced according to a unidirectional reduction system, but the lowermost assemblages contain more small debitage and bladelets, and no tools. These 110 ka layers are associated with a lower shellfish density and more diverse range of shellfish species and a higher lithic density. This points to a lesser dependency on shellfish coinciding with higher mobility in the lower layers. For the younger MIS 5c layers higher volumes of shellfish and the dominance of certain species is evident. The lithics show that all the stages of the reduction system are present and tools are produced and used on-site. This indicates a residential (provisioning of place) occupational strategy. Compared to other sites on the southern Cape coast, KRM shows exceptionally high densities in lithic artifacts while the shellfish densities are comparable to the Blombos M3 phase. The results of the analysis of the shellfish and lithic densities, technological patterns, and shellfish species exploited at Klasies River, Blombos Cave and Pinnacle Point, demonstrate a more diverse onset and expression of coastal adaptation during early MIS 5 than apparent from current literature.
... The Howiesons Poort assemblage at the site is identified by the presence of backed artefacts, high proportions of silcrete, and small flakes and cores (Mackay, 2011a). Like many sites, Klein Kliphuis witnesses very high artefact discard rates in much of the Howiesons Poort, something taken to indicate intensive site occupation (Mackay et al., 2018a). There are two departures from this pattern. ...
Full-text available
We explore the correspondence between changing palaeoenvironments, patterns of site use, and lithic technology at the rock shelter site Klein Kliphuis (South Africa) across the interval 65–55 000 years before present. This period coincides with the termination of Marine Isotope Stage (MIS) 4, and the disappearance of an iconic late Pleistocene archaeological unit known as the Howiesons Poort. Wood charcoals indicate sufficient soil moisture around Klein Kliphuis throughout the Howiesons Poort to support diverse tree species at a time when site occupation was relatively intense. At least some fuelwood‐gathering in this period may have been undertaken to support heat treatment of silcrete, which was the dominant lithology in tool production. A coherent set of changes occurs across the MIS 4/3 transition: occupational intensity declines, tracked by declining diversity of fuelwood species, an increase in the proportion of charcoals from shrubs and small flowering plants, lower prevalence of silcrete, and less heat treatment. While declining soil moisture is implied, there appears to be a significant change in behaviour relating to site usage, whereby foraging for dense fuelwoods was replaced by the construction of ‘fast fires’ that may reflect briefer visits and increased forager mobility in early MIS 3.
... More frequent residential moves during MIS 6 in the western CLP may have resulted in Yangshang shifting from a seasonally occupied residential base in GL8 (MIS 7) to a shorter-term site in GL7 and GL6 (MIS 6). In such systems, it may have been more practical to provision individuals (Kuhn, 1995) with technologies emphasizing the transport, use, and conservation of a small number of highly maintainable items rather than accumulation of workable stones and blanks (Shott, 1986;Kuhn, 1994;Mackay et al., 2018). Indeed, when groups move more frequently, especially over long distances caused by enhanced resource patchiness, heightened uncertainties of being caught short of usable stone (Elston, 1990) encourage the curation of mobile toolkits (Binford, 1979). ...
Full-text available
A multidisciplinary fieldwork and research project was recently begun at the Yangshang site (220-140 ka), a late Early Paleolithic locale in the western Chinese Loess Plateau. 1696 lithic artifacts and 337 faunal remains were recovered during the excavation. Sedimentological and paleoenvironmental investigations indicate the site preserves a relatively long and minimally disturbed archaeological sequence associated with paleoenvironmental changes during MIS 7-6. A detailed techno-typological analysis of Yangshang's lithic assemblages was undertaken to examine the influence of glacial cycles on late Middle Pleistocene hominin technological strategies in the western Chinese Loess Plateau. The results show that while the Yangshang site is dominated by quartz-based core/flake assemblages typical of most Early Paleolithic sites in North China, the lithic assemblages provide evidence that different provisioning systems existed during the penultimate glaciation. We argue that these shifts reflect changes in land use and mobility that were tied to climate change. Our results suggest that theoretically informed statistical analyses of so-called 'unchanging and crude' lithic technology can yield meaningful evidence for behavioral shifts.
Full-text available
Much has been written about Middle Stone Age hunting in southern Africa, yet there is no comprehensive overview for the development and use of stone-tipped hunting weapons. With this contribution, I use the tip cross-sectional area (TCSA) method to hypothesise about variation in weapon-assisted hunting strategies for the last 300,000 years or more. I assess and build onto previous hypotheses generated from similar approaches, introducing a larger sample from across the region. By also bolstering the standard TCSA ranges for javelin tips and stabbing/thrusting spear tips with more experimental and ethno-historical material, the method's interpretative robusticity is increased. The results indicate a general trend through time towards smaller weapon tips until reaching arrow-tip range during the MIS 4 glacial. Whereas light-weight javelins, similar to those used by African hunters today may have been in play since almost 200,000 years ago, it remains uncertain whether spearthrower-and-dart technology was ever used in southern Africa. Finally, I align the TCSA outcomes with climatic and demographic reconstructions and explain how human cognition interacts with technological adaptations such as the use of hunting weapons – demonstrating how the interplay between environment, demography, technology and cognition is integral to the development and understanding of Middle Stone Age weapon-assisted hunting strategies.
Full-text available
Open-air sites in arid and semi-arid landscapes are often subject to prolonged periods of exposure and episodes of erosion that can lead to the redistribution of artefacts and the loss of behaviorally significant spatial information. This is true along the Doring River, South Africa, where archaeologically rich sediment stacks with records exceeding 200,000 years are undergoing rapid erosion in response to modern climatic conditions and land use practices. This paper evaluates the impacts of past and future erosion on the disaggregation of artefacts from these open-air sites and the resulting loss of stratigraphic context and behaviorally significant spatial information. We use low elevation aerial images of the Klein Hoek 1 locality captured by an unmanned aerial vehicle (UAV) to develop a high-resolution local digital terrain model (DTM), which we use to model surface flow paths and quantify the potential for future sediment loss using the Revised Universal Soil Loss Equation (RUSLE). We compare the results of these analyses to the distribution of artefacts of different ages to assess artefact dispersion and to guide future research priorities at the locality. We find that some artefact clusters retain significant spatial integrity, whereas others are dispersed and likely out of primary context. The results also indicate that the geomorphic stability of a large part of Klein Hoek 1 has been compromised by erosion, with limited prospects for long-term survival given the present climate and land use practices.
Full-text available
Stone Age surface assemblages are all too often neglected in favour of stratified, datable cave sequences, thus overlooking important insights into changing behavioural patterns at a broader scale. The Olifants River Valley (Clanwilliam, Western Cape Province, South Africa) presents a rich surface lithic record alongside excavated rockshelter occupations from the early Middle Stone Age (MSA) to the Later Stone Age (LSA). Surface surveys in the Olifants River Valley mapped temporally diagnostic artefacts and their association with different topographic features in order to investigate past landscape use. Our approach refers to a hypothesis proposed by Hilary Deacon, framing the MSA within the context of earlier and later patterns of behaviour. Based on observations from sites across South Africa, Deacon described Earlier Stone Age (ESA) landscape use as ‘stenotopic’, with a narrow focus on permanent water sources, and LSA landscape use as ‘eurytopic’, using a much broader range of habitats but specifically occupying rockshelters as domestic sites. Deacon suggested that the intervening MSA, in its later stages, shows a pattern that anticipated LSA landscape use. We apply Deacon's model to the study area, observing distinctive preferences for certain locations and raw materials and approaching changing patterns of artefact discard from a technological perspective.
Full-text available
Different hypotheses identifying factors affecting the complexity of implements used to obtain food resources by hunter-gatherer groups are assessed with regression analysis. A regression model based on interaction between growing season as a proxy measure for risk and number of yearly moves fits data on the complexity of implements for 20 hunter-gatherer groups. The interaction model leads to a division of hunter-gatherer groups into two subgroups that correspond to collector vs. forager strategies for procuring resources. Implications of the interaction model for the evolution of complex implements are discussed.
Full-text available
Southern Africa presents the best-documented Middle and Later Stone Age (MSA and LSA) records in Africa, and yet significant uncertainties still exist concerning the sequence and timing of behavioral and occupational changes in the region. A recent surge in research has provided a suite of new results that indicate more intricate and complex patterns than those previously considered. This paper describes recent excavations at the archaeological site of Varsche Rivier (VR) 003 located in the poorly-researched southern Namaqualand (Knersvlakte) region of South Africa (Western Cape Province). Two seasons of excavations have revealed a long sequence of MSA and LSA cultural materials, including lithics, fauna, ostrich eggshell, marine mollusks, beads, and pigments; bedrock has yet to be reached anywhere in our excavations. Within the shelter, we have uncovered probable Howiesons Poort material, with overlying late MSA and capped by late Holocene LSA. On the slope, the deepest materials are earlier MSA, overlain by assemblages with affinities to the Still Bay and Howiesons Poort. In addition to providing descriptions of the lithic, faunal, and pigment assemblages, we report on the results of micromorphological analysis of the sediments and optically stimulated luminescence (OSL) and radiocarbon dating of the sequence. Based on the analysis of single-grain samples from both parts of the excavation, OSL age estimates suggest that the putative Howiesons Poort assemblages at VR003 were deposited 45.7–41.7 kya. While our results are stratigraphi-cally consistent, they are substantially younger than any previously published Howiesons Poort chronologies. PaleoAnthropology 2016: 100−163.
Full-text available
The austral westerlies strongly influence precipitation and ocean circulation in the southern temperate zone, with important consequences for cultures and ecosystems. Global climate models anticipate poleward contraction of the austral westerlies with future warming, but the available paleoclimate records that might test these models have been largely limited to South America, are not fully consistent with each other, and may be complicated by influences from other climatic factors. Here we present the first fine-interval diatom and sedimentological records from the winter rainfall region of South Africa, representing precipitation during the last 1400 yr. Inferred rainfall increased ~1400–1200 cal yr BP and most notably during the Little Ice Age with pulses centered on ~600, 530, 470, 330, 200, and 90 cal yr BP. Synchronous fluctuations in Antarctic ice core chemistry strongly suggest that these variations are linked to changes in the westerlies. Partial inconsistencies among South African and South American records warn against the simplistic application of local-scale histories to the Southern Hemisphere as a whole. Nonetheless, these findings in general do support model projections of increasing aridity in austral winter rainfall zones with future warming.
The South African Cape provides important evidence of behavioral and technological complexity in the period from MIS 6-2. Understanding the meaning of discontinuous temporal patterns in the distribution of technological systems is hampered by traditional culture historic approaches and culture evolutionary interpretations. These historical effects lead to depictions of the past as a series of stadial, progressive units. Evidence of variability is commonly suppressed and presumptions about what makes a technology advantageous go unquestioned. In this paper, key data used to generate existing stadial systems are considered from four sites in the Western Cape. Data are presented using the maximum available stratigraphic resolution within the constraints of the excavation systems used. Variability is shown to be a recurrent feature of technological systems. Rather than a series of discrete packages of innovation , technological change in this area is better understood in terms of the differential persistence of continually generated variation. The resulting picture is one of technologically flexible groups adapting rapidly and in some cases dramatically to changing circumstances through the Late Pleistocene.