Hearths varied a good deal. Shallow scoops and concentrated patches of
burning are found in many caves in the Middle East (Bar-Yosef 1989), such
as Kebara (Bar-Yosef et al. 1992), and Europe (Rigaud et al. 1995) and
among piles of animal bones at Molodova in the Ukraine (Chernysh 1961;
Goretsky and Ivanova 1982; Ivanova and Tseitlin 1987). However, their
construction becomes more elaborate some time before 50,000 years
ago with the stone-built hearths at Vilas Ruivas, Portugal (Vega Toscano
et al. 1999:23Ã€4). Later such stone-lined and cobble-filled constructions are
common, as at the Abri Pataud in southern France (Movius 1966; Perles
1976). In both examples the accumulation of stones defined hearths that
were repeatedly raked out and re-used.
Hearth building is also matched by alignments and arrangements. The
Abri Pataud hearths are regularly spaced in a line beneath the large rock
overhang. They form a set not only through re-use and re-modelling but also
in their relation to each other. The long stratigraphic sequence at Shelter 1B
at Klasies River Mouth, South Africa (Henderson 1992:23) shows, in levels
dated to between 70,000 and 80,000 years old a clear preference in unit PCP
for lighting, on at least five separate occasions, small fires, 50 cm in dia-
meter, in the same spot. Moreover, each occasion was stratigraphically
distinct with each visit to the rock shelter possibly separated by a few years.
Contrast this with large open air hearths, fed with mammoth bone,
at Kostenki, site I-1 and site IV-2 on the Don river in southern Russia
Did agriculture change the world? 243
(Hoffecker 2002; Klein 1969; Praslov and Rogachev 1982). These were
regularly spaced 3 m apart in the central area of the locale and surrounded
by pits and subterranean dwellings.
However, hearths are not always a focus for sets or spatial patterns as
commented on by Henderson (1992) at Klasies River Mouth, and by
Lyn Wadley (2001) at Rose Cottage Cave, South Africa (see also Kolen
1999). At Rose Cottage the oldest hearths are 31,000 years old in the Middle
Stone Age, layer Dc, and twenty ash-hearths in an area of 26 m2 have been
excavated. The ash-hearths are of variable size but on average only 40 cm
apart. People kept returning to that part of the cave and thirteen of the ash-
hearths are covered by 23,500 lithic pieces of which only 273 are retouched
tools (ibid.:Figure 2). This very high density (903 per m2 or 9,791 per m3)
masks any small-scale spatial patterning either by size or type of artefact.
Wadley sums up the result as:
un-structured camp organization with a clutter of artefacts and food waste
usually in close association with hearths. Some evidence for refuse
dumping is present but, apart from food processing or cooking, no special
purpose activity areas can be recognised.
Precisely the same conclusion was arrived at by Jan Simek (1987) at the
earlier 130,000 year old level VIII of Grotte Vaufrey in southern France
where a concentration of charcoals identified the hearth. Even when
artefacts were refitted little spatial patterning could be found (Binford 1988;
Rigaud and Geneste 1988). In addition, artefact densities were much lower
than at Rose Cottage, a mere 85 per m2, so that patterns were not being
swamped by too many lithics.
This fuzzy picture gains greater clarity once hearths are defined by
stone settings, arranged in rows and regularly cleaned out, as was the case
at Abri Pataud (Movius 1966). It was Binford (1978a) who drew attention
to the creation of circular and horseshoe patterns in the materials that
accumulate around such hearths where people sit to talk and eat.
Moreover, the materials within these rings are sorted by size; small objects
trampled by feet close to the fire, larger items in a ring up to 3 m away
Huts and shelters
There are many containers which are interpreted as huts (Desbrosse and
Kozlowski 1994; Djindjian et al. 1999; Sklenar 1976); intentionally built to
244 Origins and Revolutions
house people. The deliberate behaviour that results in circular and oval
shapes with footings and weights has, however, been questioned by Jan
Kolen (1999). He sees much of the patterning as mimicking the intention to
build, arising instead from living among materials and moving them
around a limited lifespace in a centrifugal manner, as at Molodova site I/4
(Goretsky and Ivanova 1982:Figure 8). Even at locales such as Kostenki
(Figure 8.7) (Efimenko 1958; Praslov and Rogachev 1982) and Pavlov
(Svoboda 1994), where there is general agreement that a range of shelters
have been correctly identified (Roebroeks et al. 2000), most of them are
unique projects as judged by setting, materials, shape and contents. This
has led Alexander Verpoorte (2001:131) to question if they are indeed what
we think they are and to wonder why architecture has such a low visibility in
this technological movement.
What is lacking throughout the common ground is a repeated set of
modular structures to match, for example, the frequently found hearths
spaced 3 m apart. I have puzzled over this lack of modularity (Gamble
1986), as eager as the next archaeologist to identify huts and villages to
support a Palaeolithic social revolution (Gamble 1993a). But if I am honest
these huts from technologyâ€™s common ground are unfamiliar forms,
especially when compared with the serviceable houses and villages from
the last few thousand years of the Arctic, let alone the archaeology of
farmers, and where instantly I can recognise familiar material and meta-
phorical forms (Binford 1993; Chang 1962; Damas 1984; GrÃ¸nnow et al.
1983; Morgan 1881).
Instead each of these huts or locales is a material project put together by
a bricoleur rather than a builder, assembled by repeatedly accumulating
sets of materials at a place rather than having a design (Chapter 6). The sets
these bricoleurs, Neanderthals as well as Modern humans, used in their
material palimpsests included hearths, pits, post-holes, ivory, food, orna-
ments, furs, ochre, stone and human bodies. They dug, cut, skinned,
carved, knapped, whittled, twined, wove and pounded and by these and
many other rhythmic gestures attached themselves to the locale and created
a place to live (Gamble 1999:412Ã€14).
Case study: Kostenki-Avdeevo-Pavlov-Willendorf 28,000 years ago
The open air locales of Kostenki-Avdeevo-Pavlov-Willendorf, located in
Russia, the Czech Republic and Austria are some of the richest Upper
Palaeolithic locales in Russia and Europe (Soffer and Praslov 1993;
Did agriculture change the world? 245
figure 8.7. Kostenki site 1 level 1 excavations by Nikolai Praslov (reproduced with
permission). Many pits were dug at this locale and contained a wide range of materials.
Some of these, such as shells and amber, came from well beyond the local vicinity.
Figurines, some of them intentionally fragmented (inset) were found in pits across the
locale and form a distinctive set. The pits are not entirely convincing as subterranean
houses suggesting the work of bricoleurs rather than engineers (see text).
Svoboda et al. 1996). They have produced multiple sets of instruments and
containers that link them across 1,800 km some 28,000 years ago.
These sets include the much described â€˜Venusâ€™ figurines (Abramova
1967; Delporte 1979; Gamble 1982a); made of limestone, marl, ivory, and at
DolnÄ± Vestonice in the Czech Republic, fired clay. Among the lithic
artefacts they share in common are the Kostenki shouldered knives.
Furthermore, all the locales are linked by their use of mammoth for food,
fuel, raw material for tools and building materials and an enthusiasm for
digging pits that is especially marked in the Russian locales (Efimenko 1958;
Gvozdover 1989). A recent study by Alexander Verpoorte (2001:Figure 3.22)
of the structured settlement at DolnÄ± Vestonice-Pavlov reveals the complex-
ity in the structured deposition of figurines that complements the
pyrotechnic displays (Soffer et al. 1993) and the textile containers (Soffer
et al. 2000) I have already described (Chapter 7). Bound human burials,
e.g. DolnÄ± Vestonice XVI an adult male, were found near hearths inside
a shallow scrape interpreted as the base of a hut (Svoboda 1991:Figure 6)
and e.g. DolnÄ± Vestonice III, a gracile female, under a protective covering
of mammoth scapulae (Klima 1963).
246 Origins and Revolutions
The Kostenki-Avdeevo locales were used many times. The people dug
repeatedly into the active zone above the permanently frozen ground to
make fire hollows, hundreds of shallow scoops, and pits, graves, as well as
larger â€˜pit-housesâ€™ dug to a depth of between 60 cm and 1 m (Grigorâ€™ev 1993;
Rogachev 1957) depending on the depth of the permafrost. Objects were
then planted in them. These included the following sets:
Mammoth tusks and selected bones, often in anatomical order. For
example Pithouse A at Kostenki I/1 contains 12 closely associated tusks,
ten of which have their points towards the centre of the pit (Efimenko
1958:Figure 11). At Avdeevo (Rogachev 1957:Figures 9 and 12) sets of
tusks are also arranged in pits.
Wolves and arctic foxes. Grigoriâ€™ev (1993:60) describes a pit at Avdeevo
with 12 wolves in complete anatomical order.
The deliberate placing of artefacts in pits. Gvozdover (1995:23Ã€5)
describes the three mammoth ivory female figurines placed on pur-
pose with other objects of bone and flint (Gvozdover 1989:Table 4) on
the bottom of one of the Avdeevo pits. They were placed back-to-back,
touching. In two other pits at Kostenki and Avdeevo two figurines were
Human burials. These are usually found in pits in a crouched posture,
as at Kostenki sites 14 and 15 (Praslov and Rogachev 1982:Figures 52
and 55). Sets of jewellery and instruments in ivory and mammoth bone
Eleven complete and eleven fragmented female figurines in either
ivory or chalk/marl are known from Kostenki and Avdeevo (Gvozdover
1989:Table 4). They were deliberately fragmented and then spatially
arranged to make sets. At Kostenki I/1 there is a wide scatter of marl
figurine fragments while the more complete, ivory examples are
clustered in pits at the eastern end of the locale. One of the marl
figures, excavated in 1988, came from a small pit. Both the head and
feet are missing and Nickolai Praslov (1993:166) believes it was broken
elsewhere and then intentionally placed in the pit. Figurine 10, made
of mammoth ivory, from Avdeevo (Gvozdover 1995:25Ã€6) was found
in two pieces; one at the bottom of a pithouse and the other 14 cm away
in the fill. The figurine was broken by a strong vertical blow. It still
lacks its head and both legs and is reminiscent of the marl figurine
Several sets of both containers and instruments exist (Gvozdover
1995:Table II). Among the containers are decorated needle cases
Did agriculture change the world? 247
made from bird bones; necklaces of pierced fox teeth and beads
made from the long bones of small animals; bracelets and stone
pendants (Abramova 1967). Instrument based sets from Avdeevo
include adzes or mattocks made of ivory, engraved wands and spoons,
as well as a wide range of points and awls made of ivory and
bone (Gvozdover 1995). Comparable sets come from the Kostenki
archaeological sites (Abramova 1967; Efimenko 1958; Praslov and
Case study: Blombos Cave 80,000 years ago
Blombos Cave, South Africa, is a showcase for the diversity of material life
at the start of the common ground as hominin bricoleurs constructed
metaphors for living. It is a small cave with well dated deposits. At ages in
excess of 77,000 years ago Christopher Henshilwood and his team have
recovered several sets from 13 m3 of excavated deposits (Figure 8.8).
The consumption of material culture at this locale is impressive.
The density of stone artefacts in two of the levels is even greater than
those at Rose Cottage Cave. Organic preservation is also good. The sets
Stone tools. 99 and 67 Still Bay bifacial points from levels BBC1a and
BBC1b respectively. These were made on silcrete and knapped using a
soft hammer. There was a distinct preference for this fine grained raw
material both at Blombos and other Still Bay locales (Henshilwood
et al. 2001:446).
7,914 pieces of mineral pigments of which 1,448 were 410 mm and 283
bore traces of utilisation (Table 8.3).
Two geometrically engraved ochre pieces from BBC M1.
28 bone tools, awls and points mostly from level BBC2 (Henshilwood
et al. 2001). Fifteen of the points came from 4 m3 and the rest from
8.4 m3 of the excavation.
41 perforated tick shells (Nassarius kraussianus). These carried traces
of red ochre and were found in clusters of 2 to 17 beads. In the opinion
of the archaeologists they are evidence for a necklace (Henshilwood
et al. 2004).
Containers in fragile and perishable materials. Ostrich egg shell was
worked to form water containers. Henshilwood (et al. 2001:446)
suggests that the bone awls were used to pierce leather to make bags
248 Origins and Revolutions
figure 8.8. Sets at Blombos Cave, South Africa. From the small cave a number of
artefact sets have come. The shells (upper inset) formed part of a necklace and are
therefore a container (reproduced with permission from C. Henshilwood). They relate to
nets within the landscape of habit. The Still Bay stone points (lower inset) were widely
distributed instruments and are recognised from several locales across southern Africa by
their similarity (reproduced with permission from T. Minichillo).
These sets also acted as nets across the landscape. The tick shells came
from the coast, then some 20 km away, while the source for the pigments is
some 15Ã€32 kilometres inland (Henshilwood et al. 2001:433), within the
scale of a landscape of habit.
Case study: Salzgitter-Lebenstedt, Germany c. 80,000 years ago
The Neanderthal bricoleurs who came on several occasions to the open
locale of Salzgitter-Lebenstedt had, by comparison to both Blombos
Did agriculture change the world? 249
table 8.3. The ochre sets consisting of utilised and un-utilised pieces compared
with the density of lithic material at Blombos Cave. Four stratigraphic levels,
older than 77,000 years, are shown (Henshilwood et al. 2001)
m3 density of Number of ochre Number of utilised % of utilised
Level stone artefacts pieces 410 mm ochre pieces ochre pieces
246 78 32
BBC2 2,935 85 18 21
BBC3 29,284 1,117 187 17
and the much later Kostenki site I-1, a limited number of sets. These
156 antlers from 82 individual reindeer, dominated by adult males
(Gaudzinski and Roebroeks 2000:502).
Proximal metatarsals from 44 individual reindeer, and smaller sets of
other skeletal elements (Gaudzinski and Roebroeks 2000:Table 2).
Selection of body parts with a higher food value as judged by food
quality and quantity (Gamble and Gaudzinski 2005; Gaudzinski and
5 Neanderthal body parts (two femurs, three skull fragments)
(Gaudzinski and Roebroeks 2000; Hublin 1984).
11 bone tools made of mammoth ribs and fibula (Gaudzinski 1998;
Gaudzinski 1999:Table 1).
Plano-convex bifaces made with Levallois PCT (Bosinski 1967;
Pastoors 2001), following the rule of letting artefacts go at places
which are re-visited (Pope 2002).
There were no hearths, pits, scoops or accumulations of animal remains
that suggested a hut or shelter (Tode 1953; 1982) for the autumn hunting that
took place there (Gaudzinski and Roebroeks 2000).
Does the limited evidence for bricolage by the Neanderthals of
Salzgitter-Lebenstedt point to their only making rational decisions about
what to hunt in order to meet their dietary requirements in as efficient a
manner as possible? The answer is no, and for the following reasons.
From a relational perspective a comparison can be drawn with Lewis
Binfordâ€™s (1978b) classic study of reindeer processing by the Nunamiut
people of northern Alaska. Binford provides a wealth of detail about the
250 Origins and Revolutions
â€˜maze of pathwaysâ€™ (ibid. 1978b:248) that initiates the movement and
distribution of food. He also provides a method to measure the degree of
dismemberment of the carcasses to position a locale within those pathways
by the proportions of marrow yielding bones (Binford 1978b:64). On all
scores the Salzgitter-Lebenstedt reindeer assemblage shows relatively
complete carcasses, as would be expected for a locale that stands at the
start of those chains of relationships (Gamble and Gaudzinski 2005).
However, a major difference emerges between the ethnographic and
the Neanderthal examples. The Nunamiut have a network of highly
differentiated locales in their social landscape. These are widely scattered
over a large territory and revolve around a hub, Anaktuvak Village, situa-
ted on a caribou migration route. Such a well-differentiated settlement
pattern is lacking among the Neanderthals of Central Europe (Conard
and Prindiville 2000:304; Gamble 1999). The significance of the regional
settlement pattern impacts directly upon the creation of place at locales
such as Salzgitter-Lebenstedt where the emphasis was on accumulation.
Enchainment did not extend much beyond the limits of the locale. The
Neanderthal bricoleurs brought together sets of people and animals, stone
and bone tools in a project we call hunting but which is better understood
as â€˜giftingâ€™ these categories to each other. The instruments they used to
create such associations were themselves material proxies for relationships.
Ultimately the Neanderthals understood these connections metaphori-
cally through the experiences of their bodies.
The contrast with Blombos or Kostenki is that the sets these bri-
coleurs accumulated were â€˜fitted-intoâ€™ each other, closely enchaining
childscapes with landscapes of habit and growing identity alongside the
body. The array of containers and instruments and the creation of sets
and nets through fragmentation and consumption give each of the three
locales a distinct feel, irrespective of whether Neanderthals and Modern
humans were involved or their languages revelled in a passion for either
gossip or analogy.
Feeding the body
A similar appreciation of earlier and later hominid place/land-
scape relationships is emerging from a larger sample. The competence
of Neanderthal hunters is now well attested (Burke 2000; Gaudzinski
and Turner 1996; 1999) through the multiple hunts of prime aged
animals recorded at locales such as La Borde (Jaubert et al. 1990) and
Mauran, France (David and Farizy 1994); Wallertheim, Germany
Did agriculture change the world? 251
(Gaudzinski 1992; 1995); Gabasa, Spain (Blasco Sancho 1995); and Ortvale
Klde, Georgia (Adler et al. 2006). Curtis Marean (Marean and Assefa
1999:34) has widened this sample to include Kobeh Cave in the Near East
and the Middle Stone Age levels at GvJm46 at Lukenya Hill, Kenya where
a single species focus is evident. Richard Cosgrove (Pike-Tay and Cosgrove
2002) has similar results from Tasmania where in several caves with deposits
more than 30,000 years old (Cosgrove 1999) specialised hunting of
Bennettâ€™s wallaby produces comparable sets in terms of age, body parts
and seasons to those of the reindeer, horse and bison hunters during the
common ground in Europe (for details see Gamble 1999:Tables 5.12
Katie Boyle (1990:269Ã€70 and Table 2.7), in a comprehensive study
of the use of animals in the French Upper Palaeolithic, has noted a recur-
rent pattern where the high food value anatomical parts, described by
Binford (1978b:Table 2.7) as â€˜gourmetâ€™ assemblages, are invariably asso-
ciated with the dominant animal in the fauna. By contrast, the background
fauna in such collections form â€˜bulkâ€™ assemblages made up of low value
Large mammal hunting does not, as once thought, distinguish between
hominins (Adler et al. 2006). But at the same time attention has been
drawn (Stiner et al. 2000) to increases in diet breadth during technologyâ€™s
common ground, reminiscent of Flanneryâ€™s (1969) broad spectrum
revolution that was applied to the earliest agriculture. Here the emphasis
is on the strategies prey possess to avoid predators. Mary Stiner and her
colleagues (2000) argue that fleet-footed prey such as hares, rabbits and
birds, were increasingly used from 30,000 years ago while even slow-moving
small prey, such as tortoises, show heavier predation as evidenced by their
declining size. This trend is put down to larger hominin populations and an
increase in child survivorship (Stiner et al. 2000:58). They conclude, â€˜we do
not know who in Palaeolithic societies did the inventing, but innovations in
trap, snare and net technology for hunting small prey could have been the
province of women, children, and the elderly (Stiner et al. 2000:58)â€™, since
these would be the individuals who benefited most from these small food
packages. These innovations are of course containers and, if this suggestion
is correct, we see with these material metaphors how the childscape might
have fitted into the landscape of habit. The project of growing the body
creates an understanding of other worlds before they are directly
experienced. As a result identity was changing by being extended through
252 Origins and Revolutions
table 8.4. Changing raw materials and some of the major artefact categories in the
sequence at Klasies River Mouth, South Africa (Singer and Wymer 1982:Table 7.1)