The archaeology of Western Sahara: Results of environmental and archaeological reconnaissance

Abstract

Western Sahara has one of the last remaining unexplored prehistories on the planet. The new research reported here reveals a sequence of Holocene occupation beginning in a humid period around 9000 bp, superceded around 5000 bp by an arid phase in which the land was mainly given over to pastoralism and monumental burial. The authors summarise the flint and pottery assemblage and classify the monuments, looking to neighbouring cultures in Niger, Libya and Sudan.

Full text

The archaeology of Western Sahara:
results of environmental and
archaeological reconnaissance
Nick Brooks1∗, Joanne Clarke2, Salvatore Garfi3& Anne Pirie4
Western Sahara has one of the last remaining unexplored prehistories on the planet. The new
research reported here reveals a sequence of Holocene occupation beginning in a humid period
around 9000 bp, superceded around 5000 bp by an arid phase in which the land was mainly
given over to pastoralism and monumental burial. The authors summarise the flint and pottery
assemblage and classify the monuments, looking to neighbouring cultures in Niger, Libya and
Sudan.
Keywords: Western Sahara, Tifariti, Holocene, Neolithic, Pastoral, climate, environment,
funerary monuments
Introduction
Western Sahara is a disputed, non-self-governing territory bordering Morocco, Algeria and
Mauritania. The majority of the land is currently controlled by Morocco, while the remainder
is administered by the Polisario independence movement. As a result of the territorial dispute
between the two parties, little archaeological research has been carried out in recent years
(Soler et al. 1999; Soler Subils 2004; De Buruaga Bl´
azquez 2006; Soler Subils et al. 2006).
The region is, in fact, so under-studied that it registers as devoid of sites in a recent map
showing the location and chronological distribution of North African later prehistoric sites
(Figure 1) (cf. Jousse 2004: Figure 5).
Since 2002, four seasons of archaeological and environmental reconnaissance survey,
two seasons of excavation, and one post-excavation season have been undertaken by the
University of East Anglia in the northern and southern regions of the Polisario-controlled
areas of Western Sahara (Brooks et al. 2003, 2006). The purpose of these field seasons
has been to assess the potential for detailed archaeological and environmental work in this
important region of the Sahara and to generate some preliminary results, while the long-
term aims of the project are to develop a better understanding of the relationships between
environmental and cultural changes within the broader context of Saharan archaeological
∗Author for correspondence
1Tyndall Centre for Climate Change Research, University of East Anglia, Norwich NR4 7TJ, UK (Email:
nick.brooks@uea.ac.uk)
2School of World Art and Museology, University of East Anglia, Norwich NR4 7TJ, UK
3Royal Commission on the Ancient and Historical Monuments of Wales, Plas Crug, Aberystwyth SY23 1NJ, UK
4School of Human & Environmental Sciences, University of Reading, PO Box 227, Reading RG6 6AB, UK
Received: 10 November 2008; Accepted: 30 January 2009; Revised: 17 April 2009
ANTIQUITY 83 (2009): 918–934
918

Research
The archaeology of Western Sahara
Figure 1. Location and chronological distribution of African Neolithic sites containing cattle remains (courtesy of H. Jousse).
and palaeoenvironmental research. The focus of much of the work to date has been in
the vicinity of the Wadi Tifariti, some 30km north of the east-west oriented border with
Mauritania at c. 26˚17’N, 10˚36’W (see Figures 1 and 2).
Located between the Saharan interior, the Atlas Mountains and the Atlantic coast, and
in a zone of climatic transition between the area under the influence of the African
monsoon and that dominated by the Atlantic westerlies, the region is important for
understanding prehistoric population movements, cultural transmissions and human-
environment interaction. Western Sahara is of particular relevance to the study of human
responses to the humid-arid transition centred around 4400 uncalibrated radiocarbon years
before present (bp) (Gasse 2000; based on IntCal04.14c, Reimer et al. 2004). Initial results
reported below show significant Early Holocene (eighth–ninth millennia bp) occupation of
the area at the beginning of the last Saharan humid period, with a reduction in evidence of
occupation in the drier sixth–seventh millennia bp. After the fifth millennium bp the region
was used extensively as a funerary landscape.
The study area
The study area lies in a region characterised by a series of escarpments, between which
are found ephemeral rivers and playa (wet weather) lakes. In common with other Saharan
regions, Western Sahara is characterised by an arid environment but contains numerous
indicators of past humid conditions. In the vicinity of the study area, dense networks of
drainage channels focus run-off into a number of occasionally active wadis which flow
northwards into the Saguia el-Hamra, a large ephemeral river. Other wadis are filled
919

Nick Brooks et al.
Figure 2. The current extent of the Tifariti survey area.
with accumulated sediment and appear to be permanently dry. Sand and gravel plains,
sandstone hills and elevated plateaux and extensive playa surfaces are major features of the
landscape.
Rainfall today is higher than in other Saharan regions located at similar latitudes, and
significant rainfall is not uncommon, particularly in autumn and winter. While there are
currently no meteorological stations in the vicinity of the study area, mean annual rainfall
was estimated at 30-40mm, rising to over 50mm in the elevated region to the south-west
of the study area, for the period 1926-1950 (Dubief 1953). Precipitation is sufficiently
abundant for the local Sahrawi people to have a concept of drought and to practice mobile
920

Research
The archaeology of Western Sahara
pastoralism, exploiting savannah-like vegetation after periods of significant rainfall. The
environmental evolution of Western Sahara is likely to have been complex and spatially
heterogeneous, due to its extension into the monsoon zone coupled with its proximity to
the Atlantic and the Atlas Mountains.
Our research focused on field survey along a 4km stretch of the Wadi Tifariti, within
a total area of approximately 9.5km2, which was systematically walked during two field
seasons (Figure 2). Project specific proformae were used and site/monument locations were
recorded with GPS handsets. All sites were photographed and most were sketch-planned.
We identified and recorded 411 funerary and other monuments or sites, including ten lithic
scatters, some with pottery. In addition to this, limited excavation was undertaken on two
of the funerary tumuli and at four occupation sites. The sites were found to concentrate on
an east to west ridge on the west side of Wadi Tifariti, and a south-east to north-west ridge
on the east side of the wadi (Figure 2).
This paper discusses only the Holocene material located by the project, but in addition,
there are extensive sheets of fluvial gravels containing Palaeolithic chipped stone on
the margins of major wadis and on the flanks of the surrounding hills suggesting an
extensive river system in the Pleistocene (Brooks et al. 2003: 64). Palaeolithic chipped stone
collected to date includes Levallois products and Aterian points, usually heavily rolled and
worn.
The Early Holocene (c. seventh–ninth millennia bp)
The climate of the Early Holocene is likely to have been considerably more humid than
that of today. The last Saharan humid phase is dated between approximately 10 000 and
4500 bp and associated with a stronger monsoon driven by more intense solar heat at mid-
latitudes (deMenocal et al. 2000). This humid phase was interrupted by a number of arid
episodes, whose timing and extent exhibit significant spatial variation; the most severe and
widespread occurred around 7100 bp (Gasse 2002). Accelerator Mass Spectrometry (AMS)
dating of humic material from a rockshelter at Irghayra in the north of the region yielded an
uncalibrated radiocarbon date of 6210 +
−80 bp, which is consistent with the wider Saharan
picture (Brooks et al. 2003).
There is considerable evidence for occupation of the study area in the Early and early
Middle Holocene (encompassing the Epipalaeolithic, Mesolithic and Early Pastoral periods).
Four sites in particular (WS100, 103, 104 and 107, Figure 2) were characterised by extensive
lithic scatters, some pottery, occasional hearths and in situ cultural material.
Lithics
The Early Holocene is well represented in the chipped stone assemblages from WS100,
WS103, WS104 and WS107 (Table 1). All four sites contained diagnostic microliths and
Ounan points (Figure 3). The assemblage from WS100 contained a high proportion of
blades, bladelets and microburins. Cores are mainly multiple platform, with some bipolar
on anvil cores as well. A wide variety of materials is used, including large amounts of
quartz, as well as jasper, flint and silicified sandstone, and small amounts of rock crystal and
chalcedony. All assemblages include microliths in various forms, and WS100 and WS107
921

Nick Brooks et al.
Table 1. Early Holocene chipped stone counts.
WS100 %WS104.106, 14 %WS107, 20 %WS107.21 %
Flake 839 58.1 122 61.90 62 55.40 181 47.5
Blade 29 2.0 0.00 10 8.90 42 11.00
Bladelet 197 13.7 23 11.70 12 10.70 45 11.80
Microburin 27 1.9 9 4.60 10 8.90 2 0.50
Spall 4 0.3 1 0.90 5 1.30
Chip 200 13.9 26 13.20 0.00 23 6.00
CTE 13 0.9 1 0.50 3 2.70 2 0.50
Indeterminate 117 8.1 14 7.10 12 10.70
Core 17 1.2 2 1.00 2 1.80
Total 1443 197 112 381
Tool class WS100 %WS104.106, 14 %WS107, 20 %WS107.21 %
Awl 5 2.0 0.00 1 5.30 3 4.30
Backed flake 15 6.0 4 10.30 0.00 5 7.10
Burin 17 6.7 3 7.70 3 15.80 4 5.70
Denticulate 7 2.8 3 7.70 0.00 1 1.40
Fragment 15 6.0 2 5.10 3 15.80 7 10.00
Microlith 36 14.3 6 15.40 0.00 2 2.90
Marginally retouched flake 68 27.0 12 30.80 7 36.80 18 25.70
Marginally retouched blade/let 9 3.6 1 2.60 1 5.30 14 20.00
Notch 24 9.5 2 5.10 1 5.30 2 2.90
Pi`
ece esquill´
ee 1 0.4 0.00 1 1.40
Point 31 12.3 0.00 2 10.50 9 12.90
Scraper 7 2.8 2 5.10 1 5.30 0
Strangled piece 3 1.2
Tanged piece 6 2.4
Truncation 8 3.2 4 10.3 0.00 2 2.90
Total 252 100.0 39 19 70
also include small Ounan points as well as small backed points made using the microburin
technique to form a point. The assemblages seem likely to date to the eighth or more likely
ninth millennium bp. The importance of microliths in the Early Holocene has been attested
across the Sahara, as well as south into the Sahel. These often take lunate or triangular
forms as at Wadi Tifariti, and as seen, for example, in the Acacus (di Lernia 1999b) or
in the Capsian (Lubell 2001). The presence of Ounan points is also significant, and links
Wadi Tifariti with far flung Early Holocene sites. This point form has been found across
the Sahara, most commonly in ninth-millennium sites, for example the Dakhleh Oasis
(McDonald 1982) several sites in the Basin d’Azawagh, Niger (Paris 1992) and Foum el
Alba, Mali (Raimbault 1983). Regional specialities may include the presence of backed
points, similar to ones collected by Almagro Basch at a number of coastal sites north of
Cabo Juby, such as Smeil el Leben (1946: 107-8) and found with microliths including
lunates.
Proportions of tool classes vary significantly both within some sites, and between the
sites. At WS100 one area is dominated by non-formal flake tools alongside points and
922

Research
The archaeology of Western Sahara
Figure 3. Characteristic lithic types found in the study area: nos. 1-3) Ounan points and microlithic; 4-5) blades.
microliths, and the second area contains higher numbers of blades, often forming truncation
burins. At WS104.106 there are no points, despite its similarity to the other assemblages.
WS107 has particularly high proportions of bladelets and microburins (the latter making
up 9 per cent of the assemblage). This variability presents us with good opportunities for
investigating Early Holocene use of the landscape through a suite of neighbouring sites.
Other aspects of the exploitation of this area may be present in neighbouring Wadi Kenta
where assemblages similar to those in Wadi Tifariti contain less exhausted cores showing
formal opposed blade removals at Abric del Nius (Soler Subils 2004). It is tempting to see
these as an earlier reduction stage not present in the Wadi Tifariti sites. Some of the tool
class variability in the survey area may be chronological, such as the lack of points at the
microlithic WS104.106. However, it must be noted that to the south of the Sahara, Ounan
923

Nick Brooks et al.
points are not present – at eighth- and ninth-millennium Ounjougou in southern Mali, for
example, assemblages are flake based, with microliths on quartz (Huysecom et al. 2001). The
widespread use of quartz in Wadi Tifariti, as well as substantial amounts of bipolar on anvil
reduction, is more reminiscent of these industries than those reported further east in the
Sahara.
Ceramics
The Early Holocene is also well represented in the pottery assemblages from WS100
and WS107 but less so from WS104 or WS103, where Early Holocene pottery was
found mixed with third to fifth millennia bp ceramics. The pottery assemblage in general
is typically Saharan with combed decoration predominating but also some sherds with
incised decoration (for comparisons, see Camps-Fabrer 1966: 469). Twenty-two sherds were
recovered from test pits in WS100.101 and WS100.102. Many were burnt and decorated in
tightly packed zigzags of rocker stamped impressions (Figure 4). Fabrics were dark with high
concentrations of white grits. Comparative material comes from Uan Afuda Cave (see di
Lernia 1999a: 91, Figure 6.25) and Uan Tabu rockshelter (see Garcea 1999: 159, Figure 4)
in the Libyan Sahara and further to the south where the Early T´
en´
erian pottery of Adras
Bous shows similarities in techniques of decoration (Garcea 2008: 276). Early Holocene
pottery was also prevalent at WS107 (Figure 5: 1-5). In particular, the thick-walled rim of a
bowl decorated with rocker stamping closely resembles examples from Uan Afuda where di
Lernia comments that a feature of the Early Holocene pottery is the ‘considerable thickness
of the walls’ (di Lernia 1999a: 88).
The mid–late Holocene (c. sixth millennium bp onwards)
An arid episode c. 5200 bp signalled the acceleration of a trend (beginning some centuries
earlier) towards more arid conditions in the Sahara and across the northern hemisphere
sub-tropical belt, culminating in a transition to full aridity throughout most of the Sahara
between c. 4500 and 3600 bp, depending on location, geology, topography and region
(Gasse 2000; Brooks et al. 2005; Brooks 2006). This transition occurs around 4000 bp on
the northern margins of the Sahara and c. 4500 bp in more southerly Saharan latitudes
(Gasse 2002). For regions west of 15◦E and south of 20◦N, Lancaster et al. (2002) place
the desiccation of this part of Mauritania around 4500 bp. Bulinus truncatus shells from a
playa site in the north of the region yielded a date of 4020 +
−40 bp (Brooks et al. 2003),
coinciding with the transition to aridity on the northern margins of the Sahara.
Use of the study area during this period changes, with only a few occupation sites
producing chipped stone and ceramics, and widespread scatters of Late Pastoral ceramics
possibly suggesting the presence of mobile groups. The use of the area as a funerary landscape
begins in this period, with monument building documented throughout the study area.
Occupation sites
At WS107.21 there is a greater emphasis on larger blades in an assemblage dominated by
flakes (Figure 3: 4-5). Tools include larger, more elaborately retouched points, with either
924

Research
The archaeology of Western Sahara
Figure 4. Ceramics from sites WS104 (1-3); WS100 (4); WS023 (5-12).
925

Nick Brooks et al.
Figure 5. Ceramics from site WS107.
926

Research
The archaeology of Western Sahara
fairly rectilinear edges or a somewhat more foliate shape. Many published sites across the
Sahara show a range of point types present during the fifth–sixth millennium, as at Temet
(Raimbault 1983), Et-Teyyedch´
e and Tintan (Vernet 2007), Taruma (Almagro Basch 1945-
46), and Uan Telocat (Garcea & Sebastiani 1998) where transverse, foliate and tanged
points all exist. WS107 has also produced dotted wavy line pottery on shapes with flaring
necks; a characteristic which Ponti et al. (1998: 191) suggests does not appear until the fifth
millennium bp (see Figure 5: 6-7).
WS104 produced a diagnostically different pottery assemblage from all other occupation
sites, characterised almost exclusively by parallel or cross-hatched incision or cross-hatched
impressed design (Figure 4: 1-3). Comparative material can be found much further to the
south and west in Mauritania and the Atlantic Sahara at sites such as Cap Juby (Almagro
Basch 1946: 110, Figure 42; Camps-Fabrer 1966: Plate LIII) and in the Gulf of Arguin at
Foum Arguin 38 and Jerf Sgha¨
ır (Vernet 2007: 90, 94).
Pottery was also recovered from excavations of tumulus WS023 (Brooks et al. 2006). On
and around the surface of the mound were the sherds of at least three vessels; two were plain
wheel made jars with rolled necks (Figure 4: 5) and one (comprising 43 sherds) was decorated
in carefully impressed intersecting concentric circles (Figure 4: 6). This type of decoration
appears to be relatively uncommon in the Sahara, but a similar decorative technique has
been recorded on pottery from Orub in Niger (Gr´
eb´
enart 1985: Figures 21-7) and Ain en
Nouss in Mauritania (Garcea 1998). At Orub, it occurs with other unique elements of shape
and form, as yet undocumented in the Western Sahara assemblages. Gr´
eb´
enart is unclear as
to how the motif is created on the Orub examples. One theory already suggested (Brooks
et al. 2006) is that the surface was impressed with a woven circular mat in overlapping
concentric circles. Whatever, the technique of decoration, these and the plain wheel-made
sherds are later than the pottery recovered from the occupation sites, probably dating to
after the third millennium bp.
The sixth millennium bp (Middle Pastoral period) is poorly represented by both lithics
and ceramics but its recognition may well be considerably hampered by previous collecting.
This has been noted in many parts of the Sahara and Sub-Sahara (e.g. Bedaux & Rowlands
2001; Niang et al. 2001) with Neolithic points at particular risk because of their visibility,
transportability and commercial value – points from the Western Sahara are widely found
for sale on the internet.
Burial sites
While signs of mid–late Holocene occupation are scarce, the presence of numerous stone
monuments indicates that Wadi Tifariti was used as a funerary landscape during this period.
Fourteen monument types occur within the study area (Table 2; Figure 6), including,
amongst others:
!stone tumuli (cairns of varying types)
!corbeilles (rings of closely spaced outward leaning orthostats)
!bazinas (drum cairns, round and rectangular)
!paved crescents (large boomerang shaped stone pavements with an integral cairn)
927

Nick Brooks et al.
Table 2. Preliminary overview of monument types in some Saharan regions.
N/Niger (adapted Acacus (adapted from
Monument types WS/North WS/South from Paris 1996) di Lernia et al. 2002)
Tumuli and variations
Tumuli X X X X
Crater tumuli X
Tumuli with stone alignments
(Paris 1996)
X
Platform tumuli X X X X
Paved crescents X
Low relief platforms without
tumuli
XX
Disc or Platform monument
types
X X X
Drum type monuments
Bazina X X X X
Complex monument types
Complex monuments X
Mounded crescents X X X
Double mounded crescent X
Key hole X X
Outlined monuments
Corbeilles X X
Stone rings X X X
Outlined rectangle X
Antenna monuments
V-Type with platform X
V-Type with mound (not
found in TF1 area)
X
Cresecent antennae X
Composite monument X
Other large area monuments
Large crescent (not found in
TF1 area)
XX
Axle monument (not found in
TF1 area)
XX
Ridge monument X
Goulets X
!goulets (a stone outlined alley extending out from a cairn and curving outwards and back
on itself creating two large, parallel enclosures)
!stone rings
!antenna monuments (a central cairn with radiating ‘arms’)
!standing stones
!kerb burials (elliptically shaped stone outlines with standing stones at the approximate
north and south ends – probably Islamic).
928

Research
The archaeology of Western Sahara
Figure 6. (Clockwise from top left): basic tumulus (∼1.2m in height); goulet; crescent antennae monument; complex
monument (up to 0.6m in height), Tifariti study area.
Many such monument types exhibit morphological traits similar to monuments from
Niger (Paris 1996), Mauritania, Algeria and Libya (Baistrochi 1987; di Lernia et al. 2002;
Mattingly 2003; Vernet 2007). Paved crescents represent a more regional style, having been
documented in Algeria (Gauthier et al. 1997), while ‘complex’ monuments (cairns with long
fac¸ades of orthostats extending beyond the cairn itself, and with rectangular stone outlines
at the rear) appear to represent a previously unrecorded type of monument (Brooks et al.
2006).
Basic cairns are the most common type and are at their densest on the high, rocky ground
east and west of Wadi Tifariti. Goulets and complex monuments are usually situated on
the low ground, where level space is a prerequisite, while large crescent antennae extend up
the flanks of the higher ground, and paved crescents occupy some of the highest locations.
Stone cairns with associated standing stones and/or false entrances are usually also on the
highest ground, facing eastwards, with corbeilles on lower hillside flanks. The diversity of
monument types in this small area is quite striking, and perhaps a ‘zoning’ is taking place,
both spatially and temporally. But an appreciation of this will only come about through
further fieldwork.
Dating these structures is difficult as, throughout the Sahara, monuments with human
interments range in time from c. 5600 to 1200 bp (with cattle burials occurring earlier in the
929

Nick Brooks et al.
seventh millennium bp). However, Sivili’s compilation of 14 C dates (2002: 23, Figure 3.2)
along with Paris’ dates for ‘mounded’ crescents (1996: 271, Table 42) can be used to establish
some provisional chronological correlations. Perhaps the most useful monument types for
relative dating are mounded crescents (though only three similar monuments – crescent
shaped cairns – have been found in the study area), and V-shaped antenna monuments (of
which ‘crescent antennae’, found in the study area and consisting of a large north-south
aligned, elliptical cairn, with very long arching arms facing eastwards, are a possible type).
Both of these have quite restricted date ranges (4700-3400 bp and 5400-3000 bp
respectively), placing the beginning of the appearance of these types around the transition
from the Middle to Late Pastoral period. Added to this, on associated artefactual evidence,
the tumuli in the study area may begin later than in other regions. Thus, on the limited
evidence available it appears likely that the study area was used as a funerary landscape
sometime after the climate became drier, towards the end of the sixth millennium bp. It
is likely that use continued into the Islamic period, since probable Islamic burials (kerb
burials) are found amidst the earlier monument types.
Radiocarbon dating of bone from two excavated tumuli in the study area (Brooks et al.
2006) yielded dates of 1429 +
−80 bp (site WS023) and 1394 +
−85 bp (site WS024). Low
collagen content of the samples (as specified in the laboratory report) suggests that these
results should be treated with caution, and this is supported by accompanying grave material
which appears on typological grounds to be earlier than these dates suggest. The presence
of metal artefacts in both tumuli constrains their construction to the third millennium bp
or later (Childs & Killick 1993).
Discussion
Survey work and preliminary excavation in the Wadi Tifariti have demonstrated a sequence
of occupation in the Western Sahara that includes the Early Holocene (Epipalaeolithic and
Mesolithic), and the early Middle Holocene (Early Pastoral period). Significant occupation
of the study area during the Early Holocene is indicated by lithic concentrations located
mainly on rocky outcrops above the Wadi Tifariti, often at confluence points that afford
good visibility along two channels. Considerable variability is apparent within and between
these occupation sites. Di Lernia (1999a) has pointed out the great variability of assemblage
make-up and lithic typology in the Early Holocene, and these sites contribute to a growing
picture of variability in technologies and toolkits. Further work on this suite of sites may
help us to understand how some of the different expressions of Early Holocene settlement
and technology relate to each other.
Evidence of occupation during the sixth and seventh millennia bp is less extensive,
and occurs on terraces at lower elevations, raised only slightly higher than the wadi itself,
suggesting changes in patterns of landscape use. This period encompasses the shift towards
more arid conditions around or before 5200 bp (Brooks et al. 2005; Brooks 2006), and may
have seen reduced fluvial activity in the wadi. Hunter-gatherer groups may have partially
or completely abandoned the Wadi Tifariti as conditions became more arid, or changed
their use of the landscape in ways that makes their presence less visible in the archaeological
record (e.g. increasing mobility).
930

Research
The archaeology of Western Sahara
Human use of the study area appears to change in the mid–late Holocene with the
introduction of monumental funerary and ‘ritual’ architecture (Figure 6). The diversity
of monuments is remarkable, as is the timing of this transition, with monument building
appearing after the shift to more arid conditions in the wider Saharan region, and continuing
long after the final regional desiccation. The development of funerary architecture has been
associated with the westward expansion of cattle pastoralism elsewhere in the Sahara (di
Lernia 2006; see also Jousse 2004 and Figure 1), and domesticated cattle are depicted widely
in the rock paintings and engravings in Western Sahara, the earliest of which have been
tentatively dated to the fourth millennium bp (Soler Subils et al. 2006). The replacement
or augmentation of hunter-gatherer groups by pastoralists under conditions of increasing
aridity has been described for the Central Sahara by Sereno et al. (2008). Monumental
funerary architecture has been interpreted as evidence of increasing social complexity and
stratification during the Middle Holocene in the Sahara and elsewhere, and has been linked
to social changes associated with environmental deterioration (Brooks 2006; di Lernia 2006).
The links between environmental change, social change and the development of funerary
landscapes in the vicinity of Wadi Tifariti require further study.
The work around the Wadi Tifariti represents an opportunity to compare the
archaeological record of the far west of the Sahara with records from central and eastern
regions at similar latitudes (e.g. Niger, Libya and Sudan), as well as with trajectories of coastal
occupation to the west, and of occupation on the northern Saharan fringes and in the Sahel.
Due to the relative paucity of recent work in the western areas of the Sahara, the Western
Sahara Project has had to rely to a large extent on comparisons with distant regions such as
the Acacus in Libya, some 2000km away, and the coastal Atlantic some 1000km to the south-
west. What is perhaps surprising is that these comparisons have revealed some very distinct
similarities in typology and technology of lithics, ceramics and monuments. Chipped stone
and ceramics also exhibit affinities with Sub-Saharan technologies. Nonetheless, there are
hints of distinctive local variants in e.g. backed point forms, in certain types of monuments
(specifically ‘complex’ monuments, and crescent antennae – though V-shaped antennae are
spread across the Sahara well into Libya) and in pottery forms and decoration, such as the
concentric circle pattern seen on sherds associated with WS023.
In conclusion, initial investigations in Wadi Tifariti have demonstrated the potential
importance of Western Sahara for the study of population movements, the transmission
of cultural forms and livelihoods, and the mediation of these processes by climatic and
environmental change. Future work will develop such studies, establishing chronologies
of cultural and environmental change, comparing these with similar records from other
regions, and exploring human-environment interaction in this archaeologically significant
and hitherto little-known region.
Acknowledgements
We wish to thank the people of Western Sahara, in particular the Office of the President of the Sahrawi Arab
Democratic Republic; the Minister of Culture, Khadija Hamdi; the Director of the Sahrawi National Museum,
Mr Bey Houssein; our principal coordinator in the field, Mr Bachir Mehdi Bhaua, and the Sahrawi members
of our support teams who are too numerous to mention. Thanks go to the SADR representatives in London,
Mr Yahiaoui Lamine, Mr Sidi Omar and Mr M. Limam Mohamed Ali. Thanks also go to non-Sahrawi team
931

Nick Brooks et al.
members: Matt Nichol, Alex Wasse, Jon Crisp, H´
elene Jousse and Federica Crivellaro. Illustrations were the
work of Sarah Lucas. Finally, our special thanks go to Mrs Margaret Raffin, whose dedication to the people of
Western Sahara and initiation of the Western Sahara Project has made this research possible.
The work published here was generously supported by Ophir Energy. Other funding was provided by the
British Academy, the University of Edinburgh, and through the contributions of volunteers, whose support is
gratefully acknowledged.
References
ALMAGRO BASCH, M. 1945-46. Un yacimiento del
Neol´
ıtico de tradici´
on capsiense del Sahara espa˜
nol.
Las Sebjas de Taruma (Segu´
ıa el Hamra) Ampurias
7-8: 69-81.
– 1946. Prehistoria del Norte de Africa y del Sahara
Espa˜
nol. Barcelona: Consejo Superior de
Investigaciones Cient´
ıficas, Instituto de Estudios
Africanos.
BAISTROCHI, M. 1987. Pre-Islamic megalithic
monuments of the Northern Dadrat Acacus, in
B.E. Barich (ed.) Archaeology and environment in the
Libyan Sahara: the excavations in the Tadrat Acacus
1978-1983 (British Archaeological Reports
International Series 368): 87-89. Oxford: British
Archaeological Reports.
BEDAUX, R.M.A. & M. ROWLANDS. 2001. The future
of Mali’s past. Antiquity 75: 872-6.
BROOKS, N. 2006. Cultural responses to aridity in the
Middle Holocene and increased social complexity.
Quaternary International 151: 29-49.
BROOKS, N., S. DI LERNIA, N. DRAKE, M. RAFFIN &
T. SAVAGE. 2003. The geoarchaeology of Western
Sahara. Preliminary results of the first Anglo-Italian
expedition in the ‘Free Zone’. Sahara 14: 63-79.
BROOKS, N., S. DI LERNIA, N. DRAKE, I. CHIAPELLO,
M. LEGRAND, C. MOULIN & J. PROSPERO. 2005.
The environment-society nexus in the Sahara from
prehistoric times to the present day. The Journal of
North African Studies 304: 253-92.
BROOKS, N., J. CLARKE, J. CRISP, F. CRIVELLARO,
H. JOUSSE, E. MARKIEVICZ, M. NICHOL, M.
RAFFIN, R. ROBINSON, A. WASSE & V. WINTON.
2006. Funerary sites in the ‘Free Zone’: report on
the second and third seasons of fieldwork of the
Western Sahara Project. Sahara 17: 73-93.
CAMPS-FABRER, T. 1966. Mati`
ere et art mobilier dans la
prehistoire nord-Africaine. Paris: Arts & M´
etiers
Graphiques.
CHILDS, S.T. & D. KILLICK. 1993. Indigenous African
metallurgy: nature and culture. Annual Reviews of
Anthropology 22: 317-37.
DE BURUAGA BL´
AZQUEZ, A.S. 2006. Nuevos testimonios
industriales Achelenses en el Tiris (Sahara
Occidental): presentaci´
on del taller de Gnefisa
Oum Agraid (Duguech). Veleia: Revista de
prehistoria, historia antigua, arqueolog´
ıa y filolog´
ıa
cl´
asicas 23: 405-13.
DEMENOCAL, P., J. ORTIZ, T. GUILDERSON, J. ADKINS,
M. SARNTHEIN, L. BAKER & M. YARUSINSKY. 2000.
Abrupt onset and termination of the African
Humid Period: rapid climate responses to gradual
insolation forcing. Quaternary Science Reviews 19:
347-61
DI LERNIA, S. 1999a. The Uan Afuda Cave hunter-
gatherer societies of Central Sahara (Arid Zone
Archaeology Monographs 1). Firenze: All’Insegna
del Giglio.
– 1999b. The cultural sequence, in S. di Lernia (ed.)
The Uan Afuda Cave: hunter-gatherer societies of
Central Sahara (Arid Zone Archaeology
Monographs 1): 57-130. Firenze: All’Insegna del
Giglio.
– 2006. Building monuments, creating identity: cattle
cult as a social response to rapid environmental
changes in the Holocene Sahara. Quaternary
International 151: 50-62.
DI LERNIA, S., G.B. BERTOLANI, R. CASTELLI,
F. MERIGHI & A. PALOMBINI. 2002. A regional
perspective: the surveys, in S. di Lernia &
G. Manzi (ed.) Sand, stones and bones: the
archaeology of death in the Wadi Tannezzuft Valley
(5000-2000 BP): 25-68. Firenze: All’Insegna del
Giglio.
DUBIEF, J. 1953. Essai sur l’hydrologie superficielle au
Sahara. Clairbois-Birmandreis: Gouvernment
g´
en´
eral de l’Alg´
erie, Direction du service de la
colonisation et de l’hydraulique, Service des ´
etudes
scientifiques.
GARCEA, E.A.A. 1998. From early Khartoum to the
Saharan Neolithic, ceramics in comparison, in
Actes de la VIIIe conf´
erence internationale des ´
etudes
nubiennes, 11-17 septembre 1994 (Cahiers de
recherches de l’Institut de Papyrologie et
d’Egyptologie de Lille 17): 91-104. Villeneuve
d’Ascq: Universit´
e Charles de Gaulle-Lille III.
– 1999. Aterian and ‘Early’ and ‘Late Acacus’ from the
Uan Tabu rock-shelter, Tadrart acacus (Libyan
Sahara), in M. Cremaschi & S. di Lernia (ed.) Wadi
Teshuinat palaeoenvironment and prehistory in
south-western Fezzan (Libyan Sahara). Survey and
excavations in the Tadrart Acacus, Erg Uan Kasa,
Messak Settafet and Edeyen of Murzuq 1990-1995
(C.N.R. Quaderni di Geodinamica Alpine e
Quaternaria 7): 155-82. Firenze: All’Insegna del
Giglio.
932

Research
The archaeology of Western Sahara
– 2008. The ceramics from Adrar Bous and
surroundings, in J. Desmond Clark & D. Gifford-
Gonzalez (ed.) Adrar Bous: archaeology of a Central
Saharan granitic ring complex in Niger: 245-89.
Tervuren: Royal Museum for Central Africa.
GARCEA, E.A.A. & R. SEBASTIANI. 1998. Middle and
Late Pastoral Neolithic from the Uan Telocat
rockshelter, Tadrart Acacus (Libyan Sahara), in
M. Cremaschi & S. di Lernia (ed.) Wadi Teshuinat
palaeoenvironment and prehistory in south-western
Fezzan (Libyan Sahara). Survey and excavations in
the Tadrart Acacus, Erg Uan Kasa, Messak Settafet
and Edeyen of Murzuq 1990-1995 (C.N.R.
Quaderni di Geodinamica Alpine e Quaternaria 7):
201-16. Firenze: All’Insegna del Giglio.
GASSE, F. 2000. Hydrological changes in the African
tropics since the Last Glacial Maximum.
Quaternary Science Reviews 19: 189-211.
– 2002. Diatom-inferred salinity and carbonate oxygen
isotopes in Holocene waterbodies of the Western
Sahara and Sahel (Africa). Quaternary Science
Reviews 21: 737-67.
GAUTHIER, Y., C. GAUTHIER, W. N¨
OTHER & P. LLUCH.
1997. Monuments de l’Immidir (Alg´
erie). Sahara
9: 143-8.
GR´
EB´
ENART, D. 1985. La r´
egion d’in Gall-Tegidda n
Tesemt (Niger). Programme archeologique d’urgence
1977-1981. II Le N´
eolithique Final et les d´
ebuts de la
m´
etallurgie. Niamey: Institut de Recherches en
Sciences Humaines.
HUYSECOM, E., E. BO¨
EDA, K. DEFORCE,
HD
OUTRELEPONT, A. DOWNING, N. FEDOROFF,
D. KONAT´
E, A. MAYOR, S. OZAINNE, F. RAELI, A.
ROBERT, S. SORIANO, O. SOW & S. STOKES. 2001.
Ounjougou (Mali): quatri`
eme campagne de
recherches dans le cadre du programme
international ‘Pal´
eoenvironnement et peuplement
humain en Afrique de l’Ouest’, in Jahresbericht
SLSA 2000: 105-50. Z¨
urich et Vaduz: Fondation
Suisse-Liechtenstein pour les recherches
arch´
eologiques `
a l’´
etranger.
JOUSSE, H. 2004. A new contribution to the history of
pastoralism in West Africa. Journal of African
Archaeology 2: 187-201.
LANCASTER, N., G. KOCUREK, A. SINGVI, V. PANDEY,
M.M. MEYNOUX, J.-F. GHIENNE & K. L ¨
O. 2002.
Late Pleistocene and Holocene dune activity and
wind regimes in the western Sahara Desert of
Mauritania. Geology 30: 991-4.
LUBELL, D. 2001. Late Pleistocene-Early Holocene
Maghreb, in P.N. Peregrine & M. Ember (ed.)
Encyclopedia of prehistory V1: Africa: 129-49.
New York: Kluwer Academic/Plenum.
MCDONALD, M.M.A., 1982. Dakhleh Oasis Project
third preliminary report on the lithic industries in
the Dakhleh Oasis. Journal of the Society for the
Study of Egyptian Antiquities 12: 115-38.
MATTINGLY, D.J. 2003. The archaeology of the Fazz¯
an.
Volume 1: synthesis. Tripoli & London: Department
of Antiquities and the Society for Libyan Studies.
NIANG, M., B. NAGANDO, S. SEIDOU & E. WANGARI.
2001. Le pillage de sites culturels et naturels au Niger.
Paris: Centre du Patrimoine Mondial (WHC),
UNESCO.
PARIS, F. 1992. Le Bassin de l’Azawagh: peuplements et
civilisations, du N´
eolithique `
a l’arriv´
ee de l’Islam, in
A. Marliac (ed.) Milieux, soci´
et´
es et arch´
eologues:
227-57. Paris: Karthala.
– 1996. Les s´
epultures du Sahara nig´
erien du N´
eolithique
`
a l’islamisation: 1) Coutumes fun´
eraires, chronologie,
civilisations; 2) Corpus de s´
epultures fouill´
ees. Paris:
ORSTOM.
PONTI, R., G. AURISICCHIO, R. DAMIOTTI & G. GUIDI.
1998. Pottery from the Tadrart Acacus (Libyan
Sahara): decoration, distribution and manufacture,
in M. Cremaschi & S. di Lernia (ed.) Wadi
Teshuinat palaeoenvironment and prehistory in
south-western Fezzan (Libyan Sahara). Survey and
excavations in the Tadrart Acacus, Erg Uan Kasa,
Messak Settafet and Edeyen of Murzuq 1990-1995
(C.N.R. Quaderni di Geodinamica Alpine e
Quaternaria 7): 183-200. Firenze: All’Insegna del
Giglio.
RAIMBAULT, M. 1983. Industrie lithique, in
N. Petit-Maire & J. Riser (ed.) Sahara ou Sahel?
Quaternaire r´
ecent du bassin de Taoudenni (Mali).
Marseille: Centre National Recherche
Scientifique.
REIMER P.J., M.G.L BAILLIE, E. BARD, A. BAYLISS,
J.W. BECK, C. BERTRAND, P.G. BLACKWELL, C.E.
BUCK, G. BURR, K.B. CUTLER, P.E. DAMON, R.L.
EDWARDS, R.G. FAIRBANKS, M. FRIEDRICH, T.P.
GUILDERSON, K.A. HUGHEN, B. KROMER, F.G.
MCCORMAC, S. MANNING, C. BRONK RAMSEY,
R.W. REIMER, S. REMMELE, J.R. SOUTHON, M.
STUIVER, S. TALAMO, F.W. TAYLOR, J. VAN DER
PLICHT & C.E. WEYHENMEYER. 2004.
IntCal04.14c. Radiocarbon 46: 1029-58.
SERENO, P.C., E.A.A. GARCEA, H. JOUSSE, C.M.
STOJANOWSKI, J.-F. SALI`
EGE, A. MAGA, O.A.
IDE, K.J. KNUDSON, A.M. MERCURI, T.W.
STAFFORD, T.G. KAYE, C. GIRAUDI, I. MASSAMBA
N’SIALA, E. COCCA, H.M. MOOTS, D.B.
DUTHEIL & J.P. STIVERS. 2008. Lakeside cemeteries
in the Sahara: 5000 years of Holocene population and
environmental change. PLoS ONE 3(8): e2995.
doi:10.1371/journal.pone.0002995.
SIVILI, S. 2002. A historical background: mortuary
archaeology in the Sahara between colonialism and
modern research in S. di Lernia & G. Manzi (ed.)
Sand, stones and bones: the archaeology of death in the
Wadi Tannezzuft Valley (5000-2000 BP): 17-24.
Firenze: All’Insegna del Giglio.
933

Nick Brooks et al.
SOLER, N., C. SERRA, J. ESCOLA & J. UNGE. 1999.
S´
ahara occidental: pasado y presente de un pueblo.
Girona: Universitat de Girona, Funcaci´
o privada
Girona Universitat i Futur.
SOLER SUBILS, J. 2004. Les pintures rupestres
prehist`
oriques del Zemmur (Sahara Occidental).
Unpublished PhD dissertation, University of
Girona.
SOLER SUBILS, J., N. SOLER MASFERRER & C. SERRA
SALAM´
E. 2006. The painted rockshelters of the
Zemmur (Western Sahara). Sahara 17: 129-42.
VERNET, R. 2007. Le Golfe d’Arguin de la pr´
ehistoire `
a
l’histoire: littoral et plaines int´
erieures. Nouakchott:
Parc National du Banc d’Arguin.
934