This paper synthesises the findings of three seasons of fieldwork in Western Sahara, focusing on the funerary and related archaeology of the Polisario-controlled “Free Zone”. Building on the results of a reconnaissance survey in the Northern Sector of the Free Zone in 2002, new findings from a reconnaissance survey in the Southern Sector are presented. Results from both seasons are integrated in order to provide a tentative description of the spatial distribution of stone monuments in this region. Emerging spatial patterns are described, indicating similarities but also significant differences between the Northern and Southern Sectors in terms of the archaeological record. Preliminary results from the excavation of two funerary monuments, undertaken during a third season in the Northern Sector, are also presented. These conical tumulus monuments contained articulated human skeletons and a limited variety of grave goods. The presence of metal in both tumuli suggests that they are likely to date from the first millennium BCE.
Riassunto
Vengono riportati i ritrovamenti di tre stagioni di lavoro nella Zona Libera del Sahara Occidentale, con riferimento particolare all’aspetto funerario dell’archeologia del territorio controllato dal Polisario. A partire dai risultati della ricognizione di superficie operata nel 2002 nel settore settentrionale della Zona Libera, vengono presentati i recenti risultati della ricognizione nel settore meridionale. I risultati delle due spedizioni sono integrati per cercare di fornire una descrizione della distribuzione spaziale dei monumenti in pietra della regione. Si notano determinate tipologie spaziali, che mostrano somiglianze ma anche sostanziali differenze nel record archeologico tra i settori settentrionale e meridionale. Sono inoltre presentati i risultati preliminari dello scavo di due monumenti funerari, condotto durante una terza missione nel Sahara Occidentale. Trattasi di monumenti a forma di tumulo conico, contenenti ossa umane articolate e alcuni oggetti di corredo. La presenza di metallo in entrambi i tumuli suggerisce una datazione possibile a partire dal primo millennio a.c.
Résumé
Cet article porte sur les résultats de trois saisons de travaux de terrain réalisées dans le Sahara occidental, et concerne particulièrement l'archéologie funéraire et apparentée de la "zone libérée" controllée par le Polisario. En plus des découvertes d'une prospection de reconnaissance dans le secteur nord de la zone libérée en 2002, celles d'une nouvelle prospection dans le secteur sud sont présentées. Les données des deux saisons sont intégrées afin de tenter une description de la distribution spatiale des monuments pierriers dans cette région. Les types régionaux qui en résultent sont décrits, indiquant des similitudes ainsi que des différences significatives de l'enregistrement archéologique entre les secteurs nord et sud. Le résultat préliminaire de la fouille de deux monuments funéraires, entreprise lors d'une troisième mission dans le secteur nord, est également exposé. Ces monuments tumulus coniques comportaient des squelettes humains articulés et une sobre variété d'offrandes. La présence de métal dans les deux tumuli suggère qu'ils datent probablement du premier millénium BCE.
1. Introduction
The first season of fieldwork conducted under the auspices of the Western Sahara Project took place in September and October 2002, and consisted of a geoarchaeological reconnaissance survey of the Northern Sector of the Polisario-controlled “Free Zone” (Brooks et al., 2003). The 2002 season resulted in the recording of a large number of funerary and related monuments, as well as a previously unknown rock art site at Bou Dheir. The second season of work, which took place in March 2005, extended the reconnaissance work to the Southern Sector, while the third and latest season in November 2005 focused on the excavation of two funerary monuments in the Northern Sector, in the area immediately north of Tifariti.
This paper describes the findings of the second and third seasons of work, and is divided into two main parts. The first part describes the types and distributions of funerary monuments recorded to date, focusing on the results of the recent survey in the Southern Sector. However, new results from the Northern Sector are also presented, focusing on the area around Tifariti. The results from the Northern and Southern Sectors are together interpreted within a wider Saharan context. The second part of the paper describes the findings of the November 2005 excavations north of Tifariti. The results of both excavations and survey work are synthesised in the discussion and conclusions, and are employed to develop a simple and tentative model of cultural change associated with Saharan desiccation in the middle Holocene. Although the focus of the paper is on funerary and related monuments (e.g. stone structures associated with funerary sites but not necessarily representing burials), the wider archaeological context is also addressed, with descriptions of worked stone, ceramics and (to a lesser extent) rock art.
For a discussion of the recent political history of Western Sahara and its relevance to archaeological and other research, the reader is referred to Brooks et al. (2003), Brooks (2005) and Shelley (2004). Figure 1 illustrates the de facto partition of Western Sahara resulting from the Moroccan occupation of the majority of the territory, dividing it into the “occupied territories” controlled by Morocco in the west, and the “Free Zone” controlled by the Polisario independence movement in the east.
2. Results of the reconnaissance surveys
Figure 1 shows the general locations of the main archaeological sites recorded to date. These are concentrated in the Northern Sector, north of the east-west oriented border with Mauritania, and in the southernmost parts of the Southern Sector of the Free Zone. The zone between these regions has not been subjected to reconnaissance survey to date. Most of the sites shown in Figure 1 consist of collections of stone monuments, representing a variety of functions and a wide range of construction styles. Comparison with other Saharan regions indicates that many (if not most) of these monuments are likely to be funerary in nature, and to date from the prehistoric period corresponding to the middle to late Holocene, or approximately 6000 to 2000 years before present (yrs BP) (Sivili, 2002). Contextual evidence in the form of lithics, ceramics and rock art also demonstrates the prehistoric origins of most of these monuments. However, while most of the recorded monuments represent styles found elsewhere in the Sahara, some of the styles observed (for example at Lajaud and Tifariti) are unusual within a Saharan context, at least to the knowledge of the authors.
Over the course of the three field seasons some 250 individual monuments were recorded, with basic information relating to their type, construction, geographic location, dimension and orientation noted. Figure 2 provides a typology of monuments in the form of illustrative, schematic sketches of the different monument types.
2.1. Typical Saharan monuments
Conical tumuli, platform, bazina and corbeille monuments (Figure 3) were recorded in both the northern and southern sectors, reflecting the wider Saharan tradition: such monuments are found throughout the Sahara. Tumuli range from crude mounds a few tens of cm in height to large, well-defined monuments up to 2 m high (Figure 3). In some cases tumuli are situated on natural or artificial platforms (Figures 2 and 3). Bazina monuments are constructed of layers of flattish stone blocks using a technique resembling that used to construct dry stone walls, and may be flat on top or support a tumulus-like structure (Figures 2 and 3).
A single “V-type” monument or antenna tomb (Figure 2), recorded in the vicinity of Bir Lahmar in 2002, was reminiscent of monuments recorded widely in the central Sahara, for example the Libyan Fezzan (Brooks et al., 2003) . Contrary to expectation, no further examples of this style were recorded in the southern sector in 2005, although more such monuments may well be present in areas as yet unsurveyed.
2.2. West Saharan regional styles
Goulet monuments, crescent-shaped structures, and monuments consisting of one or more conical tumuli situated on low-relief, asymmetrical, curved “platforms” (referred to from here on as “arced platforms”) reflect the general western Saharan regional tradition (Figures 2 and 4), with similar monuments having been recorded in southern Morocco, western Algeria and northern Mauritania (Gauthier et al., 1997; Gauthier and Gauthier, 2003, Milburn, 2005). The goulet monuments recorded during the reconnaissance are “long goulets” (sensu Milburn, 2005), consisting of two parallel stone arrangements (forming a “gulley” or “corridor”) extending from a tumulus or paved area. After some distance these arrangements diverge and curve back towards their origin, converging again behind the tumulus or platform to form two linked enclosures with the tumulus or platform forming the focus of the monument (Figures 2 and 4). Milburn (2005) distinguishes between these long goulets and “enclosure” monuments, which are similar in form, but consist of two separate enclosures divided by the gulley (Figure 4). Both goulet and enclosure monuments are commonly associated with low-relief stone circles, generally constructed from small to medium sized (~20 - 40 cm) stone blocks. These circles are invariably situated at the wider end of the monument, beyond the point at which the stone arrangements forming the gulley diverge.
The various types of monument representing these regional styles are not evenly distributed throughout the Free Zone. To date, goulet and arced platform monuments (Figures 2 and 4) have only been recorded in the Northern Sector, for example at site TF1 near Tifariti (Figure 9), although goulets are present to the south at Bir Lemuesat in Mauritania (Figure 1). In contrast, only two crescent-shaped monuments have been recorded in the Northern Sector, with one example being recorded on the plateau above the recently identified rock art site at Bou Dheir (Brooks et al., 2003) and the other recently recorded at Elous Lajaram (Figure 1). However, such monuments are extremely abundant in the Southern Sector, where they occur in isolation and in groups. Crescent monuments were recorded at all the sites in the Southern Sector listed in Figure 1, with the largest such monument being recorded near Lajuad. The distances between the centre of this monument and the ends of its curved arms are 180m and 270m, with the distance between the ends of the arms being 280m.
2.3. Atypical monuments at Lajuad
Four monuments were recorded at Lajuad (two at the main Lajuad site and one at a separate site to the northeast) which consist of lines of standing stones emanating from a central, low tumulus. Two of these monuments resemble structures familiar from the wider western Saharan region, which consist of a number of standing stones (typically thin, flat slabs) arranged in a line with the tallest stone in the centre so that the height of the monument decreases towards its edges (Milburn, personal communication). Such arrangements are present throughout the Free Zone, where they occur in isolation (e.g. Zoug), in association with low tumuli and corbeille-like structures (e.g. Lajuad, Azaig Bedrag), and as part of complex monuments (Tifariti).
The monument northeast of the main Lajuad site conforms to the pattern of a linear arrangement declining in height from the centre; the central stelae in this example are situated on top of a low tumulus and have collapsed. One of the three remaining monuments (at the main Lajuad site) is similar in nature (Figure 5b), although the linear arrangement runs along the edge of the central tumulus rather than through its centre. The remaining two monuments are unusual in that the stones (many of which are small rough blocks rather than vertically placed flat slabs) are arranged in a sinuous form rather than a straight line (Figure 5b). While the tallest stones occur at the centre of the monuments, in one of these monuments they curve around the perimeter of the central tumulus, forming a corbeille-like structure on one side (Figure 5b).
The three monuments at the main Lajuad site form part of a small but elaborate funerary complex situated at the base of a smooth granite hill. Two are incorporated into the perimeter of this complex, delineated away from the ridge monuments by an arc of stones extending from near the base of the hill (Figure 5a). The complex also contains a small (~2m) linear monument with a standing stone at one end, and at least three small circular tumuli with low vertical relief. This site is particularly notable for the high density of quartzite pebbles, which appear to have been deliberately deposited within the area enclosed by the perimeter arc. This complex is the smaller of two concentrations of funerary monuments at Lajuad, the other of which is very different in nature, consisting of a variety of more conventional tumuli and crescent monuments. Other isolated monuments are situated away from these two main concentrations; all the monuments at Lajuad itself are concentrated around the bases of a number of smooth granite hills which rise from a flat sand and gravel plain. One of these hills contains a cave housing numerous engravings of zoomorphic and humanoid forms, while a cave in another hill contains (presumably much later) paintings of abstract forms, Tifinagh-like symbols and zoomorphs. Lajuad is clearly an important site warranting further study.
2.4. Atypical monuments at Tifariti
A number of monuments with an unusual, complex construction were also recorded in the vicinity of Tifariti. These monuments consist of a number of elements, including a central tumulus, lines of standing stones, annexes and enclosures, all integrated into a single complex structure (Figures 2 and 6). In some cases arrangements of stones are present in the vicinity of the monument, although it is uncertain whether these are contemporaneous with the main structure. The distribution of these monuments in the vicinity of the November 2005 excavation site is shown in Figure 9b.
2.5. Standing stones
Standing stones are common in both sectors, where they are incorporated into larger monuments including tumuli and structures such as the complex of some 65 standing stones near Tifariti first noted by one of the authors (Raffin) during a preliminary visit to the area in 2000 and subsequently described by Brooks et al. (2003). To date, large (>2m high), isolated standing stones have only been recorded in the Southern Sector, in the vicinity of Azaig Bedrag (Figure 7), where two such examples were recorded. One of these standing stones (3.2 m long, 80 cm at widest point, 18 cm thick, inclined at 18º to vertical, and leaning towards the east) has a small, crude circular enclosure at its base, and is located in an area containing numerous funerary monuments, many of which are themselves associated with (often collapsed) standing stones.
2.6. Orientations of monuments
Many monuments exhibit specific orientations with respect to the cardinal points and/or prominent features within the landscape. These orientations are generally defined either by the axis of the monument or by subsidiary features associated with the monument.
Small standing stones are aligned with annexes at the perimeter of many of the tumuli (e.g. Figure 8). These annexes are common features of the conical tumuli and bazina-type monuments recorded during the fieldwork and known from other sites in the wider geographical region, for example in south-western Algeria (Lihoreau, 1993). They range from subtle and barely visible (at least to a superficial examination) to large and well-constructed (e.g. large flat slabs supported by upright stones at the sides), and generally appear to be integral to the structures with which they are associated. However, they appear to be symbolic rather than practical in nature, and do not function as physical entrances or openings into the monuments. In a sample of 27 tumuli with well-defined annexes, the bearings of the annexes ranged from 18º to 300º, with four values below 60º and five exceeding 150º. Of these nine exceptions, five are associated with a single group of monuments at Zoug, in the far south of Western Sahara, in which the “annexes” are actually secondary tumuli. These monuments should probably be interpreted as a distinct type. Generally, and in particular in the Northern Sector, these results may be interpreted as indicating a preference for an east to southeast orientation.
Preference for a particular orientation is particularly notable in crescent monuments. We define the orientation of these monuments by a line of sight from the middle of the central mound through the mid-point of a line drawn between the ends of the curved arms (Figure 2). Out of 24 crescent monuments recorded, the orientation of fourteen was determined as 110º, with a further four oriented at 100º. Orientations of the remaining monuments were 94º, 98º, 104º, 105º, 115º and 296º. These values were derived from compass bearings through a marker placed mid way between the ends of the arms, where these where well-defined, or estimated from the middle of the central mound where the ends of the arms could not be determined precisely due to disturbance or burial by blown sand. The error in the measurements may be as great as ±10º where the monuments are poorly defined, although in most cases the magnitude of the error is likely to be less than 3º. Discounting the single westerly facing monument, these data indicate that crescent monuments were deliberately constructed to face a particular direction, in most cases east-southeast and in many cases 110º, regardless of their location.
The function of crescent monuments has been a matter of some debate. It has been suggested that they may have been used to provide cover for hunting and to herd animals into killing areas, or as a means of controlling runoff from rainfall (Stefan Kröpelin, personal communication). The examples recorded in 2005 are typically located on or just below very low ridges, which provide raw material. The monuments are generally oriented with their open side facing down the slope away from the ridge, suggesting that their purpose was not to collect runoff. The cover provided by the highest, central part of the monument and the usually adjacent ridge may have aided stalking and hunting. However, the very low relief of most of the structure away from the central mound suggests that these monuments were not used as traps into which to herd animals for hunting purposes, unless they were associated with additional structural features made from perishable materials no longer present. The consistent orientations of the crescent monuments recorded in the Free Zone indicates a systematic approach to their construction that was not determined by practical considerations alone. This apparent ritual rather than functional role, coupled with the usual orientation to the east (a common feature of funerary monuments) suggests that these monuments may most reasonably be interpreted as funerary structures. Mercer (1976) mentions the excavation of such a monument at Cap Blanc, revealed as a burial monument. Further excavation in the Free Zone is required to confirm whether or not the many such monuments in this region are burials.
East-west orientations have been noted in other monuments, including the giant goulet identified near Tifariti in March 2005, subsequently described by Milburn (2005), and described in more detail below. A number of linear monuments have been recorded with north-south orientations. Both north-south and east-west directions are defined by a number of linear and circular stone arrangements (including corbeille structures and elements of the group of 65 stelae at Tifariti), in the latter case by pairs of stelae of greater height than the stones forming the bulk of the circle (e.g. Figure 3). A number of monuments are aligned with prominent features in the landscape, typically isolated hills, breaks in ridges, or other monuments. In some cases monuments are situated such that alignments with landscape features are also north-south or east-west (e.g. Figure 13). Near Tifariti, monuments align with hills which are themselves important funerary sites.
While the significance of the various alignments recorded in the Free Zone remains enigmatic, it is clear that those who constructed the monuments did so with reference to, and with a keen awareness of, the wider landscape. In considering the funerary and other monuments of this region, we should view them as part of a wider funerary and ritual landscape linked to the everyday physical and spiritual life of their prehistoric creators.
3. Excavations at Tifariti
3.1. Objectives
The principle objective of the November 2005 field season was to excavate two tumulus monuments in the Northern Sector in order to (i) retrieve material for scientific dating, in order to establish when these monuments were constructed as a first step in developing a cultural chronology for the region, and (ii) to assess burial practices and infer information about prehistoric populations through an assessment of skeletal remains and grave goods. Further survey work was conducted in the immediate vicinity of the excavation site in order to place the site within its wider archaeological and environmental context. These were the first excavations and systematic recordings of funerary monuments to be conducted in the Free Zone. To the knowledge of the authors the last professional excavations to be conducted in Western Sahara occurred in the 1970s, in the southwest coastal region (Petit-Maire, 1979).
3.2. Site and context
The excavation site was located in the extensive funerary area some 14 km north of Tifariti (site TF1). The excavated monuments were situated in an elevated area at the western end of a low ridge running approximately northwest-southeast, and sloping down to the Wadi Tifariti to the south and west and to a sand and gravel plain to the north and east (Figure 9). Today the Wadi Tifariti is well vegetated with open acacia woodland, with the slightly raised sand and gravel plains at its flanks largely devoid of vegetation (Figure 10).
Over sixty monuments, representing a variety of construction styles and constructed from locally available stone, have been recorded in the immediate vicinity of the excavations to date (Figure 9b). The vast majority of these are likely to be burials. The monuments include conical tumuli, complex monuments consisting of a tumulus and associated features such as rows of standing stones and stone annexes, corbeille structures, goulet and enclosure monuments, elongated and oval monuments, and simple mounds (Figures 2-6). It is possible that many more burials are present in this area, marked by low concentrations of stones that we might call “pseudo-tumuli”, which have not been individually recorded but which are extremely numerous. To the northwest of the excavation site is the extensive and locally well-known rock painting site of Rekeiz (Erqueiz), while the to the northeast the areas of Akchache and Rekeiz Lahmar (Figure 9a) provide evidence of both palaeolithic and Holocene occupation, and of further funerary activity (Brooks et al., 2003).
The study area was clearly an important funerary centre for the prehistoric peoples who inhabited this region. Although no secure dates are available for prehistoric human occupation in Western Sahara, by analogy with the central Sahara we may immediately conclude that the monuments are likely to date from the middle to late Holocene, being younger than about 6000 years old.
Two particularly large and notable monuments are located close to the study area, both located on raised gravel areas that would have probably remained dry when the wadis were more active than at present during the middle to late Holocene. These include an arrangement of 65 standing stones some 700 m from the excavation site (TF1.0, Figures 9 and 11), which has been described briefly by Brooks et al. (2003), and an exceptionally large goulet monument, situated approximately 1.5 km south-southeast of the excavation site (Figure 9). The latter is the largest monument of its kind recorded to date in the Sahara; a preliminary description of this monument was given by Milburn (2005). The monument is oriented east-west, aligned with an isolated flat-topped hill to the west, with the tumulus at the western end in an area of ground covered by stones to form a symbolic platform (Figures 12 and 13). To the east of the enclosures forming the main body of the monument are two large stone rings, slightly offset from the central, east-west oriented gulley or corridor (Figure 13). Beyond these are a large number of features consisting of approximately round or oval concentrations of stones, in some of which are set crude, low-relief mounds (Figure 13). A single angular rock is set in the ground some distance beyond these, aligned with the central corridor of the goulet (Figure 13). The total length of this “goulet complex” is 630 m (Figure 12). A smaller enclosure monument is situated immediately to the south, with other similar monuments situated on the gravel plain to the northeast of the excavation site (Figure 9b).
3.3. Methodology
The area containing the excavation site was known from previous visits to contain numerous monuments, and the location near the settlement of Tifariti made good logistical sense given the short duration of the field season and the need for logistical support. An important criterion in the selection of the monuments was size: the monuments selected were identified as small enough to be excavated in a 10-day period, but sufficiently large and well preserved to have retained a high degree of structural integrity, thus providing some protection to the contents from the effects of exposure to the elements, particularly water runoff. Location was also a factor in the selection of monuments: the tumuli selected for excavation occupied prominent positions on a ridge overlooking the Wadi Tifariti and adjacent areas rich in monuments (Figure 9). The monuments selected for excavation therefore form part, and are to an extent representative of, an important prehistoric funerary complex.
The initial phase of the excavation involved the careful removal of the surface layers of the tumuli, which consisted of small stones and wind-blown sediment. This was followed by the systematic excavation of one half of each monument to bedrock, exposing a cross section which left the central burial chambers intact. Sediments from the monuments were sieved throughout the excavation process in order to aid recovery of small finds (e.g. bone fragments, beads). The monuments were recorded at each stage of the excavations by means of context sheets, appropriately scaled drawings and digital photographs. Once one half of each monument had been excavated to bedrock, the central burial chambers were carefully excavated to expose their contents. The subsequent treatment of human remains was determined by their state of preservation: well-preserved skeletal material that was likely to survive removal was duly recovered, while remains unlikely to survive the removal process were assessed in situ and reburied. All skeletal remains were subject to detailed anthropological assessment, whether in situ or after removal, and were thoroughly recorded along with any grave goods. Samples of both human remains and sediments from within the burial chambers were recovered for radiometric dating and other analyses (e.g. parasitological, chemical), and dental samples were taken for genetic analysis. These analyses are currently underway, and their results will be described in future publications.
Human remains and grave goods recovered from the monuments were taken to Tifariti for recording, treatment with PVA consolidant where necessary (and after having proceeded with sampling in situ for physical-chemical analyses) and storage pending future analysis and display in the Tifariti Museum. Small finds were cleaned, photographed and carefully stored, pending more detailed assessment and conservation during future field seasons.
3.4. Results from excavation of Monument 1
The larger of the two tumuli selected for excavation (Monument 1) was roughly circular, approximately 6.5 m in diameter, 1.4 m at its highest point, and constructed of sandstone and granite readily available in local outcrops (Figure 14). A roughly oval burial chamber (with the long axis oriented E-W) was constructed in the centre of the tumulus from slightly inclined standing stones, over which were placed large (up to ~1m) sandstone slabs (Figure 14). The ground-level perimeter of the tumulus was defined by a ring of large (40-60 cm) stones; between the perimeter and the central chamber the tumulus was made up of somewhat smaller stones placed in a layer at ground level and then built up in an increasingly random fashion. The tumulus was finished with a covering of small stones, found on the top and sides of the monument. An annex resembling an entrance and constructed of large stones was located on the eastern side of the tumulus (Figure 14).
Excavation of the middle and upper layers of the burial chamber of Monument 1 (up to 30cm from the top of the chamber) yielded several fragments of human skeleton scattered in the filling (proximal metacarpal, right clavicle, coastal ribs, sternum, left pelvis, right scapula, left talus), and concentrated within the intersections of the edge stones, together with hundreds of rodent bones. A single fragment of pottery was also recovered in these layers.
Below these layers, at the base of the chamber, were the well-preserved remains of an adult individual (21-30 years), subsequently identified as female. The axis of orientation of the body was east-west, with the head pointing east (directly in line with the annex), and facing north (Figure 15). The body lay on the very bottom of the chamber, flexed on its right side, as a primary deposition, although very disturbed by termite and rodent activity. The in situ skeleton was lacking the thorax bones, with a rodent nest tunnel running in the E-W direction, following and dismantling the axis along the vertebral column. The human remains near the top of the tumulus, and the ceramic fragment, are likely to have been displaced from their original positions by animal activity. The micromammals identified within the tumulus fill and in the chamber are the Greater White-toothed Shrew (Insectivores, Crocidura russula), the North African elephant shrew (Macroscelididae Elephantulus rozeti), and the rodent species Gerbilllus gerbillus, Pachyuromys duprasi, Meriones libycus, Psammomys obesus and Jaculus jaculus, all of which occur today in arid North Africa.
The lower limbs were flexed, with the right leg crossing the left. The upper limbs were also flexed with the hands in front of the face: the right was open on the floor with the forearm forming an angle of 45° with the humerus, while the left arm was hyper-flexed towards the chin. The mandible was still articulated with the skull, collapsed towards the thorax. The cervical vertebrae were still connected to the base of the skull; compression of this connection suggests a disturbance of the burial soon after the deposition of the body.
There was evidence of red ochre in the burial chamber. All the bones, including those near the top of the tumulus, appeared to be from the same individual. A rodent burrow was evident immediately under the skeleton, where the remains of a single ceramic vessel were recovered (Figure 16), and abundant rodent and bird remains were present throughout the interior of the tumulus.
Sieving of material from above the burial also yielded three ostrich egg shell beads, a single carnelian bead, and a metal (apparently highly corroded iron) point of unknown provenance and function (Figure 16).
3.5. Results from excavation of Monument 2
The smaller monument (Monument 2) was also circular, measuring 5.5m in diameter and 1m in height. It was constructed of locally available sandstone and granite blocks and consisted of an external wall surrounding a central burial chamber (Figure 17). The wall appeared to spiral around the central chamber and was buttressed externally with flat slabs stood on end and sandstone rubble to create a more or less circular plan. This construction was covered with much smaller sandstone fragments to finish off the mound. Distinctive green blocks were noted at sporadic intervals in the stone work around the central chamber during construction. The central burial chamber consisted of a ‘beehive’ construction of massive overlapping granite and sandstone slabs. This was sealed on top by two large sandstone capstones (Figures 17 and 18). On the eastern side of the tumulus, two long stones were used in the construction of an abutting, partially collapsed structure which may originally have represented a false entrance (i.e. annex) or, at the very least, a focal point on the east side of the mound (Figure 18).
The burial chamber contained the poorly preserved remains of two humans lying one above the other, the lower an adult and the upper a young child. The lower, adult burial consisted of a partially articulated skeleton facing approximately south in a crouched position, with the head directed to the west-north-west (Figure 19). A mass of disarticulated bone, including an intact adult radius, was located around the northern and eastern sides of the chamber, suggesting significant post depositional disturbance in this area. The presence of a ‘tunnel’ from the north side of the burial chamber to the mound surface, coupled with the presence of bone fragments on the surface of the mound in this area, suggests that a burrowing animal has at some time in the past made its home in the northern part the burial chamber, thereby disturbing the contents. The portion of the lower burial lying in the southern part of the burial chamber was undisturbed. The upper burial, of a young child, was represented by a semi-articulated proximal tibia and pelvis fragment (which mirrored the position of these skeletal elements in the lower burial, thereby providing tentative evidence for a similar burial position) and numerous badly preserved, disarticulated bone fragments. Only a few centimetres of sediment separated the two burials. Indeed in some areas no clear separation between the two burials could be discerned, suggesting that the upper body was placed in the chamber at the same time as, or relatively soon after, the lower internment.
Sieving of the material above and around the burials did not reveal any artefacts, and the grave fill was excavated from around the lower burial (which was left in situ on its ‘bed’ of clean sand 8cm thick). A number of in situ grave goods were recovered during excavation of the lower burial, namely a copper earring and two shells (Figure 21). The shells were situated around the neck of the skeleton, suggesting that they formed part of a pendant or necklace. A third, smaller shell was recovered during sieving of the material directly above the burial. What appeared to be red ochre staining was noted around the head, neck and feet areas of the lower burial, being most noticeable around the lower leg bones and feet. Photographs of the shells were sent to the Natural History Museum in London for identification. The two found around the neck of the primary burial were identified as representing one or possibly two varieties of Conus, found along the West African coast and in the Cape Verde islands. In both cases the apex of the shell had been worked and the spiral structure removed, leaving a hole through which twine may have been threaded. The most likely identification for the third shell, recovered from the burial chamber fill, is Persicula cingulata, found along the West African coast. This shell had been pierced through the body to form two holes, and the central structure was missing, again suggesting that it may have formed part of a pendant. A well-preserved grinding stone, broken at one end, was exposed on the bedrock immediately adjacent to the external wall, directly south of the annex (Figure 20).
3.6. Contextual information from lithics and ceramics finds
3.6.1. Description of lithic artefacts
The lithic artefact surface finds from the Tifariti site were assessed from photographic records after the end of the field season. A variety of artefacts is presented in Figure 22, in which artefacts are numbered for the purpose of cross-referencing with the text below. A range of stone tool manufacturing techniques is represented, largely indicative of the middle and end phases of the tool-making process; and involving a variety of raw materials. The raw materials represented in the assemblage include fine-grained igneous rock such as basalt (artefacts 2, 35, 37, 38), quartz (artefact 4), quartzite (artefact 18), and a variety of cryptocrystalline or fine-grained, siliceous rocks including jasper (artefacts 10-17, 19, 21, 22, 24, 25, 30), translucent chert (artefacts 15, 20, 26, 28, 32), and grey flint (artefact 9). A range of preservation is observed; while some artefacts exhibit sharply defined flake scars (e.g. artefact 36), others are highly abraded (e.g. artefact 37). Artefact 29 shows a high degree of surface lustre across the dorsal surface which suggests prolonged exposure to wind and sand blasting. Overall the assemblage is in relatively sharp condition, though there is evidence of differential weathering due to variable degrees of exposure which may correlate to some degree with artefact age. The artefacts are likely to represent a number of successive stone tool making and/or using episodes spanning many millennia.
A variety of manufacturing techniques is represented, including
- examples of direct percussion, in which flakes of raw material have been detached artificially to create artefacts and subsequently to further shape flakes into tools
- stone grinding and polishing (artefacts 1 and 2)
- small, recurrent, bipolar blade core preparation (artefact 11)
- small, unipolar blade core preparation (artefact 17).
Blades, or fragments of blades, comprise at least 30 % of the entire assemblage of 93 recorded artefacts, and were clearly an important component of the industry during at least one phase of activity at the site.
The presence of highly reduced, diminutive cores suggest that small blades / bladelets were at least acceptable products, if not the desired outcome, of the stone working process. While further assessment in the field is necessary to determine whether blades were struck from prismatic blade cores or Levallois blade cores, the dorsal flake pattern on artefact 9 conforms to that of Levallois points. Flakes and blades appear to have been shaped into a variety of tool forms, including endscrapers (artefact 5), denticulates or ‘toothed edges’ (artefacts 6 and 23), points (artefact 30), and burins (artefact 34). Artefact 8 appears to have a retouched, tanged base, suggesting that it might have been hafted for use as a spear point.
Only 16 % of the artefacts in the assemblage have a maximum dimension exceeding 50mm. Approximately 22% of the assemblage appears to retain some trace of the natural, un-flaked, outer surface of the raw material units used (e.g. artefacts 3, 7 and 24). Considering the variety of raw materials (and therefore the number of individual nodules) represented, a higher number of flakes with remnants of the outer surface of nodules would be expected if the assemblage represented the entire reduction process. Moreover, as previously noted, some 30% of the assemblage comprises blades or blade fragments. The diagnostic parallel, dorsal flake scars on the blade products demonstrate that these items were detached from advanced phases of raw material reduction, rather than from the preliminary ‘shaping-out’ stage (which leads to more diverse dorsal flake scar orientations).
3.6.2. Interpretation of lithic artefacts
The lithics assemblage indicates that artefacts were brought to the site as shaped-out cores or individual flake and blade tools. The raw materials used were mostly selected for fine-grain quality and were presumably imported from diverse geological sources, or from a geologically diverse secondary deposit. It is possible that the range of raw materials attest to curation of lithics over considerable distances or exchange of these (valued) rock types between neighbouring social groups.
The apparent ground stone artefacts are not made of the cryptocrystalline cherts so common in the assemblage but coarser, harder rocks such as basalt. These artefacts appear not to be completely ground and polished since flake scars are still visible. Perhaps these arefacts were unfinished, as their asymmetry might suggest, or were re-flaked post-depositionally. Ground stone axes are described from Neolithic contexts in Mali such as the site of Dialaka, Kayes near the Senegal border (Konaté, 2000:38, Figure 15, b).
The tanged point (artefact 8) and the Levallois point (artefact 9) are reminiscent of the northwest African Aterian industry, which dates to between 90,000 and 20,000 years ago. Good comparative material is provided by the Aterian sites of the Zouerate region of western Mauritania, near the Western Sahara border (Ould Mohammed Naffé et al., 2000, p137, citing Pasty, 1998) However, the blade and blade fragments, which are such a common component of the assemblage, tend to be less than 50 mm in length and frequently less than 30 mm. Such small items, verging on ‘microliths’, are more characteristic of the Late Stone Age or Neolithic industries of northwest Africa. Of particular note in this regard are finds such as artefact 34, which appear to be small, obliquely retouched bladelets and, depending upon a closer examination of the shaping, may have affinities with such microlithic assemblages as that of the Neolithic site of Gobnangou in Burkino Faso (Millogo and Kote, 2000:23) and the Mékrou Valley, Niger (Idé Oumarou, 2000, p199). What the Tifariti assemblage lacks, however, are any signs of the characteristic Neolithic, tanged, arrowheads shaped by invasive, parallel retouch on thin blades that are associated with tumuli sites in the d’Akjoujt region of nearby, western Mauritania (Vernet, et al. 2000, p38). Such carefully shaped artefacts may have had very different patterns of discard to the other artefacts represented in here and if not would certainly be the first of any surface assemblage to be looted. Finely worked Neolithic arrowheads have been collected in the Free Zone by local people for sale to visitors to the refugee camps, and have also been noted by the authors at other sites (e.g. Irghrayghra: site IR1 in Figure 1b) during previous visits.
3.6.3. Ceramics
Only a very small number of sherds comprise the assemblage collected by field walking and excavation in 2005 at Tifariti. This contrasts with the March 2005 survey in the Southern Sector in which relatively large quantities of sherds were recorded. Detailed analyses of these collections will be undertaken in future seasons and a full report published in a later interim report. The purpose of this section is to comment briefly on the sherds found in association with the burial in Monument 1.
Sherds found in the burial chamber represent two different ware types, indicating the presence of at least two different vessels. The first is a body sherd from a basket impressed jar (Figure 16). This is a hand-made ware in a coarse, gritty fabric containing medium to high quantities of poorly sorted, angular mineral temper. This ware type is found ubiquitously amongst pottery collected in the surface surveys and is frequently incompletely fired to a dark red/brown with a grey core. It is distinctive for the very clear reed/straw mat or basket impressions on what would have been the outer surface of the vessel where the clay was pressed against or into the mat or basket. This technique of manufacture occurs throughout the Sahara, Egypt and south-west Asia prior to the introduction of wheel-made pottery and suggests that straw containers were widely used as moulds or shaping devices by early potters.
The second ware type is again hand-made in a light orange buff fabric, with medium quantities of small well-sorted mineral inclusions. The outer surface was wet smoothed and then incised with a series of parallel lines forming a range of linear patterns, including herringbone motif (Figure 16). A number of body sherds and a rim sherd probably represent what remains of a once complete or almost complete jar.
The most interesting aspect of the pottery and other objects collected from Monument 1 is the overall paucity of the finds. That sherds and other items found in association with the body suggest that the individual was buried with a range of personal items, including a necklace (perhaps represented by the carnelian bead), a cutting or thrusting implement (represented by the metal point) and two or more containers (represented by the pottery sherds). The fact that very few beads or pottery sherds were found in the chamber suggest that the burial was disturbed and that this disturbance led to not only the displacement of the accompanying grave goods but more interestingly the removal from the chamber of a significant proportion of the fragments. As the burial appeared to be intact this phenomenon must be a characteristic of the particular types of taphonomic activity that occur in the Sahara and which require fuller study in future seasons of excavation.
The few sherds collected through field walking in the general vicinity of the burial mounds and around the campsite are well-abraded examples of so-called "basket ware" and another as yet unidentified coarse ware.
4. Discussion and conclusions
This paper has provided a very brief overview of the funerary archaeology of Western Sahara, focusing on the Southern Sector of the Polisario-controlled Free Zone, and on a small area in the vicinity of Tifariti, in the Northern Sector. Interpreted alongside the findings of earlier work in the Northern Sector (Brooks et al., 2003), these results allow us to identify emerging patterns in the spatial distribution of funerary monuments.
The areas surveyed reveal a diverse and complex assortment funerary and associated monuments, reflecting both typical Saharan elements and features characteristic of the greater western Saharan region. In addition, survey work in both the Northern and Southern Sectors has identified unusual monuments that are likely to be the result of local innovation. While much more work is required before chronologies of human occupation, monument typology, and environmental change can be established, the combination of different monument styles does allow us to propose a tentative chronological model based on considerations of human-environment interaction, which is outlined below. This model should be viewed as a hypothesis to be tested.
The inland regions of Western Sahara today receive considerably more rainfall than do similar latitudes in the Sahara to the east, principally as a result of the region’s proximity to the Atlantic. As a consequence of this comparative humidity, many inland regions of Western Sahara are relatively well vegetated, and some locations area characterised by open acacia woodland. These conditions may be contrasted with the hyper-arid Saharan regions to the east, where significant vegetation cover is essentially absent outside the oasis and highland areas. These climatic considerations are relevant to archaeological studies, as the building of monumental funerary structures in the Sahara has been associated with increasing aridity, resulting in intensified competition for resources and increased territoriality of pastoral groups (di Lernia et al., 2002; Brooks, 2006). While the period around 5000 yrs BP may be viewed as the “hinge” between the humid early-middle Holocene and the arid middle-late Holocene, the process of desiccation was neither smooth nor spatially homogenous, with aridity setting in at different times in different locations, and with certain areas acting as refugia for remnant faunal, floral and human populations (di Lernia et al., 2002; Brooks, 2006). Considering the comparative humidity of the regions described in this paper, it is a reasonable, although as yet unproven, proposition that much of the region covered by today’s Western Sahara acted as a refuge for populations fleeing aridity to the east. Given the combination of funerary styles recorded in the Free Zone, we might postulate that (i) monuments characterised by the more familiar Saharan styles were constructed either at a time when there was considerable regional connectivity in the greater Saharan (and particularly western Saharan) region, or by incoming groups soon after the onset of regional aridity, and (ii) the more unusual monuments were constructed later, by populations isolated from the greater Saharan region by the expansion of the desert, who developed their own independent construction styles.
This model is highly speculative, but provides a useful framework within which research questions regarding linked cultural and environmental change may be formulated. The testing of this hypothesis will require the dating (through assessment of surface materials and excavation) of a number of monuments representing the proposed construction phases. As a first step in exploratory excavation, two conical tumuli near Tifariti were chosen due to their accessibility and relatively small size. Given the ubiquity of the tumulus form, it is likely that such monuments were constructed throughout the period characterised by monument building in this area. Future work will include the excavation of more complex monuments throughout the Free Zone in order to test the chronological hypothesis outlined above.
The excavations conducted at Tifariti in November 2005 have already yielded information useful for the development of cultural chronologies, with the presence of the metal point in Monument 1 and the copper earring in Monument 2 indicating that these monuments are unlikely to predate the 3rd millennium BP (1st millennium BCE), based on our current understanding of the appearance of metalworking in the Sahara (Childs and Killick, 1993). The site of Akjout in northern Mauritania is a possible source for both iron and copper. Copper mines and a small smelting site at Akjout date from the 9th to the 3rd century BCE, and there are suggestions of contact with Punic North Africa (Childs and Killick (1993) citing Lambert, 1983). Iron appears around 300-0 yrs BCE at the same site (Holl, 1997). Contact with the coast, direct or through trading networks, is also indicated by the presence of marine shells in Monument 2, while the presence of the carnelian bead also suggests that the builders of these monuments participated in active and wide-ranging Saharan trade networks focused on sources of carnelian in the western and central Sahara around this time (Mattingly 2005, p358). However, the origins of metalworking, and the nature of prehistoric trade networks, remain poorly understood in North Africa, and the possibility that the excavated monuments were constructed earlier than the 3rd millennium BP cannot be discounted.
Regardless of whether the hypothesis relating to the evolution of funerary monument styles in association with environmental change, migration and population fragmentation is correct, the number, variety and distribution of monuments suggest that monument building continued over a long period. While monuments do exist in isolation or in small groups, certain areas were clearly favoured as burial grounds, and these contain a wide variety of monument types. While the densities of worked stone and ceramic fragments vary from site to site, it appears from the preliminary survey work that in most cases funerary sites were separated from occupation sites. Exceptions to this general rule are apparent, for example at Zoug, where a group of funerary monuments is situated among abundant ceramic fragments and grinding stones. While most of the monuments at Lajuad occur in locations where other archaeological materials are scarce, a single crescent monument at the base of the hill containing the painted rock shelter is surrounded by a dense cover of decorated and undecorated ceramic fragments.
While similarities exist between the Northern and Southern Sectors, systematic differences are also apparent. Crescent monuments are much more abundant in the Southern Sector, as are grinding stones and hollows, and ceramics. Decorated ceramics are extremely abundant in the Southern Sector, and include examples exhibiting dottet wavy line patterns (Figure 23).
The rock art of the southern sector, to date recorded in two shelters at Lajuad and at an open air site at Zoug, is quite distinct in character from that of the Northern Sector. In the Northern Sector, at Sluguilla, Rekeiz and Bou Dheir, rock paintings and engravings depict the typical Saharan combination of large fauna, cattle, and human figures in often elaborate, naturalistic styles. Paintings and engravings in the Southern Sector are more schematic in nature. As well as engraved zoomorphs and anthropomorphs at Lajuad, rock art in the southern sector includes likely “proto-Berber” paintings (also at Lajuad: Figure 24) and abstract geometric forms including spiral forms and concentric circles at Zoug (Figure 25).
These differences between the Northern and Southern Sectors raise the possibility that the present-day Western Sahara straddles the boundary between at least two different prehistoric cultures, although the observed differences in the archaeological record may well represent cultural transitions or discontinuities in time as well as in space.
Future work will focus on the development of cultural sequences from surface materials and excavation at Tifariti and Lajuad, and on the identification of palaeoenvironmental indicators from which chronologies of past environmental change may be constructed. Reconnaissance work will continue in both the Northern and Southern Sectors, and will be accompanied by more detailed surveys of funerary complexes and rock art sites, and assessment of surface finds.
Acknowledgements
The authors are grateful to the Office of the President and the Ministry of Culture of the Sahrawi Arab Democratic Republic for providing the logistical support which made these seasons of fieldwork possible. Special thanks are extended to Bachir Mehdi Bhaua, whose tireless efforts in organising and coordinating the logistical aspects of the work ensured the smooth running of all three field seasons. We would like to thank all the members of our Sahrawi support teams for their hard work and enthusiasm, the Polisario staff at Rabuni, Tifariti, Dugej, Mijek, Mheires and Zoug, and all the people of Western Sahara who have shown us so much hospitality. The November 2005 field season was made possible by financial support from Ophir Oil and Gas Ltd, with contributions in the form of individual travel expenses from the Leverhulme Trust and the University of Munich (thanks to Joris Peters). Margaret Raffin funded her own participation in the November 2005 field season. Thanks are also due to Michele Raychaudhuri for her participation and support, to Pierre Mein for the identification of the small mammal remains, to the Natural History Museum in London for the identification of the shells, and to Savino di Lernia for general assistance. Figures 1 and 9 were produced using Google Earth.
References
BROOKS N,. 2005. Cultural heritage and conflict: The threatened archaeology of Western Sahara. The Journal of North African Studies 304: 413-439.
BROOKS N., 2006. Cultural responses to aridity and increased social complexity in the Middle Holocene. Quaternary International (in press).
BROOKS N., DI LERNIA S., DRAKE N., RAFFIN M. AND SAVAGE T., 2003. The geoarchaeology of Western Sahara: Preliminary results of the first Anglo-Italian expedition in the “free zone”. Sahara 14: 63–80.
BROOKS N., DI LERNIA S., DRAKE N. CHIAPELLO I., LEGRAND M., MOULIN C. AND PROSPERO J. 2005. The environment-society nexus in the Sahara from prehistoric times to the present day. The Journal of North African Studies 304, 253-292.
CHILDS S. T. AND KILLICK D., 1993, Indigenous African metallurgy: nature and culture. Annual Reviews of Anthropology 22, 317-337.
DI LERNIA S. AND MANZI G. (Eds.), 2002. Sand, Stones, and Bones. The archaeology of death in the Wadi Tanezzuft Valley (5000-2000 bp). The Archaeology of Libyan Sahara, Volume I. AZA Monographs 3, All’Insegna del Giglio, Firenze.
DI LERNIA S., MANZI G., AND MERIGHI, F., 2002. Cultural variability and human trajectories in later prehistory of the Wadi Tenezzuft, In di Lernia, S., Manzi, G. (Eds.), Sand, Stones and Bones: The Archaeology of Death in the Wadi Tannezzuft Valley (5000-2000 BP), pp. 281-302. The Archaeology of Libyan Sahara, Volume I. AZA Monographs 3, All’Insegna del Giglio, Firenze.
GAUTHIER Y. AND GAUTHIER C., 2003. Chronologie relative de trois types de monuments de l’Immidir: monuments à antennes en “V”, goulets et monuments en “trou de serrure”. Sahara 14: 155–161.
GAUTHIER Y., GAUTHIER C., NÖTHER W. and LLUCH P., 1997. Monuments de l’Immidir (Algerie). Sahara 9: 143–148.
HOLL A., 1997 – Metallurgy, iron technology and African Late Holocene societies. In Klein-Arendt, R. (ed), Traditionelles Eisenhandwerk in Afrika, Colloquium Africanum 3, Köln, 13-54.
IDÉ OUMAROU A., 2000, Préhistoire dans la vallée de la Mékrou (Niger méridional). Niamey : Institut de Recherches en Sciences humaines ; CRIAA.
KONATÉ D., 2000, Éléments d’archéologie ouest-Africaine II: Mali. Nouakchott: CRIAA.
LAMBERT N., 1983, Nouvelle contribution à l’étude du Chalcolithique de Mauritanie, pp 63-88.
LIHOREAU M., 1993, Djorf Torba, nécropole saharienne antéislamique. Karthala (éd.), Paris, 135 p.
MATTINGLY D. J., 2003. Synthesis of human activities in Fazzan. In: Mattingly, D. J. (Ed.), The Archaeology of Fazzan: Volume 1, Synthesis, Department of Antiquities, Tripoli, and Society for Libyan Studies, London, pp. 327-373.
MERCER J., 1976, Spanish Sahara. Allen and Unwin, London, 248 pp.
MILBURN M., 2005. More enigmatic stone structures of the north-western Sahara. Independent Archaeology 52: 6–8.
MILLOGO A.K. and KOTE L., 2000, Archéologie du Burkina Faso in Vernet R. (ed) L’archéologie en Afrique de l’Ouest: Sahara et Sahel. Nouakchott : CRIAA pp. 5-70
OULD MOHAMMED NAFFÉ B., VERNET R.and OULD KATTAR M., 2000, Archéologie de la Mauritanie in Vernet R. (ed) L’archéologie en Afrique de l’Ouest: Sahara et Sahel. Nouakchott : CRIAA pp. 129-204
PETIT-MAIRE N., 1979, Le Sahara Atlantique à l'Holocène. Peuplement et écologie. Mémoires du CRAPE, Alger, 340 p.
SHELLEY T., 2004. Endgame in the Western Sahara: What Future for Africa’s Last Colony? Zed Books, London.
SIVILI S., 2002, A historical background: mortuary archaeology in the Sahara between colonialism and modern research. In: di Lernia, S., Manzi, G. (Eds.), Sand, Stones and Bones: The Archaeology of Death in the Wadi Tannezzuft Valley (5000-2000 BP), Centro Interuniversitario di Ricerca per le Civiltà e l’Ambiente del Sahara Antico e Delle Zone Aride, Università Degli Studi di Roma and Department of Antiquities, Libya, pp. 17-24.
VERNET R., OULD MOHAMMED NAFFÉ B. and OULD KATTAR M., 2000, Éléments d’archéologie ouest-Africaine III : Mauritanie. Nouakchott : CRIAA.
Figure 1a. Western Sahara, showing the Occupied Territories, the Free Zone, and the approximate location of the “Berm”, the Moroccan defensive wall. Areas of particular archaeological interest, surveyed in 2002 and 2005, are show as green circles.
Figure 1b. Northern Sector in more detail, with major archaeological sites recorded to date (BD: Bou Dheir; DR: Wadi Dirt; ER: Wadi Erni; JM: El-Lous Lajaram; RK: Rekeiz; RL: Rekeiz Lahmar; SL: Sluguilla; TF: Tifariti; TR: Wadi Ternit; WT: Wadi Weyn Tergit).
Figure 2. Schematic typology of monuments recorded in the Free Zone.
Figure 3. Monuments recorded in the Free Zone reflecting types found across the Sahara. From top: conical tumulus (Tifariti, site TF1); tumulus with platform (Zoug); corbeille (Tifariti, site TF1); bazina ((Tifariti, site TF4). Note the two larger stones on the far side of the corbeille, and the marker stone in the foreground, defining a north-south direction.
Figure 4. Monuments recorded in the Free Zone reflecting types found in the western regions of the Sahara. From top: enclosure monument (Tifariti, site TF1); stone circles associated with an enclosure (Tifariti, site TF1); tumulus on arced platform (Tifariti, site TF1); crescent monument (Lajuad).
Figure 5. From top: Funerary complex at Lajuad incorporating two “ridge” monuments. Note high density of quartzite surface material; atypical ridge monument in funerary complex; same monument showing sinuous ridge or arm; linear ridge monument incorporated in perimeter of funerary complex; perimeter of funerary complex showing low mound and small linear monument.
Figure 6. Complex monuments near Tifariti (site TF1).
Figure 7. Standing stones recorded in the vicinity of Azaig Bedrag. Top: standing stone in an area containing numerous funerary monuments, described in the text. Bottom: apparently isolated standing stone recorded in transit from Azaig Bedrag to Mijek.
Figure 8. Top: Group of small standing stones approximately aligned with annex of bazina-type monument at Tifariti (site TF1). Bottom: Annex comprising two elongated boulders and granite grinding stone (Tifariti, site TF1).
Figure 10. Wadi Tifariti, viewed from excavation site, with the group of standing stones (site TF00) visible in the clear area in centre of the photograph.
Figure 11. Group of 65 standing stones (site TF1.0) near the excavation site.
Figure 12. Schematic representation of the giant goulet complex near the excavation site. A: Tumulus in area paved with dense concentration of stones. B: Poorly defined perimeter due to removal of stones. C: Parallel lines ~1 m apart, with some reuse and disturbance of stones. D & E: Stone circles. F: Area of sub-circular concentrations of stones representing possible burials. G: Possible marker stone defining end of monument complex.
Figure 13. Giant goulet complex at Tifariti. From top: View along the east-west oriented corridor forming the axis of the complex, looking west towards flat-topped hill; view from west incorporating circular structures slightly offset from main axis; one of many concentrations of stones west of the circular structures; possible marker stone aligned with corridor, at easternmost end of complex.
Figure 14. Monument 1 (Tifariti excavations). From top: monument after clearing of perimeter, before excavation; during excavation, showing large perimeter stones and burial chamber structure; annex on eastern side of monument; burial chamber photographed after removal of skeletal remains.
Figure 15. Monument 1: skeletal remains in situ.
Figure 16. Monument 1: grave goods. From left: ostrich eggshell beads, carnelian bead and metal object recovered from fill around and above skeleton; single ceramic fragment recovered from fill above skeleton, showing basket impresssions; four ceramic fragments of a different type recovered from fill underneath skeleton.
Figure 17. Left: Monument 2 (Tifariti excavations) prior to excavation. Right: Monument during excavation, showing large cap stones covering burial chamber.
Figure 19. Monument 2: skeletal remains in situ in.
Figure 20. Monument 2: fragment of grinding stone in situ at base of internal wall.
Figure 21. Monument 2: grave goods.
Figure 22. Sample of worked stone artefacts recovered from the vicinity of the excavations at Tifariti. For the purposes of references in the main text, artefacts are numbered consecutively from the top left.
Figure 23. Examples of decorated ceramics from Lajuad (top left, top right, and bottom right), and sherd from Lektaytayghra exhibiting dotted wavy line pattern (bottom).
Figure 24. Paintings at Lajuad resembling Tifinagh script.
Figure 25. Example of abstract engraved forms at the open air rock art site at Zoug.