Aridity, Cattle, and Rites. Social responses to rapid environmental changes in the Saharan Pastoral societies, 6500-5000 yr BP
by
Di Lernia, Savino
Italian Archaeological Mission in the Acacus and Messak, Libyan Sahara, Department of Scienze Storiche, Archeologiche, Antropologiche dell'Antichità, Faculty of Human Sciences, University of Rome "La Sapienza"

Introduction Although this paper deals with an individual case study – the emergence of a ‘cattle-cult’ within Holocene pastoral societies in Africa – it is worth outlining some specific theoretical and methodological aspects, which should in my view form an indispensable basis for common discussion in this meeting.

The term ‘catastrophe’ (katastrophé, from katà, down, and stréphein, turn) originally used in classical Greek tragedy to refer to the final outcome of the plot, usually a tragic one, is commonly used today to define a destructive or dangerous event. However, the term does not convey either the duration or the rapidity of the event itself. For example, the modern concentration of resources in Western economies is likely to have catastrophic effects on large parts of the world, but over a medium-long period and probably without heavily perceptible effects in the short term (for a discussion, and relations with political aspects and New Ecology see Van Buren 2001).

Viewed from this perspective, catastrophes should be considered a process, and not an event as often implicitly assumed (though the cause of the catastrophe might be an event) whose evaluation represents above all an ethic (and not emic) judgement of the final outcome: is it dangerous/negative, or not?

This appears true particularly when approaching the Holocene prehistory of the Sahara. The climatic events leading to arid conditions – severe arbiters of human cultural trajectories – short, rapid, abrupt, or however you wish to define them, lasted for centuries, and should therefore be evaluated according to a different time-scale, related to human life on a generational scale.

In concrete terms, what do we need to confront this problem? First, a common agreement on theoretical issues regarding human responses. Second, accuracy in reconstructing both climate cycles and human cultural developments.

As far as the first point is concerned, a marked and profound disagreement has recently arisen within the scientific community. We owe to Fred Wendorf and Romuald Schild (2003) the merit of having clearly (and provocatively) set the agenda on the very nature of the climate/human interface. Reviewing the book ‘Droughts, Food and Culture’ edited by Hassan (ed. 2002), outcome of a ESF workshop held in London five years ago (1998), Wendorf and Schild (2003: 132) wrote: “we believe that many of the technical and economic innovations that developed in the Sahara during the Holocene did not occur in intervals of aridity, but during periods of increased moisture, and our radiocarbon chronology (which differs significantly in many details from that of Hassan) supports that contention. We believe, for example, that the initial spread of cattle into the Egyptian Sahara (ca. 9.2 kyr BP), and their subsequent movement, with sheep/goat, into the Central Sahara (ca. 7.0-6.6 kyr BP) took place during relatively wet periods when water and grazing would be available”.

This point is of absolute importance. Other scholars, including myself, believe the opposite. Aridity, like several other ‘push conditions’, forced humans to move, for numerous reasons, which we will not discuss here; in my opinion the archaeological evidence clearly supports this position. What is important, in my view, is to look for different sources of information, in order to correctly contextualise this issue. In this direction, the ethnography and ethnoarchaeology of African pastoral societies support the idea of movement as an obliged condition, and not as a free and deliberate choice (e.g., Hogg 1990; Iro 2000). It is worth remembering Hassan (2002a: 3) when he describes how environmental change is perceived and then implemented within a social system. Thus, ethnography, or even ethnoclimatology, as I proposed years ago in London, represents a challenge for archaeologists and geologists working on this topic in the future, in the attempt to increase our capacity to decode human behaviours.

A second point is accuracy in reconstructing the timing of climatic and historical processes and events: radiometric determinations appear to be a critical topic. Notwithstanding many efforts to strengthen this field of research, we still have some serious problems. ‘One date, no date’ – someone said years ago. Yet we still rely on single radiocarbon measurements to place a phenomenon, and after several C-14 congresses, we still use conventional determinations (which are cheaper!), rather than processing our samples using AMS facilities (which are too expensive!). As a result our measurements still show hundreds of years of error, and it is all too easy to use the central date as if it were absolutely accurate: this is quite simply not the case. So, taking into consideration this (desolate) picture, I deduce that we, as a scientific community, still do not agree even on the basic phenomena which intimately affected human responses (did people move to escape aridity or to follow moisture?), and that we still carry out our research using antiquated tools (a few radiocarbon dates, mostly conventional, with sigma as large as ± 300 years… to draw an analogy, would we go to a doctor who still uses his old-fashioned hammer rather than a CAT scan?) Research on this topic still has far to go, and more systematic information must be collected, the better to redefine these crucial aspects. I wish to stress two points as this paper moves on from these brief considerations: first, ‘catastrophes’, especially rapid and dramatic climatic changes, should be considered a process; second, humans, in the absence of significant demographic growth, appear to avoid movement, or, at the very least, to dislike moving.

The problem: Middle Holocene cattle burials in the Sahara? The recent evidence from Nabta Playa (as a last, Wendorf, Schild, and associates 2001) has opened a new era in the archaeological research in the Sahara. After decades of studies almost exclusively devoted to the analysis of human adaptations to environmental changes, the existence of ceremonial sites featuring stone alignments, astronomical structures, stone covered tumuli dating back to Neolithic times really demands a different approach in our explanatory perspective.

Shortly, the entire stuff gathered at Nabta has been interpreted as evidence of a process of increasing social complexity, having as final outcome the existence of ranked society within Late Neolithic pastoral communities. Within this context, according to Wendorf and Associates, cattle played a crucial role because they “symbolise status and power”. In fact, this animal surely hold a special role in the African pastoral societies, and I wonder if is it appropriate to speak, as many colleagues do, of an ‘African cattle cult’, quoting and reusing the concept put forth by Heroswiktz (1926). I think it is a convincing and fascinating hypothesis, which probably may be strengthened by other data set, even faraway from the Eastern Sahara.

In fact, I will review this matter, adding evidence from the central Sahara, namely from cases studied in Libya and Niger. My first attempt will be to demonstrate the existing relations between the abrupt dry crisis at the end of the 7th millennium BP and the emergence of the first cattle burials in many – and distant – areas of the Sahara. Secondly, I will try to track times and trajectories of diffusion of this particular ritual. Eventually, I will draw attention to the evolution of the stone architecture connected to the practice of animal offering. Having as case study the central range of the Sahara, I will show you as stone tumuli changed their function (i.e., from animal to human burials), and how this different function change may be used as an effective tool to detect important social and ritual change within mid-Holocene pastoral societies, mostly related to rapid changes in the environmental conditions. Proxies will be drawn from settlements, rock art sites, and burials.

Background: sites with cattle remains in stone monuments The archaeological evidence gathered at Nabta is really spectacular. The cattle burials found and excavated there represent the greatest concentration in prehistoric North Africa. This phenomenon, seen as part of a complex ritual scenario related to the existence of a ceremonial centre in the Late Neolithic, should be considered prehistoric evidence of the so-called ‘African Cattle Complex’. Of interest for us are four excavated tumuli contained cattle remains, which are of two types.

On the western edge of the largest wadi to the north of Nabta we have the first of two differing types of stone-covered tumuli marking the burial sites of cattle. Seven out of the nine tumuli examined have been excavated. At E-94-1n, at the northern end of the Late Neolithic ceremonial complex, the stones covered the remains of an articulated young cow in a clay-lined chamber. Radiocarbon determination on a piece of wood gave a date of some 6400 years BP. The badly preserved cow had spine oriented north south, its head looking south.

The second type of tumulus consists of disarticulated Bos bones scattered between unshaped rocks. Sites E-94-1s, E-96-4, E-97-4, E-97-6 and E-97-16 are associated with the remains of 3, 4 (2 sub-adult, 2 young adults), 2 (1 juvenile, 1 sub-adult), 1 and 1 (sub-adult) cattle respectively. No particular body part was deliberately selected for deposition. Even if quite spectacular, the Nabta evidence is not an isolated feature. Especially in Niger, we have many evidences, reported from late 1970s to early 1990s.

Already Henry Lhote (1976, quoted in Paris 2000) reported of possible ‘cattle burial’, represented by remains of articulated cattle in the Talak area (western Air, Niger). These skeletons were located at the edges of settlements, sometimes apparently associated (at least in spatial terms) with human burials. According to Lhote, the depositional features of these animals should have to be referred to as epizootic deaths. The animals would have been immediately buried by heavy desert storms (thus preventing any possible scavenging and/or erosion), explaining why they have been found still in anatomical connection. Worthy of note, François Paris (2000) simply notes that all these animals had their head east-oriented, therefore implicating some ritual activities. Nevertheless, Paris reminds us that the presence itself of Bos burial cannot be interpreted as direct expression of ritual linked to animal (the cattle cult).

At In Tuduf, 10 km n-w di Chin Tafidet (again in Niger), Columeau (quoted in Paris 2000) surveyed a site with remains of probably Bos indicus, dated to 3500 BP. Also in this case, cattle were killed, butchered and then buried with a complex ritual. Animal lied on their right, east west oriented and with the forequarter almost always east-oriented, much similar to the humans there buried.

French archaeologists excavated three small tumuli (Site 1) in the Adrar Bous. North of Adrar n Kifi, at the edges of the Tenerean site S1, Paris and associates excavated in 1981 a small tumulus (1.4x0.4 m) containing the right forequarter of a cow. Eight years later (1989) the same team, in the same site, dug another tumulus (2 m large and 0.5 in height), discovering burnt bones of cattle, probably a whole disarticulated skeleton. Bone collagen gave 6200 ± 250 BP (Pa 753). A similar structure excavated by Roset on 1985 yielded a date of 6325 ± 300 BP (Pa 330). According to Paris (2000: 121), “it is not possible to say that they are ritual structures but they are intentional as the bones are placed in a pit closed by stones”.

A recent review of Applegate et al. (2001) shows the available literature on the argument. The most striking evidence is that initial forms of stone structures with livestock remains – be they articulated or scattered – should all backdate around 6000 years BP.

New data from the Libyan Sahara Recent research in the Messak Settafet plateau (Fig. 1) added important pieces to this picture. Some unplundered stone tumuli were excavated. In particular, three out of six yielded livestock remains, and to may be interpreted as ‘cattle’ burial or, better, ‘stone monuments build for ceremonial purposes’.

SITE 301 Site 301 (N 26°03’48. 2’’ E 12°04’56. 6’’, 815 m asl), is located on the right bank of the Wadi Tin Einessnis (Fig. 2). It is a stone tumulus inserted in a very complex scenario: apart many other stone tumuli, more than 300 hundred trapping stones punctuate the landscape. This stone, flat conical tumulus was particularly well preserved, and showed some covering boulders bearing schematical/enigmatic rock engravings. In the immediate surroundings, other stone structures were probably related to the ritual function of the monument. The tumulus is some 4.5 m in diameter and presents two small semicircular annexes.

The monument yielded lithic artefacts, one potsherd, and many faunal remains. Four hearths were present at its base, very rich in charcoal and faunal remains (Fig. 3).

Two big stone axes were also found, possibly related to a ritual deposition: according to wear traces study, the artefacts were never used (Fig. 4). The single potsherd found shows a rocker plain edge decoration, echo of a Middle Pastoral tradition: its location on the very end of the structure, again, appears to be ‘ritual’. Radiocarbon date on charcoal is consistent, giving 6080 ± 80 years BP (spot C; GX-28453).

The faunal assemblage, still under study by Francesca Alhaique (nd), consists of more than 5000 fragments (88% turned out to be unidentifiable). As a whole, we have the remains of at least 2 cows (one practically whole) and 4 domestic ovicaprines. All the individuals are rather young. The hearths show a differentiated accumulation of animal remains. In particular, Spot B yielded only ovicaprine bones, whereas Spot C (dated by radiocarbon) cattle remains. Spots A and D show a mixed composition. Modifications on bones are difficult to recognise: most of the remains are burned and poorly preserved: chop-, and other scraping marks may be interpreted as meat removal. According to the archaeozoologist, and to myself, animal bones were put in the hearths in the same moment, underlying the important quantity of animals slaughtered, and therefore the relevance of the ritual there performed. As recently reminded by Achilles Gautier, we should image that many (dozens, probably hundreds) persons took part in the ceremony, given the impressive quantity of available meat.

SITE 556, 557 Few kilometres far from site 301, we mapped a complex area in the southern part of the Wadi Tin Einessnis. The area is some 400x300 m in size, and it includes the wadi bank, a small valley bordered by two minor tributaries, and two large quartzite outcrops. Many other structures punctuated this area. The largest structures are located in the highest part of the valley, far from the wadi bank. On the bases of archaeological material (pottery mostly shows rocker plain edge zigzags), the Middle Pastoral occupation is rather frequent.

The most interesting site is a stone structure, circular in plan, some 2.8x2.3 m, composed of a ring of huge slabs set into the ground in an oblique position (Fig. 5).

It encircles a stone platform made with small and medium-sized globular and prismatic stones, varnished on the upper part (N 25° 59’ 37.7’’ E 12° 11’ 07.0’’, 756 m asl). Two standing stones were present, close to the centre of the structure, one of which was set over the bedrock with several wedging stones. A semicircular annex was present on the northwestern side of the stone ring. Two engraved slabs were recovered, one in the stone ring (with ovaloid representation), and another one some 1.6 m SE of the monument (with two cattle figures of naturalistic style). Several potsherds, mainly undecorated, but also with a rocker plain zigzag pattern, were recovered outside and inside the structure, altogether with a lot of faunal remains, many lithic artefacts, among which a cleaver, and some pieces of charcoal.

Faunal remains have been unearthed from both the stone platform and foundation pit of the engraved slab. Both spots have been also radiocarbon dated: burned bone from the platform gave 5150 ± 110 years BP (GX-28446); the bone from the pit yielded a consistent determination, of 5290 ± 40 (AMS: GX-28447). Ceramic potsherds from the structure confirm this tardy Middle Pastoral attribution.

Some two hundred fragments of animal bones come from the foundation pit (plus some 60 fragments selected for radiocarbon dating). Almost all determinable bones belong to domestic cattle. Only some parts of the animal were put in the monument, lacking parts of the cranium and of the axial skeleton. Only one animal would have been deposed, whose age remains uncertain, given the poor state of preservation.

Bones from the platform show similar preservation, and quantity. A few are surely related to domestic cattle; however, remains of a “large ungulate”, quite likely Bos, refer to other (lacking) parts. Considering the assemblage as a whole – as obvious given architectural relations and radiocarbon determinations – it is likely that a single cow as been slaughtered and put (but not complete) in the monument. Splits of animal bones are present also in the very close site (site 557), again a large ungulate: radiocarbon determination on burned bone gave an age of 5750 ± 40 (GX-28448).

I do not touch in this seat the possible relationships between 14C datings, and rock art of naturalistic style engraved on the slab. However, rock art provides spectacular evidence for the existence of a ‘cattle cult’ in the Sahara. In the Messak mountain ranges, we have scenes mostly engraved of thauropaksia, vivid descriptions of this ritual within pastoral societies. Unfortunately, their dating is still a controversial matter. Nevertheless, the same existence of an extraordinary rock art – both painted and engraved – focussed on cattle and on pastoral ideology may be considered a self-evidence of the ‘African Cattle Cult’ (as indicated by Holl 1998).

Remarks, and some hypotheses Thus, also in the central Sahara, the presence of articulated sites, dedicated to complex rituals, cattle sacrifices depicted on the wadi walls, sacred fires, and building of high-effort-demanding monuments appears to strongly resemble the evidence as known at Nabta. In the Acacus and surroundings, this phenomenon is cultural component of Middle Pastoral groups, known as transhumant cattle herders, exploiting different ecotonal niches on a seasonal basis.

But to better understand this phenomenon, we should ask ourselves: How and why did this particular cult spread so quickly and over such a large area? What is its meaning? What are the relationships with the ubiquitous megalithic architecture typical of later nomadic pastoral groups?

How and why? First attestations of ‘cattle-burial’ cult are present over a very large territory and in a rather narrow chronological interval (Fig. 6): from around 6400 years BP, we have sites sharing strictly similar ritual, but more than 3000 km distant each other. Even taking in account the bias due to archaeological research, the fact is puzzling: the idea of a fast rate of dispersal appears not unsound.

Analysing the (apparently) rapid rate of expansion of small livestock in Northern Africa, Hassan (2000b: 74) suggests that it unlikely may be interpreted as “demic (population) expansion of an advancing ‘wave’ or a ‘frontier’, but rather a ‘leap-frog’ movement by small groups”. Hassan hypothesises a possible movement of 10 km per year: if true, some 300 years would be necessary to cover the distance separating the two areas (Eastern and central Sahara).

I think it is more likely that the movement was much faster, given the environmental conditions. In fact, with conditions of high interannual variability in rainfall and increasing droughts, this idea better fits with archaeological data. As discussed before, I strongly believe that conditions of aridity are to be considered essential background to force people to move.

So, it is attractive to suggest a rapid diffusion of this ‘cult’, probably in relation to other forms of group movement connected to climate change and management of resources. The spread of megalithism in the Sahara has to be related to large-scale movements of nomadic pastoral groups: they used these stone structures to mark their territories and as aggregation signs. We should also take into consideration ways of penetration in unfamiliar landscapes, as recently argued by Rockman and Steele (2003).

However, all the initial forms seem to be related to elaborated cattle rituals, in areas characterized by a drying environment: therefore, ‘ritual’ slaughtering of cattle could have been part of an articulated scenario, related to rainmaking or other ceremonies, aimed at facing the increasingly deteriorate environmental conditions.

Literature on African palaeoclimate is too large to be recalled here: however, it is well known as the second half of the 7th millennium BP was characterised by increasing aridity. In Northern Africa, the main features of this arid spell are the patchy distribution and its length: where recorded, it seems to last around 300 years. Thus, it is noteworthy the apparent similarity in duration and chronological placing of this arid interval and the very initial spread of the ‘cattle-burial’.

Relations with later nomadic people In our view, it is likely that megalithic architecture – represented by cattle burials – spread over the Sahara in relation to mid-Holocene short, abrupt dry events. However, the ideological and cultural transformation evident in the use of these stone monuments (from animal to human burial) is yet to be stated.

This ‘megalithic’ culture, with minor differences, lasted for centuries and covered an enormous territory. Progressively, this kind of architecture moved from animals towards humans, some 5000 years ago: to my knowledge, only a couple of isolated radiocarbon dates is older 5000 years ago (5650 ± 70 years BP for Egypt: Site D5.1: Castiglioni et al. 1995; 5360 ± 200 years BP for Niger: Adrar Bous, Site 2: Paris 1997).

At the beginning, ‘human-related’ stone tumuli were generally single burials of adult males. Therefore, from ‘things fixed in the ground’ (sensu Ingold 1986) – i.e., elements of social tenure and group’ territorial markers – stone tumuli related to cattle ritual ‘became’ human tombs, changing their symbolic function and bearing the vertical expression of incipient social stratification. These monuments no longer belong to the group, as was the case for the cattle burials, but probably to lower segments of actual pastoral, semi-nomadic societies: smaller clan members or extended families.

Therefore, these monuments began to emphasise a direct relation between the group and the ancestors, and between the latter and the land where they were buried. In this way, the ancestors strengthen the cohesion of the community and its economy, claiming rights over resources. This was really a fundamental shift, whose archaeological record is still nebulous and vague, but it represented a crucial change within these pastoral societies.

It is urgent to get new data across North Africa to better understand this social trajectory. But looking at radiocarbon determinations, this ‘transition’ took several hundreds of years: timing is imprecise, and data are still poor, but it is of interest to underline that such (very long) ‘transition’ begun in apparent concomitance with another dramatic worsening of environmental conditions, the dry event at some 5000 years BP. Again, rapid climatic changes evolutions of environmental conditions push people to respond and implement these mutations innovating their social, ritual and economic structure.

Alhaique F. (Ed). Faunal remains from MSRS 2000 (Libyan Sahara). Manuscript in preparation.

Applegate A., Gauthier A. and Duncan S., 2001. The north tumuli of the Nabta Late Neolithic Ceremonial Complex. In Wendorf F., Schild R., and Associates (eds.) Holocene settlement of the Egyptian Sahara 1. The archaeology of Nabta Playa, pp. 468-488, Kluwer Academic/Plenum Publishers: New York.

Castiglioni A., Castiglioni A., and J. Vercoutter (eds.) ,1995. L’eldorado dei faraoni. Alla scoperta di Berenice Pancrisia. Istituto Geografico De Agostini: Novara.

Hassan F. (ed.), 2002. Droughts, food and culture. Kluwer Academic/Plenum Publishers: New York.

Hassan F., 2002a. Introduction. In Hassan F. (ed.) Droughts, food and culture, pp.1-7, Kluwer Academic/Plenum Publishers: New York.

Hassan F., 2002b. Palaeoclimate, food and culture change in Africa: an overview. In Hassan F. (ed.) Droughts, food and culture, pp.11-26, Kluwer Academic/Plenum Publishers: New York.

Herskovits M.J., 1926. The Cattle Complex in East Africa. American Anthropologist 28: 230-272.

Hogg R., 1990. An institutional approach to pastoral development: an example from Ethiopia. ODI papers (www.odi.org.uk).

Holl A., 1998. Social issues in Saharan Prehistory. Journal of Anthropological Archaeology 8:313-354.

Ingold T., 1986. The appropriation of nature. Essays on human ecology and social relations. Manchester University Press: Manchester.

Iro I., 2000. The characteristics of the Fulani. www.zumunta.org/Publications/Publications.html Paris F., 1997. Burials and the peopling of the Adrar Bous region. In BARICH B.E., and GATTO M.C. (eds.), Dynamics of populations, movements and responses to climatic change in Africa, pp. 49-61, Bonsignori: Roma.

Paris F., 2000. African livestock remains from Saharan mortuary contexts. In BLENCH R., and MacDonald K. (eds.) The origins of African livestock, pp. 111-126, UCL: London.

Rockman M. and Steele J., 2003. Colonization of unfamiliar landscapes: the archaeology of adaptation. Routledge, London.

Van Buren M., 2001. The archaeology of El Nino events and other “natural” disasters. Journal of Archaeological Method and Theory 8 (2): 129-149 Wendorf F., Schild R., 2003. Book review. African Archaeological Review 20 ( 2): 121-133 Wendorf F., Schild R. and Associates , 2001. Holocene Settlements of the Egyptian Sahara. Volume 1. The Archaeology of Nabta Playa, 709 pp. Kluwer Academic/Plenum.

Date received: January 27, 2004

Copyright © 2004 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # camu-24.