nabta playa and its role in northeastern african prehistory

Nabta Playa and Its Role in Northeastern African Prehistory

Fred Wendorf

Department of Anthropology, Southern Methodist University, Dallas, Texas 75275

and

Romuald Schild

Institute of Archaeology and Ethnology, Polish Academy of Sciences, Warsaw, Poland

Received October 1, 1997; revision received November 1, 1997; accepted January 10, 1998

Nabta Playa basin offers an unprecedented longitudinal view on the emergence, consolida-tion and complexification on human–livestock relationships, from the early stage of the EarlyHolocene (c. 11,000 cal. B.P.) to 6000 B.P. The problem of cattle domestication in NortheasternAfrica is considered and hopefully ‘‘solved’’ in the light of new mtDNA evidence which suggestan early late Pleistocene split between African, Asian, and Eurasian wild Bos populations. Thepaper presents a contextualized analysis of almost all the components of archaeological inves-tigation, including climatic change, culture history of Early to Mid-Holocene Nabta-Playans, thedevelopment of social differentiation, and probably ranking with ‘‘labor-consuming’’ megalithicfeatures with the emergence of characteristic features of pastoral ideology and religions. As faras the emergence and adoption of new foodways are concerned, the cultural developmentoutlined with the Nabta Playa archaeological record is important for the understanding of theHolocene prehistory of Africa as a whole. © 1998 Academic Press

INTRODUCTION

When traveling through the area, theWestern Desert of Egypt does not appearto be very promising for the study of pre-historic archaeology. On closer examina-tion, however, even the untrained eye cansee numerous scatters of lithic artifactsand other evidence of human occupationin this area which today is unoccupiedand seemingly devoid of all life. The factthat it is a complete desert, with less than1 mm of precipitation per year and is lack-ing in grasses, bushes, and trees (except ina very few places where ground watercomes near the surface), makes the ar-chaeology more visible and permits de-tailed archaeological surveys that arerarely possible in areas covered with veg-etation (see Wendorf et al. 1987a).

The Western Desert has a long historyof human use beginning at least as earlyas the early Middle Pleistocene and offersa rare opportunity to study past humanadaptation to a hyperarid environment.For example, during the early Holoceneamong the more interesting develop-ments is the appearance of presumablydomestic cattle perhaps as early as 11,000cal B.P.; the accompanying presence of so-phisticated and well-made pottery in theEarly Khartoum tradition (Banks 1980;Close 1995); the introduction of caprovidsfrom Southwest Asia between ca. 8400 and8000 cal B.P. (Gautier 1980); technologicalinnovations such as deep wells whichmade it possible for groups to live in thedesert throughout the year (Wendorf et al.1984); the emergence of a regional cere-monial center with megalithic alignments,

journal of anthropological archaeology 17, 97–123 (1998)article no. AA980319

970278-4165/98 $25.00Copyright © 1998 by Academic PressAll rights of reproduction in any form reserved.

stone circles, cattle burials, and otherlarge-scale constructions the functions ofwhich are not yet well understood; andindications of social control and perhaps aranked social system by around 7500 calB.P., several hundred years before there isevidence of similar complexity in the NileValley. This puzzling proximity of culturalinnovation and environmental stress inthe Western Desert deserves serious con-sideration by those who have interests inthe relationship between environmentand cultural processes.

NABTA PLAYA BASIN: LANDSCAPEAND PALAEOECOLOGY

Many of these interesting develop-ments in the Western Desert are best seenin a large internally drained basin knownas Nabta Playa, and located near thesoutheastern edge of the Western Desert,about 100 km west of Abu Simbel and 30km north of the Sudanese border (Fig. 1).Because of the size of the drainage area forthe basin, Nabta Playa appears to havebeen an unusually attractive locality for

FIG. 1. Map of Egypt showing location of Nabta Playa.

98 WENDORF AND SCHILD

early and middle Holocene groups, and itand the surrounding basins are one of themost important archaeological areas inthe Western Desert. Numerous archaeo-logical sites occur here, often imbeddedwithin sediments of the ephemeral pondsand lakes (playas) that filled the basins.The Nabta basin also is of particular inter-est because it has one of the longest andmost complete sequences of Holocene oc-cupations known in the Sahara. Thesesites and their stratigraphic settings arethe focus of this paper.

The Combined Prehistoric Expeditionbegan studying the Holocene archaeologyat Nabta in 1974–1975 and 1977 (Wendorfand Schild 1980: 82–165) and, more re-cently, in 1990–1992, 1994, and 1996. Thework since 1990 has not been publishedin detail; however, several short papershave appeared discussing some of themore interesting features found at Nabta(Dahlberg et al. 1995; Gautier et al. 1994;Kubiak-Martens and Wasylikowa 1994;Wasylikowa and Kubiak-Martens 1995;Wasylikowa et al. 1993, 1995, in press;Wendorf et al. 1991, 1992, 1993, in press;Wendorf and Schild 1994, 1995/1996, inpress a). A final report on the prehistory ofNabta Playa is in the final stages of prep-aration and should appear shortly.

Although the Western Desert is today arainless desert, it was not always so arid.There is good evidence that at severaltimes in the past this area received asmuch as 500 mm of precipitation per year,at which times there were permanentlakes, large springs and at least seasonalstreams. The most recent of these wet pe-riods occurred during the Last Interglacialand is dated between 130,000 and 70,000years ago by several radiometric tech-niques. During this time the area was athornbush savanna and supported nu-merous large animals such as extinct buf-falo and camels, large giraffes, and severalvarieties of antelopes and gazelles. Nu-merous Middle Paleolithic sites are asso-

ciated with the lake and spring sedimentsof this period (Caton-Thompson 1952;Wendorf et al. 1993). Still earlier MiddlePaleolithic and Final and Late Acheulean,and perhaps Middle Acheulean, artifactsare associated with playa, lake, andstream deposits of wet periods that pre-ceded those of the Last Interglacial (Schildand Wendorf 1977; Wendorf and Schild1980; Wendorf et al. 1985a; McHugh et al.1988a, b, 1989). Insofar as is known, onlyoccasional finds of heavily eroded artifactsof these earlier periods occur in the Nabtaarea.

The Western Desert was hyperaridfrom shortly after 70,000 years ago untilaround 12,500 to 13,000 cal B.P. Duringthis arid period the water table fell to alevel as low as or lower than that of today,and wind erosion scoured out numerousdeep depressions in the bedrock. One ofthose wind deflated basins was Nabta. Be-fore 12,000 years ago the summer mon-soon system of tropical Africa movednorthward as far as southern Egypt, andduring the more moist phases broughtrainfall variously estimated on the identi-fications of wood charcoal to have beenbetween 50 and 100 mm/year (Neumann1989; Barakat 1995), and on the basis ofassociated fauna between 100 and 200mm/year (Wendorf and Schild 1980: 236).Some interpretations based on sedimentsplace the rainfall much lower, around 30mm/year (Kropelin 1993). Whatever theamount, the precipitation was limited andhighly seasonal; both plants and animalsindicate that most of the rain fell duringthe summer months. The rainfall was alsounpredictable, droughts, were frequent,and some areas may have received no rainat all for long periods (Wendorf et al.1984). These limited rains during the earlyHolocene caused seasonal lakes andponds to develop in the depressions pre-viously hollowed out by the wind. TheWestern Desert was still a dry and unpre-dictable environment, with no permanent

99PREHISTORIC NABTA PLAYA

surface water and few resources. Onlysmall animals could live there, the largestof which were two varieties of gazelles,together with hares, jackals, lizards, ro-dents, and desert foxes, all of which couldexist on dew or moisture from vegetation.Cattle, regarded as domestic, were alsopresent. Limited as it was, the Holocenemoist period in the Western Desert lastedabout 5000 years, until around 5900 calB.P., and at several intervals it supportedreasonably large, but highly mobile hu-man populations who existed by large andsmall animal pastoralism, hunting, and in-tensive gathering of a wide variety of wildplants.

THE PREHISTORIC SEQUENCE

The Early Neolithic (10,800–8900 cal B.P.)

The earliest excavated sites at Nabta aredated by radiocarbon to around 10,300 calB.P. and usually are located on fossildunes that accumulated on the floors ofthe basins during the preceeding intervalof hyperaridity. These sites consist ofsmall scatters of lithic artifacts and fossilbones; there is no evidence of houses,storage pits, or wells, although most havesmall hearth areas. One of these sitesyielded charred seeds of wild millet andtwo varieties of legumes (Wasylikowa, re-port to F. Wendorf 1996). The locations ofthe sites in the lower part of the basinsand the absence of wells indicate thatthese sites were occupied when the playaswere almost dry, probably in early fall,and abandoned in the spring, the driesttime of the year when surface water wouldnot have been available.

El Adam type settlements (10,800–9800 calB.P.). These earliest sites are assigned tothe El Adam variety of Early Neolithic(Wendorf et al. 1984: 409–411) character-ized by well-made bladelet-based lithicassemblages with straight-backed pointedbladelets, perforators, and large end-

scrapers made on reused Middle Paleo-lithic artifacts. The preferred raw materialwas Egyptian flint, the nearest source forwhich was along the Eocene Plateau,about 75 km to the north of Nabta playa.Chert, agate, and chalcedony were alsoused, but less frequently. A few grindingstones and rare shards of pottery also oc-cur. The pottery is well-made and deco-rated over the entire exterior with deepimpressions in a nested chevron madewith a comb or wand. Another commondesign has closely spaced lines of combimpressions, some parallel to the rim andothers at right angles. All of the decora-tions are in the ‘‘Early Khartoum style,’’but the characteristic ‘‘dotted wavy line’’motif of that style is missing in these ear-liest ceramics but it does appear in sites ofa later variety of Early Neolithic. A puz-zling feature of this early pottery is itsrarity; it is usually limited to only a fewshards in a site, a situation which has cau-tioned us that it might be intrusive; how-ever, the shards occur in most of the ex-cavated El Adam sites and the designs areunique and limited to this period. Thefunction of this pottery is far from clear,but its rarity suggests that it was not ingeneral use as containers; they may havebeen luxury or status items (Close 1995).

The lithic artifacts in these El Adamsites, except for the pottery and the reusedMiddle Paleolithic artifacts, are closelysimilar to those found in the Arkinian inthe Nile Valley that is about the same ageor slightly older (Schild et al. 1968). Be-sides rare sherds of pottery, almost all ElAdam sites, which elsewhere in the West-ern Desert have been dated between10,800 and 9800 cal B.P., have yielded afew bones and teeth of a large bovid, iden-tified as Bos, as well as numerous bones ofgazelle and hare, plus a few bones ofjackal, turtle, small rodents, and birds,which suggests a rather poor environ-ment, comparable to the northernmostSahel today.


Our interpretation of their food econ-omy is an important aspect of understand-ing how these early Holocene groups uti-lized the Western Desert. FollowingGautier (1980, 1984) we have suggestedthat these early Holocene groups werecattle pastoralists who brought their herdsinto the desert for grazing after the sum-mer rains, coming into the desert fromsome as yet unidentified area where wildcattle were present and where the initialsteps toward domestication first occurred(Wendorf et al. 1984: 420–422; Wendorfand Schild 1994). This may have been theNile Valley, between the First and SecondCataracts, because wild cattle had beenpresent in that area (and a major preyanimal since the Middle Paleolithic; Gau-tier 1968), as were people with lithic in-dustries closely similar to those in the ear-liest Holocene sites in the Western Desert.We have suggested that cattle may havefacilitated human use of the desert by pro-viding a mobile, dependable, and renew-able food resource in the form of milk andblood. The use of cattle as a renewableresource rather than meat may be a pos-sible explanation for the paucity of cattleremains in these Saharan sites. This use ofcattle may have been closely similar tothat of modern African pastoralists, whouse the by-products from their herds, butrarely kill them for meat, and then only atimportant ceremonial occasions. Amongthese groups cattle are an importantsource of wealth and prestige. The Africanpattern of cattle pastoralism may wellhave developed in this or a closely similarsetting.

Our interpretation of the cattle remainsin the Western Desert has been highlycontraversial (Smith 1984, 1992; Clutton-Brock 1989, 1993; Muzzolini 1989, but alsosee Wendorf et al. 1987b). The objectionsto the hypothesis of an early and separatecenter of cattle domestication have beenconsiderably weakened, however, by re-cent mtDNA studies of African, Eurasian,

and Indian cattle which indicate that mod-ern African and Eurasian cattle have beengenetically separate populations for over25,000 years and that Indian cattle havebeen separate from the other two evenlonger (Bradley et al. 1996). This is strongevidence that there were three separatecenters of cattle domestication, one ofwhich was in Africa.

Other evidence favoring the hypothesisthat the Saharan cattle were domestic isseen in the restricted environment of theWestern Desert during the early Holo-cene, particularly the absence of perma-nent water. Without permanent water it ishighly unlikely that Bos could exist thereexcept under human control. Cattle needto drink almost every day and would nothave been able to move from basin tobasin as the water in those basins driedup, and as the dry season intensified, theywould not have been able to return to theNile Valley, or to move farther southwhere permanent water was present.

That the cattle were brought to thedesert under human control is alsostrongly supported by the composition ofthe other fauna that occurs with them. Afaunal assemblage consisting only ofsmall, desert adapted animals and largecows does not occur in nature. A normalpopulation would also include intermedi-ate-sized animals, such as hartebeest.Hartebeest and wild cattle were the pre-dominant game in the Nile Valley, andthey appear to have overlapping require-ments, with hartebeest being the ‘‘drier’’of the two (Kingdon 1982; Gautier 1987;Gautier and Van Neer 1989). The absenceof hartebeest in the Western desert Holo-cene faunal assemblages is difficult to ex-plain if the cattle were wild.

El Ghorab type settlements (9600–9200 calB.P.). There was a brief period of aridityaround 9800 cal B.P., when the desert ap-pears to have been abandoned. In earlierreports we proposed that when the rainsreturned the desert was reoccupied by


groups we named the El Kortein variety ofEarly Neolithic who used bifacial pointsresembling the Ounan and Harif points ofAlgeria and the Negev (Wendorf et al.1984). A preliminary restudy of the ElKortein sites, however, indicates that thiscomplex may in fact date somewhat later,perhaps the early Middle Neolithic.

The evidence now available suggests thatwhen the rains returned Nabta Playa andother basins in the Western Desert werereoccupied by groups with a lithic tool-kitthat emphasized elongated scalene trian-gles and microburin technology. Other arti-facts include grinding stones, perforators,backed bladelets, endscrapers, and a fewshards of pottery similar to that found in theEl Adam sites. This lithic industry charac-terizes the El Ghorab type of Early Neo-lithic, dated between 9600 and 9200 cal B.P.(Wendorf et al. 1984: 113–147). The associ-ated fauna are again mostly gazelle andhare, but there are a few bones of wildcat,porcupine, desert hedgehog, birds, and cat-tle. All but the cattle are desert adapted anddo not require surface water. Plant remainswere not recovered from any of these sites,but this probably is because all of the ElGhorab sites were excavated before we de-veloped proper recovery techniques. Nohouses are known for this period at NabtaPlaya, but in the Dyke area located some 200km northwest of Nabta, there are severaloval, slab-lined houses associated with lith-ics of the El Ghorab type (Schild and Wen-dorf 1977: 113–147). Again there are noknown storage pits or water wells, so it isbelieved that the desert continued to beused only after the summer rains and wasabandoned during the driest season of theyear.

Another brief period of hyperaridity,between 9200 and 9100 cal B.P., coincideswith the end of the El Ghorab Neolithic inthe desert (groups with similar lithic as-semblages apparently continued to live inthe Nile Valley after this date; Vermeersch1978). With the return of greater rainfall

around 9100 cal B.P., a new variety ofEarly Neolithic, the El Nabta type, ap-peared in the Western Desert, and theybrought with them new technologies andpossibly a new social system that en-hanced their ability to use the desert.

El Nabta type settlements (9100–8900 calB.P.). El Nabta settlements are usuallylarger than the previous Early Neolithicsites, and some of them had both largeoval huts and smaller round huts, as wellas numerous bell-shaped storage pits andlarge deep wells, sometimes with adjacentshallow basins that might have been usedto water stock (Wendorf and Schild 1980:128–140; Wendorf et al. 1984: 413–414).These El Nabta groups evidently had de-veloped the technology and social organi-zation needed to live in the desertthroughout the year. The lithic artifacts inthese El Nabta sites include numerousperforators, burins, backed bladelets(some of which are straight-backed andpointed), retouched pieces, notches, anddenticulates. Simple bone points also oc-cur, as well as pottery, the latter with sev-eral varieties of impressed designs, in-cluding ‘‘dotted wavy line.’’ Most of thevessels are small globular jars with sim-ple, constricted rims. Pottery is still rare,but more abundant than in previousphases, possibly because the settlementswere occupied by larger groups for longerperiods (but still seasonally, most knownsites are located in the lower parts of thebasins and were flooded during the sum-mer rains). The associated fauna is similarto that found in earlier Holocene sites,mostly gazelle and hare, and a few othersmall desert animals, together with an oc-casional Bos. A large series of radiocarbondates place the El Nabta phase between9100 and 8900 cal B.P.

The largest known El Nabta site (iden-tified as E-75-6) is located on a fossil dunein the lower part of Nabta Playa (Fig. 2).The site had been reoccupied many times,the first was by an El Adam group. The El


Nabta phase settlement has not beencompletely excavated, but it has at least 15houses or huts, not all occupied simulta-neously but each used several times. Thehouses are arranged in two, probablythree parallel lines, and there are three

water wells, one of which was 2.5 m deep(Wendorf and Schild 1980: 131). Adjoiningeach house are one or more large, bell-shaped storage pits. We have previouslysuggested that the arrangement of thehouses in rows indicates the presence of a

FIG. 2. Plan of Site E-75-6, El Nabta phase harvesting site at Nabta Playa. 1, Edge of playasediments; 2, limits of unexcavated features; 3, test trenches; 4, walls of features; 5, small pits andpostholes; 6, hearths; 7, later pits; 8, identification arrows; 9, possible structure.


social system with sufficient authority tocontrol the placement of houses in thevillage; however, recent excavations havedisclosed that the houses are alignedalong the edge of a deep basin, and ar-rangement of the houses may have beenstrongly influenced by this local topogra-phy. This may be clarified by furtherwork, but regardless, it is clear that therewas sufficient control over labor for theexcavation of the deep water well. Thereare, however, no indications of differencesin wealth, or even community storage fa-cilities. The site has two kinds of houses:long ovals more than 6 m long and 2.5 mwide and round structures from 3 to 4 min diameter. Stratigraphic evidence sug-gests that some of the long oval houses areearlier than some of the round houses, butmultiple radiocarbon determinations oncharcoal from both kinds of houses indi-cates that they are about the same age. Allof them appear to have been simple brushor mat covered huts, with several shallow,saucer-like floors separated by thin lensesof silt. There were from one to threehearths or burned areas on these floors,and several (sometimes several dozen)small, hemispherical ‘‘potholes’’ that werefilled with ash, charcoal, and charred ed-ible plant remains.

Since Site E-75-6 is located in the bot-tom of a large basin that was flooded eachyear with the summer rains, it is not sur-prising that the site was abandoned dur-ing these rains and then reoccupied whenthe basin became dry. It is surprising,however, that when the people returnedto the site they were able to find the pre-cise positions of the houses even thoughthey were covered by silt. The answer tothis mystery may be the structure of thehuts. A few postholes around the periph-ery of the houses apparently held uprightposts that formed the frame to hold themats, skins, or brush that presumablyformed the walls and roof of the shelter.This frame may have been left in place

when the site was abandoned each year,and when the floodwaters in the basinhad receded, these still-standing frameswould have marked the exact positions ofthe houses.

Site E-75-6 was obviously occupiedduring the dry season when manygrasses and other plants mature, and thesite seems to have been a plant collect-ing and processing locality (Wendorf etal. 1992). The house floors have yieldedover 20,000 seeds of grasses and legumesas well as tubers and fruits representing80 different morphological types, two-thirds of which have been identified asto taxonomic units of various ranks. Allof the plants are morphologically wildand grow today in the Sahelian zone ofNorth Africa. Among the more frequentof these are sorghum and several varie-ties of millets, the annual herb Schou-wia, seeds of the shrub Capparis, fruitstones of Ziziphus, and several kinds ofedible tubers (Wasylikowa et al. 1995:143–147). There are some tentative indi-cations that the sorghum may have beencultivated. Preliminary chemical analy-ses by infrared spectroscopy of thelepids in the archaeological sorghumshow closer resemblance to some mod-ern domestic sorghum than to wild va-rieties (Wasylikowa et al. 1993). Alongthis same line it is interesting to notethat the distribution of the sorghum inthe houses suggest that sorghum wastreated differently from the other seeds.The significance, however, is not inwhether or not the sorghum was wild ordomestic, but that the sorghum andother plants were being intensively har-vested and stored for future use. Onemay conclude that plant foods com-prised a significant portion of the ElNabta diet.

The numerous storage facilities associ-ated with the huts at several sites of thisphase are further testimony to the impor-tance of plants in the El Nabta economy


and indicate that these communities mayrepresent a new level of adaptation to theSaharan environment where the mobilitydemands of cattle pastoralism were some-how merged with intensive collecting ofplant foods that were harvested in signif-icant quantities, stored, eaten, and thoseremaining moved elsewhere before thesummer rains. This new adaptation mayanticipate the emergence of cultivation inthe Sahara, if it was not already underway.In this context it is undoubtedly signifi-cant that there are no traces of wheat,barley, or any other Southwest Asian do-mesticate. The barley recovered from thissite during the 1977 excavations (Hadidi inWendorf and Schild 1980: 347) is regardedas intrusive.

Previously we believed that a brief pe-riod of aridity coincided with the end ofthe El Nabta phase and the Early Neo-lithic. There is, however, strong evidencethat occupations with typical Middle Neo-lithic-styled ceramics and dated only 100years later, around 8800 to 8700 cal B.P.,occur immediately above the El Nabtalevels and without any evidence of an in-tervening episode of aridity. In the earliestof these ‘‘Middle Neolithic’’ sites theycontinued to prefer Egyptian flint formany of their lithic artifacts and there wascontinuity with the Early Neolithic in ty-pology. Scalene triangles (some verysmall, less than 15 mm in length), backedbladelets, perforators, scrapers, stemmedpoints with pointed and retouched bases,notches, and denticulates are characteris-tic tools. We now assign these sites to alater phase of the El Nabta type Neolithic.In the succeeding later Middle Neolithicthere is a shift to local rocks for lithicartifacts, with a greater use of quartz andquartzite (few of which are retouched); inaddition, bladelet technology sharply de-clines; and among the retouched tools,there are few scalene triangles or backedbladelets, while points with retouchedbases disappear and are replaced by

points made on small flakes, with convex,concave, or straight retouched bases andlightly retouched pointed tips or lateraledges.

The Middle Neolithic (8300–7600 cal B.P.)

Radiocarbon dates place the beginningof the Middle Neolithic around 8300 calB.P. The environment during this periodwas similar to that in the Early Neolithic,or slightly drier; the identification of woodcharcoal indicates fewer species of wood(Barakat 1995), and in the fauna there is anincrease in the frequency of hare relativeto gazelle (Gautier 1984). However, bothof these changes may be reflections ofgreater human presence. The missingspecies of trees may have been preferredfor firewood and were the first to be de-pleted near the settlements. A similar ex-planation may be offered for the reducedfrequency of gazelle. These shy animalswill move away from an area where theyare repeatedly hunted, while hares areless inclined to do so. The faunal assem-blages in the Middle Neolithic sites arelarger and richer than those in the EarlyNeolithic and comprise all of the speciesof animals previously noted, includingcattle, as well as several kinds of lizards,ground squirrels, field rats, hyena, sandfox, and one example of either oryx oraddax.

Around 8000 cal B.P. there was an im-portant new addition to the food economyof the Middle Neolithic. Domestic ca-provids, either sheep or goat, or both,were introduced from Southwest Asia,probably by way of the Nile Valley (al-though the oldest radiocarbon dates nowavailable for the Neolithic along the Nileare about 500 years later). Cattle and ca-provids have different herding require-ments, and in a limited environment suchas the Western Desert, the managementof the two herds must have posed a chal-lenge, but they were obviously successful.


Unlike the bones of cattle which continueto be rare in most sites, once introduced,sheep became increasingly more frequentand in later periods replaced gazelle asthe major source of meat.

Despite an intensive search at severalsites in the Nabta area, edible plant re-mains have not been recovered from Mid-dle Neolithic contexts. The absence ofplant remains is almost certainly due topreservation and makes us appreciatehow fortunate we were to find Site E-75-6.Thus far, we have not been able to find aMiddle Neolithic site in a setting that du-plicates Site E-75-6. Nevertheless, mostMiddle Neolithic sites have numerous,large, bell-shaped storage pits and abun-dant grinding stones, both of which sug-gest that plant foods, most likely the sameones found at E-75-6, were an importantcomponent of their diet.

There are often houses in Middle Neo-lithic sites. These houses are usuallyround in outline, semi-subterranean, be-tween 30 and 40 cm deep, often with slab-lined walls and sloping lateral entryways.In some sites the houses are jacal-likestructures with wattle and daub walls.Hearths are usually in the center of thefloors. Middle Neolithic sites occur in avariety of sizes and settings (Wendorf etal. 1985). Some of them are small, withonly one or two houses, and these areusually located in smaller basins. Thereare also several somewhat larger siteswith half dozen or more houses in largerbasins; other sites are located on dunesoverlookng these basins. On the sand-sheets and plateaus there are numeroussmall clusters of Middle Neolithic arti-facts, often poorly made, with a hearthand not much else. Finally, there is onevery large site with unusually deep trashaccumulation (2 m) on a dune along a highbeachline of Nabta Playa (Site E-75-8; Fig.3). This variation in settlement sizes andtheir positions in the landscape has beeninterpreted as reflecting a seasonally re-

sponsive regional settlement system inwhich the population was dispersed intosmall- and medium-sized villages locatedin the lower parts of the basins duringmost of the year, particularly the dry sea-son. During the wet season they appar-ently gathered into a large community forsocial and ceremonial purposes along thebeach of Nabta Playa, the largest basin inthe area. Houses are not known at SiteE-75-8 (Fig. 3), the supposed ‘‘aggrega-tion’’ locality, but there are numerousstone-filled hearths, and the site hasyielded the highest frequency of cattlebones of any locality in the NubianDesert. In this connection it is useful tonote that among many African pastoral-ists today, cattle are frequently sacrificedand consumed at important ceremonialoccasions to celebrate the birth or death ofan important personage and at betrothalsand marriages. The suggestion that SiteE-75-8 was where people gathered for cer-emonial purposes in the late Middle Neo-lithic anticipates the slightly later emer-gence of Nabta Playa as a regionalceremonial center similar to the regionalcenters that occur even today in Sub-Sa-haran Africa, where they serve to bindtogether groups that are often widely sep-arated in space.

The other elements in the Middle Neo-lithic settlement system include the siteson the dunes, which are believed to recordbrief occupations by Middle Neolithicpeople after they had left the ‘‘aggrega-tion’’ site and while they waited for theirbasin to dry sufficiently for them to movedown onto the playa floors, and the smallsites on the sandsheets, seen as temporarycamps by herders, possibly young boyswho were not yet skilled in stone working.These herding camps could have beenused at any time of the year, but mostusefully after the summer rains whengrazing in those areas would have been atits best.


The Late Neolithic (7500–6200 cal B.P.)

The Middle Neolithic came to an abruptend with a major period of aridity thatbegan around 7600 cal B.P. and lasted forperhaps 100 years. During this arid epi-sode the water table fell several meters,the basins were reshaped, and their floors

deflated, in some instances by more than3 m. Insofar as we can tell, the NubianDesert was not occupied during this dryperiod. When the area was reoccupied afew years later, around 7500 cal B.P., thesites are larger (except for the many smallherding camps on the plateaus) and oftenreoccupied several times, but evidently

FIG. 3. Plan of the large ‘‘aggregation’’ locality, Site E-75-8, on the north beach of Nabta Playa.1, Sandstone bedrock; 2, windblown sand; 3, playa deposits; 4, wadi channels; 5, limits of culturaldebris; 6, Early Neolithic artifact cluster; 7, Middle Neolithic hearth mounds; 8, Late Neolithichearth mound areas; 9, ‘‘calendar circle’’; 10, megalithic alignment; 11, tumuli with cattle burials; 12,excavation trenches.


not for extended periods. These Late Neo-lithic sites contain numerous shallow,oval, stone-lined and stone-filled hearths,but there are no traces of houses.Stemmed and concave based, bifaciallyflaked projectile points are common inthese sites, which may indicate increasedregional instability. There is also a newlithic technology that made extensive useof short, wide (‘‘side-blow’’) flakes, oftenused as blanks for scrapers, complexnotches, and denticulates, and a new kindof sand or fiber tempered pottery withburnished exteriors and smudged interi-ors. Impressed or incised designs are rareand limited to the rims (Banks 1980: 306–307). This new pottery is very similar tothat found in the early Baderian and Ab-kan Neolithic along the Nile, where theyare dated between 7200 and 6200 cal B.P.(Nordstrom 1972: 250–251; Hassan 1985).The source of the Abkan and BaderianNeolithic is unknown, but it was probablyderived ultimately from Southwest Asia,possibly by way of Sinai, where prepot-tery and pottery Neolithic sites have beendated between 11,000 and 9000 cal B.P.(Bar-Yosef 1985). It is interesting to notethat Terminal Paleolithic fishing andhunting groups were living in the NileValley as recently as 8100 cal B.P., appar-ently with very limited contact with theNeolithic groups living in the nearbydesert (Wendt 1966; Vermersch 1978;Wendorf and Schild 1976: 163–182). Thischanged with the Late Neolithic.

The same Middle Neolithic ‘‘aggrega-tion’’ locality at Nabta (Site E-75-8) wasalso occupied during the Late Neolithic,again presumably during the wet season,because the occupational horizons con-tinue to interfinger with playa sedimentsalong the edge of the beach. Apparentlyactivities here during the Late Neolithicwere similar to those that occurred duringthe Middle Neolithic. In addition to exten-sive and repeated occupations along thehigh beach line of the playa that resulted

in numerous bones of both large andsmall livestock left in the trash deposits,our interest in this locality as a ceremonialcenter during the Late Neolithic wasgreatly enhanced by the discovery of anorth–south oriented alignment of ninelarge sandstone blocks, set about 100 mapart, and partially imbedded in playasediments near Site E-75-8 (Wendorf et al.1994). Also, beyond the north end of thealignment there was a ‘‘calendar circle’’ ofsmaller sandstone slabs, which may havehad astronomical functions (Malville et al.1998).

The potential importance of this localityas a ceremonial center was further em-phasized by the discovery of several smallstone-covered tumuli containing the re-mains of cattle, seven of which have beenexcavated (Fig. 4). One contained a com-plete young adult cow buried in a clay-lined and roofed chamber below themound (Fig. 5); six others have yielded thepartially disarticularled remains of cattlescattered among the rocks, with probablymore than one animal in each tumulus.All of these small tumuli are located alongthe western edge of the largest wadi en-tering Nabta Playa from the north, whichwith a bit of tongue in cheek we havenamed the ‘‘Wadi of Sacrifices.’’ Thesecattle burials and offerings appear to in-dicate the presence of a cattle cult. Boththe stratigraphic and radiocarbon evi-dence place these cattle tumuli at the be-ginning of the Late Neolithic wet interval,around 7500–7400 cal B.P.

The discovery of the cattle burials led usto reconsider how these Saharan cattlepastoralists may have functioned in theSahara. Cattle require water at least everythird day and ample grass for food. Forthis reason modern cattle pastoralists liv-ing in areas that receive such limited rain-fall as even the most optimistic estimatesfor Nabta rarely aggregate into largegroups, and when they do, they gather inthe driest time of the year near the few


wells with permanent water. It is duringthis period that group ceremonies are per-formed (and thus different from the indi-cated season of ceremonial activities atNabta).

In areas of such limited and highly sea-sonal rainfall, cattle pastoralists requirean extensive range to ensure adequategrazing and access to water for theirherds. In the Sahelian and Saharan zonesof North Africa, this generally means anorth–south pattern of movement. Thereis more rainfall in the southern parts ofmost areas, and these pastoralists move tothese wetter areas during the driest timeof the year and send their herds north-ward at the onset of the summer rains.The pastoralists at Nabta had another op-tion: they could move to the Nile Valleyduring the dry season. The earliest Neo-lithic groups, who do not seem to havedug wells, may well have gone to the Nileduring the driest period of the year (Wen-dorf et al. 1984). However, during the finalphase of the Early Neolithic and the Mid-dle Neolithic, both of which have distinc-

tive ceramics, there is no evidence of suchpottery or sites that might be SaharanNeolithic in the Nile Valley between theFirst and Second Cataracts. Thus, despiteour logical expectations, the Nile Valleymay not have been a regular part of theseasonal round during the Middle Neo-lithic. This does not rule out the possibilitythat the Valley might have been used dur-ing periods of extreme drought, when theNile was the only available water. Suchrefuge sites, if they exist, could be verydifficult to detect.

In the Late Neolithic there are manysimilarities in the ceramics between theSahara and contemporary or perhapsslightly later sites along the Nile. There-fore, it is highly likely that settlements ofSaharan groups might not be identified ina preliminary survey, and for this reasonan east–west seasonal round during thisperiod cannot be ruled out. On the otherhand, the deep wells dug in the lowestpart of the Saharan playas also suggestthat groups occupied this part of thedesert during at least part of the dry sea-

FIG. 4. View of small rock-covered tumulus before excavation. Below the rocks was an oval,clay-lined chamber containing the burial of a complete young cow.


son. Even so, there must have been signif-icant seasonal movement, because cattlewill rapidly exhaust the nearby grazing insuch an environment, even when water isavailable in wells. The settlements of theMiddle and Late Neolithic cattle pastoral-ists at Nabta must have been brief.

The potential significance of Nabta asan early regional ceremonial center wasfurther strengthened by the discovery ofthree groups of megalithic structures lo-cated on an extensive and relatively highremnant of Middle Neolithic playa sedi-ments along the western edge of theNabta basin. There are approximately 30of these ‘‘structures’’ in the largest groupin an area 200 m wide and 500 m long(because some have been disarranged andothers resemble bedrock outcrops, the ex-act number cannot be determined withoutexcavation). Each of the ‘‘structures’’ con-sist of several large, roughly shaped sand-stone blocks set on edge to frame an ovalarea about 5 to 6 m long and 4 to 5 m wide,

in the center of which is a large, north–south oriented, rectangular slab (ca. 2 31.5 3 0.4 m). An interesting feature of allthree ‘‘megalithic structure localities’’ isthe complete absence of other associatedcultural debris, which is highly unusualfor the Nabta Basin, because most othersimilar areas are littered with deflatedhearths and lithic artifacts. While gener-ally similar, there are also interesting dif-ferences. One of the structures is largerthan the others and is set apart. Others arerelatively small and occur in tight inter-locking groups of up to eight units. Mostof the structures occur in loose groups ofthree or four units placed from 2 to 3 mapart and never touching. Some of thislast group are large, approaching the sizeof the isolated largest structure.

Test excavations at three of these struc-tures yielded evidence of elaborate workin stone far beyond that which was ex-pected. One of them, the largest, is of par-ticular interest. It had two large flat, hori-

FIG. 5. The young cow buried in the chamber below the small rock-covered tumulus.


zontal central stones with a third largepyramid-shaped boulder resting on them.All had their long axes aligned slightlywest of north–south. The central elementwas surrounded by large boulders set up-right on edge. Some of the stones hadbeen carefully shaped with wedges andweighed up to one and half tons. We ex-pected to find a burial pit below the cen-tral stones; instead there was only a lensof finely laminated sand and silt, restingon disturbed Middle Neolithic playa sed-iments, strongly modified by repeatedsubmergence and drying, which destroyedall traces of bedding in the original playadeposits. However, it was evident that alarge pit had been dug into the earlier playasediments and then refilled before the sur-face architecture was erected. In time the fillin this pit began to settle, leaving a shallowbasin which was filled by thin lenses of lam-inated sand and silt.

At a depth of slightly more than 1 mbelow the surface and off-set slightly fromthe center of the surface architecture wasa large, carefully shaped stone that at firstwas thought to resemble the keel of anupside-down boat (Fig. 6), but whenplaced upright, it looks vaguely like ananimal, possibly a cow. The long axis ofthe sculpture was oriented north–south,and at the north end was a rough fan-likeprojection, like the head of an animal orperson. It is slightly more than 2 m long,1.25 m wide, and 0.5 m thick, and it weighsabout 2.5 tons. One side of the stone isconvex, the other is flat; both of the uppersides are carefully smoothed, but the twounder sides are rough and unshaped.

Centered under the ‘‘sculpture,’’ andthe surface architecture, at a depth of3.5 m below the surface, was a large bed-rock mushroom-shaped table rock (Fig. 7).The table rock also had been carefullyshaped and worked into a circular outlinewith smoothed, recurved sides and a flat,smooth surface on top (Fig. 8). It has twoprojections about 40 cm wide, one to the

north, and the other to the southwest. Itwas expected that a tomb might be asso-ciated with the table rock, particularly be-yond the north projection where therewere large elongated stones standing up-right in the fill; however, excavationshowed no trace of pit or tomb (Fig. 9).

Two other megalithic structures havebeen excavated, and two others tested bydrilling. All are basically similar in theirgeneral characteristics; all built over tablerocks but only the large isolated one con-tained a sculpture. How they managed todetermine the presence of the table rocksburied from 2 to 3 m deep in heavy playaclays and silts is unknown. They may havebeen found by digging pits or by probingwith long sticks while the clay was softafter rains. The function of the structuresis not clear. It is possible that they areshrines, but we suggest that they are‘‘proxy tombs,’’ erected to honor elitemembers of the group who died else-where during their seasonal movements‘‘on the trail.’’ If so, these structures mayindicate differences in social rank, with anelite ‘‘family,’’ or kin group representedby the clustered structures, a class ofhigher ranking individuals by the groupsof larger structures, and a yet higherranked individual by the isolated struc-ture.

These limited excavations at thesemegalithic structures indicate the pres-ence of elaborate and previously unsus-pected Late Neolithic ceremonialism. Al-though the degree of social controlinvolved is not as yet established, theplanning of the structures, the work re-quired to quarry and transport the stonesculpture, the effort used to dig the 5 to6 m in diameter pits, and the time de-manded to shape the sculpture and theunderlying table rocks, represent effortsand social expenditures far beyond thatexpected from the seemingly simple cattlepastoralists represented in the living sites.


DISCUSSION

The gradual development of technol-ogies and social systems that facilitateduse of the Sahara by Neolithic groups inthe early and Middle Holocene has beenoutlined in the preceeding pages. Thepresence of social systems that permit-ted control of the labor needed to exca-vate the deep wells needed to exist in

the area in the driest season and possi-bly to determine that settlement ar-rangement was present as early as 9000cal B.P. in the final phase of the EarlyNeolithic. Regional ceremonial and (possi-bly political) systems are indicated for theMiddle Neolithic as early as 8000 cal B.P.,with the large settlement of E-75-8 andits more frequent remains of cattle. Thistrend toward social complexity reaches

FIG. 6. Carefully shaped ‘‘stone sculpture’’ found under the Late Neolithic megalithic surfacearchitecture. The longitudinal axis of the sculpture is oriented north–south and has an unshapedprojection at the north end.


its zenith with the emergence of Nabtaas a regional ceremonial center in theLate Neolithic. These developments aresignificant for African prehistory in twoareas: first, the emergence of the AfricanCattle Complex (Herskovits 1926) wherecattle serve to symbolize status andpower, and in which regional ceremonialcenters are an important component;and second, the role of African cattlepastoralists in the rise of Egyptian Civil-ization.

The source or sources of social com-plexity in Egypt has long been a topic ofdiscussion. Initially it was believed thatEgypt might have been the first to have acomplex society, but radiocarbon dateshave shown that Mesopotamia was ear-lier. It was then assumed that Egypt wasthe great borrower and that the conceptsof complexity spread from Mesopotamiato Egypt. However, it is no longer as-

sumed that social complexity can be bor-rowed or diffused from one area to an-other, instead social complexity is moreoften seen as developing from localcauses. Social complexity is frequentlyregarded as an expression of the degreeof structural differentiation and func-tional specialization evident in a society;a development can occur within the so-ciety when craft and other specialistsemerge (because when division of laborincreases, the need for control also in-creases), or from external forces wherethere are two radically different eco-nomic systems in close physical proxim-ity, as is often found where agriculturistswith a centralized political system haveclose relationships with pastoralists.These two processes of structural differ-entiation are not mutually exclusive, butmutually supportive. The pastoralistsusually live in tense harmony with their

FIG. 7. Worked mushroom-shaped ‘‘table rock’’ found under the stone sculpture at a dept of4 m. The upper figure is at the north projection. The north end of the sculpture was held in positionby the slab to the right of the figure.


village neighbors, but from time to timethey will take advantage of a weaknessand take control. It is in this setting thatthe Late Neolithic cattle pastoralists andtheir regional ceremonial center atNabta is of particular interest, because itmay well be that the Saharan pastoral-ists may have provided the basis for theexternal differentiation that stimulatedthe emergence of social complexity inEgypt.

Ceremonialism in Predynastic and OldKingdom Egypt

One of the interesting aspects of theNabta center is its possible role as a con-tact point between the early Nilotic Neo-lithic groups with their agricultural econ-omy and the cattle pastoralists in theEgyptian Sahara. The functional separa-tion of these two different economies mayhave played a significant part in the emer-

FIG. 8. Worked mushroom-shaped table-rock, showing carefully shaped sides. View is lookingsoutheast. The area below the mushroom stone has yet to be excavated.


gence of complexity among both groups.The evidence for Nilotic Egyptian influ-ence on Saharan pastoralists is not exten-sive and is presently limited to Late Neo-lithic ceramic technology, occasional shellsof Nile species, and rare stones from theNile gravels. Another way of exploringthis is by examining those aspects of po-litical and ceremonial life in the Predynas-tic and Old Kingdom that might reflectimpact from the Saharan cattle pastoral-ists. In this we have been preceeded byFrankfort (1978: 3–12) who, in his majorstudy of Egyptian and Mesopotamian re-

ligions and political systems, argued thatthe Egyptian belief system arose from anEast African substratum and was not in-troduced from Mesopotamia. To supporthis position Frankfort pointed to the sim-ilarities in religious beliefs the early Egyp-tians shared with Nilotic cattle pastoral-ists. During the Old Kingdom, cattle werea central focus of their belief system. Theywere deified and regarded as earthly rep-resentatives of the gods. A cow was alsoseen as the mother of the sun, who issometimes referred to as the ‘‘Bull ofHeaven.’’ The Egyptian pharaoh was a

FIG. 9. View of worked mushroom-shaped table-rock looking west. Note one of two largelosenge-shaped stones standing near vertical behind and to right of workmen. The positions ofthese stones suggest the possibility of a shaft.


god (similar to the Shillok king, and not anintermediary to the gods as in Mesopota-mia). He was the embodiment of twogods, Horus, for Upper Egypt, and Seth,for Lower Egypt, but he was primarily Ho-rus, son of Hathor, who was a cow. Horusis often depicted as a strong bull, and im-ages of cattle are prominent in Predynas-tic and Old Kingdom art; in some in-stances the images of bulls occur withdepictions of stars, a concept that goesback to the Predynastic (Frankfort 1978:172). Dead pharaohs were sometimes de-scribed as the Bull in Heaven. Anotherimportant Old Kingdom concept was Min,the god of rain, who is associated with awhite bull, and to whom the annual har-vest festival was dedicated.

It is interesting to note that the empha-sis on cattle in the belief system of the OldKingdom is not reflected in the economy.While cattle were known and were themajor measure of wealth, the economywas based primarily on agriculture andsmall livestock—sheep and goats. Frank-fort saw this emphasis on cattle as an in-dication that the Old Kingdom beliefswere part of an older stratum of East Af-rican concepts. It seems likely, however,that had Frankfort known that cattle pas-toralists were in the adjacent Sahara sev-eral thousands years before the Predynas-tic, he would have seen the WesternDesert cattle pastoralists as the morelikely source for the Old Kingdom reli-gious beliefs than the East African pasto-ralists. Moreover, that cattle were not im-portant among the preceding Neolithic inthe Nile Valley suggests that the OldKingdom belief system was imposed fromoutside, perhaps in the traditional fashion,a conquest by pastoralists who periodi-cally come in from their ‘‘lands of inso-lence’’ to conquer farmers (Coon 1958:295–323; Khazanov 1994). It is tempting tosuggest that the impressive cattle burialsat the A-Group site of Qustul (Williams

1986), in Egypt south of Abu Simbel, mayrelate to just such an event.

Ethnographic Data Relating to RegionalCeremonial Centers

The Nabta cattle pastoralists and theproposed regional ceremonial center alsomay contribute to a better understandingof the origins of the African Cattle Com-plex. One of the aspects of the modernAfrican Cattle Complex is the regionalceremonial center for groups that are di-vided into sections or lineages. These cen-ters serve as foci of religious, political, andsocial functions for the entire group. Sim-ilar regional ceremonial ceremonial cen-ters occur widely in Sahelian and Sub-saharan Africa, but they are usuallyassigned to the Iron Age or later. The ev-idence from Nabta suggests that regionalceremonial centers probably have a muchgreater antiquity in Africa than has previ-ously believed and suggests that weshould reconsider such a late date for thebeginning of this phenomena. The mega-lithic alignments, cattle tumuli, and cattleconsumption at Site E-75-8 all indicatethat the Nabta Basin was a ceremonialcenter, but it has not been confirmed thatthis center served to integrate separategroups, sections, or lineages.

Many African cattle pastoralists, such asthe Habana and Beni Helba Baggaratribes, who live in the hyperarid area ofnorthern Darfur, and the Gura’an in adja-cent Chad have economies in which hunt-ing and gathering are significant; or theysupplement their cattle resources andgathering activities with a symbiotic rela-tionship with a group of hunters who pro-vide meat (Nicolaisen 1968). Another so-lution is found among the Baggara tribesin northern Kordofan, who not onlygather plant foods, but also use drought-tolerant camels as well as cattle (Asad1970; Lampen 1933; Seligman and Selig-man 1918). A few pastoralists also culti-


vate gardens (Cunnison 1966). The cattlepastoralists at Nabta also must have hadsome dependence on hunting and gather-ing, and perhaps even gardens, as the as-sociated faunal and floral remains indi-cate.

Unfortunately for our purposes, themodern cattle pastoralists living 500 to800 km south of the Egyptian border, innorthern Darfur and Kordofan, such asthe Gura’an, Kababish, and Baggara, whomight be expected to share many burialand religious features with the Nabtagroup, are Moslems, and traces of theirearlier beliefs are scant (Asad 1970;Lampen 1933; Seligman and Seligman1918). Nevertheless, the tribes living innorthern Darfur use cattle for bride pay-ments, to settle blood debts, and to deter-mine wealth and prestige; they never killcattle for their meat except on ceremonialoccasions. Although most groups live inthe desert throughout the year, the Bag-gara who live in northern Kordofan havestrong ties with the Nubians living alongthe Nile near Dongola, and during periodsof extreme drought they move to the river.

The political structures of the northernDarfur tribes usually include an overalltribal leader whose position is inherited inthe male line, and who has final authorityover all disputes and issues regarding thewell-being of the tribe, but the authorityof these leaders is limited, largely becausethe tribe is divided into territorial lineageswhich can function independently. Strongleaders seem to have emerged only attimes of special need, such as warfare orother crisis, and do not seem to have beenable to maintain that authority after thatemergency had passed. Each lineage has aleader who is responsible to the triballeader and whose position is also inher-ited. Probably because of their Moslembeliefs, the ceremonial life of these north-ern Sudanese tribes does not appear toemphasize rain-making, although lineageand tribal leaders sometimes conduct

simple ceremonies seeking rain. There areshrines or sacred places, but very little isknown about them.

Almost all of the animastic tribes livingfarther south, along the Upper Nile, arecattle pastoralists. Cattle dominate theirlives: they are their primary wealth; theyare used to pay bride-payments and bloodfines, and they are the basis for prestige.Among most of these groups the rain-makers are the most common religiousfigures. These rain-makers derive theirpower from ancestral spirits and may beeither the embodiment of their high godor, more frequently, serve as an interme-diary with that god to bring rain, so thegrass will grow and their cattle will flour-ish. The rain-maker is usually the mostimportant person in the tribe; he resolvesdisputes as the final authority, and he isresponsible for all public life. Most arealso wealthy, and there is a documentedcase where an unusually powerful Nuerruler sacrificed numerous cattle and cov-ered them with an earthen mound todemonstrate his importance and wealth(Herskovits 1926: 28). The power of theserain-makers is limited and they also liveprecarious lives; they are often killedwhen rain fails to come, and they are alsokilled when they become ill or grow oldbefore they lose their power. On the otherhand, some of the East African cattle pas-toralists, such as the Shilluk, who lackrain-makers (Seligman and Seligman1932), are led by a king who is regarded asthe embodiment of their god. These kingshave much greater power and they usu-ally control larger groups than the rain-makers.

Many of these tribes in the Upper Nilebuild earthen tumuli, some of which arestill in use. They serve as deliberately con-structed regional centers for groups thatare divided into sections or lineages. Be-cause they are the foci of religious, politi-cal, and social functions for those groups,these regional centers serve to bond the


lineages together. These centers are alsoassociated with themes of sacrifice, death,and burial (Johnson 1990). In some in-stances they become the focal point ofroyal rites and the royal capital itself(Howell and Thompson 1946), althoughmost of them seem not to be connected toa particular settlement. In some instancesthe shrines include mounds built oversacrificed cattle, while other moundscover burials of prominent leaders (Bedri1939: 131; Howell 1948: 53). There are his-toric records that retainers were some-times buried with these leaders (Johnson1990: 49). Myths associated with these re-gional centers also serve to define the ter-ritorial claims of the groups identifiedwith the shrine and to legitimize the au-thority of those with spiritual power(Leinhardt 1961: 98).

An excellent example of the roleplayed by these mound-shrines today isprovided by a modern Dinka shrine,built on the border between severaltribal groups, that has become a focalpoint and national symbol for the South-ern Sudanese Liberation Movement(Johnson 1990: 53). Another examplebuilt in the early part of this century,was erected when an unusually power-ful Nuer prophet stimulated the con-struction of a huge conical earthenmound, 100 m in diameter and 15 mhigh, and surrounded around the basewith numerous elephant tusks (Evans-Pritchard 1956: 305–306; Seligman andSeligman 1932: 231). Evidently this wasdone by volunteers over a period ofmany years without conscripted labor.

There are many other kinds of shrinesused in this area, most of them simpledecorated poles, referred to as ‘‘mobileshrines,’’ but among the Bari and the Lo-tuko the rain-maker shrines consist of acircle of large upright stones with a mo-saic of smaller flat stones in the center(Seligman and Seligman 1932: 288; 330).The Nuba also have circles of large up-

right stones with smaller flat stones in thecenter that are used by the men when theyperform the new fire ceremony (Seligmanand Seligman 1932: 343–344), while amongthe Kalenjin in Kenya, tribal elders some-times sit against upright stones set in acircle (Posnansky 1966).

Most of the modern Nilotic cattle pas-toralists bury their dead in simple, shal-low graves with a small decorated stick orpole shrine nearby. Cattle are sometimessacrificed as part of the ceremony, partic-ularly for their leaders and the wealthy.Burial among the Nuba and the Moro,however, is in chambers from 2 to 3 mbelow the surface and about 2.5 m in di-ameter that are reached by shafts dugfrom the surface (Seligman and Seligman1932: 404 and 486).

Some Archaeological Ceremonial Centers

The archaeological literature for Sahe-lian and Sub-saharan Africa record nu-merous presumed regional ceremonialcenters with megalithic alignments, burialmounds, and stone circles similar to, butnot identical with, those at Nabta. Theseceremonial centers occur from Ethiopia toSenegal and north to the Maghreb(Camps 1953; Connah 1987; Desplagnes1951; Fergusson 1872; Joussaume 1974,1985; Milburn 1988; Tilner 1981). They areparticularly abundant in West Africawhere there are literally thousands of tu-muli and megaliths (Martin and Becker1974, 1984). Only a few of these tumuli andmegaliths have been dated, but they areusually assigned to the Iron Age or later.There are two older, but rejected radiocar-bon dates of 7440 and 6700 B.P. associatedwith megaliths in the Central African Re-public (Vidal 1969, Bayle des Hermens1975: 260–261).

The archaeology of the Sahara innorthern Sudan is little known (Kuper1986; Richter 1989; Schuck 1989), butnear Malha Crater in northern Darfur


there are numerous earthen mounds,some of which are very large, indicatingthat a rich ceremonial and burial com-plex existed there in the past. Many ofthese mounds occur near large, late pre-historic ‘‘cities’’ that are segmented intodistinct units and special precincts. Thearrangements of the towns suggest mul-tiple sections or lineages. Very littlework has been done at these sites, butthey are tentatively dated between 3000and 4000 B.P., when the lake sedimentsin the crater indicate an interval ofgreater precipitation (Dumont et al.1993). Of interest here is the erection ofburial mounds in special precincts awayfrom the settlements, which resemblesthe situation of the megalithic structuresat Nabta. These Malha sites could wellhave served as regional ceremonial cen-ters. It is also useful to note that thesesettlements indicate, at the very least,that large and complex groups couldfunction successfully in areas of verylimited rainfall, but we do not know ifthey were farmers or mixed farmers andpastoralists.

A Final Comment

Around 6200 cal B.P. the modernphase of hyperaridity began in the East-ern Sahara and the area was abandoned.It has been suggested that the move-ment of these perhaps better organizedLate Neolithic cattle pastoralists to theNile Valley, and the resulting turmoil,was a critical factor in the rise of socialcomplexity and the subsequent emer-gence of the Egyptian state in UpperEgypt (Hoffman 1979; Hassan 1988). Ifso, Egypt owes a major debt to thoseearly pastoral groups in the Sahara; theymay have provided Egypt with many ofthose features that still distinguish itfrom its neighbors to the east. Whiletempting, this hypothesis must beviewed with caution. Many features

which characterize the ceremonial as-pects of the regional center at Nabta areas yet unknown in the Nile Valley. Theseinclude the megalithic alignments, themegalithic structures and worked tablerocks, cattle burials in chambers built instone-covered tumuli, and calendar cir-cles. One of the fascinating aspects ofthe evidence for the working of largestones is that it seems to anticipate laterEgyptian developments. If the Saharanpeople contributed significantly to therise of complexity in the Predynastic, theprecise nature of those contribution hasyet to be defined.

REFERENCES CITED

Asad, Talal1970 The Kababish Arabs. Power authority and con-

sent in a nomadic tribe. Hurst, London.

Banks, K. M.1980 Ceramics of the Western Desert. Appendix

3. In Prehistory of the Eastern Sahara, editedby F. Wendorf and R. Schild, pp. 300–315.Academic Press, San Diego.

Barakat, Hala1995 Contribution Archeobotanique a l’Histoire de

la Vegetation dans le Sahara Oriental etSoudan Central. Ph.D. dissertation, Univ.de Droit, d’Economie et des Sciencesd’Aix-Marseille.

Bar-Yosef, O.1985 The Stone Age of the Sinai Peninsula. In

Studi de Paletnologia in Onore de Lavatore M.Puglisi, pp. 107–122.

Bayle des Hermens, R.1975 Recherches prehistoriques en Republique Centraf-

ricaine. Librairie C. Klincksieck, RecherchesOubanguiennes, 3. Universite de Paris X.

Bedri, I.1939 Dinka beliefs in their chiefs and rainmakers.

Sudan Notes and Records 22:125–131.

Bradley, Daniel G., David E. MacHugh, Patrick Cun-ningham and Ronan T. Loftus.

1996 Mitochondrial diversity and the origins ofAfrican and European cattle. Proceedings ofthe National Academy of Science USA 93:1531–1535.

Camps, G.1953 Les domens de Beni-Messous. Libyca 1:329–

372.


Caton-Thompson, G.1952 Kharga Oasis in prehistory. Athlone, Lon-

don.

Close, Angela E.1995 Few and far between: Early ceramics in

North Africa. In The emergence of pottery:Technology and Innovation in Ancient Societies,edited by William K. Barnett and John W.Hoopes, pp. 23–37. Smithsonian InstitutionPress, Washington and London.

Clutton-Brock, J. (Editor)1989 The Walking Larder. Patterns of Domestication,

Pastoralism, and Predation. Unwin Hyman,London.

Clutton-Brock, J.d1993 The spread of domestic animals in Africa. In

The archaeology of Africa: Food, metal andtowns, edited by T. Shaw, P. Sinclair, B. An-dah, and A. Okpoko, pp. 71–103. RoutledgePress, London.

Gautier, A.1980 Appendix 4. Contributions to the archaeo-

zoology of Egypt. In Prehistory of the EasternSahara. edited by F. Wendorf and R. Schild.Academic Press, New York.

Connah, G.1987 African civilizations. Cambridge Univ. Press,

Cambridge.

Coon, C. S.1958 Caravan: The story of the Middle East. Henry

Holt, New York.

Cunnison, I.1966 Baggara Arabs: Power and the lineage in a Su-

danese Nomad tribe. Clarendon, Oxford.

Dahlberg, Jeffery A., John Evans, Elizabeth Johnson,Marian Hyman, Ed Biehl, and Fred Wendorf

1995 Attempts to identify 8000-BP sorghum usingimage-analysis, infrared spectroscopy, andbiotechnological procedures. Acta Palaeobo-tanica 35(1):167–173.

Desplagnes, L.1951 Fouilles du tumulus d’El Qualedjl (Soudan).

Bulletin Institut Fondamental de l’Afrique Noire13:1159–1173.

Dumont, Henri J., and Asim I El Moghraby1993 Holocene evolution of climate and envi-

ronment, and stone ‘‘city’’ ruins in north-ern Darfur, Sudan: is there a relationship?In Environmental change and human culturein the Nile Basin and Northern Africa untilthe second millennium B.C., edited by LechKrzyzaniak, Michal Kobusiewicz, andJohn Alexander, Studies in African Ar-chaeology, Vol. 4, pp. 381–397. Archaeo-logical Museum, Poznan.

Evans-Pritchard, E. E.1956 Nuer religion. Clarendon Press. Oxford.

Fergusson, J.1872 Rude stone monuments in all countries: Their

age and uses. John Murray, London.

Frankfort, H.1978 Kingship and the gods. A study of ancient Near

Eastern religion as the integration of society andnature. Univ. of Chicago Press, Chicago.

Gautier, A.1968 Mammalian remains of the northern Sudan

and southern Egypt, edited by F. Wendorf,Vol. 1, The Prehistory of Nubia, pp. 80 –99.Southern Methodist University Press.Dallas.

1984 Archaeozoology of the Bir Kiseiba region.Eastern Sahara. In Cattle keepers of the East-ern Sahara: The Neolkithic of Bir Kiseiba, As-sembled by F. Wendorf and R. Schild, ed-ited by A. E. Close, pp. 49–72. Departmentof Anthropology, Institute for the Study ofEarth and Man, Southern Methodist Uni-versity, Dallas.

1987 Prehistoric men and cattle in North Africa:A dearth of data and a surfeit of models. InPrehistory of arid North Africa: Essays in honorof Fred Wendorf, edited by A. E. Close, pp.163–187. Southern Methodist Univ. Press,Dallas.

Gautier, A., and W. Van Neer1989 Animal remains from the Late Paleolithic

sequence at Wadi Kubbaniya. In The prehis-tory of Wadi Kubbaniya, assembled and ed-ited by F. Wendorf, R. Schild, and A. E.Close, Vol. 2, pp. 119–161. Southern Meth-odist Univ. Press, Dallas.

Gautier, Achilles, Romuald Schild, Fred Wendorf,and Thomas W. Stafford, Jr.

1994 One elephant doesn’t make a savanna.Palaeoecological significance of Loxodontaafricana in the Holocene Sahara. Sahara 6:7–14.

Hassan, F. A.1985 Radiocarbon chronology of Neolithic and

Predynastic sites in Upper Egypt and theDelta. African Archaeological Review 3:95–116.

1988 The Predynastic of Egypt. Journal of WorldPrehistory 2:135–185.

Herskovits, M. J.1926 The cattle complex in East Africa. American

Anthropologists 28:230–272; 361–380; 494–528;633–664.


Hoffman, M. A.1979 Egypt before the pharaohs. The prehistoric foun-

dations of Egyptian civilization. Knoff, NewYork.

Howell, P. P.1948 ‘‘Pyramids’’ in the upper Nile region. Man

48:52–53.

Howell, P. P., and W. P. G. Thompson.1946 The Death of the Reth of the Silluk and the

installation of his successor. Sudan Notes andRecords 27:5–85.

Johnson, D. H.1990 Fixed shrines and spiritual centers in the

upper Nile. Azania 25:41–56.

Joussaume, R.1974 Le Megalithisme en Ethiopie. Museum Na-

tional d’Histoire Naturelle, Paris.1985 Dolmens for the dead, megalith building

throughout the world. Cornell Univ. Press,Ithaca, NY.

Khazanov, Anatoly M.1994 Nomads and the outside world. Univ. of Wis-

consin Press, Madison.

Kingdon, J.1982 East African mammals: An atlas of evolution in

Africa, Volume 3, Part C, Bovids, and Part D,Bovids. Academic Press, London.

Kropelin, Stephan.1993 The Gilf Kebir and Lower Wadi Howar: Con-

trasting Early and Mid-Holocene environ-ments in the Eastern Sahara. In Environmentalchange and human culture in the Nile Basin andnorthern Africa until the second millennium B.C.,edited by Lech Krzyzaniak, Michal Ko-busiewicz and John Alexander, Studies in Af-rican Archaeology, Vol. 4, pp. 249–258.Poznan Archaeological Museum, Poznan.

Kubiak-Martens, Lucyna, and Krystyna Wasylikowa1994 Sorgo ze Stanowisak Wczesnoneolity Cznego

Nabta Playa w Poludniowym Egipcie. PolishBotanical Studies, Guide Book Series 11:109–119.

Kuper, R.1986 Wadi Howar and Laquiya. Recent field

studies into the early settlement of North-ern Sudan. In Nubische Studien. Proceedingsof the Symposium of the International Society ofNubian Studies, Heidelberg 1982, edited byM. Krause, pp. 129–136, Verlag Philip vonZabern, Mainz AM Rhein.

Lampen, G. D.1933 The Baggara Tribes of Darfur. Sudan Notes

and Records XVI:97–118.

Lienhardt, R. G.1961 Divinity and experience: The religion of the

Dinka. Clarendon, Oxford.

Malville, J. McKim, F. Wendorf, A. Mazar, and R.Schild

1998 Megaliths and Neolithic astronomy insouthern Egypt. Nature 392:488–491.

Martin, V., and C. Becker1974 Vestiges Protohistoriques et Occupation

Humaine. Annales de Demographie Histo-rique: 403–429.

1984 Inventaire des Sites Protohistoriques de la Sene-gambie. Kaolack, CNRS, Paris.

McHugh, W. P., C. S. Breed, G. G. Schaber, J. F.McCauley, and B. J. Szabo

1988a Acheulian sites along the ‘‘Radar Rivers,’’southern Egyptian Sahara. Journal of FieldArchaeology 15:361–379.

McHugh, W. P., J. F. McCauley, C. V. Haynes, C. S.Breed, and G. G. Schaber

1988b Paleorivers and geoarchaeology in thesouthern Egyptian Sahara. Geoarchaeology3:1–40.

McHugh, W. P., G. G. Schaber, C. S. Breed, and J. F.McCauley

1989 Neolithic adaptation and Holocene func-tioning of Tertiary paleodrainages in south-ern Egypt and northern Sudan. Antiquity 63:320–336.

Milburn, M.1988 A typological enquiry into some dry-stone

funerary and cult monuments of the Sahara.Scientific Reviews on Arid Zone Research (Jod-pur) 6:1–126.

Muzzolini, A.1989 La ‘‘neolithisation’’ du Nord de l’Afrique et

ses causes. In Neolithisations, edited by O.Aurenche and J. Cauvin, pp. 145–186. BARInternational Series 516. Oxford.

Neumann, Katharina1989 Zur Vegetationsgeschichte der Ostsahara im Ho-

lozan. Holzkohlen aus Prahistorischen Fund-stellen. Mit Einem Exhurs uber die Holzhohlenvon Fachi-Dogonbolo Nigher. Thesis, Univ. ofFrankfurt.

Nicolaisen, J.1968 The Haddad—a hunting people in Tchad.

Preliminary report of an enthnographicalreconnaissance. Folk 10:91–109.

Nordstrom, H. A.1972 Neolithic and A-Group sites. The Scandinavian

joint expedition to Sudanese Nubia, Vol. 3,Parts 1 and 2. Scandinavian Univ. Press,Stockholm.

Posnansky, M.1966 Kingship, archaeology and historical myth.

Uganda Journal 30:1–12.


Richter, J.1989 Neolithic sites in the Wadi Howar (Western

Sudan). In Late prehistory of the Nile Basinand the Sahara, edited by L. Krzyzaniak andM. Kobusiewicz, pp. 431–450. Poznan Ar-chaeological Museum, Poznan.

Schild, R., M. Chmielewska, and H. Wieckowska1968 The Arkinian and Sharmarkian industries. In

The prehistory of Nubia, edited by F. Wendorf,pp. 651–767. Fort Burgwin Research Centerand Southern Methodist Univ. Press, Dallas.

Schild, R., and F. Wendorf1977 The prehistory of Dakhla Oasis and adjacent

desert. Polish Academy of Sciences, Osso-lineum, Warsaw.

Schuck, W.1989 From lake to well: 5,000 Years of settlement in

Wadi Shaw (Northern Sudan). In Late prehistoryof the Nile Basin and the Sahara, edited by L.Krzyzaniak and M Kobusiewicz, pp. 421–429.Poznan Archaeological Museum, Poznan.

Seligman, C. G., and Brenda Z. Seligman1918 The Kababish, A Sudan area tribe. In Har-

vard African studies, Vol. II, pp. 104–184. Har-vard Univ. Press, Cambridge, MA.

1932 Pagan tribes of the Nilotic Sudan. Routledgeand Sons, London.

Smith, A. B.1984 The origins of food production in northeast

Africa. In Palaeoecology of Africa and the sur-rounding islands, edited by J. A. Coetzee andE. M. Van Zinderen Bakker, Vol. 16, pp.317–324. Balkema, Rotterdam.

1992 Pastoralism in Africa. Origins and developmentecology. Hurst, London.

Tilner, E.1981 Steinsetzungen, Monuments und Gruber in

der Sahara. Antikewelt 12(4):14–26.

Vermeersch, P. M.1978 El Kab II. Univ. Pers. Leuven.

Vidal, P.1969 La civilisation megalithique de Bouar, prospec-

tion et fouilles 1962–1966. Recherches ou-bangiennes I, Firmin-Didot.

Wasylikowa, Krystyna, J. R. Harlan, J. Evans, FredWendorf, Romuald Schild, Angela E. Close, andHalina Krolik

1993 Examination of botanical remains fromearly Neolithic houses at Nabta Playa,Western Desert, Egypt, with special refer-ence to sorghum grains. In The archaeologyof Africa: Food, metal and towns, edited by T.Shaw, P. Sinclair, B. Andah, and A. Okpoko,pp. 154–164. London, Routledge.

Wasylikowa, Krystyna, and Lucyna Kubiak-Martens1995 Wild sorghum from the Early Neolithic site

at Nabta Playa, South Egypt. In Res archaeo-botanicai—9th Symposium IWGP, Kiel, 1995.

Wasylikowa, Krystyna, Romuald Schild, Fred Wen-dorf, Halina Krolik, Lucyna Kubiak-Martens, andJack R. Harlan

1995 Archaeobotany of the Early Neolithic SiteE-75-6 at Nabta Playa, Western Desert, SouthEgypt. Acta Palaeobotanica 35(1):133–155.

Wasylikowa, Krystyna, J. Mitka, A. Walanus, F. Wen-dorf, and R. Schild

In Distribution of plant macrofossils within apress settlement: A case study on plant exploitation

manners in a hunter-gatherer site at NabtaPlaya, south Egypt. In Bulletin of the XIII Con-gress of the International Union of Prehistoric andProtohistoric Sciences, Forli, Italy, 1996.

Wendorf, F., A. E. Close, and F. Schild1985 Prehistoric settlements in the Nubian

Desert. American Scientist 73:132–141.1987a A survey of the Egyptian radar channels: An

example of applied archaeology. Journal ofField Archaeology 14:43–63.

1987b Early domestic cattle in the Eastern Sahara.Palaeoecology of Africa 18:441–448.

Wendorf, F., A. E. Close, R. Schild, and K. Was-likowa

1991 The Combined Prehistoric Expedition: Resultsof the 1990 and 1991 seasons. Newsletter of theAmerican Research Center in Egypt 154:1–8.

Wendorf, F., A. E. Close, and R. Schild1993 Megaliths in the Egyptian Sahara. Sahara

5:7–16.Wendorf, F., A. E. Close, R. Schild, K. Wasylikowa,R. A. Housley, J. R. Harlan, and H. Krolik

1992 Saharan exploitation of plants 8000 yearsago. Nature 359:721–724.

Wendorf, F., and R. Schild1976 Prehistory of the Nile Valley. Academic Press,

New York.1980 Prehistory of the eastern Sahara. Academic

Press, New York.Wendorf, F., and R. Schild, assemblers, edited byA. E. Close

1984 Cattle keepers of the eastern Sahara: The Neo-lithic of Bir Kiseiba. Department of Anthro-pology, Southern Methodist Univ., Dallas.

Wendorf, F., R. Schild, A. E. Close, and associates1993 Egypt during the last Interglacial. The Middle

Paleolithic of Bir Tarfawi and Bir Sahara East.Plenum, New York.

Wendorf, Fred, and Romuald Schild1994 Are the Early Holocene cattle in the eastern

Sahara domestic or wild? Evolutionary An-thropology 3(4):118–128.


1995/ Nabta Playa during the Early and Middle1996 Holocene. ANKH 4/5:33–45.In Tumuli cattle burials and society in thepressa

eastern Sahara. In Proceedings of the Forumfor African Archaeology, Rome, 1995.

In A Late Neolithic regional ceremonial centerpressb

and ranked (?) groups in the Egyptian Sa-hara. In Bulletin of the XIII Congress of theInternational Union of Prehistoric and Protohis-toric Sciences, Forli, Italy, 1996.

Wendorf, F., R. Schild, K. Wasylikowa, J. Dahlberg, J.Evans, and E. Biehl

In The use of plants during the early Holo-press cene in the Egyptian Sahara: Early Neo-

lithic food-economies. Bulletin of the XIIICongress of the International Union of Pre-historic and Protohistoric Sciences, Forli, It-aly, 1996.

Wendorf, Fred, Romuald Schild, and N. ZedenoIn A Late Neolithic megalith complex in thepress Eastern Sahara: A preliminary report. In In-

terregional contacts in the late prehistory ofnortheastern Africa, edited by L. Krzyzaniakand M. Kobusiewicz. Poznan Archaeologi-cal Museum and Polish Academy of Sci-ences, Poznan Branch, Poznan.

Wendt, E.1966 Two prehistoric archaeological sites in

Egyptian Nubia. Postilla 102:1–46.

Williams, B. B.1986 Excavations between Abu Simbel and the Sudan

Frontier: The A-Group Royal Cemetery at Qus-tul: Cemetery L, Vol. III, Part 1. The OrientalInstitute of the Univ. of Chicago, Chicago.


nabta playa and its role in northeastern african prehistory

Documents

early holoceneamong

early stage

early middle pleistocene

nabta playa basin

culture history of early

cattle burials

historic archaeology

complete desert