Nyame Akuma Issue 038

8/10/2019 Nyame Akuma Issue 038

1/74

NY ME AKUMA No 38 ecember 1992

It seems that somebody actually readst is page, i one may interpret the avalancheof contributions received during the last halfof 1992 as a response to my request forarticles in the NA 36 editorial. At any rate, Iam grateful for the outpouring ofsubmissions and apologize to authors whosearticles do not appear in t is issue; most ofthem will e included in NA 39, the June1993 issue. I also apologize for the delay in

production of the present issue, which ispartly a consequence of the chaosoccasioned by my move to the Northlandand partly a result of the logisticaldifficulties that stem from my being hererather than in Iowa (e.g., the desktopproduction is still done in Ames andprinting is now done at GainesviUe, Florida).Such problems should be substantiallyunder control by June so I hope that NA 38wil be mailed more or less on schedule.

Continuing with my optimisticassumption about the reading of t is page,and hoping that I am not pressing my lucktoo hard, I would like to issue yet anotherplea for care in the preparation of manu-script for Nyame Akuma, particularly asregards bibliography. Many articles containreferences that are incomplete or obviouslyerroneous; such sins of commission arefrequently matched by the other kinds of sinin the form of missing references. Given

Nyama Akuma s mission to publish timelyreport s of ongoing work, it seemsinappropriate to engage in extended author-editor communications about minor errors,so my practise has been to make the best Ican of the material I receive, deleting (forexample) references in the text that are notlisted in the bibliography and entries in thebibliography that are not referenced in thetext of an article. I suppose that my takingsuch editorial liberties may appear arbitrary

in some cases. Two remedies are available toauthors: 1) avoiding the manuscript errorsthat lead to my taking undue liberties in thefirst place or 2) submitting errata in the nextissue of Nyame Akuma, which I willcheerfully include on this page.

In closing, I would like to call attentionto the editorial page of the 1992 issue (Vol.10) of The African Archaeological Review, inwhich David Phillipson mentions tentativeplans for the Panafrican Congress ofPrehistory Quaternary Studies to meet inZimbabwe in 1995. An address for enquiriesabout the Congress is included in theeditorial.


2/74

N YA M E A K U M A N o 38 December 1992

BOTSW N

Kutswe Pass Segobye 1989). The presentstudy seeks to establish the distribution ofsites beyond the hills. Current land use andgeographical criteria were used to divide the

area into three landforms. These consistedoE

Landform A The grazing area south of

rchaeological Survey andExcavations in EasternBotswana

Alinah K. SegobyeDepartment o ArchaeologyUniversity of CambridgeDowning Street CB2 3 0 2Cambridge England

the hills--open plains

Landform B: Limited grazing area-TheMokgware ills range

Landform C: The cultivated area north ofthe hil ls-open plains

The environmental features of thesurvey area can be summarized as follows:

Geology. Dolerites and graniticgneisses and migmatites characterise themain hill range. These comprise weatheredand steep-sided hills with flat plateaux.Sandstone outcrops form the younger rocky

This report summarises results of landscape. These also have weathered flat

fieldwork carried out in east central plateaux where midden sites were found.

Botswana between January 1991 and April Mudstone, ironstone, calcrete, marble, shale,

1992. Archaeological survey and excavations and conglomerates occur in specific isolated

were undertaken in the Mokgware Hills locales.

area to locate, map and describe the Terrain. The highest points occur on thedistribution of sites associated with farming Mokgware range. The area is dominated bycommunities dating from the end of the undulating valleys and plains north andtenth to the fifteenth century A D Similar south of the hills. It is bordered to the westsites have been located and mapped over by the Kalahari sandveld.much of eastern Botswana Denbow 1983; Hydrology. Surface water is scarce. TheKiyaga-Mulindwa 1992). Most of these sites plains are cross cut by intermittent streamsconsist of hilltop middens with ashy soils and dry deep-cut river beds which flowcovered with Cen chrus ciliaris grass. They are eastw ards towards the Lotsane an dassociated with cattle keeping, functioning Limpopo rivers. The water table is relativelyas stock pens . Similar sit es like close to the surface and people mainlyMapungubwe and Schroda in the northern depend on wells and boreholes for regularTransvaal Hanisch 1981; Voigt 1983) have water supplies.been shown to have supported livestock Soils. Local soils are influenced by thebased economies. Some of the hilltop sites geology with reddish soils derived from theare associated with ry stone walling. Some dolerite parent rock. Tropical ferruginousmidden sites occur in the plains north and soils predominate. Claysols occur along thesouth of the hills therefore covering diverse river system interspersed with alluvialterrain in eastern Botswana. sands and loam soils. Fine white aeolian

Kalahari sands influence the western partsThe Survey fringing the sandveld.

A 40 x 20 m area was chosen for study Vegetation. cacia tree savanna coversaround the Mokgware Hills. The area was most of the southern plains with woodlanddelimi ted with the a id of 1:50,000 species occurring in the hilly areas. Opentopographic maps and aerial photographs. Colophosperrnurn rnopane bushveld and caciaAn earlier reconnaissance survey had spp. scrub bush dominates the plains northestablished the presence of sites around the of the hills. Several other species occur.


3/74


The Kalahari land system forms thewestern border of the study area. It is a largesemi-arid depession with wind-blown sandsand associated scrub vegetation.

A 5 sample was used as the basis foractual coverage. A multi-stage samplingstrategy was adopted to ensurerepresentative coverage of all the landforms.It involved the use of stratified, systematicand random sampling techniques. Inaddition, some sites were surveyed on thebasis of information from local residents;others had been mapped in an earlierreconnaissance (Segobye 1989). Forty m x

km sampling units were randomlyselected for ground coverage. These blockswere systematically traversed on footfollowing linear transects where possible.Field walkers were spaced about 20 metresapart though on hilly terrain and densevegetation areas the distance was less.

More than 80 sites were located in thesurface survey. At least 60 were ash dungmiddens covered with Cenchrus ciliaris grass.The middens varied in diameter from 25 to

over 200 metres (Table 1). Most of the siteswere not yet documented in the NationalMuseum sites register. They were allrecorded on detailed record forms designedfor this study in addition to the NationalMuseum site register forms. The sites weresketched, photographed, and mapped on1:50,000 maps cross-referenced with the aidof aerial photographs. The largest middenstended to be located on the plateaux ofsteep-sided, weathered dolerite hills.

Both the hilltop and low-lying sites wereassociated with surface occurrences ofpottery, bones (mainly on the ashy soil), ironslag, slag encrusted tuyere fragments anddaga, pole impressed daga, and grindingstones. Pottery occurred in abundance onmost sites, and diagnostic sherds werecollected for ceramic analysis. Other limitedfinds included glass beads, ostrich egg shellbeads, cowrie shell,' and metallic objects like

iron rings and copper bracelets.

inthe middle (Fig. 1). These features occuron various other sites in Botswana, likeToutswemogala, Bosutswe, Lose, Letsibogo,and Lower Shashe (Denbow 1983; Kiyaga-

Mulindwa 1992; Campbell 1991). They arealso present in the northern Transvaal onMapungubwe site (Fouche 1937) and in theTransvaal highveld and southeasternsouthveld (Maggs 1976). Though a fullregional study of these features has not beenundertaken to determine their temporal andspatial extent, some researchers have calledthem grain storage bases for grass or claygranaries. They are thought to be similar tothose used by Sotho-Tswana and Kalangapeople in the nineteenth and early twentiethcenturies to support large grain baskets, orsesigo (Campbell 1991). Those noted in thestudy area must be examined in detailbefore conclusive statements can be madeabout their function.

Excavations were undertaken at twosites in Landform C They have ash dungmiddens covered with Cenchrus ciliaris and afairly high artefact density per square metre.

The excavations were carried out to obtainmaterial for C14 dating. series of dateshave been obtained which place the sitesbetween the tenth and fifteenth centuriesA D (Table 2 . Floral and faunal remains andother artifacts were recovered which helpedin interpreting the sites' occupation record.The excavations suggest that the sites aremulti-period with several occupation levelsindicated, especially at Phate Hill site.

The xcavated Sites

Phate Hill site--Grid Reference 658929:This site is located on a low hill next to theKutswe river. It is a large site spread overthe top and gentle northern slope of the hill.Extensive terracing and stone circles areevident on the site. The central midden ismore than 150 metres in diameter. It iscovered with Cenchrus ciliaris and hasvitrified slag material in places. his slag has

been associated with burnt cow dung matterStone features associated with these as these middens are thought to haie been

sites include heaped stone cairns, circular livestock penning units . Res identialstone enclosures, and dry stone walling. The settlement remains found on the surfacestone circles varied in form. The most include daga floors with pole-impressedcommon form was a circle of approximately daga.1.5 metres diameter with a single stone set


4/74

NYAME AKUMA No. 38 December 1992

Table 1. The distribution of ash dung middens by size and landform.

Landform

Midden Diameter metres) B C Total0-50 5 7 10 2250-100 5 4 13100-150 3 1 2 6150-200 0 0 0 0200 1 1 1 3Total 14 13 17

Table 2. C14 dates from Mogware Hills age years BP).

Site: Phate Hill

a TO-3213TO-3214TO-3215TO3211

TO-3212Pta-5833Pta-5825Pta-5826Beta-51090Beta-51092Beta-51094Beta-51091

Beta-51095Beta-51093

Site: BPS 52

Pta-5827Pta-5828Pta-5829Beta-51097Beta-51096

Beta-51098

aIsotrace Radiocarbon Laboratory

b~~~~~~ -Pretoria

CBetaAna lytic Laboratory, Inc


5/74


Five trenches of x 1.5, 2 x 1 2 x 1.5,4 x 1, and x 1 metres were excavated. Theywere excavated to sterile levels with thedeepest reaching 1.6 metres. The

stratigraphy revealed several occupationlevels with three house floors identified inone of the profiles. Features like hearthremains were also found. Artifactsrecovered included pottery, ostrich egg shellbeads, metal objects, iron slag and tuyerefragments, grin ing stones, and some wholepots. Good samples of charcoal wererecovered. A human burial was found at 1.1metres in trench 5 but was left in situ as timeconstraints did not permit full excavation.

PS 52 si te -G ri d Reference 654978: Thesite is located on a low ridge in an openColophospermum mopane bushveld. It has acentral ash du ng midden of 100 metresdiameter. The midden has been disturbed byburrowing animals. One large trench and atest pit were sunk over an area with featuresand dense artifact cover. A total of thirteen 1x 1 metre squares were excavated. Thesewere excavated to a depth of 0.5 metres tosterile soil levels. Subsurface finds includedpottery, bones, charcoal, ostrich egg shellbeads, iron slag, and tuyere fragments.Other features recorded included hard-packed, floor-like levels with cementedpottery fragments. There were alsooccurrences of b urn t ashy soil withfragments of vitrified material. Sampleswere collected for further examination.

Discussion

The two excavations have providedmaterial for dating and the series ofradiocarbon dates available now place theMokgware hills sites securely within theregional framework of farming communitiessites. One of the aims of the study is todetermine whether changes in land usepatterns can be discerned in both time andspace in the study area. From the surfacesurvey, there seems to have been apreference for hilly areas for settlementlocation. Where the settlements were locatedin the plains, slightly elevated ground waspreferred. However, activity areas likeironworking sites seem to have been locatednear water sources as furnace remains wererecorded mainly along river banks. The

excavations indicate dense deposits of ashyburnt soil with vitrified slag recorded insubsurface strata to depths of 0.8 metre.

As the analyses are not completed,

conclusive remarks cannot be made aboutsite distribution patterns or the settlementand land use systems. However, preliminaryobservations suggest that settlement locationand land use strategies aimed at maximisingresource use were key in the decision-making process of the area s occupantsthrough time. The environmentalreconnaissance suggests that naturalresources like iron ores, clays, land, andwater may have been crucial within the

subsistence economies and influenced thechoice of settlement location. The Kalahariland system, which is ecologically differentfrom the study area, may have constrainedsettlement expansion westwards for theearly farming communities. This study isaimed at contributing toward understandingthe long-term management of subsistenceproduction systems and settlementorganization within the agro-pastoralsubsistence economies of eastern Botswana.

cknowledgements

This project was funded by thefollowing institutions: The University ofBotswana, The Swedish Agency forResearch Cooperation with DevelopingCountries (SAREC), The SmutsFoundation-Cambridge and The BoiseFund--Oxford. Their support is gratefullyacknowledged. Permission to undertake thisresearch was granted by the government ofBotswana. I wish to thank my fieldassistants, hosts in the field, and the staff ofKhama Memorial Museum for providinga base for this project. Several colleagues inthe University of Botswana and Departmentof Archaeology in Cambridge have helpedwith invaluable advice for this study; Ithank them. Above all, I thank mysupervisor Dr David Phillipson for hiscontinued support during the course of thisstudy.


6/74

NY M E AKUMA No 38 December ' ^

Fig 1 Stone features fr om M okgwar e sites

6


7/74

N Y M E K U M

References

No 38 December 1992

Campbell, A C.

1991 Distribution of prehistoric Kalangavillages in the Lower Shashe andLetsibogo areas. In Van Waarden, C.ed.), Kalanga Retrospect and Prospect.

Gaborone: Botswana Society:19-33.

Denbow, J

1983 lron A ge Economics: Herding Wealth andPolitics along the Fringes of the Kalahariduring the Early lron Age. UnpublishedPh.D thesis, Indiana University,Bloomington.

Fouche, L. ed.)

1937 Mapungubwe: Ancient Bantu Civilisationon the Limpopo.Cambridge: CambridgeUniversity Press.

Hanisch, E 0

1981 Schroda: A Zhizo site in the northernTransvaal. In Voigt, E ed.), Guide toArchaeological Sites i n the Northern and

Eastern Transvaal. Pretoria: TransvaalMuseum37-54.

Kiyaga-Mulindwa, D.

1992 Excavations at Lose Enclosure, CentralBotswana. In Sinclair, P., and Pwiti, G

eds.), Urban Origins in Eastern AfricaProceedings of the 7990 Workshop Harareand Great Zimbabwe. Paper No. 6.Stockholm: The Central Board ofNational Antiquities.

Maggs, T. M.

1976 Iron Age Communities of the SouthernHighveld. Pietmaritzburg: NatalMuseum.

Segobye, A K.

1989 Later Prehistoric Settlement Patterns inEastern Botswana: A Survey of the CentralMokgware Hills. Unpublished MPhil.

Thesis, University of Cambridge.

Voigt, E

1983 Map ungu bwe : A n ArchaeozoologicalInterpretation of an Iron Age Community.Pretoria: Transvaal Museum.

Les Fosses d okolo(Sud du Cameroun): Fouilleset Axes de Recherches

Christiane Atangana17 rue du Docteur Goujon75 1Z ParisFrance

Les' fouilles menbes Okolo-en aoOt1983, sous la direction de P. de Maret et enmai/juin 1984 par nous-interviement I lasuite des recherches entreprises I Obobogoen 1980, 1981, et 1983 par P. de Maret de1'Universite Libre de Bruxelles et INkometou en 1983 et 1985 p r M. Essombade la Facult6 des Lettres de Yaounde. Ladecouverte au cours des prospectionsmenees en 1983 de douze autres sitespresentant des caracteristiques semblables-

fosses apparaissant su r la sectionartificiellement cr&e de versants collinaires,ceramique d6corCe au peigne et aubbtonnet-nous imposait d'avoir uneapproche globale de ce corpus.

Ainsi, tirant parti de la relativehomogCn6it6 de ce corpus et de la serie dedatations obtenues, nous en avons extrait lesite d'Okolo, considere comme une partie du

subs trat mat6riel d'u ne soci6t6prehistorique (settlement) dont la naturereste I determiner. Dans le but d'emettrequestions et hypotheses sur le mode defonctionnement de cette soci4tC nous noussommes attach& I comprendre certainsaspects des technologies apparaissant sur cesite et le mode d'exploitation del'enviromement qu'elles supposent, autantd16Mments susceptibies d'eclairer lesinterrelations autant ces sites entre eux etd'orienter les recherches 2 venir.

Les Fouilles a Okolo

Celles-ci s'inscrivent dans le cadre d'unsauvetage des vestiges situes en bordure deroute.


8/74

NYAME AKUMA No 38 December 1992

La section du double talus pratiqukeOkolo (voir Fig. I , propice l'observationdes vestiges, a 6te faite en fonctiond'impkratifs autres qu'arch~ologiques et en

l'absence d'arch6ologues ayant pu constaterles destructions occasionnees par les travauxde terrassement. Ces informations ont et.6completees par la fouille de fosses --en1983 fouille de la fosse no 3 par P. Claes(etudiant l'Universit6 Libre de Bruxelles),et des fosses no 2 et 9 par nous-memes-et l'implantation de sondages, dont certainsont 6t6 progressivement agrandis-secteur defouille no 3, sondage no 1 et 2, en 1984.

Les donnees stratigraphiques recueilliesdans le secteur de fouille no 3 montrent quele niveau le plus ancien correspond auniveau de creusement des fosses ; il estrecouvert par un important depi3t desediments mobilises le long de la pente, lui-meme fossilis6 par l'implantation d'unhabitat sub-actuel (non etudie). Situees surle versant collinaire de l'interfluve d'Okolo,ces structures excavees sont le resultatconjugue &actions humaines et naturelles,introduisant une variabilite dans leursstratification interne, et demandent quesoient prises en compte la fois lasingularit6 de chacune des situationsarcheologiques cr&es et la recurrence decertaines de leurs caractkristiques.

L Organisation des Vestiges lmmobiliers

partir de l'examen du talus drOkolo etdes informations obtenues au cours desfouilles menees, il est possible decomprendre l'organisation generale desvestiges immobiliers. Comme Obobogo,celle-ci se compose d'alignements doublesde fosses, orient& suivant un axe nord/sud.Ainsi

les fosses du secteur de fouille no 3fosses no 16, no 17, et no 18 (non fouillke)se situent dans l'alignement des fossesno 12 et no 13 (talus inferieur). Celles-cisemblent associees des trous de

poteaux (par exemple le trou de poteausitue proximite de la fosse no 17 est leplus nettement perceptible) ayantappartenu des structures lkgeres dontla fonction et le plan restent incornus.S'agit-il de structures accompagnant

chaque fosse ou d'ensembles plus vastesexpliquant leur groupement? Le fortpendage qui affecte ce secteur nousa m h e laisser ce probleme en suspens.

les fosses no 2 i no 7 et les fosses no 8 et 9correspondent i des agencementsidentiques, le premier coupetransversalement, le second perpendicu-lairement. De la meme facon, la fosse no14 (sondage no 2) serait associee lafosse no 15 (extrCmit6 du talussuperieur). Derniere precision, cesensembles regroupent petites fossesovoydes et grandes fosses pyriformes.Comme le montrent les premieresdatations obtenues (voir plus loin), ilserait hstif de conclure, tant lacontemporaneite de toutes les fossesd'un meme ensemble, qu' l'installationsuccessive de chacun de ces ensembles.

Les Structures Excavties

La stratification interne de ces structuresse compose de trois phases de comblement.Le cas le plus frequemment rencontreassocie les phases de comblement no 1 et 2-dans les fosses no 2,3,9, et 17; la fosse no 14ne contient que des sk kn en ts de la phase no1 andis que la fosse no 16 combine les troisphases de comblement.

La Premiere Phase de Comblement

Elle se compose d'une succcession desediments meubles jaune-brun, parfoistraverses de lentilles ferrallitiques rouges etvierges; elle contient un mobilierarcheologique peu abondant et peu altere.D'apres les analyses sedimentologiquesaffectuees, la faible acidite de ces dep8ts,leur couleur, leur nature sablo-argileuse et laforte odeur amoniacale qui s'en degage,indiquent que ces depbts ont evolue enatmosph6re reductrice ou sous l'action d'unapport important de matiere organique; cecia pour consCquence la deferralisation de ces

sols et une dispersion des argiles c'est-B-diredes processus geochimiques tels qu'unealteration de la kaolinite et une dissociationde ses feuillets alumino-silicat&, marqueepar une disparition des pseudo-particules.


9/74

NYAME A K U W No 38 December 1992


10/74

NY ME KUM No. 38 ecember 1992

Cette phase de comblement presentedans toutes les fosses et cons&utive, selonnous, leur creusement, est mettre enrelation I la fonction initiale de ces

structures, ma rq uk par:un apport important de matiereorganique ayant evolue en milieureducteurun degagement de ces depats, traversesde lentilles ferrallitiques, notammentdam la fosse no 16, laissant penser unereutilisation periodique de cesstructures

Ces elements doivent etre mis en rapport

avec les caracteristiques intrinseques de cesstructures (voir plus loin).

La Deuxieme Phase de Comblement

I1 s'agit de sediments ferrallitiques decouleur rouge, contenant un importantmobilier archeologique plus alter6(ceramique, broyeurs, fragments de terrebrfilee, des noix de palmes et parfoistraverses de poches de sediments

ferrallitiques rouges vierges).Elle marque l'abandon de ces structures

s'accompagnant de l'effondrement partieldes parois, ce qui explique la presence depoches de terre stbriles, et de la mobilisationde depbts arch6ologiques, provenant desenvirons immkdiats de ces fosses, qui ontbouleverse le bassin d'epandage de cesecteur. Leur localisation sur un versantlaisse supposer un comblement rapide deces structures en creux et de ce fait, unerelative contemporaneitk avec les depotsanterieurs. Ce processus est en rapportdirect avec l'important phenomene decolluviomement occa siom ~ ar les eaux deruissellement, propre aux paysagespeneplanes de la region equatorialeinterfluves 2I versants convexes etpolyconvexes.

La Troisieme Phase de Comblement

Elle intervient lorsqu'il y a eusurcreusement de ces structures parl'homme, au cours de reoccupations de cesecteur; ces sediments se caracterisent parleur couleur rouge-brun, et l'abondance descharbons de bois. Le mobilier archkologique

est present et aussi alter6 que dans ladeuxieme phase de comblement.

L'etude d u profil general et de lastratigraphie des fosses montre que la

variabilite morphologique de ces structuresest like la transformation progressive deleurs caracMristiques intrinseques avec letemps. Ainsi parmi les fosses fouill6es,peut-on distinguer:

les fosses ovoides, de petitesdimensions: fosse no 14, profondeur:0.90 m, diametre la base: 1.10 m,diametre l'ouverture: 0.84 m; fosse no16, profondeur: 0.80 m, diametre labase: 0.87 m, cliamhtre l'ouverture: 1.28m (lie au surcreusement de cette fosse).

les fosses pyriformes, de grandesdimensions: fosse no 2, diametre 2 labase: 1.60 m, diametre l'ouverture: 2 m(partie superieure detruite); fosse no 9,diametre la base: 1.60 m (pa rtiesupe rieu re detruite); fosse no 17,profondeur: 1.75 m, diametre la base:1.60 m diametre l'ouverture: 0.90 m(la plus complete). La fosse no 3 a subi

d'importantes destructions et n'apporteque peu &informations.Ces premiers elements montrent que

l'hypothhe retenue (voir plus loin), s'appuieexclusivement sur certaines caract6ristiquesde ces vestiges irnmobiliers; les etudesvenir devraient Gtre menees de faconsurmonter le hiatus introduit entre domeesimmobilieres et mobilieres; en effet, si larelative contemporaneite des dep6tsrenfermes par ces structures sauf dans le casde surcreusements-justifie une fouille parniveau de 10 cm, leur stratificationcomplexe, l'abondance des vestigesmobiliers et une comprehension pluscomplete de la fonction de ces unites destratification requierent le recours I desmethodes de fouille plus elaborees.

Mobilier Arch6 1ogique

I1 se compose de cQamique, de materiellithique et de scories de fer:

L'important fractionnement de laceramique et son alteration entrainant116caillage de la pellicule superficielle


11/74


souvent decode de nombreux tessons,s'expliquent par la mobilisation de cesvestiges le long de la pente sous l'effet deseaux de ruissellement et du creeping

typique des paysages p6nkplan6s.Une premiere etude de la ceramique

d'Okolo montre que celle-ci rassemble desrecipients sans cols (les bords Cverses sontrares; les bords inverses bien representes),des recipients ti cols bas (everses, paralleles)et des recipients I cols hauts (everses,parallPles ou inverse%), leurs levres, dontl'extrkmite est arrondie, eff ilh , aplatie ou Bcam elu re centrale sont continues laterales(c'est-A-dire ayant une orientation differentede celle de la paroi des recipients) B un oudeux epaississements lateraux. Les fondssont generalement plats. Les tres rares6Kments de prehension rassemblent desanses, des cordons ou boutons deprehension et des trous de prehension (cesderniers uniquement dans la couchearcheologique III .

Les recipients d fOkolo ont un decorcouvrant unique ou bipartite, dont le

registre inferieur utilise un d&or camel6 aub2tonnet B pointe mousse (ces canneluressont generalement organisees en bandesparallPles ou en faisceau, plus rarement enreseau); dans les deux cas, ces ensemblesdecoratifs principaux (nous y englobons lesregistres superieurs des decors bipartites)sont souvent associes B des ensemblesdecoratifs secondaires, c'est-&dire soit deselements d'encadrement, soit des insertions.Les premiers sont faits d'impressions

simples ou pivotantes au peigne i entscirculaires ou B la spatule. Produisant desmotifs en chevrons ou des hachuresobliques, l'impression simple B la spatule estsur-representee dans la couchearcheologi que 111 au d et rim ent del'impression simple ou pivotante au peigneet l'impression pivotante a spatule.

Pour les ensembles decoratifssecondaires, ont 6t6 utilisbs l'impressioncontinue au biitomet i ointe mousse (ilsconsistent donc en cannelures uniques oumultiples), l'impression simple ou au peigne

dents circulaires (notamment dans lesfosses no 2, no 9, et no 17), l'impressionsimple au batonnet ou au peigne dentstriangulaires (dam les fosses no 2, no 9, et no

14). Les ensembles decoratifs secondairesnous semblent etre un element susceptibled'apporter de precieuses informations dansle cadre d'une seriation chronologique des

vestiges ceramiques.

Le Materiel Lithique

I1 se compose de deux fragments demeules (dont l'un provient de la fosse no 17),presentant des traces d'ensellement, defragments de broyeurs et broyeurs, de formesubcirculaire B section elliptique,subcirculaire ou plano-convexe, de formeelliptique il section sub-trapezoidale, de

forme sub-rectangulaireB

sectionquadrangulaire ou plano-convexe, et enfinde pierres B rainures et une hache polie endolerite (non en place).

Les materiaux utilises sont les rochesmetamorphiques (entre autres, le gneiss,l'amphibolite), le quartzite, mais egalementla dolerite (outre la hache polie mentionnee,une molette trouvee dans la fosse no 2 .Toutes ces roches proviennentvraisemblablement des environs immediats

du site (situe non loin d'un des rares ddmesrocheux de la dorsale Nyong-Sanaga, lemassif de Yaounde), B l'exception de ladolerite dont les principaux gisements,d' ap rb P. Claes, se situent au d e b desfleuves Nyong et Sanaga. Sous reserve del'existknce de filons isoles plus proches, onpeut y voir le temoignage d'echangesinterrkgionaux.

Les Scories de er

Sur les douze scories de fer trouveesOkolo, huit proviennent de la couchearch6ologique, trois de la fosse no 17 et unede la fosse no 9. Ces vestiges de petite taillenous r6vPlent peu d'informations sur lachahe operatoire lice B une transformationdu minerai de fer. Du moins, sont-ils desindices sur la presence probable de points detravail en rapport avec cette activite. Lesrecentes decouvertes faites B Okolo vontdans ce sens.

I1 faut signaler que les scories trouveesdans les fosses ont et6 trouvkes dans ladeuxi2me phase de comblement.


12/74

NY ME KUM No 38 December 1992

Datations et Premiers ~l 4m e nt sd lnterprdtation

L etude du site B Okolo se situe dans ucontexte regional 06 se succedent, depuis

1980, les decouvertes archeologiques, lesdernieres en date etant celles du site d'Oligapar J M. Essomba en 1989 et de Nkang par JC Mbida en 1990. Les sites da tk (Obobogo,Nkometou, Ndindan, Oliga et Okolo)appartenant B l'&e du fer, c'est-&-dire ayantlivre des artefacts de fer, couvrent uneperiode s'etendant entre le Veme si kle B.C.et le W&me iecle A D

Table 1. Les datations obtenuespour Okolo.

Fosses Datations 14c Datations 14cNo n calibrees alibres

S'il est B l'heure actuelle difficile df@treaffirmatif sur la nature du site d'Okolo etdes fosses qui s'y trouvent, du moins peut-on examiner quelques hypoth&sessusceptibles d'orienter les recherches Bvenir. De toute evidence, l'organisation desvestiges immobiliers laisse envisagerl'existence d'ensembles B vocationdomestique (et ceci au regard du mobilierretrouve dans lequel les scories de fer sont

rares); cependant, la question de la fonctionpremiere des structures excavees, reutiliseescomme depotoirs, reste entiGre: s'il etaitpossible de leur attribuer, sur la base deleurs caracteristiques intrins&ques, lafonction de fosses de conservation (termegenerique rassemblant structures destockage proprement dites et structuresutilisees B la fermentation des aliments),l'ubiquite de telles caracteristiques etl'absence de recherches palynologiquesvenant appuyer cette hypothese nousappeilent B la prudence. Celle-ci a cependantle merite d'attirer notre attention sur unensemble de points qui ne trouvent reponseque dans l'avenir.

I1 serait hiitif de conclure B la strictesedentarite des habitants dlOkolo, c'est-2-dire B une permanence de leur lieu deresidence tout au long de l'amee. L'idee de

saisomalite, associeeB

l'utilisation destructures de stockage, engage autant lesressources naturelles que les activiteshumaines et leur localisation: trois niveaux&interrogation susceptibles de nous orientersu r cette question. Pour l'instant, lesdatations obtenues pour les sites de laregion montre leur reoccupation r6@&re.

Une attention accrue aux ressourcesexploitees, permettrait de cemer la part quirevient B la veg6culture, l'agriculture etl'exploitation des plantes sauvages dansl'economie de la societe etudiee, ce quieclairerait en retour les modalitesd'implantation au niveau local mais aussiregional. Okolo, comme Obobogo, ontrouve des fruits de Canarium scheinfitrthii etdes noix de palmes (Elaeis guiniensis).

Enfin le dernier point que noussouhaitons soulever, est l'h6t6rog6neite del'espace forestier humide de cette region,

h6t6rogCneit6 tenantB

la diversite des 6co-systemes qui s'y catoient: valleesmarecageuses, versants collinaires biendrain&, mais egalement B la sylvigknGse, oitdes facteurs comrne la lurniere, l'humidit6,contribuent B faire de cet ensemble, unemosaique d'espaces B differents stadesd'4volution.

Remerciements

L'etude du site d'Okolo a 6tC realizeeda m le cadre de notre these de IIIeme cycleen Ethnologie Prehistorique B l'Universit6de Paris-I. Nous tenons B remercier lespersonnes qui nous ont permis de la mener Bson terme: MM J Devisse (Paris-I), P. deMaret (Universitk Libre de Bruxelles), J M.Essomba (Universite de Yaounde). Nousremercions Cgalement J Polet (Paris-1).

Bibliographie

Atangana C

1989 Arch6ologie du Cameroun Mkridional:Etude du Site dlOk olo Thkse de doctoratde IIIPme cycle en EthnologiePrehistorique Universite de Paris-ISorbonne.


13/74

N Y M E K U M No 38 December 1992

Essomba, J M.

1991 e Fer dans le Passe des SociitCs du Sud-Cameroun (Archeologie et Histoire).Thked Etat en Histoire Anc ienne , Universite

de Paris-I Sorbo me.

Holl, A.

1991 Age d u fer ancien. n Aux Origines deL Afrique Centrale, R. Lanfranchi et B.Clist (ss l dir. de ) CCFAC-CICIBA.

De Maret, P.

1989 Le contexte antu de l expansion Bantuen Afrique Centrale. In Expansion et

Identitt Bantu, T .Obenga (ss la dir. de),Colloque International d u CICIBALibreville, 1-6 avril 1985, 2 tomes,L Harmattan.

1985 Recent archa eological research anddates from Central Africa. Journal ofAh ca n History 26A29-48.

1982 N ew survey of archaeological researchand dates for West Central and North

Central Africa. Journal of African Histo ry231-15.

De Maret, P., B. Clist, and C. Mbida.

1983 Belgian archaeological mission inCameroun. Nyame Akuma 235 6.

Excavations a t BanockGrassfields Cameroon

Emm anuel de CritsSection de Pr6historieMu see Royal de I Afrique Centrale3 8 TervurenBelgium

Site LocationThe site of Banock, Menoua

Department, Penka Michel district (521'15N and 1011'25 E) (Fig. 1) was discoveredduring the construction of the road betweenDschang and Bafoussam. Of various pits

exposed by the road works, the one atBanock, a village situated close to anaffluent of the Membi river, occurred in theslope of the road a nd was excavated in

Fig. 1. Site location.

The site, a pit of 250 crn diameter at theopening and 170 cm deep, was capped by asurface layer of about 70 cm. The excavatedartefacts were limited to ceramics. Nothingremains of the part of the pit that wasdisturbed during road construction. The pitwas dug in a red lateritic soil. The profile(Fig. 2) presents four different colours andfour different degrees of induration (Table1). All layers contained potsherds, except forlayer 11 which was thin and probablyrepresents a fire place.

Excavations

Horizontal excavations allowed a lateralexposure of the layers. The shape of eachlayer was revealed by its colour and degreeof induration. The induration was calculatedby means of a standard shear device (typeThorvane), and the colour was recordedaccording to the Munsell Soil Colour Cards.These data were quite consistent within each


14/74


layer except for the indurations resultingfrom old fire places. The apparenthomogeneity of the layers will have to beconfirmed in textural analyses of the soil

samples. Charcoal was collected for eachlayer and will be dated.

Table 1. lnduration and colorrecorded for each layer.

Layers Soils olors IndurationWan2

~ teritic 10 R 4/8 Red 41soil

L I 5 R4/6 YellowishRed 48

L 5 YR4/4 Reddish Brown 67

L 2,5 YR 4/4 Reddish Brown 50

L I V 10 R3/3 Dusky Red 84

Archaeological Remains

The distribution of archaeologicalmaterial is different throughout the layers.Whereas layer I, the richest in material,contains sherds that allow the .refitting ofalmost entire vessels, layer 111 offers muchmore fragmented material, not allowing anyrefitting. Layer IV contained ceramicscharacterized by a variety of profiles and allkinds of handles. Pottery of an exceptionalthickness 3 cm) was uncovered in Layersand IV Their decoration consists ofoverlapping flexible roulette impressions

which are crossed by traced lines.Since the material is under study,

information on manufacturing techniquesand decoration are still preliminary.However, it is clear that the colour andtexture of the clay varies greatly, as dovessel types and decorative techniques. Themost frequent way of decorating the pots iswith twisted string roulette impressions.Impressions made with a carved woodenroulette are less common. Traced lines and

incisions are often superposed on flexibleroulette impressions. Some pots aredecorated with applied and incised coils,festoons, and pastilles Fig. 3).

Research

Further analysis of data from the pitexcavation will concentrate on revealing themeaning of the differences in colour and

texture of the layers, and on description ofthe ceramics. Pottery anlysis will focus onthe technical processes and decoration. Thiswill enable comparisons with other sitesfrom the Grassfields such as the site ofMandja near Bangangte Mbida 1992) andalso with information on ceramicmanufacturing processes observed in thisarea Gosselain 1991).

As part of the pit research, I alsoconducted an ethnographic study of a potterfrom Bafutchu-Mbu, a village at 14 krn to thesouthwest of Bamenda, NorthwesternProvince. The manufacturing processobserved is a variant of the coil technique,adding clay on a previously molded bottom.Decoration techniques are impressions madeby means of a twisted string roulette, tracedlines and stamped motifs. Zoomorphicapplications lizards nd frogs) appear onmost of the pots. The linguistic information

on Ngamambo nr 868 in Dieu and Renaud1983) collected during this enquiry has beentransmitted to the Linguistic Section of theRoyal Museum for Central Africa, TervurenBelgium).

Conclusion

The Grassfields of Cameroon is a crucialarea for African archaeology becauselinguists consider it as the cradle of Bantu

languages. The history of the Grassfields is,however, poorly known. Recent research inMbi, Shum Laka and Abeke rocksheltersAsombang 1988, de Maret et a1.1987)

provides information on the Late Stone Age,and the site of Banock may-eventually do sofor the more recent neolithic period.

Acknowledgements

Field research was conducted under the

direction of P.de Maret and funded by theUniversit6 Libre de Bruxelles. I would like tothank the Ministry of Higher Education,Computer Services and Scientific Research


15/74


NORTH PROFILE 17

Fig. 2. Banock pit north profile.

of Cameroon for research clearances, andalso Andrk Mboko, the chief of B a n d for

his hospitality and Abou Monica, potter ofBafotchu-Mbu for her cooperation.

References

Asombang, R.

1988 Bamendn in Prehistory: he Evidence fromFiye Nkwi, Mbi Crater and Shum LakaRockshelters. Unpublished Ph.D.thesis,University of London.

d e Maret, P., Clist, B. and van Neer, W.1987 RQsultats des premii.res fouilles dans

les ab ris d e Shum Laka et dfAbeke auN o r d- O u es t d u C a m e r o u n .L'Anthropologie 91:559-84

Dieu, M. and Rcnaud, P. (eds).

1983 Si tuat ion L ingu is t ique en Afr ique

Ce ntm le, lnventaire Pr liminaire: LeCameroun . Atlas l inguistique dc1 Afrique Centr ale (ALAC), ACCT,CERDOTOLA-DGRST. Yaound6, Paris.

Gosselain, 0 P.

1991 Technologie ckramiqu e et cthno-arch6omQtrie: resultat pr6liminaires d cla mission 1990 au Camcroun. N si 8-96-15.

Mbida, C.1992 Archaeologic al res earc h in So uth

Cameroon: Preliminary rcsults on thc1990 ficld season. Nyarne kuma 37:2 4.


16/74


Fig. 3. Pottery sherds from Banock decorated with a) incised lines, b) twisted stringroulette double), c) carved wood roulette, and dl applied and incised coil.


17/74


Malgre les resultats encourageants d6jBobtenus, le centre et le sud du Camerounrestent des regions insuffisammentprospectees et couvertes arch6ologiquement.C'est dans ce sens qu'une mission dereconnaissance archeologique du site deMadoumba a eu lieu le 30/06/91. Le but decette operation etait de localiser, de repereret d'identifier le site, puis de determinerl'importance des structures en place et lanature du site pour des fouilles hentuelles.

Le village de Madoumba est situe surl'axe lourd Douala Yaounde ii 60 km deYaounde. l appartient l'arrondissement deMatomb dans le departement de Nyong etKelle, province du Centre Cameroun. Ladecouverte du site a kt6 faite 2 partir del'observation directe du sol. En effet, devantla paroisse de l'eglise EPC de Madoumba,

apparaissent sur le sol des structuresnoiratres de forme circulaire et des scoriesdispersees. Les indices de metallurgie de ferfurent port& notre comaissance par lePasteur Song. Le site est actuellementoccupe par les populations Bassa.

Soper, R C Le site se localise a la cour de l'eglise

198 Roulette decoration on African pottery: 8 m du batiment principal. I1 est limite au

Technical considerations, dating and nord par une piste, au sud par l'eglise, a Irest

distribution. The frican rchaeological par la case du catechiste, etB l'ouest par le

eview 329 51. foyer de la paroisse. Les conditionsclimatiques semblent expliquer la mise aujour d u si e. En effet, c'est une regionforestiere qui connait des pluies abondantes

archeologique et les autres elementsp6dologiques. Elle a un diametre de 1.5 mSes vestiges s compasent de deux scoriesapparaissant en surface. La scorie a 10 cmde longueur alors que la scorie B a 12 cm. Lastructure B, qui se presente aussi commeune fosse, decrit sur le sol une formecirculaire. La couleur est grise et moins noireque la stucture A. Elle a un diamittre de 1 m.

La composition des vestiges est variee. Onpeut distinguer des scories et des briquesrouges. Les scories de surface sont aunombre de trois. Elles mesurentrespectivement cm de longueur, 10 cm et4 cm. Elles sont concentrees l'est de lafosse. Les briques au nombre de trois aussi,se concentrent ?I l'ouest de la fosse etmesurent respectivement 10 cm, 13 cm, et5 cm de longueur. La structure C est uncreuset de pr&s de 10 kg, mesurent 47 cm delongueur sur 34 cm de largeur. Le centreforme un creux, poli et profond de 5 cm,puis circulaire. Ce creuset est form6 dansson assemblage de dechet de fonte. Ladifference geometrique des structuresimplique des differences fonctionnelles.

Identification du SiteMetallurgique de MadoumbaCentre Cameroun: LesPremiBres Donn6esArch6ologiques

Jean Paul Ossah Mvo ndoCharge de CoursUniversitb de Yaoundb ENSDkpartement dfHistoireBP 7 YaoundeCameroun

en aoQt-septembre-octobre. Elles ont apritsplusieurs amees erode la roche argileuse,dont le lessivage a atteint 8 cm deprofondeur. Les structures metallurgiquesse sont donc retrouvees en surface. Ceterosion du sol atteint certains endroits dusite 20 cm de profondeur. Trois types de

structures apparaissent en surface au niveauspatial: deux structures circulaires et uncreuset.

Description Spatiale des Structures deSurface.

La structure A se presente comme unefosse de forme circulaire. Elle decrit sur lesol un cercle de couleur noire, alors que lesol nature1 est en argile de couleur rouge.

Les Conditions de DBcouverte du Site Cette difference de coloration du sol marquede Madoumba la differenciation entre structure


18/74


L'observation spatiale de la repartition desvestiges offre un espace triangulaire. Lasurface B est 1.5 m de la structure A. Cettestructure est 1.6 m de la structure C. La

structure occupe le centre du trianglespatial. On peut remarquer une equidistanceentre la disposition des deux fosses et lecreuset. Les autres vestiges recoltes sontdeux molettes calcides de couleur noire, lapierre ayant certainement brQle par le feu.La premiere molette mesure 9 cm sur 6 cmla seconde 10 cm sur 6.5 cm.

Ces premiers vestiges de surface du sitemetallurgique de Madoumba peuventpermettre d'avancer quelques hypotheses detravail. La presence des scories, et leurconcentration, la forme circulaire des fosseset la presence du creuset indiqueraient unatelier de fonte. La presence des briquesconstituent des elements de l'architecture dufoumeau.

Malgre la richesse des evidencesarch6ologiques rCcolt6.s en surface, lesprobkmes restent po s sur la technique detransformation, l'architecture du foumeau,

la repartition du travail et l'organisationfonctiomelle de l'espace. Dans le m@mearrondissement des sites metallurgiquesfurent aussi decouverts. Les resultatsobtenus apres l'etude du materiel peuventeclairer les premieres domees de surfaceprelevees sur le site de Madoumba.

En effet, deux sites Pan Mangueda etPan Nsas furent identifies, sondes et trait .(Essomba 1991: 270-318). Pan Mangueda alivrC parmi les vestiges de surface des

scories un fourneau et deux anciensemplacements de cases. Le fourneau estdetruit la partie superieure, et construitl'aide de petites briques de terre cuite,mesurant 4 5 cm de longueur et 2 cmd'epaisseur. Le diametre interieur estcompris entre 40 cm et 60 cm.

On note aussi une fosse de 3 cm delargeur sur 2 cm de large, puis 2 cm deprofondeur. Les sondages effectuesl'emplacement du fourneau ont livre ducharbon de bois. La fouille du site deMadoumba permettra de savoir si la base del'emplacement du fourneau avait une fosse.De plus l'extension des prospectionspermettra peut-@tre de reperer aussi des

restes de cases qui participaient laproduction du fer.

Le site de Pan Nsas, s i t d au sommet dela colline a livre des vestiges constitues par

un foumeau de 3.8 m de hauteur et de 2.5 mde circonference. La forme est conique, lapartie terminale arrondie. I1 est difficile desavoir dans l'etat actuel des travaux siMadoumba les populations ont utilise lesmi2mes types de foumeau. Le dkpotoir descories, autres vestiges de ce site, comprenddes fragments de scories de dimensionsvariees, comprises entre 5 cm et 2 cmd'epaisseur, puis des fragments de briques.

La particularite des vestiges deMadoumba est la presence d'un creuset deforme ovale parmi les scories. I1 pose leprobleme de son utilite et de sa fonctionda m le processus technologiques dans cetteregion.

Un des problemes que pose le site deMadoumba est la chronologie. Les sitesvoisins de Pan Mangueda et Pan Nsan sontdates respectivement de 330 130 P 200120 B P Ces datations font penser que

l'activite metallurgique dans l'arrondisse-ment de Matomb s'est effectuee de fasoncontinue probablement du XV siecle auXWI I siecle de notre ere. Peutdtre que lesresultats attendus du site de Madoumbav iendron t conf irmer ce tt e d u d ed'occupation ou alors ils apporteront denouvelles series chronologiques l'iige dufer de la region de Matomb.

Le probleme important que pose cesvestiges de surface, est aussi celui de lacaracterisation et de la connaissancephysico-chimique des scories et du minerai.Ces analyses sont entreprendre. partirdes vestiges des sites de Pan Mangueda etPan-Nsan, des caracteristiques se degagentconcernant le region de Matomb. En effet,les analyses chimiques de scories de PanNsan montrent des scories silicieux avec60.86 de S102 et des elements en etat detrace tels que: le barium, le cobalt, crome,cuivre, et le nickel.

Ces analyses diffractometriquesindiquent la presence du quartz, cristobaliteet l'hkmatite. Les analyses du minerai dePan Nsas, par diffraction des rayons Xindiquent la presence de la gibbsite, du


19/74


quartz et du goethite, temoin d'un minerairiche en fer. L'analyse prochaine des scorieset minerai du site de Madoumba permettrad'avoir des donnees comparatives au niveau

physico-chimique et compl6tera les premiersrksultats d4jh obtenus dans cette region. Latemperature de reduction du ferMadoumba reste determiner, mais on peutavancer 11hypoth6se d'une temperature dereduction variable allant de 1200 1300,cornrne l'indique les tempkratures obtenuesda m le site voisin de Pan Nsas.

Perspectives

L'identification du site Madoumba posebeaucoup de problemes qu'il faudrarksoudre. Pour cela il faudra des fouillesprogrammes et des sondages. L'etude dusite de Madoumba permettra d'etablir dessirnilaritbs avec les sites de l'arrondissement.I1 peut permettre de dkgager et d'etablir descontacts culturels, des differences spatiales,technologiques et temporelles dansl'arrondissement. Les analyses permettrontd'elargir le champs de connaissance de la

metallurgie de fer dans le sud cameroun.Les recherches prochaines doivent s'orientervers les fouilles, et les datations, puis larecherche des autres e lkmentscaractckistiques de la technologie de fer damle Nyong et Kelle. Enfin, il est utile deproceder une etude ethnoarcheologiquesur la metallurgie de fer dans ledepartement du Nyong et Kelle, afin d'avoirdes indicateurs sur la technologie anciennede fer dans le sud Cameroun.

Remerciements

Nous remercions le Pasteur Song quinous avait indique le site, puis M. Akameavec qui, en compagnie du Pasteur, nousavons fait la reconnaissance archeologiquedu site de Madoumba.

Bibliographie

Essomba, J M

99 Le Fer dans le Passt des Socittts du SudC a m e r o u n These d gtat, UniversiteParis-I.

EGYPT

A Further Report on FarafraWestern Desert, Egypt):

New Research andRadiocarbon Dates

M . AlessioDipartimento di FisicaUniversity of Rom e La SapienzaPiazzale Aldo Moro 500785 Rome, ltaly

B. E. BarichDipartimento di Scienze StoricheArcheologiche e AntropologichedelllAntichitaUniversity of Rome La SapienzaVia Palestro 600 785 Rome, ltaly

G. BelluominiCentro di Studio per la Geocrono logia eGeochimica delle Formazioni Recenti

Consiglio Nazionale de lle Ricerchec/ Dipartimento di Scienze della TerraUniversity of Rome La Sapienza00785 Rome, ltaly

F. A. HassanDepartment of AnthropologyWashington State U niversityPullman, W A99 64U.S.A.

A. A. MahmoudDepartment o GeologyAin Shams U niversityCairo, Egyp t

L. Man raDipartimento di Scienze della TerraUniversity of Rome La Sapienza00 8 5 Rome, ltaly

A. StoppielloScuola Nazionale di Archeologia

Rome, ltaly

The Research

The Rome University Mission to FarafraOasis (Fig. 1) started fieldwork in the winter


20/74


of 1987 and work has continued on a regularbasis in the subsequent three years. The nextseason is planned for summer 1992.Preliminary results of the 1987 season have

been published in this journal (Barich andHassan 1987). Elsewhere, the methodologyand a more detailed account of the resultshave been presented (Barich and Hassan198487; 1990).

The Archaeological Project in the FarafraOasis is directed by Barbara E. Barich incollaboration with Fekri A. Hassan. A. A.Mahmoud assists with the geo-morphological study; G Belluomini is incharge of radiocarbon dating; C. A. Amici isresponsible for the topographical study. Weacknowledge the participation in the 1987-91 field campaigns of A. A. Stoppiello, G.Catania, R. Sebastiani, A. Abderahmann, M.Serwi, M. Bashendi, M. Said, and K ElSendiouni. Mr. Sayed Yamani, to whom wewish to express our gratitude for help andsupport, is a representative of the EgyptianAntiquities Organization. Geological mapshave been contributed by F. A. Hassan.Drawings of the stone materials are byC Placidi.

Our work sheds lights on the prehistoryof a region that hitherto has been virtuallyunknown. knowledge of the prehistory ofFarafra Oasis is particularly significantbecause of it s central location in theEgyptian Sahara, as well as its intermediateposition between Libya and the Nile valley.Following our preliminary investigations,which revealed the importance of the

distribution of water resources in humanactivities, we have focused on theassociation of settlements with potentialprehistoric sources of water. We have alsobegun to examine change in economic andsocial structures related to the shift fromhunting-gathering to food production. Wehave also continued to investigate the placeof Farafra Oasis in the prehistory of theSahara, its connection with other Egyptianoases, and the possibility of interactions

with the communities of the Nile valley.In order to construct a general model for

the use of the territory and its resources, andfor the distribution of settlements, we haveoperated over a broad area. Not only havewe examined places with a high

concentration of artifacts-those tradition-ally defined as siteso-but we have alsoinvestigated their environmental setting andassociated traces of habitation. The sites are

characteristically positioned in relation toremnants of Holocene playas. Hassan(Barich and Hassan 1984-87: 139-40)identifies at least three wet phases which arelinked to the wet period of the earlyHolocene (9300-8800 or 8600-7100 bp) andto the wet intervals of the Middle Holocene(5900-5000 and 4800-4600/4500 bp).

Remnants of the earliest playa havebeen found in the vicinity of Bir Abu Nuss.The playa is close to the centre of the oasis.The archaeological remains are meagre. Anarea of debitage was also located in a morerecent playa deposit, for which the followingdates have been obtained: 6950 0 bp (R-1894) and 6670 95 bp (R-1895). Theradiocarbon datings are summarized inTable 1, which includes both conventionaland calendric ages. (The text givesconventional ages calculated on the basis ofLibby's (1955) half-life.)

We have also identified an areaapproximately 15 kilometres from the centreof the oasis where remains of open campsoccur. The area is more than 10 m ong andincludes Ain Kifrein, Ain e-Raml, and AbuKasseb (Fig. 2). Ain e-Ram1 was examined in1988 and 1989. A systematic collection wasmade within 200 square metres subdividedinto 10 x 10 m grid units. Archaeologicalrecovery was restricted to a depth of 10-15cm. Hearths were located. sample of

carbon collected from one of them provideda date of 9650 190 (R-1983) radiocarbonyears bp.

The debitage, which includes a largecomponent of blades, is dominated bysplinters from unipolar cores. Amongst thetools may be noted denticulates, scrapersand pieces with foliate retouching. Of someinterest is the presence of blades withbacked retouch. The presence of some burinspalls is sigruficant.

The lithic material collected both at Aine-Ram1 and at other nearby settlements isassociated with pottery. The sherds, whichare definitely of prehistoric date, haveheavily weathered surfaces due to long


21/74

NY ME AKUMA No 38 December 1992


22/74

NYAME AKUMA No. 38 ecember 1992

exposure, so they cannot provide detailsabout manufacturing techniques anddecoration.

Another important center of settlement

has been found approximately 130kilometres northwest of Qasr Farafra,around Ain Dalla. The artifacts areassociated with sabkhas and areextraordinarily rich, but they differ fromthose of Qasr Farafra. The raw materialconsists of dark brown, very shiny chert. Nopottery is known from Ain Dalla. sampleof ostrich eggshell collected in one of theconcentrations yielded a radiocarbon date of7000 10 years bp (R-1909).

Midway between Ain Dalla and QasrFarafra near the summit of Quss Abu Said, amajor settlement was located (Fig. 3).Investigations at Bahr Playa began in 1988and are still in progress. The settlementconsists of numerous concentrations andrepresents an ideal subject for an analysis ofcontinuity and change in human occupationduring the Holocene. The playa has asurface area of approximately one squarekilometre. Before collecting surface material,a genera l inspection of the playa wasundertaken: longitudinal t ransectsapproximately 100 metres apart were aimedat discovering the distribution of variousconcentrations of artifacts. A total of fiveareas of high density were recognized andmapped. Archaeological features at BahrPlaya include: manufacturing stations,alignments of hearths, possible buildingfoundations, dwellings, and lithic artifact

assemblages. The latter include a variety oftools, such as scrapers on flakes with lateralbulbs, knives/saws, and disks withmarginal retouch. A tranchet with unifacialretouch on the dorsal surface was found inone area (Fig. 4). Samples from two differentareas in the playa provided two radiocarbondates: one is on charcoal gathered from ahearth in area 2,8080 s 70 years bp (R-1902);the other date on ostrich eggshell, from area5, is 6730 60 years bp (R-1901).

The Rajih-Bir Murr area, approximately80 kilometres east of Qasr Farafra, s anotherlocus of prehistoric settlement. Situated onthe eastern side of the depression, itrepresents an outlier. Today the Rajih-BirMurr area is separated from the center of the

oasis by an eolian sand sheet and is cut offfrom the Nile valley by a plateau. It lies onthe caravan route which leads from Siwa,via Farafra, to Assiut in the Nile valley.

Investigations in 1988 and 1989 revealedconcentrations of artifacts associated withsmall playas at Rajih. Of some significance isthe presence of a blade industry whichincludes a particular type of arrowhead withunifacial retouch (Ounan-Arif type). Thepresence of thin-walled pottery withimpressed decoration, quite different fromother types in the depression, is alsoimportant. Both the ceramic and lithicindustries show affinities with the

archaeological complex of Bashendi inDakhla Oasis. A radiocarbon date from anartifact concentration investigated duringthe 1988 field campaign is 5380 110 (R-2006).

adiocarbon ates

Seven early radiocarbon dates areavailable from the Farafra Depression.(Another date, obtained on charcoal from a

recent hearth, is 320 0 years bp, R-1983. Afurther six dates on samples collectedduring the 1991 Mission are under study.)The early dates are summarised in Table 1,which also includes details concerningmaterial dated and the chronologicalsigruficance of the radiocarbon dates.

The method of radiocarbon dating isreported in detail by Alessio et al. (1985).Charcoal, after having been examined with astereoscopic microscope, was treated in anacid-alkaline solution according to thetechnique described by Alessio et al. (1970),which is now generally employed. Ostricheggshell fragments were processedaccording to Bada and Protsch s procedure(1973).

Charcoal and ostrich eggshelltransformation (as well as that of wood,peat, bone etc.) into benzene is obtainedwith Broecher et al. s technique (1959)

partially modified by Belluomini et al.(1978). Measure of 14C residual activity wascarried out using liquid scintillator counters.Counts were taken automatically every 30minutes in 20-hour periods; each benzene


23/74

NYAME KUM No 38 December 799

Q A S RF A R A F R A

A l n

l b s e i O A b uK a s s e b

Fig. 2. Farafra Oasis Egypt). Archaeological sites i n the Abu Nuss region.Key: Esna shale, 2 Farafra limestone, 3 Abdalla limestone,4 depression floor, Khoman chalk, 5 playa deposit.


24/74

NY ME KUM

Fig.

No 38 ecember 1992

3. Farafra Oasis Egypt). Bahr playa and Ain Dalla region.Key: Khoman chalk, 2 playa deposit, 3 Abdalla limestone,4 Esna shale, 5 Farafra limestone, 6 Dachla shale, 7 El-Hefuf,8 aeolian sand, 9 fault.


25/74

N YA M E A K U M A No 38 December 1992

sample was counted for a total of 60 hours(Alessio et al. 1970).

iscussion

According to present data, the 14Cdatings place the occupation of the oasis in arange that includes both Early and Mid-Holocene phases. Thus the datings supportthe climatic sequence put forward byHassan and mentioned above. Our datingsalso show general agreement with thosefrom other gyptian oases, such as Baharia,Kharga, Siwa, and Dachla (Hassan 1978a;Wendorf and Schild 1980:193; Hassan 1978b;

McDonald in press).Minor occupation areas, related to pointsources of water (outlets of wadis, springdischarges, solution depressions fromrunoff, etc.), can be identified inside theFarafra depression. At present thearchaeological remains near the centre of theoasis are rather scant, partially because ofhuman disturbance in recent times. Forexample, since 1987 the area surroundingQasr Farafra has undergone substantial

transformations and some of the sites wehad previously recorded (e.g., Bir Nuss)have disappeared. On the other hand, themost recent research (1988-91) hasemphasized the importance of the BahrPlaya-Ain Dalla archaeological unit, at thenorthern end of the depression.

On the basis of the 14C datings we inferthat the occupation at Bahr Playa startedduring the phase of climatic stability in theEarly Holocene, approximately between8600 and 7000 bp. Water resources, whichare fairly abundant in this subregion, mighthave favoured the development of a semi-sedentary settlement pattern, determiningthe subregion's autonomy from others. Therecent discovery of a village at Bahr Playa,containing proper residential structures,supports the above-mentioned indications.

As for the subsistence pattern, intensivegathering of wild plants can be inferred

from the abundance of sickle-blades andgrinding equipment. Along with this type ofeconomy, which could be said to be proto-agricultural, sheep-goat herding wasprobably practiced. The drastic fall inprecipitation that occurred during the

seventh millennium bp put an end to thecondit ions for stable set t lement.Consequently, Farafra became a morefrequent destination in the itineraries of

pastoral nomadic groups, as documented bythe abundant remains of fireplaces. On theother hand, the more stable groups wereforced to spread their settlement moreextensively inside the depression, reachingits external limits, as indicated by the Rajih-Bir Mum sites.

We hope that the continuation ofresearch by the University of Romearchaeological mission at Bahr Playa andRajih may add further data to our

knowledge of the occupation of FarafraOasis, specifying further its Holocenecultural sequence.

References

Alessio, M., Allegri, L., Bella, F., Belluomini, G.Calderoni, G. Calderoni, G., Improta,S., Manfra, L., Petrone, V. and Turi, B.

1985 Datazioni con il Carbonio-14: risultaticonseguiti, attivita' recenti, futurisviluppi. Contributi del Centro LinceoI n t e r d i s c i p l i n a r e d i S c i e n z eMatematiche e Loro Applicazioni.G io m ta di Studiosul tema: Archeometria69:81-134.

Alessio, M. Bella, F., Irnprota, S., Belluomini, G.,Cortesi, C., and Turi,B.

1970 Report on the equipment and activitiesof Rome University's Carbon-14 DatingLaboratory. Quaternaria 13: 357-76.

Bada, J. L. and Protsch, R.

1973 Racemization reaction of aspartic acidand its use n dating fossil bones.Proc.Nat Acad Sci. 70 (5):131-1334.

Barich, B. E., and Hassan, F. A

I1 Sahara e le Oasi: Farafra nel DesertoOccidentale Egiziano. Sahara 3: 53-62.

The Farafra Oasis ArchaeologicalProject. Nyame Akuma 29: 16-21

The Farafra Oasis ArchaeologicalProject (Western Desert, Egypt)-1987Field Campaign. Origini 13: 117-91.


26/74


Fig. 4. Farafra Oasis Egypt). Lithic tools wit h abrupt and folia te retouchfrom Bahr Playa. Key: 1, 7 tranchet-axes; 2 ,3 drills; 4 , 6 perforators;5 arrow-head; sickle-knife; 9 gouge.

6


27/74

Ta

le

a

c

R

o

b

Ac

Maea

a

C

o

a

C

bae

ra

L

aoy

lo

c

am

g

(9

% c

d

a

sm

e

s

e

u

5

Z

y

Tl25

y)

nm

R2

RJH

O

c

5

1

P

553Ot110BP b

4

-3

BC

e

h

(1

4

R-1

F

X

O

c

6

6

1

5

5

BC

e

h

(3

5

R1

B

H 5

O

c

6

6

6

5

5

BC

e

h

(4

R-1

FA

O

c

6

715

+

6

-5

BC

e

h

(4

7

R-1

D X

b

O

c

7

7

4

6

5

BC

e

h

(1

3

R

1

BAH 2

C

c

(1

8

8

Z

7

-6

BC

R

1

AR e

C

c

(82

9

9

Z

9

-8

BC

aC

o

a

weccaeo

hb

sobhLb

s(1

h

eo5

# 3

y

a

Osoea

(1

a

Osoa

K

es(1

h

eo5

4y

T

d

ee

bw

hw

dec

ep

oa

3pc

O

sa

dd

ao

s

ince

A

aenceabo

epe

BP)unAD 1

ah"pc

"

C

C

baea

oce

cy

w

eoan

a

dn

oKcnca

1

Kom

ca

(1

a

Suv

ea(1

dE

a

aereu


28/74


Barich, B. E., Hassan, F. A., and Mahmoud, A. A.

n.d. From settlement to site: formation andtransformation of archaeological traces.Scienze delllAntichita 5 in press).

Belluomini, G., Delfino, A., Manfra, L., andPetrone, V.

1978 Benzene synthesis for the radiocarbondating and study of the catalyst usedfor acetylene trimerization. Int. J Appl.Radiat. Isotopes 29: 453-59.

Broecker, W. S., Tucek, C. S., and Olson, E.

1959 Radiocarbon analysis of Oceanic C02.

Int. I. Appl. Radiat. Isotopes7 : 1-18.Hassan, F. A.

1978a Prehistoric studies of northern Egypt.Baharia Oasis and West Delta. PullmanReport to the Egyptian AntiquitiesDepartment.

1978b Archaeological explorations of the SiwaOasis region, Egypt. CurrentAnthropology 19 1): 146-48.

Klein, J., Lerman, J. C., Damon, P. E., and Ralph,E K

1982 Calibrat ion of radiocarbon dates.Radiocarbon 24 2): 103-50.

Kromer, B., Rhein, M., Bruns, M., Schoch-Fisher,H., Munnich, K. O., Stuiver, M., andBecker, B.

1986 Radiocarbon calibration data for the6th to the 8th millennia BC Radiocarbon28 28): 954-60

Libby, W. F.

1955 Radiocarbon Dating. 2nd ed. Chicago:The University of Chicago Press.

McDonald, M. A.

n.d. Dakhleh Oasis Project. Holoceneprehistory: Interim report on the 1988and 1989 seasons. The Journal of theSociety for the Study of EgyptianAntiquities in press).

Olsson, I. U., Karlen, J., Tumbull, A. H., andProsse, N. J. D.

1962 A determination of the half-life of C14with a proportional counter. Arkiv for

Fysik 22: 237-55.

Olsson, I. U., and Karlen, J.1963 The half-life of C14 and the problems

which are encountered on absolutemeasurements on beta-decaying gases.Radiocarbon Dating, Proceedings of aSymposium, Athens, Nov. 1962. IAEA:pp 3 11.

Stuiver, M., Kromer, B., Becker, B., and Ferguson,C. W.

1986 Radiocarbon age calibration back to13,300 years B P and the C14 agematching of the German oak and USbristlecone pine chronologies.Radiocarbon 28 28): 969-79.

Wendorf, F., and Schild, R.

1980 Prehistory of Eastern Sahara. New York:Academic Press, p. 193.

Radiocarbon Dates fromAsante Ghana) Sites

P. L. ShinnieDep artment of ArchaeologyThe University of Calgary2500 University Drive N.WCalgary AlbertaCanada T2N N

The report on the 1991 Asante ResearchProject S h i ~ i e nd Vivian 1991) did no t

contain information on radiocarbon date ssince at the time of writing they were notavai lable . The samples have now beenprocessed and the resu l t s a re in thefollowing table:


29/74


Table 1

Lab No

Gd 6540

Gd 5996

Gd 6545

Gd 6541

Gd 6537

Gd 6546

Gd 5998

Gd 5997

Gd 6543

Gd 6542

Asante radiocarbon datesSite Uncalibrated age P

Adansemanso 310 _ 70

Adansemanso 680 50

Adansemanso 740 70

Adansemanso 1050 00

Ad me ma nso 1110 110

Anyinam 170 70

Anyinam 550 50

Esiease 230 0

Esiease 190 60

Esiease 450 80

These dates were supplied by Dr. M F.Pazdur of the Radiocarbon Laboratory atGliwice, Poland. The chronology suggestedby these dates is congruent with whatarchaeological evidence had suggested, withoccupation at Anyinam and Esiease beingcomparatively recent whilst Adansemansohas evidence for an earlier occupation withfive of the d ate s suggesting activityconsiderably before European contact. Atthis site lack of imports and smoking pipeshad already suggested at least partialabandonment before the seventeenthcentury.

References

Shinnie, P. and Vivian B. C.1991 Asa nte research project. yame kuma

36:2 6.

Recent Findings of MiddleStone Age Material fromEast Turkana

Allison J. KellyDepartment of AnthropologyRutgers UniversityP. 0 Box 270Ne w Brunswick NJ 08903-0270U.S.A.

John W .K. HarrisDepa rtment of AnthropologyRutgers UniversityP. 0 Box 270Ne w Brunswick NJ 08903-0270U.S.A.

Paleoanthropological fieldwork, underthe coordination of Dr. Richard Leakey,began at Koobi Fora, east of Lake Turkana inthe late 1960s Leakey and Leakey 1978). TheKoobi Fora region contains rich fossiliferousand artifact bearing deposits which span thetime period from the Plio-Pleistocene to theHolocene. Although the Plio-Pleistocene andHolocene archaeological remains have beenwell documented Barthelme 1985; Harrisand Isaac 1978), very little is known aboutthe Late Pleistocene archaeological record. Adetailed s tudy of the Middle Stone Agehereafter MSA) archaeology of Koobi Fora

was therefore undertaken by one of usKelly in prep)

Earlier preliminary fieldwork by one ofus 0 W K. H.) yielded one minor scatter ofMSA artifacts FwJi 1--see Fig. 1 associatedwith the hominid femur KNM-ER 999 Kellyand Harris in prep). A partial craniumrecovered in the 1980s within severalkilometers of FwJi 1 has also been identified

recently as belonging to H o m o sapiensBrauer et al. 1991).

This paper will provide a preliminarydescription of the location, geologicalcontext, and distribution of new artifactfinds documented in fieldwork undertaken


30/74


by one of us A. K.) during 1990, 1991 and1992. A brief characterization of theassemblages follows together with apossible explanation for the low density of

the East Turkana MSA in comparison toother MSA assemblages from Kenya.

Site Location and Geological ontext

The artifacts were recovered from threeseparate areas in the East Turkana regionsee Fig. 1). Two occurrences are found

within a kilometer of one another along anerosional escarpment just to the south ofIleret Police Post. The northern occurrence,

designated as FwJi 2, consists of an erosionaloutcrop which extends for approximatelykilometer. Surface collection of tworelatively high density 5 x 5 m areas wasundertaken in 1991. During the followingsummer, a x 2 m geological trench wasexcavated in the proximity of one of thesurface samples to determine whether theMSA artifacts occurred in a discrete horizon.A clear artifact horizon could not bedistinguished within the predominantly

sandy sediments.The southern occurrence, designated

FwJi 3 is also found along an erosionaloutcrop, although the extent of the MSAdeposits is considerably more restricted andthe artifact horizon was significantly moredistinct. The sediments which make up theartifact horizon consist of sands, pebbles andsmall gravels and appear to be colluvial inorigin. Thirty-four square meters were

surface collected and a lxlm area wasexcavated to confirm that the artifacts wereeroding from the colluvial horizon.

Two more MSA occurrences exist on thebackslopes of the Karari Escarpment. Thesesites have been designated FxJj 61 and FxJj66. The site of FxJj 66 consisted of a 500 mlong strip of eroded outcrop. Six 5 x 5 msurface collections were made in areas ofhigher artifact density during the summer of1991. Additional research during thesummer of 1992 indicated that excavation inthis area would not be feasible because ofthe topography of the deposits. The site ofFxJj 61 is composed of about 1 kilometer oferosional outcrop upon which MSA artifactswere found. During the summer of 1991, six

random 5x5m surface samples were takenalong the outcrop. To ascertain whetherthere was a discrete artifact horizon andwhether it could be traced along the entire

outcrop, seven geological trenches were alsoexcavated during this season. The sedimentsconsist mainly of calcium carbonate richsands and occasional silts. Excavation wasinitiated in 1992 at the northern end of theoutcrop where the highest surface density ofartifacts was located. Artifacts wererecovered in relatively low densities fromthe 12 square meter excavation. To increasethe sample size, additional artifacts weresurface collected beginning at the excavation

and extending approximately 30 m to thesouth along the outcrop.

The final MSA occurrence, designatedGaJj 17, is located approximately 15kilometers east of the Lake Turkanashoreline. The artifact bearing sediment iscomposed of an aerially exposed coarse tomedium-grained sandstone which overliesfine-grained sands. During the summer of1991, three 5 x 5 m areas were surface

collected and a x m excavation wasundertaken. The excavation confirmed thatthe artifacts were restricted to the overlyingindurated sandstone.

Site Distribution

The three areas in which MSAoccurrences have been located in the EastTurkana region appear to differ ecologically.The Ileret occurrences FwJi 2 and 3) exist

near the present shoreline and near a majorriver channel. The vegetation along theriver, namely large trees and thicker shrub,is scarce within the lake basin. The rivermay not have been present during the LatePleistocene, but if it was, it would haveprovided resources e.g., plant food, riverinefauna, large cobbles) which wereunavailable in other areas of the basin.

The inland sites to the east of the Karariescarpm ent i.e., FxJj 61 an d 66) lie

approximately 23 kilometers from thepresent lake margin, but are closer to thebasin margin than any of the other sites. It isat the basin margin that significant sourcesof chert and chalcedony outcrop C. Fiebeland C. Nelson pers comm). The proximity of


31/74

NY ME KUM No 38 December 1992

dKilometers

.

* *

.

Si t e s

- - - Rivers1 .- - 2000 ontour

Interval9

Fig 1 Physiogeographic Ma p of East Turkana Region Indicating Location of Midd leStone Age Sites


32/74


these raw material sources may have madethis area an at tractive place.

Finally, the site of GaJj 17 lies on whatappears to be a paleo-beach Feibel pers

cornm). Its probable location near the shoresof Lake Turkana would have provided easyaccess to lacustrine resources such as fishand hippo.

Although analysis of the East TurkanaMSA assemblages is not complete,preliminary research indicates that theassemblages from these three areas alsodiffer. It may be tha t part of the intra-regional variation in these assemblages canbe related to activity differences amongvarying habitats.

Lithic and Faunal ssemblages

A total of 121 artifacts along with somefragmentary faunal remains were recoveredfrom the surface samples and geologicaltrench at FwJi 2. The majority of the artifactsconsists of flakes and flake fragmentsmanufactured on chert and chalcedony

o t h e r r a w m a t e r i a l s i n c l u d ephonolite/basalt and quartz). Additionally,several Levallois-type cores and oneretouched point were recovered. Themajority of the faunal remains that could beidentified were fish e.g., Nile perch), turtle,and crocodile, although there were somelarge bovid i.e., size 2 and 3) and hipporemains.

The artifacts n 232) recovered fromFwJi 3 were similar in character to those

found at FwJi 2. Again the majority of theassemblage is composed of flakes and flakefragments of micro-crystalline silicates

o t h e r r a w m a t e r i a l s i n c l u d ephonolite/basalt, ignimbrite and quartz),along with several Levallois-type cores andmiscellaneous retouched pieces. The fauna issomewhat different from that recovered atFwJi 2. Although there are still some fishremains, there is a greater number ofmammalian remains, including the lowerright M3 and upper right M2 ofMegalotragus the distal end of a femur froma small primate, and several lower limbelements from size 3 bovids.

At FxJj 66 and 61, the MSA lithics appearto be somewhat larger overall than those

from FwJi 2 and FwJi 3. Although chert andchalcedony are still the main raw materialsat these occurrences, there is a higherpercentage of phonolite/basalt artifacts.

hisincrease in volcanic artifacts at FxJj 61and 66 may contribute to the apparent

increase in overall artifact size. Theassemblages n 194 at FxJj 66; n 509 atFxJj 61 are mainly composed of flakes andflake fragments, although several pointshave been recovered from both sites andthere are a few examples of Levallois typecores. Faunal preservation at both sites ispoor. There was no fauna recovered fromthe surface at FxJj 66. At FxJj 61 examples of

identifiable faunal remains include crocodileteeth, hippo tusk, a lower molar of a size 3alcelaphine, and fragments of a size 4tragelaphine horn core.

Lastly, at GaJj 17 a total of 378 artifactswas recovered. Again the assemblage ispredominantly manufactured on chert andchalcedony, with phonolite/basalt an dignimbrite being less common. As with theother sites, flakes and flake fragments makeup the majority of the assemblage andLevallois-type cores and points exist. Thechert and chalcedony is very different incolor from that found at FwJi 2 and 3,suggesting procurement from differentsources. In overall artifact size, theassemblage appears to be similar to thosecollected from FxJj 61 and 66. Theidentifiable faunal remains from the site arepredominantly of hippo, crocodile, turt leand fish, although a few mammalian

remains, including a suid upper tusk, werealso found.

Site Density

An interest ing aspect of t he MSAoccurrences in the East Turkana region isthat they are all relatively low in artifactdensity compared with occurrenceselsewhere in East Africa. The low densitymay in part be du e to collection of mainly

surface artifacts i.e. the assemblages havebeen winnowed by water and other postdepositional factors). However, excavationsat FxJj 61, FwJi 3, and FwJi also suggestdiffuse artifact scatters. This pattern is inmarked contrast to the pattern found forother Kenyan MSA assemblages where


33/74

N YA M E A K U M A No 38 December 1992

artifact density is high, such as at ProspectFarm, Prolong s Drift, Muguruk, andLukenya Hill (Anthony 1978; Merrick 1975;McBrearty 1986). Some of the difference may

be due to variation in depositionalenvironment between the East Turkanaregion and regions further to the south.However, we suggest that the frequency,duration, and number of MSA occupationsat Lake Turkana differed from those in theCentral Rift Valley and Lake Victoriaregions and that this may also explain thevariation in artifact density.

The present environment at EastTurkana is arid and supports a limitedvariety and number of plants and animals.Although the lake level has fluctuatedperiodically in response to climatic changes(Feibel 1988), the surrounding landscapeappears to have remained relatively stable.If this reconstruction is correct, then it isreasonable to suggest that the climate nearLake Turkana was arid during the earlyLater Pleistocene. The region wouldtherefore not have been an area to whichMSA hominids were frequently attracted.Furthermore, even if there were populationsliving permanently in the area, the carryingcapacity of the environment must have keptthe numbers of MSA hominids low. Suchconditions could partially explain the lowdensity of MSA artifacts within the EastTurkana region.

The environments surrounding LakeNaivasha/Nakuru, Lake Victoria, and onthe Athi Plains, however, were more

hospitable. Vegetation and animals weremore abundant and variable and rainfallwas more frequent. Although climatic shiftsduring the Late Pleistocene had an effect onthe vegetation, fauna, and lake levels ofthese regions (Messerli et al. 1980; Street andGrove 1976; Washbourn-Kamau 1967), theconditions were almost certainly never asharsh as those at East Turkana. As such, thehigher density of artifacts at the sites remotefrom Lake Turkana s to be expected.

Conclusion

The research on the MSA from EastTurkana has not only provided new andunique lithic assemblages with which future

comparisons can be made, but also providesnew insights into the range and variabilityof MSA hominid activities within the region.Furthermore, the contrasts between the

diffuse MSA occurrences at East Turkanaand the high density sites in other regions ofKenya, makes it possible to argue that MSAhominid populations adapted theirbehaviors to the varying landscapes.Whether the low density sites in EastTurkana reflect infrequent use of the area orlow population levels remains problematic,but it is clear that MSA horninids were welladapted to exploit a range of habitats in amanner not unlike the Later Stone Age

peoples.

cknowledgements

We would l i e o thank the Governmentof Kenya for granting us permission toconduct research in Kenya and the NationalMuseums of Kenya an d Kenya WildlifeServices for their help and support whileundertaking fieldwork and laboratoryresearch. For their financial support, we

would like to thank the L. S. B. LeakeyFoundation, the Boise Fund, the Holt FamilyFoundation and the Graduate School,Rutgers University. Additionally, we wouldlike to thank the Koobi Fora Field School(Harvard University and National Museumsof Kenya) staff, crew and students for theirlogistical support, help in the field, andadvice.

References

Anthony, B.

1978 Th e Prospect Indu stry: Definition.Unpublished Ph.D. dissertation,Haward University.

Barthleme, J.

1985 Fisher-hunters and Neolithic Pastoralistsin East Tu rk an a Ken y a . BARInternational Series 254:l-364.

Brauer, G. Leakey, R., and Mbua, E.

1991 A first report on the ER-3884 cranialremains from Ileret/East Turkana,Kenya. In G. Brauer, and F. Smith (eds),Con tinu ity or Replacement? Controversies


34/74


in Homo sapiens Evolution. Rotterdam:Balkema.

Feibel, C

1988 Paleoenvironments of the Koobi ForaFormation, Turkana Basin, NorthernKenya. Unpublished Ph.D. dissertation.University of Utah.

Harris, J. W. K. nd Isaac, G.

1978 Archaeology. In Leakey, M. G. andLeakey, R. eds), Koobi Fora ResearchProject, Volume 1 . Oxford: ClarendonPress, pp. 64-85.

Kelly, A.

in prep The Middle Stone Age of East Turkana,Kenya. Unpublished Ph.D. dissertation,Rutgers University.

Kelly, A., and Hams, J. W. K.

in prep The Middle Stone Age of East Turkana,Kenya: New insights into variability inMSA hominid land-use strategies.Paper to be presented at The

International Congress in Honor of Dr.Mary Douglass Leakey's OutstandingContr ibution in Paleoanthropology: FourMillio n Years of Hominid Evolution inAfrica. August 8-14, 1993, Arusha,Tanzania.

Leakey, M. G., and Leakey, R. eds)

1978 Koobi Fora Research Project, Volume :The Fossil Hominids and an Introduction

to their Context, 1968-1974. Oxford:Clarendon Press.

Merrick, H.

1975 Change in Later Pleistocene LithicIn d u s t r i e s i n Eas t e rn Af r i ca .Unpublished Ph.D. dissertation,University of California, Berkeley.

Messerli, B., Winiger, W., and Rognon, P.

1980 The saharan and east African uplandsduring the Quaternary. In Williams,M., and Faure, H. eds) The Sahara andthe Nile. Rotterdam: Balkema, pp. 87-118.

McBrearty, S.

1986 The Archaeology of the Muguruk Site,Western Kenya. Unpublished Ph.D.dissertation, Harvard University.

Street, F. and Grove, A.

1976 Env iron men tal an d clim aticimplications of late Quatemary lakelevel fluctuations in Africa. N a t u r e261:385-90.

Washboum-Kamau, C

1967 Lake levels and Quaternary climates inthe Eastem Rift Valley of Kenya. Nature

Sangoanlechnologyandabitat a t Simbi

Sally M cBreartyDepartmen t of AnthropologyBrandeis UniversityWaltham MA 2245U.S.A.

It is well known that implements of theSangoan industry overlie Acheulian artifactsand underlie those of the MSA at a numberof equatorial African sites. The Sangoan issignificant because the shift from Acheulianto MSA technology seems to indicate a largescale change in hominid behavior, whichmay coincide with the first appearance ofmodem humans in Africa McBrearty 1987 .

The precise nature of the Sangoanindustry, its environmental circumstances,and its age have been problematic since thefirst assemblages were described nearly 70years ago Wayland Smith 1923 .Primarily our ignorance is due to the dearthof excavated samples from reliablestratigraphic contexts, conditions in thehumid tropics that conspire against thepreservation of bone or other organicremains, and the lack of reliable datingtechniques for thi s range of time.Application of laser technology, however,


35/74


36/74


Fig 1 Map of the Simbi Region

Fig 2 Geologic Map of the Simbi Region


37/74

N Y M E K U M No 38, December 1992

sequence; as many as five minor paleosolsdeveloped upon superimposed landsurfaceswere observed in a thickness of m ofsediment in excavation in units 4 and 5 Fig.

5). Isotopic analysis of paleosol carbonatesand organic matter, as well as of fossil toothenamel, indicates a consistently open,grassland C4) environment throughout thesection.

The direct evidence of fossil plants,inferred habitat preferences for fossilungulates, and the isotopic content ofpaleosols and fossil teeth are all consistentwith a grassland paleohabitat throughoutthe Simbi sequence. The data do not supportthe predictions of Clark 1960,1964a, 1964b,1965,1972,1982,1988) hat the Sangoan is awoodland adaptation. Simbi Sangoanextractive technology may thus be seen as anadaptation to a habitat not unlike thatknown for earlier East African Plio-Pleistocene sites, but the precise function ofth

Documents

Nyame Akuma Issue 038