NYAME AKUMA No. 70 DECEMBER 2008

0.2 Jan-Berend Stuut, Universität

TH Bremen, Germany. Paleoclimate on ABSTRACTS OF THE 19 geological timescales: African climate BIENNIAL CONFERENCE OF since the Late Neogene. THE S OCIETY OF AFRICANIST In this presentation I will try to give a geolo- ARCHAEOLOGISTS, gist’s perspective on how past changes in e.g. tec- tonics, orbital forcing, and global climate change, FRANKFURT, GERMANY, shaped the African continent during the past 35 mil- SEPTEMBER 8-11, 2008 lion years. I will start with an overview of the present-day climate systems that act on the African continent 0. Plenary Session. Session Chair: and then go back into geological time to see what Diane Gifford-Gonzalez. these climate systems were like during geologic his- tory, and how these could have been important for 0.1 Nicholas Conard, Eberhard Karls the four large stages in human evolution and cultural history: Universität Tiibingen, Germany. Did behavioral modernity evolve exclusively •Hominid evolution between 6 and 2 Ma BP in Africa? •Appearance of modern man c. 100 Ka BP This paper provides an overview of patterns •Pleistocene/Holocene transition and the settling of cultural innovations in Africa, Eurasia and Aus- of the Sahara tralia. While Africa is certainly the source of anatomi- •Late Holocene aridification and the development cally modern humans, the archaeological record is of agriculture. less clear on matters concerning the paleogeography of cultural innovations. Several key innovations, par- ticularly in the areas of organic and symbolic arti- 0.3 David Killick, University of Arizona, facts, are documented outside of Africa prior to their USA. Metals in African history and appearance in Africa. Unless one rejects the current prehistory: a synthesis and some new empirical record of the Old World as highly biased by variable preservation or differential intensities of re- directions. search, the possibility of polycentric, non-exclusively The first century of research on African metal- African origins of behavioral modernity must be taken lurgy was dominated by two topics: (1) the social seriously. This paper presents data from inside and construction of recent African iron smelting; and (2) outside of Africa and argues for a mosaic polycentric whether metallurgy in sub-Saharan Africa was inde- model for the rise of behavioral modernity. Several pendently invented. The writings of Africanists on key innovations seem to appear at the geographic the first topic have attracted considerable interna- interface between anatomically modern and archaic tional interest, but there is little possibility of further populations. This suggests that some innovations field research. Debate on the second topic is at an developed outside of Africa in the social contexts impasse, as almost all of the earliest radiocarbon dates associated with expansion of modern populations fall within a ‘black hole’ in radiocarbon calibration (c. into territories previously occupied by archaic 800-400 BCE). Resolution must await new research populations. In some settings, Darwinian competi- using other dating methods (thermoluminescence or tion between populations appears to have triggered cross-dated imported materials), and close coordina- new cultural behaviors that provided expanding tion of results with those from North Africa, Nubia groups of modern humans with biological advan- and the Arabian Peninsula. tages in relation to indigenous archaic groups.

58 NYAME AKUMA No. 70 DECEMBER 2008

New directions in research include: the explo- et celle d’un Laboratoire de Préhistoire en 1980-1981 ration through archaeometallurgy of the extraordi- ouvre de nouvelles perspectives pour la recherche nary diversity of past African iron smelting techniques; archéologique au Niger car, non seulement deux serv- quantitative estimates of the production of iron at ices au sein de l’Université, l’IRSH et l’Ecole des major production sites; the transfer of metallurgical Lettres et Sciences Humaines, font des recherches techniques to Africa through the Islamic world sys- archéologiques, mais aussi le problème de formation tem; the rise of indigenous tin-smelting and bronze- d’archéologues nigériens trouve ainsi sa solution, making industries in southern Africa; and changes in permettant ainsi a l’avenir une participation de plus the organization ofproduction of iron as a conse- en plus importante des spécialistes nigériens dans quence of the Atlantic slave trade. A particularly im- ce domaine de recherche jusqu’a présent totalement portant technical development is the developing use extraverti. of lead isotope analysis in reconstructing trade in metals. This is being used to track the earliest trans- Saharan trade and to trace the late trade in southern 0.5 Karega-Munene, United States African tin. In future it will play an important role in International University Nairobi, Kenya studying past trade in non-ferrous metals, glass and and Peter Schmidt, University of Florida, ceramic glazes around the margins of the Indian USA. Postcolonial archaeologies in Ocean. Africa. A plenary address on postcolonial 0.4 Oumarou Amadou Idé, Université archaeologies will provide a continent-wide overview Abdou Moumouni de Niamey, Niger. of the varied success of archaeology in Africa to Evolution de la recherche archéologique decolonize the practice of archaeology. Such an over- au Niger des independances à nos jours. view will examine: what colonial legacies continue to influence and guide explanations for change; the si- S’interroger sur l’histoire de la recherche lencing and marginalization of African archaeologists archéologique au Niger des indépendances a nos who challenge established paradigms; the tensions jours, est un exercice complexe et passionnant a la between the privileged world of archaeologists and fois. Une approche délibérément chronologique the impoverished world of people among whom we pourrait aboutir à l’énumération des événements les work; the systems of patronage that restrict access plus significatifs de son évolution sur le territoire. to archaeology and “eat the young”; the embracing Les recherches archéologiques au Niger ont souffert of an activist archaeology that engages local com- pendant longtemps d’une disproportion munities, their social/economic problems, and their géographique par rapport à l’ensemble du territoire human rights issues; and an assessment of what the national. Jusqu’en 1976, le seul organisme nigérien future holds for an African archaeology that is still de recherche s’occupant d’archéologie était l’Institut struggling to find an African voice. These unspoken de Recherches en Sciences Humaines (IRSH), avec and mostly silenced issues deeply affect and inform un service d’archéologie crée en 1966. A partir de the practice of archaeology in Africa today. This ple- 1976, l’introduction progressive d’un enseignement nary address will fittingly and compellingly bring them de Préhistoire et d’Archéologie au Département to the surface for fulsome discussion and reflection d’Histoire de l’Ecole des Lettres et Sciences Humaines of an appropriate if not necessary condition as we de l’Université de Niamey, jusqu’a la création d’une come to the half-century mark for independence of option Préhistoire et Archéologie en année de maitrise most African countries from colonial rule.

59 NYAME AKUMA No. 70 DECEMBER 2008

1 The Gobero Archaeological Project: rock common in Holocene tool kits in Niger has a Early and Middle Holocene Human local origin. That rock is a microcrystalline felsite, Adaptations in the Sahara-Sahel Border- which has been used since the Pleistocene and be- came the exclusive source rock for Tenerean disc land. Session Chairs: Elena A.A. Garcea knives during the mid-Holocene. A chronology for and Oumarou Amadou Idé. Gobero was established on 9 OSL dates on palaeodune sand and 78 C 14 AMS dates on human 1.1 Abdoulaye Maga, Université Abdou burials, harpoon points, ceramics, midden materials, fauna and sediments. Direct dating of human skel- Moumouni de Niamey, Niger / ECOWAS etons, fauna and artifacts was based on the bioapatite (Economic Community of West African component of enamel and bone. The record at Gobero States), Nigeria. Circuits et facteurs was divided into four occupation phases extending internes du trafic des objets from the Late Pleistocene to the late Holocene. The archéologiques nigériens et ouest- most important are the early and mid-Holocene occu- africains. pation phases, which are separated by an arid inter- val recognized across the central Sahara. These two In West Africa the countries that are the most pulses of semi-sedentary occupation are clearly dis- affected by the plundering and trafficking of archaeo- tinguished by material culture and funerary practices logical artifacts are mainly Nigeria for the reputed as well as the distinct physical features of the occu- beauty of its bronzes, brasses, and terracottas; Mali pants. The taller early Holocene occupants (-7500 for the terracottas in its interior delta; and Niger for BCE) are buried in tightly bound, hyperflexed pos- its dinosaur fossils and funeral statues from the Bura tures that compose the earliest cemetery in the Sa- Site. Countries such as Guinea, Senegal, and Burkina hara. Most closely related to ‘Mechtoid’ populations Faso have also been affected by rings of plunderers. in Mali and Mauritania and to early inhabitants of In general, it is very difficult to find a country in the Maghreb, these early occupants have long, low Africa, which does not face these international rings crania and are associated with a Kiffian industry char- of cultural object traffickers. Even art collectors, art acterized by microliths and harpoon-fishing. They traders and Western museums selling such objects abandon the site complex under arid conditions and are often victims of robberies. are replaced by a more gracile people (-4600 BCE), with deeper, more prognathous crania, buried in less flexed posture often with grave goods. Notable buri- 1.2 Paul C. Sereno, University of als include an adult male seated in a mud turtle cara- Chicago, USA. Gobero: an exceptional pace, an adult male with skull resting on a half ce- record of early and mid-Holocene human ramic vessel, and a triple burial composed of an adult adaptation in the southern Sahara. female and two juveniles in symbolic pose with un- derlying pollen clusters indicative of the presence of A minimum of 200 human burials are preserved flower heads of woolflower (Celosia). Grave goods on the edge of a paleolake in central Niger and pro- include an upper arm bracelet and an incised ivory vide an exceptional record of human occupation in pendant as well as hollow-based and tanged projec- the southern Sahara under severe climate fluctuation tile points that have been associated with a Tenerean during the Holocene. The Gobero site complex, lo- tool kit. Middens close in age and proximity to the cated on the northeastern rim of the Chad Basin, was burials suggest continued reliance on hunting and discovered in 2000, briefly revisited in 2003, and in- fishing in the shallow waters of the Gobero Paleolake. vestigated during field seasons in 2005 and 2006. An The final phase recorded at the Gobero site complex international team excavated 84 burials and mapped, lacks burials but is represented by isolated sectioned and sampled the central paleodune burial undecorated pottery, suggesting the transient pres- sites and adjacent paleolake deposits. In 2005, we ence of late Tenereans. Gobero chronicles the rapid discovered a Holocene quarry 160 km north of Gobero pace of biotic and cultural change in response to on the edge of the Air massif, showing that the green profound humid-arid fluctuation during the Holocene.

60 NYAME AKUMA No. 70 DECEMBER 2008

1.3 Carlo Giraudi, ENEA (Ente per le 1.4 Anna Maria Mercuri and Isabella Nuove Tecnologie, l’Energia e l’ Massamba N’siala, Universita degli studi Ambiente), Italy. Late Upper di Modena e Reggio Emilia, Italy. Veg- Pleistocene and Holocene hydrological etation and plant use at Gobero (Central balance in the GOBERO temporary lake Niger) inferred from pollen. (Niger). Pollen is generally known to be an effective The Gobero depression formed as a conse- tool to reconstruct plant landscape evolution, as it quence of strong deflation affecting Cretaceous con- permits to study diachronically environmental tinental sediments gently dipping toward the south. changes. Within the multidisciplinary studies carried Inside the closed basin, every time the hydrological out at the archaeological site of Gobero, palynological balance of the catchment became positive, a lake was analyses can help to reconstruct past environments formed. The lacustrine sediments are preserved in giving details on the flora and vegetation cover of small patches in the Gobero depression due to the this Saharan region during the early and middle wind deflation and the erosional phases that affected Holocene. Moreover, in archaeological contexts, as them since their sedimentation. The sediments have in the case of Gobero, pollen can provide information been deposited in a temporary lake having a maxi- on the past relationships between human populations mum extent of about 50 km2, and a maximum depth of and the environment. During the 2006 field season, about 6-7 m. The geological-geomorphological sur- pollen samples were collected from spot areas in the vey carried out on the Gobero basin allowed the iden- burials in the Gland G3 cemeteries, and from a short tification of lacustrine features such as ancient shore- sequence in the Gobero Paleolake (GO 1), a desic- lines, wave-cut platforms, and two different outflows. cated lake near the burial sites. Pollen spectra reflect The first outflow was near the SE border of the former low biodiversity and were dominated by Sahelian taxa lake. The altitude of the sill level that conditioned the (Poaceae + Cyperaceae), whereas Saharan taxa lake level was about 6.5-7 m above the bottom of the (Chenopodiaceae + Asteraceae) were less repre- lake. The second outflow was at the SW border of sented. Data from the burials give fairly concordant the basin and formed later. In a first period the sill information on the Gobero plant landscape which, at elevation was about 5.5-6 m above the lake bottom, the time of use of the cemetery, was largely an open but erosion due to the outflow lowered the sill until it environment, herb-dominated by grassland or reached the elevation of about 4.5-5 m above the lake shrubland vegetation. Several hydrophytes and al- bottom. Wave-cut platforms were also found at the gal elements (Concentrycistes) testify to the pres- top of some ridges and at the top of archaeological ence of permanent water, such as ponds or lakes, and sites, and wave-cut small terraces or platforms have the presence of fresh water environments in the area. been shaped during the lowering of the lake level. The occupation of the Gobero archaeological sites 1.5 Christopher M. Stojanowski, Arizona was strongly conditioned by the presence of the lake and by the lake level oscillations. For example, some State University, USA. Biological struc- sites lying at about 2.5 m above the lake bottom could ture and population affinity of burial not be occupied when the lake level was higher. The components at Gobero: intra-site and lake water submerged the sites three times: the Kiffian regional perspectives. settlements were interrupted twice, and the Tenerean settlements on one occasion. Lacustrine sediments This paper will address one of the more cen- formed during Late Upper Pleistocene wet periods tral issues of biological anthropology of the south- have been recognized, but at the present no direct ern Sahara by considering the following issues: (1) datings are available for them. Who were the earliest inhabitants at Gobero (c. 8500- 7300 BP) and are they ancestral to the later inhabit- ants (c. 5500-4500 BP) of the area that buried their dead in the same dunes? (2) What are the continental affinities of each burial component as compared with published data from Aterian, Iberomaurusian, and

61 NYAME AKUMA No. 70 DECEMBER 2008

Capsian sites, crania from Columnata, crania from sites 1.6 Elena A.A. Garcea1, C. Lemorine, E. along the Atlantic coast of Mauritania, and crania Cocca2 and G. Mutre2: The lithic and from interior Mali? ceramic assemblages from Gobero. (1. Descriptively, crania at Gobero become shorter Universita degli Studi di Cassino, Italy; through a re-structuring of the occipital bone, most 2. Universita degli Studi di Roma “La notably the disappearance of the occipital bun that Sapienza”, Italy). is present in the Kiffian population. The overall size of the cranial vault decreases through time while the The lithic and ceramic assemblages from the height of the vault increases slightly. The mandible Gobero archaeological area presented in this paper becomes shorter, narrower, and less deep as well. come from surface collections and excavations at the Although there is a decrease in cranial robusticity two main sites, GI and G3. Even though these sites through time, the Kiffian sample does not exhibit the were intensively used as cemeteries by Kiffian late same degree of rob usti city seen in Late Palaeolithic foragers and Tenerian pastoralists, the large majority human populations in North Africa. Craniometric of the lithic and ceramic artifacts are not associated analysis using mean-based principal component ex- with the burials. They are included in the anthropic tractions as well as population genetic R-matrix meth- deposit in which the burials had been dug and there- ods indicate the two burial components at Gobero fore indicate that the sites were used for living or are as distinct as the Late Pleistocene/Early Holocene performing some specific activities on the spot as Iberomaurusian and Capsian populations were from well. the preceding Aterians. This structure is documented Pottery was the first archaeological marker that using both principal components and R-matrix meth- could be used to identify the cultural groups who ods and holds under differing scenarios of demo- occupied the Gobero area. In fact, two distinct ce- graphic expansion. Because the latter transition ramic assemblages are present; one shows impressed (Aterian-Iberomaurusian) is interpreted as popula- decorations made with an evenly serrated comb, pro- tion replacement, the compressed time frame at Gobero ducing dotted zigzags and dotted wavy lines, which may suggest a similar explanation of biological diver- are typical ofthe early Holocene late foragers of the sity. Comparison with the Howell’s craniometric Sahara-Sudanese region and locally associated with dataset using FORDISC 2.0 allocates none of the in- the Kiffian industry; the other, locally called Tenerian, dividuals from Gobero to any African population in- features decorations made with a double-pronged cluded in this database. Although comparison of re- tool, producing alternately pivoting stamped motifs mains of the age of Gobero with the largely modern which typify the early pastoral period in the Sahara. series in the Howells database is problematic, the A wide variety of raw materials used to produce the results are interesting in allocating the Kiffians to lithic assemblage is a remarkable feature of Gobero. Ainu, Japanese, and Chinese populations while the Seven different types of rock could be first distin- Tenerians were allocated to more diverse populations guished by macroscopic inspection of the artifacts, such as Easter Island, Eskimo, and Guam. That sev- and fourteen samples have been submitted to petro- eral of the individuals from the Tenerian phase were graphic analysis of thin sections under a binocular also allocated to Ainu/Japanese populations suggests microscope in order to identify the mineralogical fea- the diversity present at Gobero during the Kiffian tures which characterise the different types of rock. phase was still present during the Tenerian, but sig- Functional analysis of the lithic industry indicated nificant immigration of allochthonous populations macro-traces of use with edge-removals that could may have occurred. This interpretation is supported be observed on some artifacts under a by a Relethford-Blangero analysis, which indicates stereomicroscope. Edge-removals were recorded on significant extra-regional alleles were introduced into three categories of knapped tools: arrowheads, geo- the population at Gobero during the Tenerian occu- metric lunates, and end-scrapers. pation. Therefore, a pattern of in situ evolution with significant immigration of extra-local individuals may best explain the biological diversity within Gobero’s cemeteries.

62 NYAME AKUMA No. 70 DECEMBER 2008

1.7 Helene Jousse, Naturhistorisches 1.8 Kelly J. Knudson and Christopher Museum Wien, Austria. Patterns of M. Stojanowski, Arizona State Univer- animal exploitation in a lacustrine envi- sity, USA. Recent bioarchaeological and ronment by the two human populations of biogeochemical research at Gobero: Gobero (Niger) during the Holocene. palaeodiet and residential mobility in the The newly discovered Holocene site called early and middle Holocene. Gobero in central Niger provides a rich faunal record The early and middle Holocene site of Gobero, indicating change in palaeoenvironment and dietary Niger, offers a unique opportunity to understand customs for two successive phases of human occu- Holocene palaeo-diet and residential mobility through pation. During the first occupation phase in the early bioarchaeological and biogeochemical approaches. Holocene, between 7600 and 6300 cal BC, fish domi- Preliminary strontium isotope analysis of archaeo- nates the faunal remains, especially large species liv- logical human remains from Gobero shows homoge- ing in deep and permanent waters linked to the Chad neity in geographic origins and does not demonstrate Basin. Nile perch and large catfish were taken with a high levels of mobility within individuals’ lives or variety of harpoons and hooks made of bone; wild between different occupations of the site. Palaeodiet mammals that may have been hunted include hippos, and weaning behaviors are inferred based on carbon small carnivores, and many antelopes living in forest and oxygen isotope analysis of archaeological hu- savanna to drier landscapes as well as turtles and man remains from Gobero, as well as frequencies of crocodiles. A large mammal rib was found with dental pathologies. Preliminary carbon and oxygen cutmarks, and two barbed bone points were found in isotope data suggest little difference in carbon iso- a midden dated to this phase. Many other barbed tope signatures between the early and middle bone points come from the area surrounding the site, Holocene occupations, while the variability in oxy- including some found in situ in lake sediments. They gen isotope data is influenced by weaning behaviors show a large variety of size and morphology, includ- and environmental changes. Dental caries are rare ing points and harpoons with one and two rows of but show an increase through time, suggestive of barbs. the adoption of a more cariogenic lifestyle as climatic A second occupational phase in the mid- conditions deteriorated throughout the Holocene. Holocene, from 4000 to 2800 cal BC, remained heavily dependent on aquatic resources. Midden remains 1.9 Hannah M. Moots and Paul C. suggest shallow-water net fishing with catch domi- Sereno, University of Chicago, USA. nated by small clarids and tilapia, mud turtles and mollusks. Wild ungulate bones associated with this Holocene ornaments from Gobero. phase include dry savanna species such as warthog, Excavation of 84 burials and collection of arti- giraffe, elephant, tortoise, reedbuck, oryx and ga- facts from paleodune and lakebed surfaces at Gobero zelles. Domesticated cattle are present within this resulted in the recovery of a broad range of orna- phase from 3900-2600 cal BC as a minor component ments made of ivory, eggshell, bone or stone. We in the middens and surrounding paleolake deposit. divide these artifacts into those recovered in situ as The significance of the faunal assemblages at Gobero grave goods and those found out of context on de- is compared to other Holocene sites in the central flation surfaces. Ornaments found as grave goods Sahara. The occupation and use of the Gobero site is include one upper arm bracelet and two necklaces. discussed, proposing the hypothesis of a particular The upper arm bracelet is made of hippo ivory and functionality and/or seasonal practices, in the light was found on the left humerus of an immature female of other archaeological and analytical data. (approximately 11 years old) that was dated to the mid-Holocene occupation phase (4250±40 BP). The bracelet has an oval shape with narrow internal and external margins. The enamel surface shows a tessel- late fracture pattern under ultraviolet light, a pattern that appears to be unique to hippo tusk enamel. The two necklaces were associated with mature individu-

63 NYAME AKUMA No. 70 DECEMBER 2008

als and were made of similar materials which include palaeodunes provide a chronological framework that ostrich eggshell beads and larger beads made from a clearly identifies two successive human occupational green metamorphic rock. The more complete of these phases at the Holocene site of Gobero in central Niger. necklaces was found in place on a mature female Due to degradation of bone collagen under arid con- skeleton dated to the younger occupation phase ditions, direct dating of human remains has rarely (4860±40 BP). In addition to eggshell beads and one been achieved in the Sahara from sites of early and larger green rock bead, there are nine beads and a middle Holocene age. We used as an alternative the pendant made of hippo ivory. All of the beads are more stable carbonate component in bioapatite. Di- drilled from each side, some have asymmetric center rect dating of plant temper in ceramic sherds also holes, and a few are square rather than uniformly allowed chronological control for this important com- round. The pendant is subrectangular and has a char- ponent of human culture. Direct dating of bone har- acteristic’ crows foot’ pattern of incisions along each poons found in place in palaeolake sediments shed long border. The incisions are in alternate positions light not only on the age of the artifacts but also on on each side, such that in lateral view the edge has a the age of the lakebed. Bioapatite dating techniques serpentine form. Ornaments found out of place on and results are discussed and tested in two ways. the paleo dune or surface of the palaeolake include First, we dated and compared multiple samples from several amazonite and red and yellow carnelian beads different source materials from a single individual in a variety of sizes and shapes. One amazonite bead specimen, such as enamel, cortical bone, and interior is partially drilled, and there are many amazonite cores bone from an individual human skeleton. Second, we that remain unworked. The most unusual amazonite dated and compared multiple samples of different ornament is a lobate pendant, 44 mm in length. Other materials from a single archaeological structure, such ornaments include the distal phalanx of Gazella dama as bone, ceramics, mollusc shell, and sediment from a with the collateral ligament pits drilled through, a single midden. The dates are generally very consist- pendant made from an external section of a hippo ent. Small differences may be due to isotopic exchange tusk with two holes for attachment, and a pendant after burial, variance in laboratory pre-treatment pro- made from an elongate curved cranial bone from a cedures, and variance during AMS calculation. Man- large Nile perch (Lates niloticus). No ornaments were ganese analysis underscores the influence of inun- found in any of the burials from the early Holocene dation in bone and ceramics, imparting a dark color occupation phase, and so it appears likely that most to most bones older than 4000 cal BC, including all of the deflated ornaments were also associated with skeletons of the early human occupation. A subset the mid-Holocene occupants. among the first human occupants is shown by their chronological coincidence and close spacing to com-

1 pose the oldest cemetery in the Sahara, dating from 1.10 Helene Jousse , Jean Francois 7500 cal BC (8640 ± 40 BP). Some sherds with Kiffian 2 3 Saliège , Paul C. Sereno , Elena A.A. decoration are dated directly to 7200 cal BC (8150 ± Garcea4 & Thomas W. Stafford5. Cross- 40 BP). The chronology at Gobero is compared to checking of dating methods for discern- occupational sites elsewhere in the Sahara. ing the succession of human occupation and palaeoenvironmental change at 1.11 Elena A. A. Garcea, Universita Gobero. (1. Naturhistorisches Museum degli Studi di Cassino, Italy. Gobero: Wien, Austria; 2. Université Pierre et secular or sacred place? Marie Curie Paris 6, France; 3. Univer- This paper attempts to provide an overview of sity of Chicago, USA; 4. Universita degli the various papers on the archaeological, anthro- Studi di Cassino, Italy; 5. Stafford Re- pological, and environmental evidence from Gobero, search Laboratories, USA). an early and middle Holocene site in the Tenere desert, Niger, presented at this conference. The ample set of More than 80 AMS radiocarbon dates obtained data on the funerary practices, biological variation, from human and faunal bioapatites, barbed bone diseases and traumas, diet, and ornaments of the points, ceramics plant tempers, charcoal, mollusks people buried at Gobero provides a previously and organic sediments, as well as 9 OSL dates from

64 NYAME AKUMA No. 70 DECEMBER 2008

unknown panorama of anthropological features and of oral traditions is one avenue for the reconstruc- cultural traditions of both Kiffian late foragers and tion of population movements and makes it possible Tenerian pastoralists who used the site as a cem- to examine the contexts in which scattered etery. populations regrouped into new ethnic entities. Ar- chaeology reveals that the plain had more significant At the same time, the data on the zoo- settlement before the arrival of new populations than archaeology and the lithic and ceramic artifactual afterwards, and that there are correspondences be- assemblages indicate that the site also functioned as tween these old sites on the plain and the villages in a living place for the same human groups. Therefore, the surrounding mountains. Further data were ob- the sacredness of the site, as evidenced by the selec- tained through the study of traditional pottery and tion of the location of the site in a barren area of the the feature of standing stones. In sum, all evidence Tenere desert for the interment of the dead, is linked combined suggests an ancient cultural continuum, with the secularity of living and coping with the chal- which receives further confirmation from linguistic lenges of an environment at the moving border be- data. My hypothesis is that the ancestors of today’s tween the Sahara and the Sahel. Its ecological condi- mountain dwellers moved away from a common point tions, as evidenced by the palaeoclimatic and of origin, probably on the plains bordering their geoarchaeological data, provide reasonable hints that present territories. The Faro valley is economically a Gobero had been elected as both an exceptional secu- far more advantageous location, which leads to the lar and sacred place. After reviewing the major fea- question why they should have retreated into the tures of the archaeological data from Gobero, this mountains. The present occupation of the river val- paper assesses it in the general framework of the ley by the Foulbe and the Bata suggests that this early and middle Holocene prehistory of North and economically strategic zone was acquired by force at West Africa. their arrival in the area. A climate of insecurity may best explain the abrupt demographic and social changes, with the original river population, probably less well organized in military terms, retreating to the 2. The Archaeology of Central Africa: more easily defensible mountain tops. Eventually, Its Current State Session Chairs: their new settlement pattern resulted in the formation Manfred K.H. Eggert and Conny of new ethnic entities, however, with a shared cul- Meister. tural past explaining the similarities observed today.

2.1 Alice Lucie Mezop Temgoua, 2.2 Manfred K.H. Eggert, Eberhard Universite Libre de Bruxelles, Belgium: Karls Universitat Tiibingen, Germany. Archaeology, oral traditions, ethnogra- Fortified settlements in the inner Congo phy, linguistics, and the settlement dy- Basin: a multi-faceted approach namics in the Faro area (Cameroon and During archaeological reconnaissance work Nigeria). along major rivers in the inner Congo basin a number The mountain dwellers of the Faro area are gen- of fortified settlement sites were discovered. They erally considered to be refugees in the region, but in consisted usually of a circular ditch with a diameter fact, very little is known concerning their past. They between 50 and 100 m. Exploratory excavations in live in isolation on several mountains, yet their cul- one of them yielded ceramics belonging to the time ture is quite similar. Events in the past may explain of early European exploration from the mid-lS70s on- this peculiar pattern. Because little information is ward. According to oral tradition collected during available, an attempt was made at combining oral tra- reconnaissance these settlements were a widespread dition, archaeology, ethnography, and linguistics in phenomenon in those days. They had to be given up order to examine the past of the Faro plain and the because the colonial administration forced the popu- encircling mountains, at least from the 17th century lation to do so. In addition to archaeological and onward, and elucidate the processes that account ethnohistorical evidence there are some references for settlement in the area over that period. The study to these villages in early missionary literature. We

65 NYAME AKUMA No. 70 DECEMBER 2008

are thus able to elucidate a specific historical phe- 2.4 Marie-Claude Van Grunderbeek, nomenon by means of archaeological, oral, and writ- Musees de la Ville de Bruxelles -Maison ten evidence. du Roi, Belgium/ Environment in Rwanda in the Early Iron Age: a case 2.3 Alexandre Livingstone Smith and study. Anneleen van der Veken, Musee Royal Several aspects of the environment and hu- de l’Afrique centrale, Belgium. Crossing man interference study in Early Iron Age Rwanda Borders II: new data on material and were published in distinct articles. This communica- immaterial culture in the Luba area tion aims to show how Emile Roche, Hugues (Katanga, DRC). Doutrelepont and Marie-Claude Van Grunderbeek proceeded, relating conclusions to come to a whole Central Katanga (DRC) has been the focus of picture. This will be completed by (hopefully) new scholarly attention since the arrival of the first ex- results on phytolith analysis. plorers up until today. Thanks to extensive archaeo- logical surveys and excavations, the archaeological sequence goes back to the 9th century AD. This re- 2.5 Conny Meister, Eberhard Karls gion is also home to one of the most famous ancient Universitat Tiibingen, Germany. Re- savannah kingdoms of south central Africa, that of marks on the first Early Iron Age funeral the Luba. According to historians using oral histori- tradition of southern Cameroon. cal data, the Luba kingdom goes back to at least the 18th century AD. However, the connection between Since 2004 the Research Unit 510 of the Ger- the people living in the area today and those who man Research Foundation (Deutsche inhabited central Katanga in the past is not clear. Forschungsgemeinschaft) has been exploring the Indeed, the material culture uncovered during the changing interrelationship of environment and cul- excavations displays both elements of continuity and ture in the forest-savanna regions of West and Cen- dramatic changes throughout the sequence, particu- tral Africa. The project’s Central Africa team has con- larly as regards ceramic morphological and ornamen- centrated on survey work as well as excavation of tal styles. To explore the recent history of the area archaeological features related to the settlement of and the cultural variations observed in the archaeo- the rain forest of southern Cameroon between 2500 logical sequence, we are developing an approach and 1800 BP. Deep pits, discovered during based on the integration of archaeology and histori- prospection of road profiles and vegetationless zones cal linguistics. To do so, we have decided to study are the most common feature. They contain mainly the spatial distribution of contemporary ceramic tra- ceramics, stones, macrobotanical finds and charcoal, ditions, pottery manufacturing processes and vo- but also slag, tuyere fragments and small iron ob- cabulary as an interface between the present and the jects, such as rings, axes or knives. Apart from these past. structures, elongated shallow features with rich iron and ceramic furnishings have been found at This paper presents the ongoing research of Akonetye and Campo. Features with a similar array our project. We will briefly outline the distribution of of finds have also been discovered in other parts of technical traditions in the area to show how some of the forested regions of southern Cameroon, e.g.in the primary shaping technologies were identified on Mouanko-Lobethal and Mpoengou/Kribi. It is a well- archaeological vessels going back to the 10th century known fact that bones are rarely preserved in tropi- AD. We will then outline the first results of our cal soils. Their destruction is caused by the high dialectological study. Finally, we will consider the acidity of the latter. It is therefore not surprising that connections between the Upemba depression and only scant evidence of bone has been recovered. the broader framework of south-central Africa. However, we are confident that the structures under discussion represent burials. The shape of the fea- tures as well as the arrangement of iron and ceramic artifacts found within them supports this interpreta- tion. Furthermore, the iron objects may well have been

66 NYAME AKUMA No. 70 DECEMBER 2008

grave goods. It seems evident that these features 2.8 Pascal Roger Nlend Nlend, embody the first Early Iron Age funeral tradition thus Universite de Yaounde I, Cameroon. far discovered in southern Cameroon. According to Synthese preliminaire de l’etude du site radiocarbon dates it existed between about 2000 and 1700 BP. archeologique de Bwambe (Kribi, Litto- ral du Cameroun). 2.6 Christophe Mbida, Direction du La région du littoral sud-camerounais com- Patrimoine Culturel du Ministère de la mence a être connue sur le plan archéologique. Des prospections et fouilles menées depuis les années Culture, Cameroon. Preliminary study of 2000 ont mis en évidence de nombreux sites parmi soil samples from Iron Age Campo lesquels celui de Bwamhe; il apparait comme une necropoles. référence dans la zone, parce qu’il a connu une re- cherche interdisciplinaire (archéobotanique et No abstract. archéologie). La céramique du site de Bwambe présente principalement deux traditions céramiques 2. 7 Bienvenu Gouem Gouem, qui semblent avoir coexistées pendant le stade Université Libre de Bruxelles, Belgium. néolithique et dont l’une a perdurée jusqu’a l’Age New perspectives on early farming du fer ancien. L’apport de la carpologie et la palynologie permet également d’entrevoir le contrôle settlement patterns on the southern par ces populations sur les espèces végétales telles coast of Cameroon. que Pennisetum glaucum, Elaeis guineensis et Since June 2001 we have been carrying out Canarium schweinfurthii. archaeological research in the southern littoral re- gion of Cameroon; initially along the layout of the 2.9 Richard Oslisly 1, Pierre Kinyock2, Chad-Cameroon pipeline (between 200 I and 2004) as Pascal Nlend Nlend3, Francois Ngouoh & part of the archaeological program of the Chadian 4 Project of Export Oil, subsequently (between 2004 Olivier Nkonkonda . Archéologie de la and 2005) in the immediate vicinity of this layout for région de Douala (Cameroun): premiers a PhD program. These archaeological investigations résultats de la fouille de sauvetage du have revealed a remarkable number of sites. Essen- site de Dibamba (1. Institut de recherche tially composed of pits, these sites belong to the first pour le développement (IRD), village communities who inhabited in the southern Cameroon; 2. Université Libre de forest region of Cameroon c. 3000 BP onwards. An effort was made to understand the original infilling Bruxelles, Belgium; 3. Université de process as well as the primary and secondary func- Yaoundé I, Cameroon; 4 COTCO, tions of the pit structures. The analysis of ceramics Cameroon). from the sites of Bissiang, Dombe, Bidou, Bidjoka, Dépuis 2000, les recherches archéologiques sur Talla, and Mpoengu enabled the definition of three le littoral sud du Cameroun (Kribi-Campo) ont mis en major pottery traditions. The oldest of these tradi- évidence l’existence d’un peuplement qui s’étale de tions appears to be an extension of the Obobogo l’Age de la pierre récent à l’Age du fer. A l’exception Tradition, identified in the Yaounde region at the for- d’une attention sommaire accordée à la région de est-savannah border. Also, a set of iron objects as- Mouanko sur l’estuaire de la Sanaga, la partie nord sociated with overturned vases (most likely funerary de la côte camerounaise n’a suscite que peu d’intérêt. remains) were found at the Mpoengu site. This ex- Dans le cadre d’une archéologie préventive, les en- ceptional phenomenon has already been observed virons de Douala qui ne disposaient d’aucune source at other sites in the forest area of the country. These archéologique, révèlent un riche potentiel confirme results yield important elements for understanding par la découverte du site de Dibamba. Les études the settlement patterns and the ancient lifestyles in préliminaires permettent d’ébaucher un premier cadre southern forested Cameroon, and hence for the whole chronologique de près de 3000 ans, se résumant en: of the Central African forest area.

67 NYAME AKUMA No. 70 DECEMBER 2008

• un stade néolithique défini par des pots bilobés the sub-actual and ancient populations. The major- et des outils polis sur basalte; ity of the excavated structures in the pays tikar seem to date to the late Stone Age or Iron Age. These pits • un Age du fer qui se distingué par une céramique are the subject of studies and several debates with aux formes et décors complexes avec la présence regard to their functions. Among the studied regions, de la roulette et une grande maitrise de la the Tikar region is still poorly understood from this métallurgie du fer; point of view; this region remains particularly inter- • une période précoloniale caractérisée par des esting with regard to the regional distribution of pit récipients en faïence et de nombreuses perles de sites in the central and southern part of Cameroon. verre bleues importés d’Europe. The presence of such sites is a major feature in the reconstruction of chrono-cultural sequences in Cen- tral Africa. With a presentation and description of 2.10 Marie Juliette Leka, Universite de these pits, we will try to add to the available data Yaoundé, Cameroon / Université de from pit sites in southern Cameroon (e.g. Avoh, Paris, France. Etude préliminaire des Ndindan Nkang, Nkometou, Obobogo and Okolo), sites archéologiques en pays tikar: especially regarding functions and the role played description et analyse d’une série de by these excavated structures in the economy ofIron fosses (Preliminary study of archaeologi- Age sub-Saharan populations. cal sites in the Tikar region of central Cameroon: description and analysis of a 2.11 Alfred Ngomanda, Katharina series of pits). Neumann, Stefanie Kahlheber and Alexa Hohn. Goethe-Universitat Frankfurt, Dans la partie méridionale du Cameroun, les recherches visant à la reconstruction du peuplement Germany. What happened in the Central a travers l’archéologie remontent à plusieurs African rain forest during the 3rd millen- décennies. Parmi les régions étudiées, la vallée du nium BP? Mbam, encore appelée pays tikar, reste méconnue. During the third millennium BP, a dramatic deg- Dans cette partie du Cameroun, comme dans d’autres radation of the evergreen rain forest and an exten- régions plus méridionales, l’essentiel des structures sion of pioneer formations and savannas occurred in reconnues sont de grandes fosses qui ont fait l’objet Central Africa. Although the serious disturbance of de nombreux débats concernant leurs fonctions the rain forest is thought to have favoured the immi- primaires. La majorité des structures excavées gration of early Bantu agriculturalists, the knowledge reconnues dans le pays Tikar semblent dater du début about the timing of this event and the underlying de l’Age du fer. Les ‘sites a fosses’ identifies causal factors is still limited. We present new localement semblent ainsi rapprocher les anciennes multidisciplinary, high-resolution palaeo- occupations du pays Tikar de celles des régions plus environmental data from Central Africa for the late méridionales (cf. les sites d’Avoh, Ndindan, Nkang, Holocene. A combination of marine data with diatom, Nkometou, Obobogo et Okolo), bien davantage que pollen and archaeobotanical evidence shows that the de celles reconnues au nord du pays. Après une perturbation of the rain forest occurred after 2500 cal présentation et une description de ces fosses, nous BP and was probably induced by a much more ac- tenterons d’apporter de nouveaux arguments au centuated seasonality. A climate with a distinct boreal débat concernant la fonction de ces structures winter dry season seems to be linked with abrupt emblématiques des occupations anciennes du Cam- warming ofthe subequatorial Atlantic Ocean, associ- eroun méridional. In the central part of Cameroon, ated with enhanced NE trade winds (harmattan). We research aiming at the reconstruction of the past propose that the Central African rain forest crisis through archaeology goes back several decades. between 2500 and 2200 cal BP created favourable Among the regions studied, the Mbam valley, also conditions for farming and paved the way for a major called pays tikar, remains neglected. In this region of expansion of Bantu-peaking populations. Cameroon, as in several regions of sub-Saharan Af- rica, pits are studied as part of the reconstruction of

68 NYAME AKUMA No. 70 DECEMBER 2008

2.12 Benjamin Smith, University of the 3. 2 W. Paul Adderley, University of Witwatersrand, South Africa. ‘The Flesh Stirling, Great Britain and Carlos Eaters’. A prehistory or the pygmies of Magnavita, Goethe Universitat, Frank- Central Africa? furt, Germany: Multi-use ditches or The subject of this paper is the social con- simply mud? Geoarchaeological investi- struction and deconstruction of African pygmyness gations at Gajiganna Culture sites in by those living, mostly, outside of Africa. It concerns northern Nigeria. academic discourse, social values, ethnocide and ar- Past archaeological studies around the com- chaeology. plex of sites at and around the town of Gajiganna in northeastern Nigeria have focused upon examining the Neolithic communities in the southern Lake Chad basin. The sites suggest a dynamic period of settle- 3. Trajectories to Complexity in West ment between the early second and the middle of the and East Africa: A Session in Honor of first millennium BC. Within that wider complex of Graham Connah. Session Chairs: Detlef sites, several have been examined. They have shown Gronenborn and Scott MacEachern. demonstrable phases of activity with differences in finds and settlement structure that reflect changes in 1. Emerging complexity in the Neolithic material culture and socioeconomic developments. In parallel with these temporal issues, the location of and Earlv Iron Age. the sites produces questions concerning settlement organization and societal complexity both at intra- 3.1 Randi Haaland, Universitetet i site and regional scales. Successful examination of Bergen, Norway: Changing food ways as these issues requires a geoarchaeological under- indicators of emerging social complexity. standing of the landscapes surrounding the early From the Neolithic agro-pastoralists to settlements to address questions relating to interac- tions between the natural environment and past so- the Meroitic civilization. cieties. If the earliest Gajiganna Culture sites are con- Based on archaeological material from the Su- sidered as a starting point, the period to be examined danese Nile Valley one can observe that the typical offers a pronounced contrast in settlement activities, African food system based on porridge and beer from a period of initial settlement within a relatively emerged from c. 6000 bp. In this area beer and por- pristine natural landscape, to the later period where ridge were made from summer-growing crops such settlement dynamics will have been more strongly as sorghum and millet. Over time the emergence of an influenced by factors related to human activities in, increased variety of pottery types (small vessels, such and in the vicinity of, each site. This paper considers as cups) used for the serving of liquid foods and the distinct deep encircling ditches around the late drink developed. This is seen as related to increased Gajiganna sites of Zilum and Rungwa. Such ditches social differentiation where drinking was part of so- are considered unique in this period of African pre- cial display. Beer seems especially important for use history. Using geoarchaeology methodologies such in the ritual spheres from the early part of 4th millen- as sediment thin-section micromorphology and in situ nium bp to the later Meroitic period. Many of the jars OSL dating to examine the ditch sediments, allows a from the Late Meroitic period had the capacity to consideration of their function. Key issues are to contain several hundred liters of beer. The finer what use or uses these ditches were put by people Meroitic pottery, the drinking cups and goblets ap- during site occupation and whether the ditch-fill rep- pear to have been used in special contexts of ritual resents abandonment or simply the absence of a need consumption, which also include ceremonial break- to (re)dig ditches. Evidence for possible multiple uses ing of the vessels. of these ditches will be presented and discussed in respect of historical decision-making, landscape re- sources, and the wider environmental context. This methodology allows consideration of more generic

69 NYAME AKUMA No. 70 DECEMBER 2008

issues of complexity within archaeological and an- vine rulers. There are thus no clear-cut separations thropological settings, and of adaptive systems; in from the more simple manifestations of the sacred this case, the stability sought by local actors when dimension of power to the most elaborate kind of initially developing sustainable settlements can be divine kingship. seen to be addressed within opposing constraints presented by environmental and societal factors.

3.3 Nicole Rupp and Peter Breunig, IIa Emerging complexity in West Africa Goethe-Universitat Frankfurt, Germany: 1, Anne Mayor1, Recent studies on the Nok Culture, 3.5 Eric Huysecom Barbara Eichhorn2 Central Nigeria. and Genevieve Perreard1. : Dourou-Boro and Pegue: a For decades the Nok Culture of Central Nigeria re-evaluation of ‘Toloy’ constructions in has been well known because of its terracotta figu- the Dogon country (Mali). (1. Université rines, representing the earliest sculptural art in sub- Saharan Africa. In addition, Nok plays a prominent de Genève, Switzerland; 2. Goethe- role in the emergence of iron technology, providing Universität Frankfurt, Germany) some of the earliest evidence of iron smelting in West In the Dogon country of Mali new discoveries Africa (500 BC). Despite its scientific importance, made in 2007 and 2008 demonstrate an increase in little archaeological fieldwork has been devoted to socio-economic complexity from the beginning of the the Nok Culture. In 2005, investigation of the Nok 1st century AD and make it possible to re-evaluate the Culture became part of the German DFG Research chronology, the technique and the meaning of ‘Toloy’ Group 510 “Ecological and Cultural Change in West constructions. In the Dourou region, along the Cliff and Central Africa”. Surveys and excavations on Nok of Bandiagara, circular constructions placed side by sites have lead to discoveries comprising various side, located in several horizontal faults in a small aspects of the Nok Culture: patterns in site location, valley perpendicular to the Cliff of Bandiagara, have economy, iron technology, pottery, stone tools” and, walls erected by superimposition of clay coils some- of course, terracotta, discovered in contexts that will times plastered by a clayey coating, and tempered illuminate the latter’s still enigmatic function. This with small millet chaff or leaves. A small semi-circular paper will present some of the new discoveries. opening is placed against the rocky ceiling. Similar constructions discovered at Pegue near Sanga have II Emerging complexity in the first and been interpreted until recently as granaries from the rd nd second millennia AD 3 - 2 centuries BC, built by the ‘Toloy’ population, then re-used as burial places during the 11th century by the ‘Tellem’ population (meaning “those who came 3.4 Pierre de Maret, Université Libre before” in Dogon). Twelve dates obtained at Dourou- de Bruxelles, Belgium: From kinship to Boro, and seven new dates from Pegue-Sanga, as kingship: an African journey into com- well as a detailed analysis of the archaeological ma- terial indicate that these were buildings conceived plexity, or is it perplexity? from the start as burials, built between the 3rd century Since Frazer’s seminal work, the notion of di- BC and the 13th century AD. The archaeological ma- vine kingship has attracted much interest. It has gen- terial from Dourou-Boro notably includes beads of erated an enormous amount of literature having close semi-precious stones and glass paste imported from relationships to both religion and politics. In Africa, the Middle East, indicating the existence of an in- one can find examples of small-scale societies with- crease in complexity with the presence of social out a state-like or even a chiefdom structure where hierarchization and long-distance contacts from the nevertheless some individuals are vested with cru- beginning of the 1st century AD, well before the ar- cial ritual functions, similar to those of sacred or di- rival of the Dogon.

70 NYAME AKUMA No. 70 DECEMBER 2008

3.6 Anne Mayor, Université de Genève, Despite the intense pillaging at Gao-Saney, two ar- Switzerland. The Toloy-Tellem sequence eas with undisturbed deposits were excavated in 2000 (Dogon country, Mali) revisited through to a depth of approximately 7 m. Both produced a series of radiocarbon dates between 1250-1100 BP ceramic and iron objects studies: a (uncalibrated). complex history ofpopulation revealed during the pt millennium AD. At Gao-Ancien, extensive excavations have been underway since 2003 in the area known locally Before the research program “Palaeo- as ‘the mosque of Kankou Moussa’. Several large environment and Human population of West Africa”, stone and/or brick buildings have been uncovered. archaeological works conducted in several caves of The primary early stone construction phase appears the Bandiagara cliff near Sanga showed a cultural to date to the 9th and lOth centuries CEo Basal depos- sequence characterized by a ‘Toloy’ occupation (3rd- its have provided dates between 1380-1180 BP 2nd centuries BC), a ‘Tellem’ occupation (11th - 15th (uncalibrated). This area was intensively occupied, centuries AD) after a hiatus of more than one millen- with much reuse of earlier stone materials, as well as nium, followed by a Dogon occupation from the 16th clearing and dumping of refuse and building materi- century AD. A lack of stratigraphy as well as chang- als. This paper will present the architecture of the ing functions of the same spaces depending on peri- Kankou Moussa site and the challenges of interpret- ods of time or users (alternation of domestic, ritual, ing its chronology. Finds from both Gao-Ancien and and funerary uses) made it difficult to understand Gao-Saney will be discussed and compared, with pre- the ceramic assemblages, often mixed up. For ten liminary observations on the potential of these sites years now, the excavation of several new sites on the relating to the development of trade and complex plateau, at the summit and at the bottom of the cliff, society on the eastern bend of the Niger. as well as in the Seno plain, gives a different image of the population of the Dogon country during the 15th millennium AD, with evidence of metallurgical (Fiko), 3.8 Kevin C. MacDonald, University domestic (Promontoire, Songona), ritual College London, Great Britain: Cities (Dangandouloun) and funerary (Dourou-Boro) ac- and the Invisible State: lessons from the tivities during the whole period previously regarded oral history and archaeology of Segou. as a cultural hiatus. The contrasted ceramic assem- blages reveal a territory at the crossing of influences Over the past few decades most of the archae- from the Mande, Gur and Proto-Songhay spheres, ology of the Middle Niger has focused on long-term suggesting the cohabitation of different groups. Cer- occupations, particularly urban tells. Such work has tain ceramics, e.g. highly decorated three-footed also given rise to models of social organisation which, bowls, as well as iron objects, be they pieces of jew- while undoubtedly appropriate to urban clusters, ellery, arms, or agricultural tools, also show a high cannot be easily applied to the relatively short-lived local technical knowledge of pre-Dogon blacksmiths polities which surround these cities. Recent research and potters. on the heartland of Bamana Segou provides some new insight into the dynamics, typology, and mor- phology of settlements founded or modified by the central organizing force of the state. Local oral histo- th 3.7 Susan K. McIntosh and Mamadou ries of place, concerning the geography of 18 and 19th century Segou, are complemented by the inter- Cissé, Rice University, USA. Recent pretation of archaeological surface survey from 2004- excavations at Gao-Saney and Gao 2005. Ancien (Mali): preliminary results. Since Tim Insoll’s pioneering work at Gao in the 1990s, new research has been undertaken at both Gao-Saney (located 8km from Gao) and Gao-Ancien.

71 NYAME AKUMA No. 70 DECEMBER 2008

3.9 Anne C. Haour, University of East the Kanem-Borno kingdom. This paper traces these Anglia, Great Britain: Garumele, former developments using some examples from recent and Kanem-Borno capital? past archaeological research around Lake Chad. A polity with a continuous presence in the Chad Basin of perhaps nine hundred years, Kanem-Borno 3.11 Detlef Gronenborn, Romisch- represents one of sub-Saharan Africa’s best-known Germanisches Zentralmusem Mainz / examples of social complexity. One of its most endur- Johannes Gutenberg Universität Mainz, ing enigmas remains the site of Garumele, Republic Germany and Scott MacEachern, of Niger. Garumele has been many times described, Bowdoin College, USA:A tribute to and is said by oral tradition to have served as the Graham Connah. capital of Kanem-Borno after the exile of its leaders from Kanem, in modern Chad. The capital is then said to have shifted to Birnin Gazargamo, in the Yobe Val- 3.12 Scott MacEachern, Bowdoin ley (Nigeria), known through the excavations of College, USA: Wandala and the DGB Graham Connah in the 1960s. In the Nigerian part of sites: political centralisation and its the Chad Basin, archaeologists have shed light on alternatives north of the Mandara past occupation sequences relating to Kanem-Borno, Mountains. and have developed detailed ceramic sequences; well-known sites include Daima, Birnin Gazargamo, The DGB sites (named after the Mafa term Diy- and Ngala. In stark contrast, Garumele, and the gid-biy), with their striking dry-stone architecture, Nigerien side of the Chad Basin as a whole, have are the largest and among the earliest archaeological remained largely unexplored. Accordingly, preliminary sites in the northern Mandara Mountains. Even given archaeological work was initiated by Anne Haour and their size and internal complexity, they provide only Boube Gado at Garumele in 2005, with special atten- ambiguous messages about power and authority in tion to issues of ceramic traditions and chronology. the region five centuries ago. At the same time, these This paper will suggest how Garumele may fit, cultur- sites are situated in proximity to the heartland of the ally and chronologically, within the activities of Wandala polity, which was first noted in European Kanem-Borno, and it will widen its scope to consider and Arabic sources at the time of DGB occupation. what the traditions, which speak of successive capi- Over the next centuries, Wandala progressively dif- tals for the polity, may be trying to say regarding ferentiated itself from neighbouring Chadic-speak- socio-cultural complexity in this part of the world. ing communities, adopting the political appurte- nances and expansionist tactics of an Islamic Sudanic state. This paper will discuss the implications of this 3.10 Carlos Magnavita, Goethe- geographic and political proximity for both Wandala Universitat Frankfurt, Germany: and the occupants of the DGB sites. Displaying power: mudwalled settle- ments and fired-brick architecture at 3.13 Detlef Gronenborn, Romisch- Lake Chad. Germanisches Zentralmusem Mainz / Over the last few years, it became apparent Johannes Gutenberg, Universität Mainz, that major permanent settlements to the southwest Germany and James Ameje, National of Lake Chad were probably once surrounded by Commission for Museums and Monu- protective mud walls or ramparts. This trend began by about the middle of the first millennium BC and ments, Nigeria. Durbi Takusheyi – persisted up to the 19th century AD. Alongside their citadels without cities? practical function, these structures probably also dis- The 13th/14th century AD elite burial site of Durbi played the power of the communities they enclosed. Takusheyi in present-day Katsina State, northern th th From the 12 - 14 centuries AD onwards, fired-brick Nigeria, is situated mid-way between the two early buildings additionally became part of the symbolic Hausa city states of Daura and Katsina. No remains repertoire of power in the territories dominated by of any associated larger settlement have hitherto been

72 NYAME AKUMA No. 70 DECEMBER 2008

identified. Despite its apparently isolated location, of Segu between the 17th and the mid-19th century the burials contain rich goods of local manufacture AD. The malleable character of Somono identity is but also imported material from distant regions of the further attested by their incorporation of people from Islamic world. With this pattern - known from vari- virtually any background, such as Bamana, Soninke, ous regions world-wide - the Takusheyi burials re- Songhay, Mossi, and Peulh. The Somono have also flect a characteristic behaviour of emerging elites and been portrayed as namakalaw, a term that describes also throw light on the constitution and organisation occupationally defined artisans and bards who guard of initial state-level societies in the west-central Bilad their professional secrets through endogamy and aI-Sudan. esoteric ritual procedures. Recent studies of women craft specialists in the Inland Niger Delta, who iden- tify themselves as Somono and numusow (black- 3.14 Akin Ogundiran, University of North smiths’ wives), hence Somono/numu, might throw Carolina, USA. Aspects of everyday new light on the origins of the Somono and their material life in Ede-Ile (southwestern status as nyamakalaw, which is further attested by Nigeria), c. 1600-1830s. the region’s long sequence of occupation. Most ar- chaeologists, however, have been reluctant to project A combination of material culture analysis and the techniques used by today’s ethnic groups into settlement archaeology is applied to study the dy- the past and to recognize the same groups at removes namics of everyday life in Ede-Ile (southwestern Ni- of a millennium or more. Nevertheless, in this paper I geria), a colony of the Oyo Empire during the will illustrate that our investigations at Dia, located th 1ihthrough the early 19 century AD. Excavations in at the western edge of the Inland Niger Delta, offer the northeastern quadrant of the settlement allow us promising clues to carry out an ethno-historical study to identify some of the major activity areas and the of Somono/numu origins. By combining archaeologi- spatial patterns of artifact distribution. The cal, ethnographic and historical data I will explore the depositional patterns and analysis of the artifacts likelihood that the Somono might have emerged prior reveal the dynamics of material life in relation to pro- to the Bamana State of Segu. duction, consumption, regional exchange networks, imperial colonization, migrations, and everyday ex- istence. The presentation will emphasize the impor- 3.16 Judy Sterner, Alberta College of Art tance of linking approaches of material culture analy- and Design, Canada. Gender and mean- sis with settlement studies. The methodological and ing in Basketry: the view from the theoretical implications of these approaches, I argue, Mandara Mountains. offer archaeology a critical niche in making important contributions to the social history of quotidian life The importance of basketry as a potential during the era of Atlantic economy and regional im- source of information on society and culture has yet perialism in West Africa. to be generally accepted. In this paper I describe, using Willeke Wendrich’s classification, the varie- ties of basketry made and used in the Mandara Moun- 3.15 Noemie Arazi, Heritage Manage- tains of Nigeria and Cameroon. This class of material ment Services, Brussels: Fact or fiction? culture, manufactured by both men and women, cor- The presence of Somono/numu in the relates with gender and ethnicity very differently from Inland Niger Delta of Mali before the other, specialist-made, artifact classes. Everywhere rise of Bamana Segu. there is a strong association of technique with gen- der rather than other social categories. Male-made Together with the Bozo, the Somono are basketry in the Mandara Mountains is clearly linked Mande-speaking fishing populations of the Inland to architecture, and a link between coiled baskets Niger Delta. Even though they share the same family and coiled pottery might be argued. But unlike pot- names, the fundamental difference between them is tery where types are often strongly associated with that the Bozo have been identified as an ethnic group, one or the other gender, the highly gendered differ- while the Somono have been described as a distinct ence in techniques is not mirrored in use. Baskets occupational group associated with the Bamana State rarely figure in ritual or as markers of status. There

73 NYAME AKUMA No. 70 DECEMBER 2008

are hints that they, like granaries, may sometimes have excavations indicate that this pattern extends back gender thrust upon them, but, unlike pots, baskets into the first millennium AD, and probably earlier. are not, or only very rarely, people. Consideration of Although likely engaged in localized trade, the coast the small number of other case studies of African was at the extreme periphery of wider trade networks basketry suggests that in anyone region at anyone of the northern forest and savanna. Increasing ur- time gender and basketry technique are highly corre- banization is the most striking change in African so- lated, although there is variability both across space cieties on the coast during the post-European con- and through time. Second, there is considerable vari- tact period. This began as a gradual process during ation in the extent to which symbolic thinking per- the 15th and 16th centuries AD, but culminated during vades African basketry. the following centuries. Beginning with the found- ing of Castelo de Silo Jorge da Mina by the Portu- guese in 1482, Europeans established numerous out- 3.17 Christopher R. DeCorse, Syracuse posts on the central Ghanaian coast. African towns University, USA. Complexity in the era adjacent to these European forts and castles became of the Atlantic World: perspectives from centers of commercial activity and the conduits the Central Region Project, coastal through which European goods flowed to the inte- Ghana. rior, as well as points from which enslaved Africans were transported to the Americas. With the advent of This paper examines socio-cultural transforma- the Atlantic trade, many of the earlier, smaller coastal tions in coastal Ghana from the first millennium settlements were abandoned and towns associated though the 19th century AD, with particular emphasis with European trade entrepots and the capitals of the on trade, technological change, and the emergence states in the immediate hinterland, such as Asebu, of complexity during the era of the Atlantic World. Efutu and Eguafo, expanded in size and importance. While historians have examined many aspects of this Archaeological research from both the coast and hin- period, the paucity of written sources for much of the terland will be used to illustrate the impact of the time period constrains assessment of the transfor- European trade and the way in which this impact is mations that occurred. Archaeology affords a time represented archaeologically. Fieldwork undertaken depth not accessible through the documentary record during 2007 is reported on, including the ongoing or oral traditions, in this way providing the most valu- survey of terrestrial and underwater sites in Central able source of information for evaluating African so- Region, and excavations at the British fort at cieties over the past 500 years. This paper briefly Komenda. examines archaeological data from across West Af- rica and then focuses specifically on developments in southern Ghana between AD 1400 and 1900. The central Ghanaian coast can be characterized by its IIb Emerging complexity in eastern and comparative lack of urbanization prior to the advent southern Africa of the Atlantic trade. In the late 15th century AD small fishing and fanning villages were scattered along the coast and the adjacent hinterland. Evidence for the 3.18 Paul J. J. Sinclair, Uppsala pre-European contact occupation of the coast is rep- Universitet, Sweden. The Urban Mind: resented by surface scatters and midden deposits long-term urban dynamics in eastern and along low rises just above the shoreline or adjacent southern Africa. to lagoons and, in some locales, hilltop sites. In some instances, scatters of surface material extend for al- This paper explores a new concept, the ‘Urban most a kilometer along the shore, suggesting signifi- Mind’, from combined humanities and natural sci- cant pre-15th century AD coastal occupation. While ence studies of the development of urbanism and testament to the importance of coastal resources climate change in eastern and southern Africa. The during the pre-European contact period, this material ‘Urban Mind’ concept is being formulated as part of likely represents small, shifting settlements occupied the global IHOPE (Integrated History and Future of over long periods of time rather than single, continu- Peoples on Earth) initiative. In regard to Africa spe- ous settlement. Recent archaeological survey and cific focus will be made upon the urban landscape

74 NYAME AKUMA No. 70 DECEMBER 2008

dynamics and environmental change in the Shashi 3.19 Peter Robertshaw, California State Limpopo Basin and the Swahili coast. The develop- University San Bernardino, USA. “Two ment of urbanism has been a global phenomenon tonnes of excavated potsherds”: reflec- unfolding as co-evolving human-environmental sys- tems, unfolding over millennia but taking radically tions on state formation in western different fonns in eastern and southern Africa. Cru- Uganda. cially, in addition to their physical characteristics Graham Connah once remarked that two tonnes towns are ideological constructs: as we invent them, of excavated potsherds were unlikely to tell us any- so we believe in them. Drawing upon Bateson’s early thing about the Bacwezi dynasty. In response to re- approach to human-environmental interactions in the cent critiques, I offer a more heterarchical and less ‘Ecology of Mind’ as well as contributions from the neo-evolutionary perspective on the archaeology of new field of Historical Ecology, this paper proposes the Bacwezi and the early state in western Uganda a combined humanist and natural science explora- than I have done previously, while further clarifying tion, and delineation of the cultural and environmen- the region’s chronology and settlement history, as tal dynamics of the ‘Urban Mind’ as part of the IHOPE reconstructed from not only ceramics, but also ar- initiative. Towns in southern and eastern Africa arose chaeological surveys and sediment cores. In doing from existing patterns of human settlement, over the so I reflect upon some of the major economic, politi- last two millennia. GIS coverages from the region will cal, and climatic forces relevant to understanding the be used to illustrate the development of urbanism. development of complexity in this part of the world. Towns as spatially ordered, demographically dense This discussion touches upon the results of settlements or clusters of settlements, and their affili- Connah’s excavations at Kibiro, as well as new evi- ated communities add a cognitive dimension to the dence from glass bead analysis. landscape. A landscape is seen as a multidimensional mosaic that relates social constructivist and bio-geo- physical aspects of reality. The complex interactions 3.20 Andrew Reid, University College of urban factors at different scales often defy sim- London, Great Britain. Animals and plistic models of linear development. Towns contain power in the archaeology of Great Lakes a critical mass of people often from differing cultural Africa. backgrounds, producing a variety of ideas, goods and services. Towns are primarily attractors of hu- When Graham Connah produced his important mans but also support a diverse range of other ani- 1987 volume on African states, the coverage of the mals and plants. Urban dynamics in eastern and Great Lakes region was necessarily thin. Despite the southern Africa are characterized by a variety of lo- significant amount of work undertaken in the last 20 cal and regional socio-economic interactions and are years, there still remains an absence of detail on also influenced by forcing mechanisms deriving from economies, one of the main building blocks of Connah’ long-distance effects of changes in world system and s approach. This paper will review the information physical climate as shown in palaeoclimatic modeling we now have available for animal exploitation in the work of Bryson and Bryson and recent work on high region. Derived from a wide range of sites, these not resolution climate change in eastern and southern only highlight the economic role of certain animals, Africa by Holmgren et al. Taken together, we aim particularly cattle, but also their significance in po- ultimately at a genuinely culturally and environmen- litical and religious activity. tally informed approach to the cognitive dimensions of urbanism, an increasingly prevalent form of societal organization in the region today.

75 NYAME AKUMA No. 70 DECEMBER 2008

3.21 Jeffrey Fleisher, Rice University, 4 Tracing the Diversity of African Plant USA. Daily practice in early East Afri- and Animal Exploitation Contributions can coastal houses: two burned 7th – 9th of Natural Sciences. Session Chairs: century houses from Pemba Island, Katharina Neumann and Stefanie Tanzania. Kahlheber Excavations of two burned 19th century houses at Tumbe on Pemba Island, Tanzania, provide unique 4.1 Geeske H. J. Langejans, University insight into the daily lives of early coastal people. of the Witwatersrand, South Africa. That these houses burned, collapsed and buried a Much ado about nothing? Fact and relatively intact assemblage of household materials fiction in organic residue analysis. offers a rare opportunity to examine the contents of early coastal houses, and to begin to understand the Residue analysis aims to identify microscopic spatial setting in which people lived and carried out remains or traces that are left on a tool’s surface after their day-to-day practices. This paper will consider use. Analysts have identified organic remains such the structural arrangement of the house, as well as as plant starch grains, fibres, tissue cells, resin, ani- forms of production indicated by the material record mal blood films and red blood cells, muscle tissue, abandoned within the houses themselves. By recon- connective fibres, bone fragments, hairs, fish scales, structing household activities, the goal of this paper and shell. Besides organic residues, it is also possi- is to begin articulating the importance of practice to ble to distinguish inorganic deposits such as ochre. the constitution of social relations along the Swahili Residue analysis is conducted with a light micro- coast. Household archaeology offers one way of scope, using a range of magnifications, and a large doing this, by understanding that the material record experimental comparative collection. of the house and household activities are the means This paper discusses research issues in resi- through which social life was constituted and lived. due analysis. Using examples from ancient samples from Sterkfontein and Sibudu Cave and from a large 3.22 Adria LaViolette, University of experimental sample, the author reviews issues such Virginia, USA. Urban Swahili households as the role of contaminants, how to distinguish use- related residues from contaminants and the influence from Chwaka, Tanzania, AD 1000-1600. of sediment pH, sediment moisture, rock type and Households and their associated midden and other variables on residue preservation. Favorable activity areas were the focal point of a study of large and poor circumstances for biological, mechanical and small Swahili settlements in northern Pemba, Tan- and chemical decay are discussed. This leads to pre- zania. We addressed houses as both units of com- dictions for residue preservation, which can be used parison between settlements of different times and to select sites at which residue analysis is expected scales in one settlement region, and as the key to to be fruitful. For example, it appears unlikely that understanding daily life, subsistence practices, so- open air sites and sites with high bioactivity will pre- cial status, specialization, and change within each serve residues. There is also an apparent bias in the settlement. The longest-lived and largest settlement preservation of different residues types. First, some in the study was Chwaka, an urban settlement span- residues are more durable than others. For example, ning AD 1000-1600. There, after targeting several muscle tissue and starch are more tragi Ie than plant stone-built houses for excavation, we conducted ex- tissue and bone. Secondly, different circumstances tensive subsurface testing to recover remains of affect residue types in varied ways. For example, deeply buried earthen houses in the town. This pa- starch grains preserve best in a dry alkaline environ- per will provide a comparison of the houses exca- ment and in a waterlogged acidic environment, but vated at Chwaka in terms of the diversity of domestic resin preserves best in a neutral environment with activities and clues to socioeconomic status revealed, fluctuating soil moisture. against the backdrop of Chwaka’s civic/religious ar- chitecture and life history.

76 NYAME AKUMA No. 70 DECEMBER 2008

4.2 Amanda L. Logan, University of Kintampo culture. Finds of fruit tree species like Michigan, USA and Catherine d’Andrea, Canarium schweirifurthii and Vitex cf. doniana Simon Fraser University, Canada. Iden- stress the importance of wild resources. The combi- nation of cereal and pulse based cropping systems tifying sorghum using phytoliths and with oleaginous fruit exploitation is characteristic for starch grains: a case study from Sudan. the prehistoric economy of the Sahel and Sudan zone, Phytolith and starch grain analyses have con- and we suggest that it was dominant all over the tributed a significant amount of data to understand- West African savannas at the beginning of the Iron ing the development and spread of agriculture in the Age. tropics; however, few of these microbotanical stud- ies have focused on the African continent. In this 4.4 Christine Sievers, University of the paper, we report on the development of an identifica- Witwatersrand, South Africa. Experi- tion methodology for sorghum that employs both phytoliths and starch grains. This method was ap- mental carbonization of fruiting struc- plied at Dangeil, a Meroitic site in northern Sudan, tures beneath hearths. and yielded evidence of sorghum in residues trom Various fruits, nuts and seeds from South Afri- several different types of artifacts. The presence of can trees and shrubs were buried at pre-determined domesticated sorghum at the site is further corrobo- depths and distances from the centre of experimental rated by macroremains. Future application ofthis fires. The rruiting structures were buried in coarse method holds promise for tracing the spread of sor- sand and a fine ashy matrix respectively, and differ- ghum throughout Africa and beyond. ent types and quantities of wood were used for the fires. The cold ashes of the fires and the area around 4.3 Stefanie Kahlheber, Goethe- and below them were excavated using standard ar- Universität Frankfurt, Germany. Feeding chaeological techniques and dry-screening. The ex- cavation of the various experimental hearths revealed Nok - subsistence data for the Nok the same structure and color changes seen in some Culture, Nigeria, 500 BC to 200 AD. Middle Stone Age hearths at the rockshelter of Well known for its terracotta art and providing Sibudu, KwaZulu-Natal, South Africa. The fruiting some of the earliest evidence for iron technology in structures recovered from beneath the experimental subSaharan Africa the Nok Culture of Central Nigeria hearths were affected to varying extents: those bur- remained for a long time virgin territory in archaeo- ied 5 cm below the centre of the fire, in anoxic condi- logical regards. Since 2005 a research project at the tions, were carbonized (became black throughout); Goethe University in Frankfurt (German DFG Research those buried 10 cm below the centre of the fire Group 510 “Ecological and Cultural Change in West changed color to darker shades of brown; and those and Central Africa”) has been focusing on the cul- 5 cm and 10 cm below the surface at the perimeter of tural background of the terracotta art, investigating the fire appeared unaffected. Size, shape, moisture or various aspects of the material culture as well as the oil content of the original fruit or seed did not appear economical and environmental context, settlement to influence the process. Temperatures recorded 5 patterns and chronological development of the Nok cm below the experimental fires indicate that carboni- culture. Samples taken for archaeobotanical studies zation occurred at or before a maximum temperature during various test excavations and at the site of of 328 DC, and also at lower maximum temperatures Ungwar Kura yielded few but sufficient plant remains (not above 152 DC) that were maintained for long for a preliminary reconstruction of economy and en- periods. Thus, the experiments demonstrate that even vironment. Based on the results of rruit and seed a relatively low temperature, if it is maintained for remain analysis we tentatively reconstruct a plant long enough, can cause carbonization of botanical exploitation system based on mixed cropping of pearl material. Furthermore, the experiments indicate that millet (Pennisetum glaucum) and cowpea (Vigna material that may enter archaeological deposits by unguiculata). This practice is also known from the various routes, such as human, animal or natural Nigerian Chad Basin in the same period and might means like wind, may at some later stage become have been already applied by the 2nd millennium BC carbonized by chance when a fire is lit above the area

77 NYAME AKUMA No. 70 DECEMBER 2008

where it lies buried. The implications of the experi- 2005 field season at Saïs (c. 4200 BC) in the Nile Delta, ments are that accidental carbonization of material in an Early Neolithic fish midden was excavated. A pos- cave or shelter environments may be far more com- sible parallel with the Fayum Neolithic, with an em- mon that previously thought, that carbonization of phasis on seasonal fishing, was first suspected. This seeds introduced by non-human routes of entry is year, the fish remains of the midden were analysed to possible in these environments, that interpretations test this hypothesis. Animal bone remains from two of carbonized remains should take these issues into Neolithic sites in the Eastern Desert of Egypt, Sodmein account, and that it is essential to sample below Cave (c. 6200-5000 BC) and the Tree Shelter (c. 5700- hearths and not only the hearths themselves. 3700 BC), have recently been reinvestigated. The re- sults confirm the early presence of small domestic stock there. Dung pellets found at both sites, in the 4.5 Joanna Casey, University of South case of Sodmein Cave in the shape of very large pack- Carolina, USA. An ethnoarchaeological ages, have also been examined more carefully. It was study of shea butter production in north- tried to identify the animal species that produced ern Ghana. them, as well as to establish the period of the year in which these animals visited the site. These new faunal This paper looks at the material correlates of data from the Neolithic in Egypt, in combination with shea butter production in Northern Ghana. Currently data from the literature, point to regional differences shea butter is achieving global recognition as an im- in timing of the earliest stock keeping, its relative portant ingredient in the exotic cosmetics industry, economic importance, the domestic animal species but historically it was also an important resource in involved and the mobility of the human communi- th West Africa, making it into 15 century AD travelers’ ties. These differences, and their underlying causes, accounts, appearing in markets as far away as Mo- have implications for the neolithization of Africa as a th rocco and transforming landscapes before the 18 whole. century AD. As the only indigenous oil in the sa- vanna, shea is used for everything from cooking to conditioning wood, skin and hair, as a lubricant and 4.7 Josephine Lesur-Gebremariam, in medicine and ritual. Although the carbonized re- Musée National d’Histoire Naturelle, mains of shea have been found at very few archaeo- France. Climatic changes in Holocene logical sites, it was undoubtedly an important re- Horn of Africa and animals exploitation source in the past. This paper looks at the materials involved in the production of shea butler as a means by human societies. of suggesting how we might be able to recognize this The Horn of Africa presents a great diversity industry in the archaeological record. of landscapes that have favored the installation of many different natural environments. Several recent 4.6 Veerle Linseele, Katholieke studies of lacustrine sediments, pollen sequences and geomorphologic data have contributed to a re- Universiteit Leuven, Belgium. New construction of the palaeoclimate of the Horn of Af- faunal data from the Neolithic in Egypt. rica, especially of the Rift Valley lakes. These data Since 2006, new excavations are ongoing at show that, until around 8500 BP, humid conditions Kom K and Kom W (c. 4500 Be), probably the two prevailed in the region. Then a dry phase began with most famous Fayum Neolithic sites. The excavations an optimum around 7500-7000 BP, which was followed have yielded exceptionally large faunal assemblages between 6500 and 4500 BP by a new wet phase, and that allow throwing new light on the sites’ economy. finally an arid phase that continues until today. A For the moment, more than 100,000 animal bones have short humid episode is noticeable around 2500-1000 been studied. Some domestic livestock, cattle, sheep, BP. These palaeoclimatic phases have led to the goat and pig, and wild game could be identified, but recomposition of biodiversity and ecological pres- the vast majority of the remains are from fish. The sures variable according to the different environ- species spectrum as well as size reconstructions of ments. Evolution of human societies during the the fish document seasonal activities at the sites, Holocene presents therefore many patterns charac- which may reflect a seasonal occupation. During the terized by these environmental changes.

78 NYAME AKUMA No. 70 DECEMBER 2008

To illustrate these evolutions and human ad- 4.8 Gbèkponhami Monique Tossou, aptations, we will present archaeozoological results Université d’Abomey-Calavi, Benin, from three archaeological sites occupied during the Didier N’dah, Université d’Abomey- second half of the Holocene. The first one is the rockshelter of Moche Borago, located on the south- Calavi, Benin et Akpovi Akeogninou, western edge of the Ethiopian highlands. It has pro- Université Nationale du Benin, Benin. vided evidence of human occupations during the Role de l’homme dans la mise en place Holocene. This includes more than 30,000 animal du Dahomey-Gap: apport bones, which have allowed us to reconstruct the way palynologique et archéologique. animals were exploited by humans from the 6th millen- nium BC until the first half of the 1st millennium AD. Le Dahomey-gap est la discontinuité qui sépare Results show that humans hunted, mainlyfor con- les deux grands blocs forestiers guinéen et congolais sumption, a great diversity of mammals, especially en Afrique. II est caractérise par la présence de bovids. The great stability of the faunal spectrum savanes qui descendent jusqu’a la cote. Les données during all occupation phases suggests that climatic palynologiques obtenues dans cette zone d’étude changes did not strongly affect this mountainous montrent que la période située vers 2500 ans BP et area. On the other hand, in the semi-desertic plain of l’actuel est considérée comme celle ou l’action Gobaad in Djibouti, two archaeological sites, Ali Daba anthropique s’est réellement manifestée. En effet and Asa Koma, occupied respectively during the 4th malgré les conditions favorables pour une reprise de and the 2nd millennium BC, present a very different la foret dans cette coupure phytogéographique, on situation. Indeed, the fauna from Ali Daba suggests constate plutôt une progression des savanes. the presence of a humid environment with species L’homme aurait joue et continue de jouer un rôle dans living in aquatic environment and woodland savan- cette dynamique de la végétation. La présente com- nah. Two thousand years later, results from Asa Koma munication vise à évaluer ce rôle humain sur la show that aridification had already strongly changed végétation, a la lumière de nouvelles données the landscape with the presence of desertic fauna archéologiques au Benin et surtout au Sud-Benin. and the disappearance of aquatic large species. En effet, de nombreux sites surtout Moreover, this site is the first one in the Horn of archéometallurgiques ont été découverts et en partie Africa where domestic animal bones have been found étudies ces dernières années. Les résultats des and their diffusion could be linked with those cli- recherches archéologiques obtenus dans la sous- matic changes. In conclusion, the Horn of Africa, région ouest-africaine seront également mis à contri- because of its geographical and environmental di- bution dans cette analyse. versity, has been affected in many different ways by the climatic changes during the Holocene. Human 4.9 Katharina Neumann1, Ahmed communities living in these regions had to adapt to 2 3 those variations partly by a strong mobility to find Fahmy , Laurent Lespez & Aziz 4 resources necessary for their survival but also by Ballouche . The Early Holocene the adoption of new ways to exploit animals, such as grasslands of Ravin de la Mouche husbandry. (Mali): pottery production and its palaeoenvironmental background. (1. Goethe-Universität, Frankfurt, Ger- many; 2. Helwan University, Cairo, Egypt; 3 Université de Caen, France; 4. Université d’Angers, France). At the site of Ravin de la Mouche (Mali) early Holocene ceramics have been found that are among the oldest in Africa. We applied a multi-proxy ap- proach to the sediments of Ravin de la Mouche to reconstruct the palaeoenvironmental conditions un-

79 NYAME AKUMA No. 70 DECEMBER 2008

der which the important innovation of pottery pro- southwestern Arabia. Arguments for African compo- duction developed. Phytoliths, pollen, charcoal, nents have often been based on affinities with palynofacies and soil micromorphology give evi- neighboring regions of the continent rather than on dence of open grassland mainly consisting of annu- strictly local features. Recent research has revealed als and only rarely affected by fire. Among the evidence that implies more strictly local parallels in phytoliths, the grass-subfamily Panicoideae is well earlier contexts. It also appears that elements indicat- represented. Numerous wild grasses with edible ing overseas connections were a highly variable oc- grains belong to the Panicoideae. We suggest that currence. The paper concludes by suggesting ways the early Holocene expansion of panicoid grasses in which cultural-stratigraphic nomenclature might was a triggering factor for the invention of ceramics. more clearly reflect these new interpretations. As grass grains must be heated to facilitate diges- tion, cooking with pots was a most successful adap- tation of human populations for the effective exploi- 5.2 Peter R. Schmidt, University of tation of the vast early Holocene West African Florida, USA. The Ancient Ona Culture grasslands. of Eritrea: ritual and subsistence - a way of de-homogenizing the Pre-Aksumite? Research in Eritrea between 1998 and 2003 5 Pre-Aksumite Ethiopia and Eritrea. documented hundreds of ancient settlements around the margins of Asmara, Eritrea, dating to the early Session Chairs: David W. Phillipson and and mid-l st millennium BCE and ranging from single Peter Schmidt. family dwelling to small and large hamlets, small and large villages, and small towns. Such sites, because 5.1 David W. Phillipson, Museum of of their chronological attribution, would convention- Archaeology and Anthropology Cam- ally be identified as Pre-Aksumite. (Pre-Aksumite is a bridge, Great Britain. The last millen- term that some find awkward for the very reason that people identified with these places would have had nium BC in the highlands of northern no way of seeing themselves as living in an era be- Ethiopia and south-central Eritrea: the fore Aksum came to its florescence.) We refer to these need for revised terminology. settlements and ritual and multiple use locales as Ancient ana sites, using a local Tigrinya term for During the past fifty years, varied classifica- ‘ruin’ and modifying it to remove any connotation of tions have been applied to the societies of the last relatively recent ruins. Beyond these classificatory millennium BC that have been revealed by archaeo- issues, our findings around Asmara for the first time logical research in the highlands of northern Ethio- testify to considerable subsistence, economic, and pia and south-central Eritrea. Almost invariably the ideological variation within a region of 10 x 15 km. terms have not followed the standard archaeological These variations are linked to different kinds of ter- practice of being taken from a site where the entity rain, soil, water, and otherprecious resources such as concerned was first recognized or most comprehen- gold. Archaeological evidence indicates that beneath sively described. Instead, they have reflected an ill- the homogenizing label of ‘Pre-Aksumite’ or ‘Ancient defined relationship with the Aksumite civilization of ana’ there is a rich variability that testifies to different the first millennium AD. Furthermore, different us- subsistence strategies in the well-watered, open ba- ages are now developing on either side of the mod- sin to the west of Asmara (wheat, barley, flax; cattle) ern frontier between Eritrea and Ethiopia. The paper compared to the uplands north and east of Asmara compares the various terminological schemes that (lentils, flax, tef; goats/sheep). Furthermore, those have been proposed and argues that none adequately communities situated in close proximity to the gold reflects the levels of diversity now recognized within workings north of Asmara seem to have experienced the region’s archaeological record, nor the interpre- rapid growth and access to a wider variety of faunal tations which that record supports. In some areas, and agricultural products - possibly through intra- undue emphasis has been placed on connections regional exchange. Of particular interest are the dif- with overseas regions largely, but not exclusively, in ferent patterns of ritual life, with a major ritual center

80 NYAME AKUMA No. 70 DECEMBER 2008

located at the Sembel Kushet site, west of Asmara, nium AD). The site, though known since the early and radiating, lesser ritual locales within a range of 19th century, was more extensively excavated in 1960, approximately eight kilometers. For the first time, then, 1971, 1972 and 1973 by the Institute of Ethiopian Ar- we can point to distinctive ways of life within a small, chaeology and the French Archaeological Unit in tightly bounded region, distinctions that sustain Ethiopia, under the direction of Francis Anfray. On possible ritual/political differentiation but no clear the basis of these excavations in the 1960s Anfray social hierarchy until the 4th century BCE. divided the culture historical sequence of central Eritrea and Tigray into three main periods: Pre- Aksumite or ‘Ethio-Sabean’ Period (mid-1st millennium 5.3 Matthew C. Curtis, University of BC) when a South Arabian influence was stronger, California Santa Barbara, USA. Relat- Intermediate Period (late 1st millennium BC - early 1st ing the Ancient Ona Culture to the wider millennium AD) when local cultural traditions emerged northern Horn: discerning patterns and again, and Aksumite Period (1st millennium AD) when problems in the archaeology of the 1st a new ‘Aksumite’ culture consolidated. In 1971 and millennium BC. 1972 the author participated in the fieldwork at Yeha and was charged with studying the ceramics. The This paper argues that the Ancient Ona Cul- results of that work were published in 1980 and re- ture area of the central highlands of Eritrea, while flected the state of art at the time. The pottery se- distinct in a number of ways, represents an important quence was divided into three phases: Pre-Aksumite and early regional manifestation of a wider macro- 1, when the ceramics of Yeha were different from those regional culture often called ‘Pre-Aksumite’. Ar- from Matara; Pre-Aksumite 2, when the same types chaeological material culture, subsistence, and set- of ceramics were found at Yeha and Matara, and Pre- tlement data from the Greater Asmara area are as- Aksumite 3, when the ceramics from these sites were sessed in relation to wider data for the ‘Pre-Aksumite’ again different. These remarks were later used to out- period of the early to mid 1st millennium BC in the line the ‘Pre-Aksumite Culture’ as a distinctive ar- attempt to discern how the Ancient Ona Culture area chaeological culture, characterized by South Arabian relates to other coeval localized regional culture elements emphasizing the status of the elite and local groups, and how it might provide important insights elements typical of the rest of the population. Today, for exploring ancient social organization and com- recent fieldwork at Aksum, Shire, Gulo Makeda plexity. Suggestions for new terminology concern- (Tigray) and the Greater Asmara region (Eritrea) points ing the archaeology of the 1st millennium BC northern to a much more articulated (and complicated) picture. Horn are presented as well. This evidence suggests that a PreAksumite culture actually did not exist and the South Arabian elements (in turn, transformed into local cultural symbolic fea- 5.4 Rodolfo Fattovich, Universita degli tures as Andrea Manzo will demonstrate in his paper Studi di Napoli “L’Orientale”, Italy. at this conference) characterized a few sites scat- Reconsidering Yeha, Tigray (Ethiopia), tered in a mosaic of different (archaeological) cul- ca. 700-400 BC. tures in the region in the 15th millennium BC, and thus may support the interpretation that Daamat was a Yeha (Tigray) is the most impressive site with federative state as the Sabean title ‘mukarib’ of the evidence of a South Arabian influence dating to the last kings strongly suggests. In this paper this hy- st 1 millennium BC in the northern Horn of Africa pothesis will be tested on an updated revision of the (Eritrea, Northern Ethiopia). Since the late 19th cen- ceramics at Yeha and their comparison with those tury, this site was regarded as the best evidence of a from Matara and other sites of the 15th millennium BC Sabean colonization of the region in early historical in Tigray and Eritrea. time. In particular, the occurrence of the ruins of a temple in South Arabian style and an elite cemetery with rich shaft tombs suggested that this site was the capital city of a state (Daamat) preceding the King- dom of Aksum (late 1st millennium BC - late 1st millen-

81 NYAME AKUMA No. 70 DECEMBER 2008

5.5 Laurel Phillipson, Museum of 5.6 Andrea Manzo, Universita degli Studi Archaeology and Anthropology di Napoli “L’Orientale”, Italy. Capra Cambridge, Great Britain. Lithic nubiana in Berbere sauce? For a artifacts as a source of cultural, social constructionist approach to and economic information: the evidence Pre-Aksumite art. from Aksum, Ethiopia. Several sites dating to the first half of the 1st A decade of research into the lithic industries millennium BC in Tigray and Akele Guzai are charac- of Aksum and its near hinterland has progressed from terized by the occurrence of South Arabian elements, the straightforward recovery and recording of exca- mainly evident in monumental architecture, sculp- vated and surface-collected artifacts to the demon- ture and inscriptions. The indisputable presence of stration that these were a significant, varied compo- South Arabian elements was first regarded as the nent of all periods of Aksumite material culture, and proof of a Sabean colonization of the region in early to an understanding of their economic and social historical time which could also explain the origin of significance within that culture. It is now possible to the Semitic languages spoken in Ethiopia and Eritrea. demonstrate chronological changes in the forms of In the 1970s new investigations conducted by some lithic tools and in the composition of excavated Fr. Anfray and R. Fattovich for archaeology and by assemblages and surface collections, in response to A.I. Drewes and R. Schneider for epigraphy showed the developing needs of Aksum’s monetized, urban that the South Arabian elements had to be regarded economy. It is also possible to identify the uses to as markers of elite status. In the meantime the pres- which some of the lithic tools were applied and the ence of elements representative of the local culture social patterns involved in their production and use. of the rest of the population was stressed. In this Some examples of this will be illustrated. It is also paper some artworks going back to this phase will be now apparent that two distinct lithic traditions were reconsidered. It will be suggested that they are char- present during the Pre-Aksumite period. Of these, acterized by a fully aware use of exogenous elements the dominant industry at Aksum is morphologically of different origins (not just Sabean, not just South and technologically related to and derived from ear- Arabian) assembled in a local and original syntax lier autochthonous aceramic and ceramic Mode 5 in- and expressing local social and symbolic messages. dustries. This microlithic Pre-Aksumite industry de- Finally, a new assessment of the art of this phase will veloped in an unbroken sequence throughout the be attempted in the light of a constructionist time from the Pre-Aksumite to its demise towards the identitarian approach, and a tentative explanation for end of the Late Aksumite or during the Post-Aksumite its rise and development will be proposed. period. No such continuity is demonstrable for a sur- prisingly archaic-looking macrolithic Pre-Aksumite 5.7 Catherine d’Andrea, Simon Fraser industry which comprises the only knapped lithics recovered from Hwalti, south of Aksum, and from a University, Canada. Pre-Aksumite and few minor occurrences closer to Aksum. The pres- Aksumite settlement of Gulo Makeda, ence in the same area of two very different Pre- eastern Tigray, Ethiopia. Aksumite lithic industries, one of which appears to Systematic archaeological survey and excava- have been an entirely local development and the other tion conducted in Gulo-Makeda, northeastern Tigray, intrusive, may have a significant bearing on any con- have produced new insights into the Pre-Aksumite sideration of Pre-Aksumite settlement patterns and and Aksumite settlement history of northern Ethio- the rise of the Aksumite civilization. Examples of these pia (800 BC - AD 700). Results of settlement data, two industries will be shown. Auditors may find it ceramic, and lithic artifact analyses from Gulo-Makeda helpful to have a written summary of R. Fattovich’s indicate that the region experienced continuity in site chronology used in this talk: Pre-Aksumite (700-400 occupations through time, suggesting a degree of BC), Proto- Aksumite (c. 400-150 BC), Early Aksumite political and economic stability in contrast to the (150 BC-AD 150), Classic Aksumite (AD 150 - c. 400), Aksum- Yeha regions. Sites in Gulo-Makeda are stra- Middle Aksumite (c. AD 400 – 550), Late Aksumite tegically located along historically known trade (AD 550-700) and Post-Aksumite (after AD 700). routes in areas with moderate to high water flow

82 NYAME AKUMA No. 70 DECEMBER 2008

potential, suggesting that control of trade and high tions - consisting of two separate sets of connecting agricultural productivity were factors in the develop- rectangular tanks (or vats) excavated in protruding ment of elite groups in the region. Cultural links to natural boulders – were probably intended, on the Eritrea are evident in ceramic evidence dating to Pre- analogy of broadly comparable features in Nubia, for Aksumite and later Middle to Late Aksumite times. processing grapes and storing wine in the mid- Gulo-Makeda grew from having a culturally periph- Aksumite period. As a sign of the adoption of elite eral role in the Pre-Aksumite period to a center of tastes, that would be a further reflection of Aksum’s some importance during Classic Aksumite times, evi- opulence generated through extensive commercial denced by the presence of artifacts of trading elites. and cultural contacts - with or without a gold factor. Results of recent excavations at the site of Mezber are presented, focusing on new data generated on Pre-Aksumite rural agricultural economy. 6 The Diversity of Foraging and 5.8 John E.G. Sutton, University of Food-Producing Communities in Oxford, Great Britain. Aksum: the Holocene East Africa. Session Chair: question of gold. Mary Prendergast. The assumption that the Aksumite empire val- 6.1 Loretta Dibble, Rutgers State ued gold and exported it, among other regional prod- ucts, to Egypt and the Roman Mediterranean is not University of New Jersey, USA. Fishing new. But firm evidence is scant - notably gold coins and land use during the mid-Holocene at struck from the 3rd century AD on, occasional traces Koobi Fora, Lake Turkana. in monumental tombs, and the imprecise testimony Since 2000, investigations into land and re- of Cosmas Indicopleustes in the 6th century AD. Sig- source use by mid-Holocene and modern fishers at nificantly, the Periplus Maris Erythraei, a system- Lake Turkana, Northern Kenya, have been conducted atic gazetteer of Red Sea trade in the 1st century AD, by the Koobi Fora Research and Training Project, a does not list gold among the merchandise from Aksum collaboration of National Museums of Kenya and and its harbor, Adulis. On the other hand, Laurel Rutgers University. Various models of site location Phillipson - at SAfA 2004, followed by an article in preference by mid-Holocene fishers have been de- Azania XLI (2006) - has proposed that gold was col- veloped using analogs of prey fish behavioral ecol- lected and refined at Aksum itself, and would help ogy and ethnoarchaeological studies of fishing peo- explain the siting, by the 1st century AD, of this emerg- ples along Lake Turkana. The input variables for the ing capital city. If that should prove correct, the im- model are drawn from studies of fishing encampments plications for rethinking the history, as normally told, conducted at Lake Turkana by Diane Gifford- of both Aksum itself and its empire should be obvi- Gonzalez, and ethnoarchaeological work done by ous enough. However, the case rests partly on rumors Purity Kuira. These site preference variables are then of alluvial gold and place-name interpretations which mapped onto a reconstruction of the palaeo- appear unsubstantiated by the local mineralogy and environments of 9000 to 4000 BP. geology. Moreover, two distinct types of enigmatic ‘megalithic’ installations to the north-west of Aksum town, which (by the same argument) could have 6.2 Darla Dale, Washington University served for washing alluvial gold, appear, on practical St Louis, USA and Ceri Z. Ashley, Uni- considerations, more plausibly designed for very dif- versity College London, Great Britain. ferent functions. First, the massive rectangular slabs with rimmed edges (mistah werki), set horizontally Kansyore hunter-fisher-gatherers: new on stone mounds (with variants carved into immove- research from western Kenya. able rocks), may, conceivably, have had a religious In this paper we contribute to a better under- purpose (nevertheless, their known distribution, all standing of the timing and socio-economic activities to one side of the town, seems odd, indicating per- of the Kansyore mid-Holocene hunter-gatherers who haps some special activity there). The other installa- lived near the shores of Lake Victoria. Kansyore

83 NYAME AKUMA No. 70 DECEMBER 2008

huntergatherers are of considerable anthropological which confirm Dale et al.’s theory in several aspects. interest because they are associated with highly As Dale observed, the sites are remarkably similar to decorated and abundant ceramics prior to the intro- those located along the Nile in the Sahara-Sahel zone duction of food production (c. 5000-4000 BP) in East in terms of possible sedentism, use of ceramics, and Africa. They are also associated, at some sites, with economy. A more surprising result is evidence for the a relatively intensive lacustrine-based subsistence consumption of domestic animals within the afore- and relatively intense occupations, presenting a con- mentioned hunter-gatherer tradition. I propose that trast to socio-economic patterns associated with these foragers’ semi-sedentary, moderately delayed- other hunter-gatherers from this region. The Kansyore return lifestyle may have enabled them to more easily archaeological entity is associated with dates rang- adopt aspects of food production, while other East ing from c. 8000-3000 BP. The wide range of dates for African foragers resisted changing their economic this entity, mixing of levels at some sites, and unreli- system until relatively recently. able dating materials was called into question the timing of Kansyore. We use archaeological materi- als, especially ceramics, and direct dates obtained 6.4 Kennedy K. Mutundu, Kenyatta from our excavations at Siror and Haa - large, open University, Kenya. The diversity of sites located near rapids in western Kenya – to argue foraging and food- producing communi- for an Early and Late Kansyore chronology. We dis- ties in Holocene East Africa. An cuss migration, interaction, and exchange as possi- ethnoarchaeological ji-amework for the ble factors relating to the identity of the makers of identification and distinction of late Kansyore pottery, and Kansyore decoration proc- esses, inter-site variability, and socio-economic pat- Holocene hunter-gatherer/pastoral terns as a way to broaden our understanding of Neolithic communities in East Africa. hunter-gatherer lifeways in East Africa’s past. The advent and development of pastoral economies over the past 5000 years is one of the 6.3 Mary Prendergast, Harvard Univer- relatively well studied topics in the archaeology and sity, USA. Forager variation and transi- prehistory of East Africa. In this regard, one of the interpretive problems that have attracted the interest tions to food- production in secondary of archaeologists is the difficulty of distinguishing settings: the view from Lake Victoria. between archaeological remains of local hunter- Studies of so-called ‘complex’ hunter-gather- gatherers who had access to domestic stock, and ers have mainly focused on the Americas and Asia, those of immigrating pastoralists who practiced a with little application to Africa south of the Sahara. It subsistence strategy that may have included wild has been assumed that most foragers represented by animals. Lack of knowledge of how these two inter- the East African archaeological record would have acting socio-economic groups might be distinguished had economic and social systems similar to those of archaeologically impedes our better understanding the immediate-return foraging societies documented of prehistoric hunter-gatherer interactions with in sub-Saharan Africa over the last century. How- pastoralists and the secondary adoption of food pro- ever, a paper published in 2004 by Dale, Marshall and duction. In this paper, I use ethnoarchaeological ob- Pilgram proposed that a moderately complex, de- servations among the historic Mukogodo hunter layed-return foraging system existed in the Lake Vic- gatherers and the pastoral Maasai of Kenya to pro- toria basin some 8000-2000 years ago. This theory pose an interpretative framework that may be used to was based largely on ethnographic analogy, site size, distinguish between archaeological sequences as- density of occupation and the presence of ceramics, sociated with different subsistence economies dur- but no study had been made of those archaeological ing the ‘Pastoral Neolithic’ in East Africa or similar remains most pertinent to the economy: plant or ani- contact settings. My study shows that site features mal remains. Here I present zooarchaeological data and subtle patterns in faunal assemblages can be from several mid-Holocene sites near Lake Victoria, useful in the identification and distinction of relevant sites, and in addressing some of the interpretive dif-

84 NYAME AKUMA No. 70 DECEMBER 2008

ficulties that archaeologists have encountered in the 6.7 Paul Lane, University of York, Great study of late Holocene archaeological sites in East Britain. Pastoralist landscapes and Africa. burial practices in Samburu and Laikipia, Kenya. 6.5 Emmanuel Ndiema, Rutgers State This paper will present the results of ongoing University of New Jersey, USA, Carolyn collaborative research with Charles Hilton and Bilinda Dillian, Princeton University, USA and Straight on the changing nature of pastoralist settle- David Braun, University of Cape Town, ments, burial practices and landscapes on the South Africa. Spatial dimension of early Leroghoi and Laikipia Plateaus, Kenya. The focus of pastoralist adaptations at Koobi Fora the discussion will be on the evidence from recently (Kenya): evidence from geochemical excavated stone cairns and pastoral Neolithic and pastoral Iron Age sites in the vicinity of the village of analysis of obsidian sources and arti- Baawa, situated c.30 km south of Maralal, and how facts. this can inform current understanding of (a) the es- The initial foray by pastoralists into new tablishment of pastoral economies in this area of hunter-gatherer-fisher populations’ niches at Lake Kenya, and (b) the changing nature of pastoralist Turkana, Kenya, probably involved competition, net- conceptions of landscape and memory. working and exchange with indigenous populations. We adapt a landscape approach to investigate early 6.8 Elisabeth A. Hildebrand, Stony pastoralists in East Africa in terms of ranging pat- Brook University, USA and Steven A. terns, exchange networks among groups as an adap- tive strategy, and the associated selective advan- Brandt, University of Florida, USA. tage of pastoralism when early pastoralists made their Holocene archaeology of Highland Kafa, initial entry into this new ecological setting in the SW Ethiopia. Turkana Basin. There the Galana Boi Formation rep- Highland southwest Ethiopia is an interesting resents a rare opportunity to study the dynamics of context for exploring diversity in Holocene foraging pastoralists’ ancient lifestyles during periods of in- and food-producing economies. Cool, moist condi- creased climatic variability. Using XRF and LA-ICP- tions, steep topography, and compressed ecological MS techniques we investigate the geochemical sig- zones make for an unusual array of potential envi- natures of obsidian artifacts, with the aim of deter- ronments and resources for prehistoric foragers and mining raw material sources in relation to proximity farmers. Southwest Ethiopia has long been consid- to archaeological sites. We incorporate geomatic tech- ered a probable domestication center for several in- niques such as least-cost path analysis to model land digenous Ethiopian crops requiring moist conditions: use patterns and delineate possible landscape facets enset (Ensete ventricosum), coffee (Coffea arabica), that were utilized by competing economies. Results yams (Dioscorea spp.) and others. Later prehistoric indicate that there are three types of obsidian arti- and early historic periods saw the formation of com- facts from different sources and some yet undocu- plex societies such as the kingdoms of Kafa and Yem, mented sources. Abundance of obsidian artifacts at which depended on local crops and products, and archaeological sites indicate that there was an elabo- constituted important southern counterpoints to the rate social network that acted as a fallback during evolving Ethiopian empire based in the northern Horn times of nutritional stress. Future efforts in our on- of Africa. Despite these obvious points of theoreti- going project will help us formulate questions on cal interest, southwest Ethiopia saw little archaeo- why some sites were used whereas others were not. logical fieldwork until 2004, when the Kafa Archaeo- More importantly, we will address the question of logical Project (KAP) initiated a program of survey whether populations co-existed practicing different and excavation of caves and rockshelters in Kafa economies or whether they were populations with Zone. Survey yielded more than 20 caves, ten of which ‘three arms’. were subject to test excavation. Here we present the results of fieldwork and preliminary analysis of lithics and ceramics. Large, well-preserved faunal and

85 NYAME AKUMA No. 70 DECEMBER 2008

archaeobotanical assemblages were also recovered colonial tree-planting schemes and to early European but will require extended analysis. Data from these observations of a largely treeless landscape. Moreo- rockshelters should eventually shed light on local ver, the area’s reputation as a centre for iron produc- transitions from foraging to food production, and tion in the late precolonial period suggests at least the emergence of social complexity, and the shifting the possibility that any deforestation that occurred strength of regional connections between southwest resulted from local extraction of fuel wood, and in- Ethiopia and neighboring areas in Sudan, Kenya, the vites questions regarding the clearance of forests for Ethiopian Rift, and the northern Ethiopian highlands. agriculture, particularly since several markets and set- tlements are recorded as having developed as a di- rect result of the area’s role as a supply station for 6.9 Daryl Stump, University of York, ivory caravans during the 19th century. Drawing on Great Britain. Archaeological recent and ongoing archaeological fieldwork this perspectives on indigenous conservation paper will examine the veracity of these various his- in pre-colonial Pare, Tanzania. torical narratives, and in doing so aims to highlight the degree to which conservationist initiatives rely The landscape within the North Pare Moun- on perceptions of environmental history. tains of northeastern Tanzania includes large areas of terraced agricultural fields, hundreds of water stor- age reservoirs, an extensive network of irrigation fur- 6.10 Karega-Munene, United States rows, and pockets of woodlands protected locally International University Nairobi, Kenya. by their status as sacred groves. Together these fea- ‘Grave digging’: contestations of the tures contain a rich archive of information concern- value of archaeological research. ing the history of local resource exploitation, and have been cited as examples of ‘indigenous conser- Archaeologists working in Kenya are often vation’ on the grounds that the sacred groves repre- accused by the general public of not engaging them sent preserved remnants of extensive pre-colonial in their researches. Consequently, the general public woodlands and that the irrigation furrows and agri- tends to identifY archaeologists as ‘grave diggers’, a cultural terraces attest to the existence of effective group of educated people who are more concerned soil and water conservation measures that predate with ‘the dead’ than with ‘the living’, treasure hunt- European contact in the mid-19th century. Thus, Pare ers, or a group of people who make a living out of the reservoirs and furrows have been the focus of a se- material remains of the dead (including human skel- ries of rehabilitation programs from the 1990s on- etal remains), without adding value to the aspirations wards, first as a joint project by the Dutch and Ger- of a ‘developing’ country. This paper explores ways man development agencies SNV and GTZ and now in which archaeologists working especially on the as part of an ongoing initiative by a local agency, the Holocene can engage the public. The paper explores Traditional Irrigation and Environmental Develop- how archaeologists can use their research findings ment Organization (TIP). A recent initiative by TIP is on the varied subsistence strategies, improved tech- also promoting the virtues of agricultural terraces, nologies, environmental and climatic changes, and whilst the area’s inclusion within the eastern mitigation of environmental risk that characterized Afromontane bioversity hotspot gives added impe- the Holocene to aid better understanding of current tus to the maintenance of local traditions which act human conditions induced by environmental and cli- to limit further incursions into the remaining forested matic changes in the region. It is hoped that such an areas. There are, however, potential alternatives to understanding can, in turn, help advance sustain- these historical narratives which question the antiq- able use of natural resources. uity of the sacred groves by reference to records of

86 NYAME AKUMA No. 70 DECEMBER 2008

7 Genetics of African Domestic evidenced by tomb and rock paintings. By 4000-3500 Animals: A Symposium and Panel BP there were sheep in Eastern Africa. Rock art shows Discussion. Session Chair: Diane that these sheep were fat-tailed. Climatic changes in northern Africa are believed to have influenced the Gifford-Gonzalez. southward migration of sheep into eastern and south- ern Africa. The spread of sheep into southern Africa 7.1 Diane Gifford-Gonzalez,University took a long time because of the tsetse disease barrier. of California Santa Cruz, USA. The earliest southern African sheep remains to have Introduction to the session. been found are dating to after 2000 BP. Genetic char- acterization can be used as a tool to elucidate centers of domestication and migration routes. Initial mito- 7.2 Daniel G. Bradley, Trinity College, chondrial DNA (mtDNA) studies indicated that there Ireland. Genetic signatures of migration, were two distinct lineages. To date up to five line- selection and domestication in African ages have been identified. By exclusion of other Old cattle. World wild sheep species and partial match to feral European Mouflon, the Asian Mouflon has been pro- Whereas cattle herding in Africa is ancient, its posed as the wild progenitor of Ovis aries, the do- origins are unclear. Archaeological evidence for local mestic sheep. Further, the studies have supported domestication is inconclusive. Genetic investigation the Near East as the centre of domestication. reveals different layers of ancestral influence in local Amongst the sub-Saharan African sheep, micro sat- breeds, including Near Eastern, South Asian, and a ellite data has confirmed that there are two distinct stratum that seems particular to Africa. Mitochon- genetic groups: the fat-tailed sheep of eastern and drial DNA diversity has been used to argue for a southern Africa and the thin-tailed sheep of western separate origin from Near Eastern and European Africa. This matches the archaeological evidence for cattle. This is discussed here together with new data two separate entry points for African sheep. The re- from the autosomal genome which illustrates the an- sults also indicate that genetic exchanges are likely cestral affinities of African cattle and which points to have occurred between breeds from eastern and toward genes, for example milk proteins that may have western Africa along the Sahelian belt and that the been under selection as a consequence of domesti- southern African sheep are genetically related to the cation. eastern African sheep. Whether or not sheep also migrated from western Africa to southern Africa re- Species histories in Africa mains unclear. Our results are providing a new in- sight into the diffusion patterns of sheep pastoral- ism within the African continent. 7.3 Anne Muigai, Jomo Kenyatta University of Agriculture and Technology, Kenya, Olivier Hanotte, 7.4 K. Ann Horsburgh, Stanford International Livestock Research University, USA. The spread of Institute, Kenya and Miika Tapio, domesticated dogs across Africa: University of Oulu, Finland: Origin and contributions from ancient DNA. migration of the sheep of Africa. Our understanding of the spread of domesti- cated dogs across Africa has been hampered by the African domestic sheep originated from out- osteological similarities between domesticated dogs side the continent. They were likely first domesti- and several wild canid species from wolves and cated in the Fertile Crescent, with the earliest domes- golden jackals in the Near East to black-backed jack- tic sheep remains dated c. 9000 BC at Zavi Chemi als in southern Africa. Summarized here is our cur- Shanidar in Iraq. The earliest domestic sheep in Af- rent state of knowledge about domesticated dogs in rica are likely to have entered the continent via the Africa, and recent work dedicated to the identifica- Isthmus of Suez and/or the Southern Sinai between tion of domesticated dogs in the southern African 7700 and 7000 BP. These sheep were thin-tailed, as archaeological record.

87 NYAME AKUMA No. 70 DECEMBER 2008

7.5 Jianlin Han1, J.M. Mwacharo2, V.A. 7.6 Diane Gifford-Gonzalez, University Mobegi3, T. Amano4 & O. Hanotte3: of California Santa Cruz, USA and Origin and geographic distribution of Albano Beja-Pereira, Research Center in maternal lineages in African indigenous Biodiversity and Genetic Resources, chickens (1. Chinese Academy of Vairao / Universidade do Porto, Portugal: Agricultural Sciences (CAAS), China; The multiple African origins of domestic 2. Kenya Agricultural Research Institute, donkeys. Kenya; 3. Intemational Livestock Molecular genetics has resolved a long ar- Research Institute, Kenya; 4. Tokyo chaeological and zoological debate over the origins University of Agriculture, Japan). of the domestic donkey (Equus asinus) from wild ancestors. Because the ancestral species of the don- The origin and history of African indigenous key, Equus africanus, is found in sub-specific forms chickens remain unclear as archaeological data are across arid zones of Africa and the Arabian Penin- relatively sparse and inconclusive. Using the first sula, and because some earlydomestic forms were 397 BP long mitochondrial DNA D-loop region recovered from Dynastic Egypt, it was open to de- sequenced in 557 individuals from 37 populations bate whether the species was domesticated in Africa sampled in 12 African countries, the possible mater- or the Near East. Phylogenetic analysis of wild and nal ancestors of African chickens were determined. domestic mtDNA lineages of the species and its close 76 haplotypes are identified. Phylogenetic analysis relatives from 52 countries has identified two highly incorporating these 76 and 12 reference haplotypes divergent phylogenetic groups, the Nubian wild ass from putative centers of chicken domestication in (Equus africanus africanus) and the Somali wild ass Asia provides strong evidence for the presence of (Equus africanus somaliensis) as representing an- two major and one minor maternal lineages in African cestral lineages of the domestic ass, excluding Asi- chickens, derived from the Indian subcontinent atic wild asses as progenitors. Details of the esti- (54 haplotypes), Indonesia (16 haplotypes) and China mated dates of domestication and the dispersion of (6 haplotypes). The geographic distribution of these this highly useful and ubiquitous equid are provided. lineages varies significantly across the continent. The Eastern Africa region has all three lineages with the one from the Indian subcontinent occurring with the highest frequency followed by the one from In- donesia. The three lineages are also observed in 8 African Metal Working: Organization Southern Africa with the lineage from Indonesia hav- of Production and Diffusion Patterns ing a higher frequency. This Indonesian lineage is (Metal, Objects, Knowledge). Session absent in Western and Northern Africa, where the Chairs: Caroline Robion-Brunner and lineage from the Indian subcontinent predominates. Sebastien Perret. In all regions, the lineage from China is the least rep- resented. Partitions of genetic variation among re- gions reveal interesting geographic patterns. The lin- 8.1 Sylvain Badey, Université de Paris I, eage from the Indian subcontinent is the most di- France. Exploitation minière du cuivre verse in the eastern part of the continent, while the dans la région de Nioro-du-Sahel (Mali Indonesian lineage is the most diverse in the south- actuel) a l’époque des empires ern part of Africa. In conclusion, our results reveal médiévaux subsahariens. two major entry points for domestic chickens into the African continent: coastal Eastern Africa, a legacy Dans le commerce transsaharien à l’époque des of ancient Indian Ocean trading between Africa and empires soudanais, le cuivre est une matière première the Indian subcontinent; and the southeastern part dont l’importance - symbolique et financière - est of Africa following likely the arrival and establish- affirmée par tous les textes arabes et confirmée par ment of people of Indonesian origin in Madagascar. différentes fouilles archéologiques. On a longtemps The origin of the Chinese lineage remains unclear. considéré l’afflux de ce métal selon une seule voie: un acheminement du Maghreb (entre autres les mines

88 NYAME AKUMA No. 70 DECEMBER 2008

du Sud marocain) vers les Etats au sud du Sahara 8.2 David Killick1, Lisa Molofsky1, Simon pour y être échange contre l’or. Cependant les Hal12, Shadreck Chirikure2, Robert ressources en cuivre de l’Afrique subsaharienne ont, Heimann3, John Chesley1, Joaquin Ruiz1 elles aussi, été exploitées, notamment à Akjoujt 1 (Mauritanie actuelle), Agades et Takedda (Niger and Dana Drake Rosenstein . The actuel) ainsi que dans la région de Nioro-du-Sahel (a development of an indigenous tin and la frontière entre le Mali et la Mauritanie) et ceci, bronze industry in southern Africa during parfois, de manière concomitante a l’arrivée du métal the second millennium CE (1. University maghrébin. La région de Nioro-du-Sahel, en tant que of Arizona, USA; 2. University of Cape zone d’exploitation mini ère et de production du cuivre, Town, South Africa; 3. Oceangate est connue depuis longtemps mais son importance a été souvent sous-estimée par les historiens et les Consulting GbR, Germany). mines elles-mêmes n’ont jusqu’à lors jamais fait Bronze and tin artifacts have been recovered l’objet d’une étude dédiée. L’exploitation, au XIVème in archaeological sites from southern Africa in con- siècle, de mines de cuivre locales dans l’ouest africain texts dating from c. 1200 CE; prior to this time neither est mentionnée chez deux auteurs de langue arabe tin nor bronze appears to have been known. The only (Ibn Battutah et Al Umari) et plusieurs chercheurs (R. precolonial tin mines discovered to date in southern Mauny 1961, E.W. Herbert 1984) évoquent les mines Africa are in the Rooiberg valley in South Africa, de Nioro-du-Sahel. Malgré cela, la tendance est de where at least 2 x 105 kg of tin ore (cassiterite) are mettre en avant l’apport massif de cuivre maghrébin thought to have been mined. In this paper we (1) comme source d’approvisionnement des empires summarize our recent work on the technology of tin ‘médiévaux’ subsahariens. En fait ce déséquilibre est production at Rooiberg from c. 1650 CE to 1850 CE, du au manque d’information lie essentiellement a and (2) demonstrate by lead isotope analysis that l’absence de travaux archéologiques spécifiques Rooiberg tin, and bronze made from it, was widely concernant les sites cuprifères de la région. Nos traded within southern Africa. travaux en cours nous ont permis, pour la région de Nioro-du-Sahel, d’établir un inventaire quasi exhaustif des sites d’extraction du minerai de cuivre et de les caractériser, ceci grâce a trois missions de 8.3 Thomas R. Fenn1, Abdoulaye Maga2, prospections et de sondages archéologiques entre Oumarou Ide3, David Killick1, John 2005 et 2007. Chesley1 and Joaquin Ruiz1. Technology Une première synthèse de ces travaux et des and trans-Saharan commerce: medieval résultats obtenus peut d’ores et déjà être effectuée: metals trade in the middle Sahel zone, c’est ce que nous proposons de faire dans la présente communication. sub-Saharan West Africa (1. University of Arizona, USA; 2 Université Abdou Moumouni de Niamey, Niger / ECOWAS (Economic Community of West African States), Nigeria; 3. Université Abdou Moumouni de Niamey, Niger). The factors promoting urbanization in the Sahel zone of sub-Saharan West Africa have been the source of much debate and attention in recent years. The development of trans-Saharan trade in the late first millennium AD has often been credited with this sub-Saharan socio-cultural and economic efflores- cence. Archaeological investigations in Senegal, Mali, Burkina Faso, and Niger in the past two dec- ades have helped to examine these factors and the

89 NYAME AKUMA No. 70 DECEMBER 2008

relationship to trans-Saharan commerce. However, For the purposes of the research, ten of the one of the underlying problems in these investiga- Later Iron Age sites were chosen for excavation tions is trying to chronologically link the archaeo- across the three areas. Initial analyses from one site logical evidence to trans-Saharan commerce. As the have highlighted the high level of variation that can origins and early development of trans-Saharan com- exist within a very small area: two areas of smelting merce is poorly understood and studied, establish- remains separated by only twenty metres have re- ing this relationship is still somewhat problematic. vealed two very different technical approaches to This paper proposes that examining metal technol- the production of iron at the site. Both clusters of ogy and metals trade, both before and after the ad- iron production activity demonstrated the use of a vent of major trans-Saharan commerce, can contrib- copper-rich iron ore, however, at one cluster a sec- ute to our growing understanding of West African ond, manganese-rich material was being added to Iron Age urbanization. The examination of copper the smelts, either intentionally or inadvertently act- metallurgical debris and objects recovered from sites ing as a flux and resulting in an increase in technical linked to trans-Saharan commerce in Niger will pro- efficiency. vide a proxy for examining economic (and urban) de- As this research and these analyses progress, velopment in the region, as metals and metal trade a broader picture will develop as to the nature and were a major component of this regional growth and causes of such technological variations across the expansion. Evidence will be presented, which prima- kingdom, allowing for a greater understanding of the rily examines north-south connections across the mechanisms under which these iron production tech- Sahara as well as east-west connections from Niger nologies operated. across the middle Sahel zone of sub-Saharan West Africa. 8.5 James Ameje, National Commission 8.4 Louise Iles, University College for Museums and Monuments, Nigeria. London, Great Britain. The iron Traditional iron working in parts of the industries of Bunyoro- Kitara: recent ‘Nok Culture’ area: notes and posers archaeometallurgical research in western from preliminary investigations. Uganda. This paper results from some still preliminary attempts in the study of early traditional iron work- Archaeological survey conducted in the latter ing in the ‘Nok Culture’ area of the north-central re- half of 2007 across three areas of western and south- gion of Nigeria. Research findings so far have pro- ern Uganda has revealed plentiful evidence for iron voked observations relating to the origins of iron production relating to the pre-colonial kingdom of working in the Nok area, as well as the relationship of Bunyoro-Kitara. This paper will discuss the findings the present inhabitants of the area with the iron-work- from the subsequent excavations and the preliminary ing remains found there. The possible connection of archaeometallurgical analyses, which have enabled the Nok iron metallurgy with that of other established aspects of those technologies to be reconstructed, iron-working traditions around it (notably lower revealing more about how these industries operated Benue valley, Anambra valley, and Daima) has been within their social, political and physical environ- examined, and it has been observed that there is as ments. Extensive pragmatic yet systematic survey yet no firm basis for linking them up. It has also been was undertaken across three zones within the extent tentatively held that, although the present inhabit- of influence of the former kingdom: Mwenge, Kooki ants of the area have some relationship with these and Masindi. In total, over 200 new sites were re- iron-smelting remains, they are unlikely to be ulti- corded during this survey, mostly relating to the Late mately responsible for them. Iron Age (LIA), but also including one Stone Age Site in the Kooki area and several Early Iron Age sites in Mwenge. Many sites consisted of large con- centrations of slag blocks and other smelting debris, frequently also containing the preserved remains of furnace bases, whereas other sites were comprised of mining pits for the procurement of iron ore.

90 NYAME AKUMA No. 70 DECEMBER 2008

8.6 Pamela Eze-Uzomaka, University of increased, are intrinsic questions to be answered if a Nigeria, Nigeria. New dates for iron true appreciation of sociotechnological histories is smelting in Lejja, Nigeria. to be aimed for. This paper presents the results of a large-scale project concerned with iron production Lejja has an archaeological heritage of iron in Southern Rwanda, combined with a review of pre- metallurgy that is very important to the understand- vious work. Based on the documentation of over 50 ing of the Iron Age in Africa. Many iron-smelting newly located iron production sites, and six inten- furnaces have been located. There are different types sively investigated smelting locations, and combin- of furnaces in Lejja. They range from large furnaces, ing theoretical, ethnographic and materials-science small furnaces and slag pits still containing cylindri- based approaches, issues such as adoption, adapta- cal slag blocks. The copious amount of slag blocks tion, expansion, and impact of prehistoric iron pro- stacked in this area revealed the extent of iron smelt- duction in Rwanda are beginning to be addressed. ing in Lejja. Excavations carried out in Lejja revealed Rwanda possesses a long and well-documented his- that the slag blocks, which have been stacked in their tory of iron production, beginning with the earliest various village squares, were residues of an iron- ‘decorated brick’ furnaces that have ensured smelting culture that may have spanned thousands Rwanda’s place within the archaeometallurgical his- of years. Many of the ethnographic recollections tory of Africa. Associated with the still elusive Urewe show that iron smelting is entrenched in myths, leg- pottery-using societies, very little is known of the ends and ways of life. Pottery collected shows simi- groups organizing themselves to spend significant larities with Early Iron Age pottery from other sites. time on the resource-thirsty job of producing iron. Radiocarbon dates show that iron metallurgy in this Through the examination of an early smelting site, area may have commenced before some of the dates alongside the metallographic analyses of iron ob- which had been published earlier in this region. jects found in an Urewe burial, and incorporating this information with the growing corpus of knowl- edge regarding this period in Great Lakes History, 8.7 Jane Humphris, University College further light will be shed on early technological tradi- London, Great Britain. Rwanda made tions. However, this initial phase of production is not iron - iron made Rwanda? The role of the only momentous period of Rwanda’s metallurgi- iron in pre-colonial Rwanda. cal past. From around the 16th century AD, with the gradual move towards centralised power, the Across the Great Lakes region it would seem Nyiginya Kingdombegan to emerge. As the kingdom that iron played a central role throughout the last grew in size and power, the demand for iron increased two thousand years, becoming embedded within ori- dramatically. Tools for land clearance and field main- gin myths, kingship, and ultimately becoming essen- tenance, and weapons for the continually busy and tial to the fertility and survival of many societies. vast armies of the kingdom, were constantly needed. While work is ongoing and the story far from com- Until now little was known of how the regional iron plete, a general understanding of the evolution of production technologies developed and adapted the wider area on social, economic, political and cul- during the evolution of one of the most successful tural levels has been tentatively reached. The devel- polities of the wider Great Lakes region, or how a opment of agriculture and pastoralism, the gradual continual iron supply was ensured throughout the transformation of societies into kingdom and state following centuries. The results presented are pro- organizations, the role of healers and belief systems, viding the first glimpse into the long-term techno- are all issues that have been accessed. The resulting logical history of pre-colonial Rwanda, and highlight- picture provides the context for exploring the role ing the remarkable recent metallurgical past of this and place of iron within this complex history. How or region. why iron was produced, the effects of this produc- tion on society and on the industry itself as demand

91 NYAME AKUMA No. 70 DECEMBER 2008

8.8 Silja V. Meyer and Manfred K.H. 8.10 Nicholas David, University of Eggert, Eberhard Karls Universität Calgary, Canada. Ricardo and von Tubingen, Germany. Iron production in Liebig in the Mandara Mountains: southern Cameroon: new evidence in iron, comparative advantage, and comparative perspective. specialization. Since 2004 the German Research Foundation A reconstruction of iron production, distribu- (Deutsche Forschungsgemeinschaft) has been fund- tion and consumption amongst casted groups in the ing a project entitled ‘Environmental and cultural northern Mandara Mountains during the 1930s and change in West and Central Africa’, in which archae- 1940s is considered in the light of David Ricardo’s ologists, archaeobotanists and geographers from the Law of Comparative Advantage and Justus von Universities of Frankfurt and Tiibingen participate. Liebig’s Law of the Minimum. Ricardo suggests that During the last field season from early January to it is not specialization per se but rather its absence early March 2008 the southern Cameroon team of the that requires explanation. This leads to likely expla- University of Tubingen detected, inter alia, four cir- nations for differential production of commodities in cular structures of burnt clay at the village of Nkpwala- the region. However, it was a comparative advantage Esse in the vicinity of Sangmelima. Two of these shal- in marketing that led to Sukur’s preeminence as a low bowl structures, measuring about one meter in producer of smelted metal. Ricardo’s law cannot ac- diameter, were excavated. They were obviously re- count for the nature of specialization in this region. lated to iron production. Similar structures had been Instead it is argued following von Liebig that time, excavated in the 1980s in the inner Congo basin by particularly the window of time available for farmers an archaeological team also headed by M. K. H. to weed, was the ‘deficient nutrient’, and that this Eggert. This paper will present a brief account of the favored the delegation of a variety of specialist tasks newly excavated Cameroonian features, which will to a smith-potter caste. Finally, the collapse of the then be discussed in the light of the Congolese struc- ancient smelting industry is explained in terms of the tures and some ethnohistoric and ethnographic evi- minimal scale of iron consumption which facilitated dence. replacement by byproducts, scrap, the debris of ‘mod- ernization’ as the opportunity costs of iron produc- tion became too great, and, in a developing cash 8.9 Olivier Langlois, CNRS (Centre economy, people turned to groundnut cultivation and National de la Recherche Scientifique), migrant labor. France and Otto Thierry, Laboratoire d’écologie fonctionnelle, France. An iron 8.11 Shadreck Chirikure, University of smelting observed near Molkwo (eastern Cape Town, South Africa and Thilo piedmont of the Mandara Mountains, Rehren, University College London, northern Cameroon) in the late 1980s: Great Britain. Traversing the liminal: from the building to the use of the fur- versatility and variability in African iron nace. smelting practices. In this paper we will present the main steps of Since the seminal work of Cline (1937), the a furnace construction and a smelting session that study of indigenous iron metallurgy has gradually we observed in the late 1980s near the massif of developed into a lush field for studies of pre-colonial Molkwo, i.e. in an area that has been one of the main technology in sub-Saharan Africa. Over the years, iron-production centers in the Mandara Mountains. research themes bordering on the origins, process technology and socio-cultural aspects of iron pro- duction and use have been repeatedly engaged with. For all their diversity, existing studies of iron produc- tion seem to converge around the fact that the tech- nology employed in much of sub-Saharan Africa was

92 NYAME AKUMA No. 70 DECEMBER 2008

invariably the bloomery process. This blanket term, ‘castes’ in West Africa hold a capital place in current however, masks some fundamental differences in as research. The issues linked to the technical evolu- far as different groups and regions utilized different tion and the transmission of knowledge and social smelting recipes and achieved significantly different change are important. Indeed, even if the current situ- results. Curiously however, studies of iron metallurgy ation is rather well understood, little is known about in Africa have somewhat avoided delineating some the social and historical reasons of craftsmen caste of the factors that promoted product and process formation, and the reasons for the presence or ab- variation across space and time. Using case studies sence of such systems in a society. Moreover, we do drawn from widely separated regions of sub-Saharan not know whether the emergence of castes led to Africa, this contribution seeks to understand the major technical changes and if these can be evi- variability in African bloomery process and enumer- denced by the study of archaeological remains. In ate the decisions and constraints that promoted or the Dogon country, three castes have been distin- stifled it. Particular attention will be paid to the na- guished: the Jeme-irin on the Bandiagara plateau; ture of raw materials and their availability, system- the Jeme-na on the Seno plain; and the Jeme-yelin on driven factors, local innovations, organization of pro- the southern plateau. From the oral traditions, I was duction and the cultural and social contexts in which able to specify, in my PhD thesis, the historical, geo- iron was produced and used. It is hoped that such an graphical and social diversity of the Jeme-irin, and approach can generate fresh perspectives on the define the identity of ironworkers. The study of their subject of iron metallurgy. settlements allows establishing a certain correlation between the surnames and the Dogon languages. Indeed, the distribution of patronymics appears to 8.12 Eileen Kose and Marc Seifert, be a good indicator to distinguish the linguistic zones Universität zu Koln, Germany. Your belly where blacksmiths arrived and/or where they acquired is my smithy - cooperative field research their status of specialized craftsmen. The reconstitu- to fathom African metallurgy of the tion of migration patterns based on patronymics recent past. shows that the Jeme-irin possess their own history, though depending on the farmers’ one, character- The main focus of ACACIA (Arid Climate, Ad- ised by a non-monolithic process of formation. They aptation and Cultural Innovation in Africa), a 12-year belong to a non-hermetic caste whose marital rules multidisciplinary research project at the University change according to economic, political and social of Cologne, was laid on geographical movements and circumstances. This caste constituted itself in the human-environment relations in African arid climate Dogon country from indigenous populations, suc- zones during the past 12,000 years. Based on a case cessively joined by foreigners originating from the study about the reintroduction of iron metallurgy interior Niger. Originally, all of these peoples did not along the Central Kavango in northern Namibia dur- belong to blacksmith castes. Native farmers, Dogon, ing the Late Iron Age period, we will present an initial foreigners, and slaves became iron specialists to fulfill attempt at a multidisciplinary conception of environ- economic needs. These transformations were volun- mental adaptation that brought along cultural loss tary or forced by the Dogon. and innovation in metallurgy. The disciplines in- volved are archaeology, oral history, cultural anthro- pology, and comparative motif research of folktales. 8.14 Elisée Coulibaly, Université de Paris I, France. La transmission des savoirs et savoir-faire de la sidérurgie 8.13 Caroline Robion-Brunner, directe en Afrique Occidentale: Le cas Université de Geneve, Switzerland. The des métallurgistes du Bwamu (Burkina Jeme-irin, an example of the develop- Faso, Mali). ment of blacksmith clan (Dogon area, Mali): social status and iron working. La question de la transmission des savoirs et des savoir-faire en matière de recherche sur l’art de la The discussions about the specialization in métallurgie directe du fer en Afrique reste un sujet iron metallurgy and the formation of blacksmith encore très peu étudiée. En effet, si quelques auteurs

93 NYAME AKUMA No. 70 DECEMBER 2008

comme Amadou Hampate Ba, ou Bruno Martinelli les derniers détenteurs de ces savoirs et savoir-faire ont déjà aborde quelques aspects de la transmission sont en voie de disparition. des techniques anciennes liées au travail du fer dans leurs travaux, force est de reconnaitre que cette ques- tion a retenu très peu l’attention des chercheurs en 8.15 Sebastien Perret, Université de paléometallurgie africaine. C’est ce constat qui justifie Fribourg, Switzerland, Caroline l’intérêt du thème propose. Grace aux données Robion-Brunner, Université de Geneve, nouvelles actuellement disponibles sur l’ histoire de Switzerland and Vincent Serneels, la métallurgie du fer en Afrique, on sait aujourd’hui Université de Fribourg, Switzerland. The que les anciens métallurgistes ont su mettre au point, socio-economic organisation of the iron au cours des siècles passes, des savoir-faire fort élabores sur l’ensemble de la chaine opératoire, allant production: examples from the Dogon de la prospection du minerai, a la transformation du area (Mali). métal en objet fini a la forge. Ainsi les études sur les In traditional societies, iron is a key material, procèdes de la forge, même si ceux-ci sont également improving agricultural productivity and warfare effi- peu développées, elles ont pour intérêt principal ciency. It is a valuable trading good that enters the d’apporter des informations précieuses sur les tech- exchange networks. Control over the iron produc- niques mises en œuvre dans la fabrication des objets, tion can increase the influence and power of a group en l’occurrence la maitrise des techniques de of people. Overall, one has to consider iron metal- soudures, d’aciérage ou de nitro-carburation du fer. lurgy as an important structuring element for the so- Les analyses physico-chimiques pratiquées sur les ciety, having a strong economic and cultural impact. archéo-matériaux et les objets archéologiques ou In the Dogon area, the blacksmiths are organized in ethno-archéologiques, attestent que la mise en œuvre more or less hermetic castes and play an important de telles techniques fait appel a des connaissances role in several key aspects of the Dogon society. et des gestes techniques précis que l’artisan du fer Although their knowledge is required at most stages sait faire intervenir de manière judicieuse au cours de of the metallurgical process (particularly those de- l’opération de fabrication, dans le but d’obtenir manding high technical skills), the Dogon farmers volontairement un résultat donne. La nitrocarburation are also deeply involved in the iron production. The nécessite des procèdes précis, en matière de temps ironworkers’ status (field ownership, holders of tech- de chauffe du metal, de dosage et de composition nical and magical knowledge) and matters relating to des substances organiques qui interviennent dans control (of natural resources, workforce, products) l’opération d’acierage. Tous ces savoir-faire represent central issues as to the organization of the s’accumulent avec Ie temps pour constituer un iron production. As can be seen from the examples patrimoine immatériel, une culturel technique ‘non we will present, several very different modes of or- écrite’, qui va finalement se transmettre de génération ganization can coexist in a relatively small area, re- en génération, dans l’espace et le temps, entre les flecting specific local needs and socio-economic con- détenteurs de ces savoirs et les jeunes générations figurations. We have chosen sites with varying pro- qui doivent en hériter et perpétuer leur transmission. duction scales, ranging from small-scale iron smelt- Us se posent alors d’importantes questions ing sites oriented towards local consumption to large- auxquelles se trouve confronte le chercheur: qui était scale metallurgies with huge surpluses. The organi- détenteur des savoir-faire métallurgiques? Qui zation of the work and workforce management on pouvait transmettre ces savoir-faire? Quelles étaient these sites differs in many fundamental aspects. For les conditions et les modes de transmission de ces all steps of the whole ‘chaine opératoire’ (from the savoir-faire? Quel entait la place du secret mining and charcoal production to smelting and professionnel, mais aussi du rituel religieux autour bloom-smithing), we will show what segments of the de ces savoir-faire qui masquaient souvent les codes population are involved (farmers, blacksmith, origin) rationnels? Voici autant de questions autour to which extent, and address questions relating to desquelles nous essayerons d’organiser notre status and control. As far as possible, we will dis- ret1exion sur cette question importante de la trans- cuss the consumption and diffusion mechanisms of mission et du transfert des savoir-faire de la these productions (local consumption, export, trade). métallurgie ancienne du fer, en ayant a l’esprit que

94 NYAME AKUMA No. 70 DECEMBER 2008

8.16 Charlize C.L. Tomaselli, Arab records show that gold was being traded University of Witswatersrand, South out of the Shashe-Limpopo valley at the time of both Africa. A social analysis of metal Schroda and K2. There is, however, no archaeologi- cal evidence that gold was used at these times within distribution in the Shashi-Limpopo the society. Gold was originally seen as a means to Valley, CE 900-1300 - a preliminary wealth and later as wealth itself. The ownership of report. gold would have been highly controlled by the Mapungubwe elite. Most previous research has fo- The Shashe-Limpopo confluence is an area of cused on the technology of metal production, and archaeological research that has provided profound little on the social role of finished objects. This project insight into the change from kin-based to class-based focuses on the distribution and identification of metal society. Evidence for this complexity is found in many products, in order to understand the role different archaeological aspects, including spatial layout, set- metals played in the socio-political context of the era. tlement size, differentiated settlements, artifacts of ritual and symbolic importance and imported items from the East African trade. The rise of complex soci- 8.17 Anneleen van der Veken and Koen eties in part stems from intensive agriculture, im- Bostoen, Musée royal de I’Afrique proved technology and control of long-distance centrale, Belgium. A reconsideration of trade. A result of this transition is the specialization of activities. Where, in the past, most people did a lexically based assumptions on the early little of all the chores associated with their age and diffusion of metal working in Central gender groups, people were now focusing on dedi- Africa. cated activities. Several activities became the respon- Reflecting on the historical implications of his sibility of the commoners as the ideological system Comparative Bantu studies, Guthrie (1970) concluded of the era became more and more class-based. A way that “the speakers of the proto-language probably for the nascent elite to control the socio-political sys- knew how to forge iron before the Bantu dispersion tem was to use material objects representing sym- began”. This conclusion was of major historical im- bolic and ritual power of the elite in ceremonial portance, since it was assumed to provide lexical evi- events. Ritual symbols were also used in the con- dence for the then widely held belief among histori- struction of epic architectural landscapes. The con- ans, archaeologists and ethnologists that the diffu- cepts of power and wealth were thus expressed sion of iron metallurgy was linked to that of the Bantu through objects, through a process known as ‘mate- languages. Ever since, the history of metal working rialization’ (Earle 1997). In southern Africa, ritual ob- has continued to arouse linguistic interest, not at jects were often made from iron, copper and gold. least because direct archaeological evidence is still Their symbolic power has been misconstrued be- scarce and dates for the earliest emergence of metal cause of the ethnography applied to Central Africa. working remain indecisive, especially in Central Af- As a result, the elite was thought to be metal work- rica. More recent publications relying on lexical evi- ers. There is, however, no direct evidence to support dence threw a new linguistic light on the diffusion of this hypothesis in southern Africa. Bone shafts have metallurgical knowhow, and changed our understand- been found at Mapungubwe with traces of iron stains. ing of the role this technology played in the Bantu These have been interpreted as arrow tangs. These language dispersal. Nevertheless, many issues are tools are interesting in that they represent a blending still unexplained. Moreover, a recent linguistic study of pre-metal-using and metal-using hunting technol- by van der Veken on the transmission of a modern ogy. The use of bone for the fore-shaft may indicate metallurgical technique, i.e. aluminum casting in West that iron was a rare and valuable commodity. Ivory Africa, whose origin and diffusion are quite well objects that appear to be wrist guards, commonly known, casts some new doubts on the possibility to used among archers worldwide, were also found at draw far-reaching conclusions on the ancient his- Mapungubwe. The association of the ivory wrist tory of metallurgy from lexical data. In the present guards with the arrows would suggest that these paper, the so far lexically based assumptions on the were royal archers who were considered to be of high origin and early diffusion of iron working in Bantu- status as the use and distribution of ivory wastightly speaking Africa will be reviewed in the light of these controlled by the upper class. new insights.

95 NYAME AKUMA No. 70 DECEMBER 2008

9 In and Out of Africa: Archaeological, our research on the Hyblean Plateau territory that Palaeoanthropological, shows geological and geographical features differ- Palaeoenvironmental, Genetic and ent from the rest of Sicily. In this work we propose to start from a zero level, beginning with the Linguistic Perspectives on Late palaeoecological factors of this territory and select- Pleistocene Population Dispersals. ing the suitable areas for future research. Session Chair: Steven Brandt. 9.2 Daniel Richter1, J. Moser2, 9.1 Maria Rosa Iovino, Centro M. Nami3, J. Eiwanger4, A. Mikdad5 and Internazionale di sperimentazione, di R. Hutterer6. The Middle Palaeolithic in documentazione e di studio per la NW Africa as another area of early preistoria e l’etnografia dei popoli modern behavior. New data from the primitive, Italy and Corrado Marziano, sequence of lfri n ‘Ammar, Rif Oriental, ICAZ (International Council for Morocco. (1. Max Planck Institute for Archaeozoology), Italy. Palaeoecological Evolutionary Anthropology, Germany; diversity of southeastern Sicily during 2. Germany; 3. Centre d’lnventaire et de Neogene/Quaternary and its implications Documentation du Patrimoine (CIDP), for territory capabilities, behavior and Ministere de la Culture, Morocco; adaptation of (early?) human odysseys. 4. Deutsches Archaologisches Institut The research carried out in the Neogene/Qua- (KAAK), Germany; 5. I.N.S.A.P., ternary of Sicily produced numerous data related to Morocco; 6. Zoo1ogisches several branches such as geology, stratigraphy, Forschungsmuseum Alexander Koenig, palaeoceanography, palaeoecology, biogeography, and palaeontology. Although in the past a certain Bonn, Germany. amount of research took place, the understanding of In current archaeological research the early human colonization(s) of Sicily is still an open populations of anatomically modern humans, which question, with conjectural data ranging from Lower were responsible for the worldwide dispersal of this Pleistocene up to the Last Glacial Maximum. The two species, are often linked to the origin of modern be- main theories of early population dynamics of Sicily haviour. However, such a link is difficult, if not im- are (a) considering the arrival of human beings from possible, to prove. Additionally, the definition of the Italian peninsula through the Messina strait, and modern behavior is not, and probably cannot be, (b) coming from Africa through a Siculo-Maghrebin clearly defined. But in general items such as pierced bridge formed during regression periods. The last shells are taken as evidence of modern behavior, e.g. decades of palaeoanthropological research in Sicily in South Africa, leading to a consideration of South have been dominated by the debate on the fossil Africa as the origin of anatomically modern humans. specimen MDS-AG 2840, known as the Mandrascava If such a concept of linkage is generally accepted, skull, and other research by Bianchini in the area then NW Africa with its frequent occurrence of around the town of Agrigento situated in central pierced shells in Middle Palaeolithic / Middle Stone southern Sicily, just in front of Africa. Unfortunately Age assemblages containing tanged lithic tools has the impact of this research has been irrelevant both to be added to the picture. Here we present an analy- because it was rejected by scholars as fake and be- sis of the current chronometric information available cause in the area of the supposed discovery no sci- for the Middle Palaeolithic of NW Africa. Addition- entific research has subsequently been attempted. ally, new thermoluminescence (TL) dates for heated Human migrations depended on many parameters flint artifacts from Ifri n’ Ammar provide evidence for among which the most relevant are surely the topo- another early example of modern behaviour in the graphic, geographic, and climatic factors. The dis- form of intentionally pierced shells, together with the covery of a Homo sapiens skull, found recently by use of a colorant. Furthermore, the double alternat- the authors at Giovanna Cave, stimulated a focus of ing sequence of Middle Palaeolithic industries con-

96 NYAME AKUMA No. 70 DECEMBER 2008

taining and lacking tanged tools at Ifri n’ Ammar technological analyses call into question interpreta- shows the inapplicability of a chronoevolutionary tions of specialized point production, the notion that concept of the development of tanging. Tanged tools Nubian Type 1 and 2 and radial Levallois techniques have to be viewed as a technical phenomenon, rooted represent distinct technologies, and the existence of in a specific response of hunter-gatherer populations the Lower Nile Valley Complex. to achieve or improve certain aspects of subsistence and mobility. This can be taken as a reflection of cognitive skills related to contact and exchange of 9.4 Philip van Peer, Katholieke groups. The technique of tanging can therefore also Universiteit Leuven, Belgium. An be viewed as a part of behavioral modernity of the interpretation of Upper Palaeolithic anatomically modern humans, which are responsible origins in Northeast Africa. for Middle Palaeolithic assemblages in NW Africa as it is evidenced e.g. in Jebel Irhoud. This paper examines the late Middle Stone Age archaeological record in northern Africa from a de- mographic perspective. It is argued that population 9.3 Shannon McPherron1 Harold L. dynamics in the context of changing environmental Dibble2/3, Deborah I. Olszewski3, Utsav conditions during MIS 5 have laid out the conditions A. Schurmans3, Laurent Chiotti4 and for technological and social change in subsequent Jennifer R. Smith5. Revisiting the Nile MIS 4. One particular trajectory of change has led to the emergence of an Upper Palaeo lithic lifestyle in corridor and Out-of-Africa model: early the Lower Nile Valley. Consequently, this area is pro- modern humans and modern behavior in posed as a core area for the long term historic proc- the Egyptian High Desert (1. Max esses reflected in the MIS 3 archaeological records Planck Institute for Evolutionary of western Eurasia. Anthropology, Germany; 2. University of Pennsylvania, USA; 3. Penn Museum, 9.5 Pierre Vermeersch, Katholieke USA; 4. Museum National d’Histoire Universiteit Leuven, Belgium. ‘Out of Naturelle, France; 5. Washington Africa 2’ and Egypt – an evaluation. University St. Louis, USA). Many authors accept that AMH originated One of the major routes out of Africa for early (mainly) in Africa at the end of the Middle modern humans was along the Nile Valley corridor. Pleistocene. There is still some discussion which Previous investigations of Middle Palaeolithic set- geographical corridor was used to reach Asia. Based tlement systems focused on a small number of sites on MtDNA and on Y-DNA most authors suggest in the terraces of the Nile Valley, the desert oases, that AMH first left Africa crossing de Bab-el-Mandab. and the Red Sea Mountains. Research suggested The Nile Valley is considered as not important the presence of two groups of early modern humans because no haplotypes of the early AMH are present - the Lower Nile Valley Complex and the Nubian Com- in the actual population. The implications of such a plex. The Nubian Complex, in particular, was inter- model will be evaluated against the data provided by preted as a radiating settlement system that incorpo- research in the Egyptian Nile Valley. Very few human rated specialized point production. Recently, sys- fossils have been found in the Nile Valley and in tematic survey by the Abydos Survey for Palaeolithic Arabia. Taking in account that ‘Out of Africa’ is gen- Sites project has recorded Middle Palaeolithic arti- erally situated around 70 ka BP, the climatic and fact density, distribution, typology, and technology geomorphologic conditions in the Arabian and the across the high desert landscape west of the Nile Egyptian area during this time period will be Valley in Middle Egypt. High desert data indicate analysed. In Arabia, the absence of good compara- that the Nubian Complex associated with early mod- ble Middle Palaeolithic sites reduces the potentiali- ern humans along this route out of Africa reflects a ties of understanding the human occupation in the circulating, rather than a radiating, settlement sys- critical period. An important reduction of the human tem. Moreover, extensive lithic artifact refitting and population in the Nile Valley occurred during the later

97 NYAME AKUMA No. 70 DECEMBER 2008

Middle Palaeolithic and the Upper Palaeolithic. Along 9.7 Katerina Harvati, Max Planck the Red Sea humans are absent during OIS 3. During Institute for Evolutionary Anthropology, OIS 3, dunes from the Sahara had invaded the Egyp- Germany. Affinities of Upper tian Nile Valley, damming the valley at several points. Large intra-valley lakes were created along which Late Palaeolithic Europeans: African origin? Palaeolithic fishers have settled. The catastrophic The current modern human origins debate events (dam breaching) in the Egyptian Nile Valley at centers on the possibility and degree of admixture the start of the Tardiglacial (Bolling) resulted in the between indigenous archaic humans and modern disappearance of a human population in the Egyp- human populations migrating out of Africa into Eu- tian Nile Valley. The repopulation of the Egyptian rope and Asia in the Late Pleistocene. Evidence for Nile Valley started with the increased dryness of the such admixture must be sought in the earliest fossil Sahara around 5500 BP. Therefore, the present Egyp- record of modern humans outside Africa, as it is those tian population should not be used for DNA studies populations that would have encountered, and pos- related to ‘Out of Africa 2’. Finally one should take sibly interbred with, archaic hominins. In the case of into account the presence of a very early Upper Europe, several aspects of Upper Palaeolithic mor- Palaeolithic in the Egyptian Nile Valley. phology have been proposed as evidence of at least partial Neanderthal evidence. A 3-D geometric 9.6 Geoffrey C.P. King, Institut de morphometric analysis of the cranial morphology in available Upper Palaeolithic specimens is conducted Physique de Globe de Paris, France. The using a large comparative sample of recent human role of active tectonics and volcanism in geographic populations, Neanderthals and Middle- hominin dispersal. Late Pleistocene fossil hominins from Europe, Africa and the Levant. The aim of the study is to establish The relative environments of western Arabia the morphological affinities of Upper Palaeolithic compare with those of the Afar region of Djibouti Europeans and to evaluate its morphology for evi- and Ethiopia, providing a similar suite of tectonic dence of admixture between Neanderthals and early and volcanic conditions which extend along the Le- modern humans. Preliminary results do not support vant coastal region, the Golan region and north into the hypothesis of admixture. Rather, they highlight Georgia. It is suggested that these similarities made the relative morphological cohesion of Eurasian early the crossing from Afar to Arabia a favored dispersal modern humans and point to their cranial morphol- route from Africa both in the later Pleistocene and ogy as a generalized, ancestral modern human mor- earlier periods. The geographically restricted Galilee phology, from which later geographic groups di- region contains 65% of all reported Palaeolithic sites verged. in Israel and includes the most important. It is pro- posed that tectonic activity on the Dead Sea and Haifa faults together with active volcanism has played 9.8 Nick A. Drake, King’s College a major role in making this a favored habitat that has London, Great Britain. Was the Sahara influenced patterns of dispersal throughout human a barrier out of Africa? Biogeographic history. In this paper we will examine the role of tectonics and volcanism both within and outside and palaeoclimate evidence. Africa, drawing on a range of spatial and temporal The Sahara desert has long been thought of as examples. While acknowledging the effect of tecton- a barrier to the movement of hominins and animals ics and volcanism in promoting the preservation and out of Africa, yet it is clear that hominins managed to visibility of fossil sites, and their occasional destruc- get out a number of times during the Quaternary. To tive consequences, we emphasize their predominantly explain this apparent paradox a number of different constructive role in creating ecologically diverse and routes have been proposed such as across the Bab well-watered local landscapes that can offer a wide el Mandeb Straits or the Straits of Gibralter, along the range of foraging and hunting opportunities and Nile corridor, or around the coast. The evidence for moderate the effects of climatic oscillations, provid- the Sahara being a barrier is, however, contradictory. ing important zones of refugia and dispersal. For example, Hooghiemstra et al. (1992) show that pollen from arid taxa exist in Atlantic Ocean cores

98 NYAME AKUMA No. 70 DECEMBER 2008

throughout their entire length and from this concluded fossils and their depictions in rock art reveal a trans- that a belt of aridity always separated the humid re- Saharan distribution suggesting that these animals gion in the north from that in the south. However, roamed throughout the Sahara in the Holocene. There the biogeography of the Sahara suggests that many is evidence that people followed these waterways animals, including such water-loving creatures as with the spatial distribution of bone harpoons and fish, crossed it quite recently. Their spatial distribu- Nilo-Saharan languages closely following it. Thus as tion, having population centers both north and south the fish and other aquatic life moved across the Sa- of the Sahara with small relict populations in central hara, the fishermen appear to have followed them. regions, suggests a trans-Saharan distribution in the Our understanding of the pre-Holocene palaeoclimate past. This paper evaluates the palaeohydrology and of the Sahara will be reviewed to reveal other time biogeography of the Sahara and shows how water- periods when it might have been possible to cross loving animals and hominins could have crossed it. the Sahara. It will be shown that they correspond to A remote sensing (RS) and digital elevation model many proposed ‘out of Africa’ migrations. (DEM) analysis of the Sahara reveals numerous palaeo-channels with a high drainage density. When these deltas are located on the margin of a river’s 9.9 Roger M. Blench, Kay Williamson catchment their distributary channels can link adja- Educational Foundation, USA. The cent river systems, thus allowing aquatic life to move ‘Green Sahara’ and the dispersal of from one basin to the next. For example, an inland Nilo-Saharan fishing cultures. delta in southern Darfur links the river Nile to the Chari River, thus connecting the Nile basin and Chad Recent palaeoclimatological data for the Sa- basin and explaining the similarity in their aquatic hara made it possible to draw a long-term picture of flora and fauna. RS and DEM analysis shows the the waterways that have crossed the region. During Sahara also supported some truly giant lakes, eight the Pleistocene, the Sahara seems to have been a of which were larger than 20,000 km2, the largest be- significant barrier to both human and faunal disper- ing Lake Megachad with an area of 360,000 km2. When sals, although some traces remain of earlier ‘green’ these lakes were full, they overflowed, linking adja- periods. With the Holocene, dense new networks of cent catchments. For example, Lake Megachad waterways and lakes developed, creating a major ex- overspilled into the Benue River linking the Chad pansion of resources, both for aquatic fauna and the and Niger River basins, creating the world’s second humans that could exploit it. The evidence for this is largest river system (5,041,532 km2). With the Chad found in climatological, biogeographical, archaeologi- basin also linked to the Nile via an inland delta, this cal and linguistic sources, which testify to the dis- creates the world’s biggest interconnected waterway persal of a variety of aquatic species in the Holocene, at c. 9,000,000 km2. Further rivers, lakes and inlands evidence for which can be found in animal bones and deltas to the north make it possible to cross the Sa- rock paintings. In addition, the synchronic biogeog- hara from east to west and north to south during raphy of fish, amphibians and other species provide humid periods and always be in or close to water. pointers to this process. The Nilo-Saharan languages This explains why the fish species of the Sahara (and are found across this region today, although frag- the Nile) are so similar, forming a single biogeographic mented by the subsequent expansion of Berber. Their group. Some of these fish are cichlids that are known greatest area of diversity is in the Ethio-Sudan bor- to speciate rapidly. As they do not appear to have derlands where they may have existed as foragers for done so, it appears that the Saharan inland waterway a long period. The paper argues that their dispersal was functioning during the African humid period. It was strongly associated with these new resource will be shown that the distribution of freshwater mol- opportunities, and that a west-east movement at this luscs, some frogs and toads, and a freshwater turtle period can be tracked through finds of bone har- also suggests trans-Saharan water connections. Fur- poons and through fossils and rock paintings of thermore, by plotting the historical range of hippo- aquatic species such as hippos in what are now arid potamus, crocodiles and elephants alongside the regions. spatial distribution of reported sightings, Holocene

99 NYAME AKUMA No. 70 DECEMBER 2008

9.10 Chantal Tribolo1, N. Mercier1, H. 9.11 Geoffrey N. Bailey, University of Valladas2, J.L. Joron3, P. Guibert1, Y. York, Great Britain. The Red Sea basin Lefrais1, M. Selo4, P.-J. Texier5, J.-Ph. as habitat and dispersal corridor. 5 6 7 Rigaud , G. Porraz , C. Poggepoel , J. In this presentation I shall examine current ideas 7 5 Parkington , J.-P. Texier and A. about human dispersal across the Red Sea, particu- Lenoble8. Old age estimates for the larly at the southern end, and the likely role of the Howiesons Poort and Stillbay assem- wider region as a zone of settlement and dispersal, blages at Diepkloof rockshelter (1. taking account of the results of new work on the I’Institut de Recherche sur les Farasan Islands and the adjacent mainland of Saudi Arabia. Palaeoenvironmental reconstructions in- Archeomateriaux - Centre de Recherche formed by new data on sea-level change, new en Physique Appliquée à l’ Archéologie, palaeoclimatic data, and new evaluations of the ar- CNRS (Centre National de la Recherche chaeological record now make possible a more fo- Scientifique) - Université de Bordeaux, cused and critical discussion of alternative dispersal France; 2. Laboratoire des Sciences du hypotheses. Climat et de l’Environnement, CEA, France; 3. Laboratoire Pierre Sue, CEA, 9.12 Alison S. Brooks, George Wash- France; 4. Laboratoire d’Etude de la ington University, USA, C. Tryon, Matiere Extra-terrestre, CNRS, France; George Washington University, USA and 5. Institut de Préhistoire et Géologie du J. Yellen, National Science Foundation, Quaternaire, CNRS - Université de USA. Background to ‘Out of Africa’: Bordeaux, France; 6. CEPAM, CNRS, human behavioral responses to France; 7. University of Cape Town, demographic stress during the MSA. South Africa; 8. Département de The Middle Stone Age archaeological records Préhistoire du Museum National of both eastern and southern Africa are character- d’Histoire Naturelle – CNRS. ized by long term trends towards more specialized technologies, increased resource breadth, greater Excavations have been undertaken at imports of exotic raw materials, and use of coloring Diepkloofrockshelter (DRS; South Africa) since 1999. materials, incised decoration and other possibly sym- It is one of the very few sites where Howiesons Poort bolic artifacts. As we have argued elsewhere, these and Stillbay assemblages can be collected from the factors would have increased the potential for demo- same archeosequence. These Middle Stone Age graphic expansion, both by reducing risk and by in- techno-complexes are particularly interesting for their creasing subsistence potential and reliability. Evi- affinities with the much younger Later Stone Age dence for these trends is patchy and discontinuous, facies, and their association with evidences for sym- and the patterning suggests multiple hiatuses and bolic behavior. Establishing their chronology is there- reversals especially in the best documented records fore particularly important for the understanding of of South Africa. Here, we demonstrate that these the apparition and the evolution of the so-called ‘mod- trends are related in several eastern African case stud- ern’ behaviors. Data already available suggest ages ies, as specialized projectiles in localities lacking fine- ranging from 55 to 80 ka for the Howiesons Poort and grained raw materials are preferentially made on ex- from 70 to 80 ka for the Stillbay. The thermo- otic raw materials, suggesting the development of luminescence dating undertaken at DRS on 22 stone long-distance networks. Other specialized technolo- samples coming from the whole stratigraphic record gies such as bone points are preferentially associ- indicate 10 to 50 ka earlier intervals for these techno- ated with increased diet breadth, e.g. fishing. The complexes at DRS. Possible caveats in the dating fact that many of these trends are not found in the process are examined and rejected; the archaeologi- earliest industries of Australia suggests that the cal implications of these new results are discussed. patchiness of the MSA record is not due to a deficit

100 NYAME AKUMA No. 70 DECEMBER 2008

in human cognitive capacity but rather to situational 9.14 Jeffrey Rose, Oxford Brookes responses to demographic stress. New climate University, Great Britain: Arabian records suggest that different regions of Africa expe- refugia in the Late Pleistocene rienced non-congruent fluctuations in rainfall, which would have profoundly impacted the potential for and implications for modern human human settlement and demographic expansion. Us- expansion. ing the Smithsonian’s new online data base, we show Within the past few years, southern Arabia has that site numbers also fluctuated dramatically during been transformed from terra incognita to an emerg- the African MSA. ing and important area of Palaeolithic research. This trend is linked to an increased understanding of the 9.13 Steven A. Brandt, University of genetic record, which suggests the ‘Arabian Corri- Florida, USA, Elisabeth Hildebrand, dor’ served as a significant conduit for hominin dis- persals throughout the Late Pleistocene, coupled with Stony Brook University, USA and Erich developments in infrastructure that have facilitated Fisher, University of Florida, USA. Were logistical accessibility to previously remote areas. the Ethiopian highlands a major center Stratified sites have been unearthed; at long last, we for aggregation and dispersal of late can begin to place some Arabian lithic assemblages Quaternary hunter-gatherer within a chronological framework. This paper con- populations? siders the results from South Arabia’s first dated Palaeolithic sites within their geographic and tempo- Over the last decade, many archaeologists and ral contexts. Contrary to the traditionally envisioned geneticists have identified the ‘Southern Corridor’ scenario of a heavily trafficked corridor of human as the most likely dispersal route of anatomically migration out of Africa, the emerging picture sug- modern humans out of Africa, across the Red Sea gests a complex picture of autochthonous develop- and into Arabia at ~50-70 kya. However, little atten- ment over the course of the Late Pleistocene, with tion has been paid to the environmental and social distinct populations tethered to different refugia contexts from which these African founder throughout the region. This new understanding of populations emerged. Immediately prior to these mi- human occupation in southern Arabia contradicts grations, the hyperarid and cold conditions of MIS 4 some pre-existing models of modern human expan- made much of northern and eastern Africa uninhabit- sion out of Africa. able. We have previously posed the hypothesis that during this time, the relatively moist SW Ethiopian highlands served as a palaeoenvironmental refugium 9.15 Anthony E. Marks, Southern that attracted culturally diverse hunter-gatherer Methodist University, USA and groups from surrounding regions. In this paper, we Hans-Peter Uerpmann, Eberhard Karls further explore the implications of technological and Universität Tubingen, Germany. New social change in a refugium context. We suggest that Palaeolithic finds in southeastern Arabia contact among these culturally diverse foraging and the question of their possible East groups may have stimulated technological and so- cial innovations that were further developed as cli- African origins. matic conditions ameliorated during MIS 3. Armed This paper examines the archaeological impli- with these technological and social skills, hunter- cations for the recent genetic studies suggesting an gatherer populations would have been able to suc- eastern route out of Africa into southern Arabia by cessfully adapt to a wide range of conditions as they early moderns. It considers what archaeological cul- migrated out of SW Ethiopian highlands across and tures might have been carried by such people and out of Africa. After discussing the archaeological data what we might reasonably expect to find in southern that will be necessary to test this hypothesis, we Arabia that might confirm this model. Then, recent conclude by considering how the principles of this finds from southeastern Arabia, at Faya I, will be c hypothesis apply to parallel situations during other onsidered in this East African context. While studies time periods such as MIS 2/1. of the Arabian materials are still preliminary, they

101 NYAME AKUMA No. 70 DECEMBER 2008

suggest that the prevailing expectations may well (AAR) dating techniques have given a range of ages need re-evaluation. for Bed V from younger than the 27 ka highstand to 65 ka. Given the problems associated with 14Cdating and uncertainty concerning the uranium-series ages, 9.16 John R.F. Bower, University of the site’s chronology remains ambiguous. Here we California Davis, USA and Audax Z. P. report the preliminary results of an optically stimu- Mabulla, University of Dar es Salaam, lated luminescence (OSL) program to improve our Tanzani. Territorial exclusion in the late understanding of the chronology of Bed V at Mumba Middle Stone Age of northern Tanzania. rockshelter. OSL dating is a technique that measures the elapsed time since luminescent minerals (such as The Loiyangalani site (Heldl), a late Middle quartz and feldspar) were last exposed to sunlight. Stone Age (MSA) occurrence at the western edge of The burial age is estimated from the energy stored in the Serengeti Plain, Tanzania, has produced substan- the crystal lattice as a result of the supply of ionizing tial evidence of isolation from nearby, essentially radiation from naturally-occurring radioactive ele- contemporaneous MSA occurrences situated in the ments in the deposit. East African quartz has previ- Lake Eyasi basin to the south and at the eastern mar- ously proven troublesome for OSL dating techniques, gin of the Serengeti Plain. The evidence consists because quartz of metamorphic and igneous origin mainly of marked differences in lithic typology, espe- commonly does not possess the same favorable char- cially as regards point form, and the absence of ex- acteristics as sedimentary quartz. We have examined otic lithic material, mainly obsidian that is present in sand-sized grains of quartz from Mumba rockshelter, the nearby contemporaneous occurrences. In this and found that they too are dominated by the ‘poorly paper, we consider ecological, cultural and demo- behaving’ components of the OSL signal. These com- graphic scenarios that might explain the indications ponents are not ideal for dating, so a variety of ex- of territorial exclusion in the western Serengeti Plain periments and analytical methods have been per- during the late MSA. In addition, we consider the formed with the aim of removing these unwanted com- implications of such exclusion for demographic ex- ponents. In this presentation, we will report on the pansion, perhaps contributing to ‘Out of Africa II’, progress achieved in meeting this aim, and the prob- the spread of Homo sapiens into Eurasia and be- lems still to be overcome. The implications of this yond. revised chronology on the origins and dispersal of modern human behaviour will be discussed briefly in 9.17 Luke A. Gliganic1, Z. Jacobsl, R.G. our talk. Roberts1, M. Dominguez-Rodrigo2. OSL dating investigations of the Middle and 9.18 Amanuel Beyin, Stony Brook Later Stone Age at Mumba rockshelter, University, US. The Coastal Oasis Tanzania: progress and problems. (1. Model and new data from the Gulf of University of Wollongong, Australia; 2. Zula (Asfet), Red Sea Coast of Eritrea. Complutense University Madrid, Spain). This paper presents results of recent archaeo- Mumba rockshelter in Tanzania has one of the logical investigations in the Gulf of Zula and Buri most complete and continuous archaeological se- Peninsula, Red Sea Coast of Eritrea. The findings quences in East Africa. It extends from a MSA level include abundant Later Stone Age (LSA) sites from of about 130 ka in Bed VI to the present, including an inland and coastal landscape. Test excavations on early LSA lithic assemblage in Bed V. The timing of three sites, Asfet, Gelalo NW, and Misse East uncov- the MSA-LSA transition is of archaeological inter- ered LSA lithic artifacts and mollusk shells in close est, as the development of LSA technology is widely association suggesting human exploitation of ma- viewed as a product of modern human behaviour. rine resources. Backed tools and prismatic cores char- The top of Bed V is fairly well constrained by a beach acterize the lithic assemblages at the LSA sites. The deposit which has been correlated with other local excavated sites have produced radiometric dates lake highstands at 27 ka. Previous studies using ranging between 5000 and 8000 years BP. These dates 14C,uranium-series, and amino acid racemization coincide with the early mid-Holocene intermittent dry

102 NYAME AKUMA No. 70 DECEMBER 2008

periods. Such adverse scenarios may have triggered this intricate web of interests, and the positions taken population diversification and periodic migrations by and forced upon archaeology and archaeologists within the Horn of Africa and beyond. One possible in this context. The monumental Merowe (formerly explanation for the existence of early to mid-Holocene Hamdab) Dam at the Fourth Nile Cataract, the largest settlements along the Eritrean coast is that humans hydropower project currently under construction in were attracted to coastal habitats during the dry pe- Africa, will result not only in the flooding of a unique riods as resources in the interior of the Danakil De- cultural landscape in a 200 km stretch of the Middle pression deteriorated. Human coastal settlements Nile but also in the destruction of the livelihoods and have been widely documented for this time period identities of its about 55.000 modern inhabitants. The from various regions of Africa and Eurasia. The re- paper outlines the conditions of archaeological sal- sults of this project bring new data on Holocene vage operations within this particular context, explor- coastal settlements in the Horn of Africa. ing how the complex web of interests connected with the project - paradoxically - led to the expulsion of archaeological missions from the research area by local opinion leaders and thus to the irretrievable 10 Which Pasts for What Future? loss of cultural heritage. Political, Ethical, and Scientific Dimensions of Salvage Archaeology and 10.3 Marek Chlodnicki, Poznan Cultural Heritage Management in Africa. Archaeological Museum, Poland. Session Chairs: Cornelia Kleinitz and Hamdab Dam Rescue Project. Activity of Claudia Naser. the Poznan Archaeological Museum at the 4th Nile Cataract. 10.1 Cornelia Kleinitz and Claudia The Nile’s 4th Cataract, or at least a significant Naser, Humboldt-Universitat zu Berlin, part of it, would not find its way to archaeologists’ Germany: Which pasts for what future? interest if not for the plans to build a dam on the Nile. Introduction to the session. In this hard-to-reach area, considered archaeologi- cally unattractive, archaeological excavations were 10.2 Claudia Naser and Cornelia never conducted before. In 2003 work on the right Kleinitz, Humboldt-Universitat zu Berlin, bank of the Nile, between Es-Sadda and Gebel Gurgurib, was undertaken by a Poznan Archaeologi- Germany. The good, the bad and the cal Museum expedition. Back then this area was a ugly: a case study on political, ethical white mark on the archaeological map of Sudan. Five and scientific dimensions of salvage seasons of exploration allowed us to uncover the archaeology from northern Sudan. rich heritage of this region, settled currently by Manasir and Rubatab tribes. Archaeologists’ activ- Recent events in northern Sudan have brought ity highly influenced local interest in the past, which up a host of questions concerning political, ethical aside from positive aspects also caused negative re- and scientific dimensions of salvage archaeology actions. Protests against re-settlement politics, un- connected to major development projects in Africa dertaken by the Manasir community, made rescue and beyond. Archaeological salvage missions regu- excavations harder and harder until they were com- larly find themselves in an impasse between their pletely impossible by the end of 2007. Population scientific responsibilities, requests and restrictions waiting for relocation often deals with robbing exca- put on them by their academic institutions and fund- vations, even though the Manasir as a community ing bodies, political and economic interests of differ- protect the local heritage. Another situation occurred ent stake holders, including the administration of the in the area settled by Rubatab. This community, which development projects, government institutions, hu- will not be re-settled, often offered help and advice in man rights groups, various local communities as well finding important archaeological sites. Close interac- as (local) opinion leaders. Using a recent case study tion with the school in El Ar showed us the possibili- this paper aims at exploring the wider implications of ties of cooperation in protecting the archaeological

103 NYAME AKUMA No. 70 DECEMBER 2008

heritage of the region when archaeologists leave. 10.5 Henriette Hafsaas, Universitetet i Rock art galleries deserve special protection, since Bergen, Norway and Alexandros Tsakos, they can serve as tourist attraction but require con- Humboldt Universitat zu Berlin, stant monitoring and protection from thieves. Only the local population can save vast tumulus cemeter- Germany. Rescuing the cultural heritage ies and box graves from destruction, by planning or preserving the cultural landscape? new buildings and cropping fields in such a way as Natural environment and national not to damage monuments of the past. This hope is territory in Sudan and Ethiopia. created by educating the youngest generation. Re- sults of the archaeologists’ work should not only The wishful leap of African countries to the find their way to specialist literature, but should also modern world necessitates economic development, be made available to local communities in a more popu- which is mainly linked to the ultimate exploitation of larized form. natural resources for the sake of national progress. In the Sudan, this is primarily employed through the construction of a series of dams for the purpose of 10.4 Tim Karberg, Humboldt-Universität producing electricity and improving irrigation zu Berlin, Germany. Development and through the control of the waters of the Nile. As a the looting of antiquities: a case study. consequence, the people of northern Sudan are ex- periencing a dramatic alteration of their natural envi- The looting of antiquities is a world-wide phe- ronment through the flooding of large parts of the nomenon. Its causes and effects have been dis- Middle Nile Valley. Subsequently, this causes a mass cussed in some detail for various parts of the African destruction of the cultural heritage and landscape. If continent. Today, looting in connection with large- the salvage of the former is being secured through scale development projects in Africa could be caus- the works of the National Corporation for Antiquities ing an increasing threat to our ability to gain any and Museums (NCAM) as well as appeals to foreign knowledge on the past of areas affected by such (archaeological) missions, the latter will disappear development projects. The recent salvage campaigns despite being vital for the self-identification of the at the Fourth Nile Cataract in northern Sudan have riverine Sudanese. been accompanied by the large-scale looting and destruction of archaeological sites by local (and The Merowe Dam on the Fourth Cataract has other?) people. The few years available for archaeo- already given a clear example of the uprooting of logical fieldwork during the construction of the local communities and the loss of their customs, tra- Merowe Dam quite often saw what may be described ditions, and ways of life. On the same token, the as races between archaeologists and looters. Often, Kajbar Dam, under construction in the Third Cata- archaeological sites found intact during a survey were ract area, represents a very real threat to the continu- found completely plundered in the next year render- ity of the Mahas habitation present there at least ing any archaeological excavation obsolete. Only in since the medieval period; while the proposed dam at very few cases the National Corporation of Antiqui- the Dal Cataract might bring under water the most ties and Museums of the Sudan was able to take important archaeological sites of Soleb, Sedeinga, significant action after looting was recorded. The situ- Sesibi, and Sai. This past is not important neither for ation at the Fourth Cataract was made even more the present Islamic government of the Sudan nor for complicated by the fact that the phenomenon of loot- the multi-national enterprises responsible for the crea- ing cannot be fully understood without taking note tion of a new national identity based purely on capi- of political controversies between the local commu- talist principles. Presently, the pre-Islamic periods are nities and the authorities of the Sudan, especially systematically ignored in Sudanese schools and thus about the issue of the resettlement of the population in the consciousness of the future citizens of the in the area affected by the new dam and locals’ rights state. In neighboring Ethiopia, in contrast, the most to the ‘treasures’ of their land. This paper presents important sites of the late antique and medieval past, some examples for local and organized looting at the mainly Aksum and Lalibela, constitute an integral part Fourth Cataract, and it also outlines issues to be ad- of the national and religious identity and are con- dressed by authorities as well as archaeologists in tinuously venerated by the Ethiopians. In this sense, similar circumstances.

104 NYAME AKUMA No. 70 DECEMBER 2008

we find it revealing that these sites remain on the multinational oil companies and global financial in- margins of archaeological research - but neverthe- stitutions. In addition, the connection between CRM less in the heart of tourist interests and development and academic archaeology remains a significant ques- of related facilities. So, the bulk of this research is of tion in this area. This paper briefly describes the de- limited scope focusing mainly on Early Man and the velopment of these issues over time, as well as some (pre-) Aksumite past. It is to be noted that the field- of the possibilities and challenges that archaeolo- work is almost exclusively done by foreign missions. gists may face if such programmes become more wide- The larger number of archaeological excavations and spread in Africa. research in the Sudan can function as a support and not as a holdback for issues of rescuing and promot- ing the cultural landscape of the Middle Nile. 10.8 Nathan Schlanger, INRAP (Institut national de recherches archéologiques preventives), France. Preventive ar- 10.6 Khidir A. Ahmed, Nilein University chaeology in Africa - the call of Khartoum, Sudan. Encounter at the Nouakchott. cataracts. Preventive archaeology - planning and acting The importance of viewing salvage archaeol- ahead to avoid the bulldozers looming over the ves- ogy from a proper perspective cannot be too strongly tiges with frantic but helpless archaeologists nearby emphasized. In the Nile Valley successive salvage - actually has its origins in Africa, with the first phases operations have undoubtedly transformed our un- of the Aswan Dam, a hundred years ago. Since then, derstanding of the ancient cultures of Nubia. Con- the ‘World Heritage’ conception first applied to ex- siderations of greater agricultural and industrial pro- ceptional finds has broadened to include also the ductivity have certainly played their part in the con- more ‘ordinary’ archaeological remains, whose sig- struction of dams. In the Sudan a series of dams are nificance is otherwise important. Protection meas- being built or planned for construction at the sites of ures taken in Europe with the Malta convention of a number of Nile cataracts. These activities are being 1992 are worth heeding to, especially when the aim is carried out at exceptional times in the modern history to integrate the protection and enhancement of the of the country. It is within this context that present archaeological heritage, at national, regional and com- and prospective rescue work should be examined. munity levels, with considerations of spatial plan- ning, economic growth, tourism and sustainable de- 10.7 Scott MacEachern, Bowdoin Col- velopment on the African continent. lege, USA. Archaeology and the Chad- Cameroon oil pipeline: challenges and 10.9 Merrick Posnansky, University opportunities. of California Los Angeles, USA. Between 1999 and 2004, an extensive cultural Justifying Imperial sites: a dilemma for research management programme was developed as ex-colonialists. part of the Chad Export Project, a large-scale con- How does one justify excavating, publishing struction effort involving the development of oil and promoting sites of the Colonial and Imperial eras fields in southern Chad and the delivery of that oil to in Africa? Several rational answers can be posed. the Atlantic coast of Cameroon, via a 1070 km long oil Such sites are often substantial, important aspects pipeline. This CRM program involved efforts by of the monumental landscape. They indicate interac- Cameroonian, Chadian and foreign archaeologists, tion between overseas powers and African commu- and resulted in the discovery of over 450 archaeo- nities, and their excavation raises questions of how logical sites and the production of a great deal of processes of change took place. Some of the key new knowledge concerning Central African prehis- information on their existence is reliant on foreign tory. The gestation of this program was extremely sources and most importantly their excavation is im- complex, involving a variety of different stakeholders portant for understanding the later history of Africa. in both countries as well as their relations to both A further reason can also be proposed that they help

105 NYAME AKUMA No. 70 DECEMBER 2008

tourist promotion and many, being related to the Slave signed, perceived and evolved from the colonial pe- Trade, provide an essential link between Africa and riod to the present. I will insist on the 1990s and the world-wide African diaspora. Examples of both 2000s which are characterized by an epistemological the importance of such sites and the value of their shift that contrasts the colonial period. This will per- excavation are drawn from the West African coast mit us to show the power relations involved in the and from northern Uganda, and in particular from the production of history as well as the complexities of 19th century Egyptian fort of Dufile, excavated in 2006- the politics memory in Senegalese museums. 2007, that it is hoped to be conserved as a major sustainable historical monument and a regional in- formation center catering to the need of both Uganda 10.11 Rahim Rajan, Aluka, USA and education and tourist promotion. Heinz Ruther, University of Cape Town, South Africa. The documentation and conservation of African Cultural Herit- 10.10 Ndèye Sokhna Gueye, Université age Sites and Landscapes. Cheikh Anta Diop de Dakar, Senegal. Museum exhibitions in Senegal: Aluka (www.aluka.org), an initiative of JSTOR instrumentation of history or politics of (www.jstor.org), and Heinz Riither at the University memory? of Cape Town (South Africa) are working together to document and build a digital library about African West African museums, especially in Senegal, Cultural Heritage Sites and Landscapes. This data- remain essentially marked by their colonial and post- base is intended for the use and adoption of the glo- colonial past. The Senegalese museums are offspring bal scholarly and heritage communities. In Africa, of the colonial enterprise. This raises questions about access to both the Aluka and JSTOR databases is their legitimacy and recognition by local populations. freely available to academic, government, and re- Although sites of memory and cultural heritage, they search institutions. Thus far, heritage sites and land- were modeled according to colonial ethnographic scapes documented include: Djenne (Mali), Timbuktu views and therefore reflected colonial perceptions of (Mali), Elmina (Ghana), Asante Temples (Ghana), the Africans. Those models were challenged at the end rock-hewn churches of Lalibela (Ethiopia), Aksum of the colonial era as independences in the 1960s (Ethiopia), Lamu Archipelago (Kenya), Kilwa Kisiwani ushered a search for a new national consciousness. (Tanzania), Engaruka (Tanzania), Great Zimbabwe This was achieved through rewriting African history, (Zimbabwe), and a number of rock shelters and caves promoting national culture, values, and heroes. In- in South Africa’s Cederberg Mountains. The digital dependent Senegal searched to promote national library also includes a growing image archive of Afri- museums; unfortunately, the exhibits remained elitist can rock art from two important repositories in Africa as they continued to focus on the remote traditional - the Trust for African Rock Art in Nairobi, Kenya, past sometimes to promote nationalist or political and the Rock Art Research Institute in South Africa. views. That ideological orientation contributed to Our presentation will (1) introduce the methods and local indifference in respect to postcolonial muse- technologies deployed in the documentation of these ums. Instead of curating collective memory, muse- heritage sites and landscapes; (2) identify preserva- ums were locally viewed as the continuation of the tion and environmental challenges facing these cul- African past invented by the colonizers and later re- tural sites and landscapes; and (3) discuss the value produced by the post-colonial state in a nationalist of Aluka’s digital library and how this information and elitist manner. Significant changes are however may assist researchers, educators, conservators, and noted as, by the 1990s, museums’ specialists began site managers. to initiate public outreach programs that searched to decolonize and adapt exhibitions to local needs and demands. This outlined trajectory is here illustrated at the three major museums in Goree Island (Sen- egal): the Historical Museum, the Slaver Warehouse, and the Women’s Museum Henriette Bathily. Our objective is to describe how the museums were de-

106 NYAME AKUMA No. 70 DECEMBER 2008

11 New Thoughts on Modern Cognition years ago. An Earlier Upper Pleistocene model places in the Middle Stone Age. Session Chair: the origin at the Acheulian/MSA boundary 250,000 Lyn Wadley. years ago, and a Later Middle Pleistocene model places the origin nearer to the end of the Middle Pleistocene at OIS 6/5. Another model sees the evo- 11.1 Marlize Lombard, University of the lution of modern behavior as a gradual process de- Witwatersrand, South Africa. Direct veloping sometime in the Middle Stone Age. Ar- evidence for the use of ochre in Middle chaeological evidence forms the backbone of mod- Stone Age adhesives. els on the origins of symbolic behavior, but knowl- edge from other disciplines should constrain hypoth- This contribution provides direct evidence for eses based on archaeological remains. There is clearly the use of ochre in adhesive recipes during the a need to discuss the biological implications of mod- Stillbay, Howiesons Poort and post-Howiesons Poort ern behavior models, especially with regard to brain of South Africa. Stone points and segments ITomtwo evolution. However, evidence from brain sciences KwaZuluNatal sites were microscopically analyzed needs as careful evaluation as information from the to document ochre and resin occurrences. These mi- archaeological record. A number of principles, for cro-residues show a clear distribution pattern on the example, whether current cognitive neurological func- tool portions that are associated with hafting. Re- tioning is relevant to the evolutionary history of such sults from separate quartz and crystal quartz seg- functioning and whether the evolution of symbolic ment samples may indicate that different adhesive behavior is tied to a mutation or co-evolutionary proc- recipes were applied to different raw materials. A pos- ess must be considered. If it is assumed that sym- sible functional application for ochre in association bolic abilities are the result of a co-evolutionary proc- with Later Stone Age mastics is also explored. The ess in which behaviors drive the direction of brain evidence and suggestions presented here expand our development, then modern symbolic behavior may understanding of the versatility, use and value of have been a behavioral trait of the ancestor of pigmentatious materials in prehistory; it is not viewed Neanderthals and modern humans. There are indeed as an alternative or replacement hypothesis for its some suggestions that modern-type behaviors are possible symbolic role during the Late Pleistocene. associated with multiple species. Other models sug- gest that a neural mutation at 50,000 years ago, per- 11.2 Sarah Wurz, Iziko Museums of haps involving FOXP2 gene involved in speech and language, was important in the evolution of modern Cape Town / University of Cape Town / symbolic behaviors. These models would place less University of Stellenbosch, South Africa. emphasis on co-evolution. This paper will discuss Debating modern behavior in South these issues by integrating perspectives from musi- Africa. cology and developmental psychology with South African Middle Stone Age material cultural expres- South Africa is one of the geographical areas sions in the Upper Pleistocene. where some of the earliest evidence of what is de- fined as symbolic behavior has been discovered. Finds of engraved ochre and shell beads from 11.3 Christopher Stuart Henshilwood, Blombos Cave, engraved ochre from Klein Kliphuis, University of the Witwatersrand, South and engraved ostrich eggshell from Diepkloof are Africa /Universitetet i Bergen, Norway. surprisingly early expressions of symbolic behavior. The >100 ka levels at Blombos Cave, These finds contribute to the trend to separate dis- cussions on the evolution of modern symbolic southern Cape: early pointers to modern behavior from the European Upper Palaeolithic. The human cognition? evolution of symbolic capabilities is increasingly seen Blombos Cave is some 100 m from the coast as an earlier rather than later phenomenon. In 2003, and 35 m above sea level. The cave is a wave-cut Henshilwood and Marean for example discussed three bench in Mio/Pliocene Wankoe Formation aeolian ‘earlier modern behavior’ models which place the ori- deposits. Interior cave deposits, including those in gins of modern behavior at various times before 40,000 recesses, cover more than 80 m2. About 20 m2 of the

107 NYAME AKUMA No. 70 DECEMBER 2008

MSA has been excavated to a depth of c. 2 m below represents the source of behavioral modernity. While the original surface. Excavations carried out since a case can be made for this point of view, this paper 1991 at the site provide snapshots of life in the Mid- examines the strengths and weaknesses of this hy- dle Stone Age (MSA) in the southern Cape, South pothesis and ultimately rejects the idea of a Africa. Three phases of MSA occupation have been monocentric southern African origin in favor of a identified named Ml, M2 and M3. Dating by the opti- model for a historically contextualized, mosaic, poly- cally stimulated luminescence (OSL) and centric rise of cultural modernity. This model for thermoluminescence (TL) methods have provided Mosaic Polycentric Modernity (MPM), while not de- occupation dates for each phase. Of particular inter- nying the unique role of Africa in the later phases of est are the previously undated lower levels. OSL re- human evolution, better matches the sults just received provide a c.143 ka date for the globalarchaeological record than do models empha- lower M3 phase (MIS 5e/6) with a c.100 ka (MIS 5d) sizing strictly African origins. for the upper M3 phase. A hiatus period of c. 23 ka separates the upper M2 phase of the deposits that is associated with the Stillbay. It is possible that this 11.5 Lyn Wadley, University of the represents a hiatus at the site and closure of the cave Witwatersrand, South Africa. Compound mouth. Blombos Cave was probably occupied spo- adhesives as evidence for modern cogni- radically and for relatively short periods of time. In tion in the Middle Stone Age. the M3 phase silcrete is dominant but there are fewer retouched tools. A preliminary study of the M3 phase Plant resin, iron oxides and sometimes fat were lithics suggest they do not fall within the MSA I or used together for compound adhesives at Sibudu MSA II phases of the MSA as described for the Cave at least by 70 ka. Replication of compound ad- Klasies site. Striated ochre, particularly in large chunk hesives (using various combinations of Acacia gum, form, is common in these levels. Ochre processing iron oxide ground from naturally occurring nodules tools include lower and upper grindstones and ham- and beeswax) demonstrates the complexity of the mer stones. Dense shellfish middens characterize the process. Making reliable adhesive involves a calcu- lower layers with very large hearths. Faunal remains lated manipulation of disparate ingredients such as from this phase show that a wide range of terrestrial Acacia gum and iron oxide ground from naturally resources were exploited. Fish bones, marine shells, occurring nodules. After lithic inserts are attached to and seal and dolphin remains attest to extensive ex- their hafts with wet adhesive, the composite tools ploitation of aquatic resources. Four human teeth must be slowly and carefully dried near a fire, using were recovered that represent gracile and robust in- controlled heat. If heat is not used, the adhesive takes dividuals. Of particular interest is whether the M3 nearly a week to set; if the weapons are placed too phase reflects any signs of modern human cognitive close to heat the adhesive forms air pockets and is processes that are present in the M1 and M2 Stillbay weakened. The plasticity of plant gum, the aggre- phases. gate-size of ground iron oxides and the heat of fires are variable factors that require a slightly different recipe and processing procedure for each hafting 11.4 Nicolas Conard, Eberhard Karls event. The method is sufficiently complex that it seems Universität Tiibingen, Germany. A to provide circumstantial evidence for modern work- critical view of the evidence for a ing memory capacity (WMC). Effective glue allows southern African origin of behavioral flexible placement of lithic inserts. At Sibudu, differ- ent sizes of Howiesons Poort segments were rotated modernity. differently on their hafts - for example, transversely, The Middle Stone Age of southern Africa has longitudinally, diagonally - resulting in a variety of produced a wealth of new data that points to this end-products for disparate tasks. This act, requiring region as a center for cultural development during mental rotation of a single shape to achieve worth- the Middle and Late Pleistocene. In South Africa it while end-products (different from rotating cores has become scientifically fashionable and politically during knapping), is a complex span task of the kind expedient in recent years to emphasize these impor- that psychologists use today to measure WMC, that tant findings and to suggest that southern Africa is, the ability to control attention to keep relevant

108 NYAME AKUMA No. 70 DECEMBER 2008

representations in active memory or easily retriev- 12.2 Wazi Apoh, Binghamton University, able from inactive memory. Individual differences in USA. German colonial residues and WMC correspond to differences in circuitry associ- legacies in Ghana: archaeological insight ated with the prefrontal cortex, which is related to complex behaviors. into Kpando-Todzi site (Volta region of Ghana). In comparison with the robust archaeological study of British and French colonial sites in Africa, 12 Historical Archaeology and the assessment of the archaeology of German colo- Ethnoarchaeology. Session Chair: Diane nial rule and legacies in Africa has not received equal Lyons. attention. The intellectual void created as a result of this inattention has necessitated this study as an 12.1 Chris Ehret, University of California attempt at problematizing this practice. The archae- ology of German colonial history in the tropics needs Los Angeles, USA. Validating linguistic to be emphasized so as to generate comparable case evidence: when is it trustworthy and studies aimed at assessing commonalities and varia- when is it not? tions in intercultural entanglements and agency in colonized hinterland regions of the world. Through Over the years numerous Africanist scholars dissertation improvement supports from the National of language and culture history have presented ar- Science Foundation and the Wenner-Gren Founda- chaeologists with histories built ostensibly on lin- tion for Anthropological Research, multiple eviden- guistic evidence. But how is an archaeologist to judge tial sources were explored between June and Decem- the validity of these proposals? The fundamental re- ber 2005 to document how practices of Kpando peo- quirement in using language evidence for history is ple in Ghana were impacted by pre-colonial and Ger- that the scholar be working from a systematic histori- man colonial political economic pressures. Data gath- cal linguistic reconstruction of the language family, ered from oral histories, archival documents and eth- or the branch of a family that contains the relevant nographic information revealed that Kpando was first evidence. The systematic formulation of sound settled in the 16th and 17th centuries by a community change history in a language family constitutes the of Akan and Ewe-speaking migrant groups. These essential analytical apparatus for determining whether sources also reveal how Gennan missionaries, mer- the surface similarities between two words of like chants and colonial officials (1847/1886-1914) estab- meaning are due to chance, to borrowing, or to actual lished settlements at Kpando and other stations in common derivation from the same root word. If one the Volta region of Ghana (Gennan/British Togoland) does not undertake a rigorous historical linguistic and worked to cultivate new markets for their Euro- reconstruction first, or does not make use of an exist- pean products and Christianity. The reverberations ing reconstruction, it does not matter how sophisti- of these varied encounters led to the monetization cated one’s understanding of the possibilities and and restructuring of the local economy, which im- pitfalls of historical linguistics is. One can make edu- pacted gendered divisions of labor, led to new forms cated guesses and, often, guesses with a high prob- of specialization and indigenous reactions to new ability of being correct, but guesses nevertheless. products. Complementary data from archaeological Many linguistic-based proposals of the past five test excavations at the Kpando-Todzi site provides decades that matter for African archaeological corre- insights into the residues and materiality of these lation do meet the requirement of being founded on a political economic encounters. The findings from this systematic historical linguistic reconstruction. A great archaeological investigation will contribute to a re- many more do not. This paper systematically identi- examination of the silences on German colonialism. fies which particular linguistics-based works of the Above all, this research will enhance the establish- past fifty years attain that standard and therefore ment of a proposed district museum project at the compel serious attention from Africanist archaeolo- Kpando-Todzi site for the promotion of education, gists, and which do not. tourism and cultural resource management of Ger- man colonial relics at Kpando and the Volta Region of Ghana.

109 NYAME AKUMA No. 70 DECEMBER 2008

12.3 Philip Allsworth-Jones, University of 12.5 Diane Lyons, University of Calgary, Sheffield, Great Britain. Canada. End of an era - diversity in Ethnoarchaeology at former feudal landscapes: an IITA lbadan. ethnoarchaeological review of powerful Excavations were conducted at the Archaeo- places in eastern Tigray, Ethiopia. logical Reserve, established on the site of the fonner Domestic rural houses in eastern Tigray rarely village of Adesina Oja, in the International Institute survive for more than a century. Consequently the of Tropical Agriculture Ibadan in 1980-1986. An ac- rural built landscape that can be observed today is count regarding abandoned house no. 2 was pub- partly constituted by building practices of the last lished in Azania in 2004. This report concerns the decades of Ethiopia’s feudal state which ended in midden mound excavated at the same time. It yielded 1974 when Emperor Haile Selassie was deposed. This a wealth of information about Yoruba life in the first paper presents an ethnoarchaeological study of some part of the 20th century, some of which was summa- of the elite houses that remain in the Eastern Admin- rized in hitherto unpublished projects prepared by istrative Zone of Tigray Region. At the time of study, students of the University of Ibadan under the su- many of these houses were still occupied by the origi- pervision of the author. The objective is to ascertain nal builders or their direct descendents. In addition what light the remains of material culture excavated to their social memories, this time period has the added at a village peripheral to a great urban centre can advantage of historic documentation of political throw on Yoruba life at the time. events and feudal practices that provide a rich gen- eral context for the study. The architectural diversity 12.4 Caleb Adebayo Folorunso, Univer- and use of prominent settings provide archaeolo- gists with an extant (albeit incomplete) model of feu- sity of Ibadan, Nigeria. Updating the dal landscapes of power. cultural landscape of Old Oyo. The site of Old Oyo had been the subject of 12.6 Innocent Pikirayi, University of archaeological investigation for over half a century. Pretoria, South Africa and Anders Recent research in the past five years, including re- connaissance, mapping, and excavations on the site, Lindahl, Lunds Universitet, Sweden. is adding more information to our knowledge of the Ceramics and the Ethnographic Present: cultural landscape of the site. While some of the rel- the social context of pottery ics that were reported in the past are fast deteriorat- production and distribution in the ing, new insights are being gained into other relics Limpopo Province of South Africa. that hitherto had not been reported. We also present a comparison of our recent mapping of the Old Oyo This research is based on the study of modern walls using GPS with earlier mapping on the site. Dig- pottery makers in the Limpopo Province of South ital photography and GPS readings have also been Africa. By identifying their production techniques used for quick recording of the cultural landscape as well as distribution of the finished ceramic prod- features, which include wells that have been identi- ucts, we seek to inform the Iron Age archaeological fied in some parts of the site, grinding hollows found record, especially in the middle Limpopo valley, an on rock outcrop surfaces, and seemingly complete area associated with the rise of social complexity in pottery vessels that are found partially buried in many the early second millennium AD. The objective is to parts of the site. Recently, a portion of a potsherd provide alternative explanations on the distribution pavement was excavated outside the walling system pattern of archaeologically defined ceramic/cultural of the site, and more evidence for potsherd pave- units. ments is revealed on the site surface in the general area where the first pavement was excavated. These and many more cultural relics litter the landscape of the Old Oyo site, and this paper provides inventories using GIS to locate the relics.

110 NYAME AKUMA No. 70 DECEMBER 2008

12.7 Moustapha Sall, Université Cheikh gists with an interest in the field of ceramics. And Anta Diop de Dakar, Senegal. Culture since the questions have everything to do with the matérielle ceramique et identités en pays ways we approach past and present people’s rela- tionship to the material world, they are of particular Fogny (Gambie, Casamance). interest when seen in light of recent academic preoc- Dans la zone historique du Fogny, qui englobe cupation with the interface between ‘indigenous aujourd’hui le Sud-Ouest de la Gambie et la Casamance knowledge systems’ and scientific knowledge. This (Sud-Ouest du Senegal), plusieurs sites debate raises the important issue of what social sci- archéologiques furent découverts. Ces derniers ence is and should be when facing rationalities that constitués d’amas coquilliers et d’anciens villages are different in various aspects from Modernist think- témoignent d’une longue pratique de la cueillette des ing. Bearing these considerations in mind, this paper mollusques et des interactions entre plusieurs will follow the various stages in the ‘life histories’ of groupes culturels. Cependant, l’essentiel de ces sites clay in households, aiming to demonstrate some im- sont attribués aux seules populations Diola qui portant differences in people’s relationship to clay constitue le peuplement le plus important de cette and potting practices between two Tswana contexts sous-region. Contrairement à cette thèse, les in southern Botswana. The examples will provide the recherches ethnoarchéologiques portant sur les background for some theoretical and methodologi- pratiques de recoltes des mollusques et sur les cal reflections on how we approach the materiality of comportements socioculturels et techniques de fab- clay and ceramic technology in material culture stud- rication des poteries (les principaux artifacts trouves ies. In turn, these reflections will be related to the title sur ces sites) ont montre que d’anciennes question: what do we learn from such studies? populations Baynouk ignorées ou negligees auraient, d’une part, constitue le peuplement le plus ancien du 12.9 Jonathan R. Walz, University of Fogny, et d’autre part, legue un heritage technique à plusieurs populations qui revendiquent la paternité Florida, USA and N. Thomas Hakansson, de plusieurs sites archéologiques. Lund University, Sweden / University of Kentucky, USA. Gonja revisited: ar- chaeology and regional political ecology 12.8 Per Ditlef Fredriksen, Universitetet in the South Pare Hills and Eastern i Bergen, Norway. What do we learn Pangani Valley, Tanzania. from studying clay and pottery in house- holds? Two differing examples from Renewed interest in the long-term histories of southern Botswana. mainland northeast Tanzania has focused attention on the people of the Eastern Arc Mountains and ar- From ethnographic and ethnoarchaeological eas first informally investigated by archaeologists accounts of sub-Saharan ceramic practices we learn more than forty years ago. Many archaeological and that this form of technology has been deeply embed- historical studies of human-environment relationships ded in a framework of thought which may be referred in eastern Africa have taken for granted that long- to as a ‘thermodynamic philosophy’. Central here is term ecological processes were only the result of im- an ambivalent relationship between potting and hu- migrations and variations in population densities. man reproductivity, most often found expressed as Such studies are locked in a view that production pollution ideas and prohibitions closely associated was the sole result of local subsistence requirements with female bodily experience: menstruation, sexual and use values. This inability theoretically to con- activity and pregnancy. Potentially polluting states ceive of an exchange-based economy, together with and actions were regarded as hot while cleansing the construction of a pre-colonial/colonial duality of agents were cool. But how do these rather abstract use values versus commodities, still permeates many thought-patterns relate to our experiences during archaeological and historical analyses. Based on field field studies of material culture, when we as research- visits to the South Pare Hills and archaeological and ers are faced with the everyday of present domestic ethno-historical research in the vicinity of Gonja (since life? Do they relate to them at all? These are primary 2003), we argue to the contrary that regional and in- questions to anthropologically oriented archaeolo-

111 NYAME AKUMA No. 70 DECEMBER 2008

ternational exchange systems contributed to the de- A detailed archaeological mapping of material cul- velopment of landesque capital beginning in the early ture and features was done only for the Mbajeng site second millennium. Reconnaissance near Gonja iden- because it had more finds than the other four. Ethno- tified more than twenty new archaeological sites con- graphic studies of some present-day Wimbum mate- centrated along the skirt of the South Pares dating to rial culture also revealed aspects of continuity and as early as the later first millennium AD. Surface and change in the making and use of the material culture. excavated evidence of shell bead manufacture and, In conclusion, the paper observes that though the later, intensive iron production and agricultural ter- corroboration of the sources provided a much clearer racing accompany indications of coastwise interna- picture of a more peaceful centralized society con- tional exchange. Emergent political ecologies tied to trary to the conflict-ridden one propagated in most population shifts toward mountain fringes, the ar- colonial documents, not all the finds found at the site rival of plains-based pastoral communities, and fluc- could benefit from this corroboration. For example, tuations in political power bolstering rainmaking oral traditions were silent on the stone buildings up elites, influenced local land use and altered land- the hills while there was none in existence in the eth- scapes. Although originally reinforced by the boom- nographic present forcing us to generate exclusively ing caravan trade of the mid to late 19th century, es- archaeological interpretations based on their con- tablished political ecologies (rooted in chiefly power) text. and small-scale centralization eventually succumbed to environmental stresses and the effects of colonial demands on labor. This case study emphasizes how political and economic circumstances at particular 13 Trade and Exchange - Terrestrial and junctures of history and at different scales account Maritime. Session Chair: Sonja for long-term social change in the vicinity of Gonja. A Magnavita. historical anthropological approach, integrating a range of source material, demonstrates how regional 1 2 and world systems integrations and interactions 13.1 Sonja Magnavita , Oumarou Idé , shaped land use beneath the South Pare Hills since Abdoulaye Maga3, Carlos Magnavita1 the later first millennium AD. and Gerhard Breyl. Sahelian connec- tions: new studies on Marandet, Niger 12.10 Richard Talla Tanto, University of (1. Goethe-Universitat Frankfurt, Ger- Buea, Cameroon. Ethnoarchaeological many; 2. Université Abdou Moumouni de investigations of Mbajeng in the Niamey, Niger; 3. Universite Abdou Wimbum land of Donga Mantung Divi- Moumouni de Niamey, Niger / ECOWAS sion, Northwest Province of Cameroon. (Economic Community Of West African States), Nigeria). Most colonial literature on the culture history of the Wimbum and most parts of Cameroon was The site of Marandet in central Niger has been written from an alien perspective and the conclusions tentatively equated by some researchers with the his- arrived at were generally misrepresentations and dis- torical Maranda, obviously one of several important torted since they were not versed with the societies. stations mentioned by the first Arabic documents Based on the above shortcomings, this work sets describing an ancient trade route linking Egypt with out to employ different lines of evidence: oral tradi- the Kingdom of Ghana. Preliminary archaeological tions, ethnographic and archaeological data to see investigations involving mapping, test excavations how they may substantiate each other for a better and analysis of cultural materials recovered at the reconstruction ofthe Wimbum early culture history site were conducted by a cooperative research project using the Mbajeng archaeological site as a case study. of the Universities of Niamey, Arizona, and Frank- Oral tradition assisted in tracing the migratory his- furt. The first results suggest that Marandet was both tory of the Wimbum and also in the archaeological a producer and consumer site for goods widely dis- survey of the region, which revealed five sites - tributed in West Africa: copper alloys were produced Mbajeng, Konchep, Mbadung, Mbiribo, and Ngulu. directly on the site for export, while domestic and

112 NYAME AKUMA No. 70 DECEMBER 2008

luxury goods reached the place coming from several 13.3 Liza A. Gijanto, Syracuse Univer- different regions. Indeed, it appears that the site was sity, USA. Trade, interaction and change part of an exchange network that connected North during the Atlantic Trade on the Gambia Africa, the Sahara, the Lake Chad area, and the East- ern Niger Bend from at least the 7th to the 9th century River. AD. The Gambia River was connected to many of the major trade systems in West Africa, with direct 13.2 Sam R. Nixon, University College ties to the Saharan and Atlantic networks. Prior to the arrival of the Portuguese, Mande traders traveled London, Great Britain. New archaeo- from the inland Niger delta to the coast trading gold logical insights into the Early Islamic and other commodities for salt. In the mid-15th cen- trans-Saharan gold trade: gold coin tury, Portuguese merchants initiated commerce along moulds from the West African Sahel the river, ultimately reorienting trade patterns in the (Essouk-Tadmekka archaeological site, Senegambia from interior land-based to Atlantic mari- Republic of Mali). time commerce. This paper presents preliminary find- ings from archaeological and historical investigations This talk presents new archaeological data from at the Niumi trade center of Juffure as well as periph- the West African Sahel to provide fresh insights into eral villages of San Domingo and Lamin. The overall the development of the Early Islamic trans-Saharan focus of these investigations is to examine potential gold trade. The talk is focused by the study of a set changes in daily life due to shifts in interactions and of 9th/10th century AD gold coin mould artifacts re- commerce from the 15th to the 19th century. cently excavated from the site of the celebrated Early Islamic trans-Saharan trading town of Essouk- Tadmekka, situated in the north of Mali. The talk com- 13.4 Rachel L. Horlings, Syracuse mences by looking at the limited documentary his- University, USA. The environment of torical and archaeological data sets related to the historic maritime trade: shipwreck site Early Islamic gold trade and the over-simplified and formation processes in Ghana. potentially problematic ideas which have been con- structed upon these. Following this, and the provi- The investigation of shipwreck site formation sion of a brief contextual account of the documen- processes sets the stage for the interpretation of tary history of EssoukTadmekka and the recent exca- events from an historical maritime past and supports vation of its archaeological site, the account of the the investigations of wider cultural phenomena. In a analysis of the Essouk Tadmekka gold coin mould region such as coastal Ghana, where relatively little artifacts is provided. The account of the analysis is known concerning the maritime environment and explains the process of their identification against its effects on historic trade and on consequent sub- the received wisdom concerning this artifact type in merged cultural resources, an understanding of site the West African Sahel and the chemical and teclmical formation processes is foundational in terms of how analysis carried out following their identification. The sites are investigated and interpreted. This paper is a final part of the talk is a discussion of the impact of discussion of an integrative and interdisciplinary the Essouk- Tadmekka findings on our understand- approach to investigating these processes within the ing of the trans-Saharan gold trade, focusing on three natural and historical cultural settings of a select ship- issues: the chronological development of the trans- wreck site in Ghana. Data from this shipwreck site, in Saharan gold routes; the development of the sources conjunction with data from experimental control ar- of the gold; and the development of the processing eas, provide insights into historic trade in the region, and shipping of the gold. and make available a comparative data set for future investigations on a regional scale.

113 NYAME AKUMA No. 70 DECEMBER 2008

14 West Africa and the Sahara - New 14.2 Cameron D. Gokee, University of Insights. Session Chair: Eric Huysecom Michigan, USA. Of time and the river: and Donatella Usai. investigating late prehistoric socio- economy in the central Faleme Valley. 14.1 Cornelia Kleinitz, Humboldt- Universitat zu Berlin, Germany. ‘Enig- matic markings’? Adiachronic perspec- Extant archaeological research along the Faleme River in eastern Senegal and western Mali has docu- tive on multi-sensory engagements with mented a palimpsest of human occupations ranging rock surfaces in sub-Saharan West in age and scale from the Palaeolithic tool scatter to Africa. the modern agro-pastoral village. In the winter of 2008, the Central Faleme Archaeological Project launched So-called ‘cup marks’, ‘grinding hollows’ and a field program focused specifically on the long-term ‘sharpening’ or ‘polishing grooves’ are common fea- socio-economic dynamics of the Late Holocene pe- tures in natural and worked rock surfaces in Africa riod (c. 4000 BP to present). Preliminary analyses of and beyond. Such forms are often attributed to one survey and excavation data suggest that people liv- or another utilitarian function, but they have also ing in the central Faleme corridor experienced both been suggested to have played various symbolic cultural continuity and diversification in the face of roles. Due to their sustained presence in the land- broad climatic, economic, and political changes tak- scape such ‘enigmatic markings’ are likely to have ing place across West Africa. Settlement patterns been re-used and/or reinterpreted over time, and thus among numerous small sites associated with a may have formed persistent components of site his- microlithic tool assemblage document a regional oc- tories. Their frequency and durability as well as their cupation by mobile to semi-sedentary foragers dur- apparent variability in size and shape, in techniques ing the Late Stone Age. The subsequent Iron Age of manufacture and in location contexts call for a occupation marks a shift towards more sedentary vil- more detailed discussion of such forms. At rock art lage life, a food-producing domestic economy, and sites in sub-Saharan West Africa cup marks and re- increasing craft specialization - concomitant with the lated forms frequently occur in larger numbers, often formation and expansion of larger polities along the in close spatial relationship to painted rock art and Senegal River to the north and the Niger River to the occasionally in locations that preclude their use in south and east. Ceramic data from regional surface day-to-day circumstances. Their formal variation is collection and two excavation probes at the village described here and a range of possible utilitarian and/ site of Pathe Djimba suggest that Iron Age communi- or symbolic manufacture and use contexts are as- ties had become enmeshed in a wide array of cultural sessed on the basis of attributes of form and manu- and economic networks. Finally, archaeological re- facture techniques, and their location at the sites and mains from the past several centuries provide a mate- in the landscape. These markings are argued here to rial testament to socio-economic patterns - described have resulted from traditions of subtractive engage- in various textual sources - fundamental to the emer- ment with rock surfaces spanning several millennia gence of contemporary cultural practices and identi- and different cultural contexts. While such forms ties. These and future data will ultimately provide a contain only limited visual information, their making case study for investigating the interplay between involved multi-sensory performative engagements regional socio-economic processes and more local- with rock surfaces - in particular a series of gestures ized culture historic trajectories in West Africa over - which could point to concepts motivating the crea- the past several millennia. tion and use of such petroglyphs at specific locales.

114 NYAME AKUMA No. 70 DECEMBER 2008

14.3 Katie M. Manning, University of 14.5 Daouda Keita, Université de Oxford, Great Britain. Dynamics of an Bamako, Mali. Kokolo, un site d’habitat agro-pastoral economy during the mid- protohistorique en Pays dogon. late Jrd millennium BC in the Lower Situe sur le Plateau dogon, a environ 15 km au Tilemsi Valley, eastern Mali. nord-est de Bandiagara, le site de Kokolo est un des The Lower Tilemsi Valley has long been her- nombreux sites qui compo sent le gisement alded as a key region in the development of agro- d’Ounjougou. Les recherches archeologiques pastoral economies along the Sahara-Sahel border- menees depuis 2004 dans le cadre du Programme lands in West Africa. And yet, the precise timing and Peuplement Humain et evolution paleoclimatique en nature of such developments has remained relatively Aji-ique de l ‘Ouest ont permis de découvrir des restes unknown. This paper sets out a new chronological de structures d’habitat qui se compo sent framework for the occupation of the Lower Tilemsi essentiellement de blocs de pierres seches entourant region and presents evidence for the exploitation of des dalles disposées horizontalement. A ces struc- domesticated pearl millet in the mid-3rd millennium tures, il faut ajouter la découverte d’un espace re- BC. Furthermore, this paper considers the regional serve aux activites domestiques (telle que le travail dynamics of agropastoral organization, contemplat- de mouture, etc.). Dne quantité importante de mobilier ing inter-site variability in subsistence evidence at archéologique a été découverte au cours de ces Karkarichinkat North and contemporaneous, neigh- travaux parmi lequel de l’outillage lithique, de la bouring sites. ceramique et des objets en fer. La présente communi- cation a pour principal objectif de présenter les resultants des trois campagnes de fouilles menées à 14.4 Annabelle Gallin, Universite de Kokolo. Provence, France. Peopling of the Sahel during the late Neolithic: new data from 14.6 Gabriele Franke, Goethe- Kobadi site (Malian Sahel, 1700-1400 Universitat Frankfurt, Germany. Cul- BC). tural change in the mid first millennium During the 3rd and 2nd millennium BC, the BC - new evidence from the Nigerian aridification of the southern Sahara is considered to Chad Basin. have caused the migration of Neolithic people to- wards the Sahelian belt. One of these migrations has Archaeological research in the Nigerian Chad been hypothesized to led from Azawad, Hassi el- Basin over the last 20 years has led to new insights Abiod well area, to the Inland Delta of Niger. In this into the cultural development during the second and framework, the Kobadi site (1700-1400 BC), located first millennium BC. In the past years, work has fo- in the Merna area, has been regarded as one of the cused on large settlements that seem to appear sud- first settlements of the Neolithic fishers/hunters from denly in this region from the middle of the first millen- Azawad. The technical and stylistic study of the nium BC onwards. Excavations and geomagnetic sur- whole ceramic corpus excavated in Kobadi, performed veys at Zilum and neighbouring sites in the Gajiganna by the author during her doctoral research, and the Culture area have produced evidence for radical review of Hassi el-Abiod collections permit to test changes in settlement size and structure as well as the hypothetical link between Azawad and Merna. socio-economic systems, e.g. ditches surrounding The existence of various technical traditions and the settlements and the cowpea as new cultigen. To decorative styles in the Kobadi corpus and their ech- find out whether these changes only occurred lo- oes in the Inland Delta of Niger, Azawad and Niamey cally or were a regional phenomenon, the research areas enable us to draw a new picture of the peopling was extended to the south of the Gajiganna Culture of the Sahelian belt, around the Niger river, during area. Test excavations were carried out at two sites in the late Neolithic (3rd and 2nd millennium BC). 2004 and 2005. In this paper I will present the analy- sis results from one of these large sites (Malankari, c. 30 hectares), dating to about the fourth century BC. A complex settlement structure as shown by the

115 NYAME AKUMA No. 70 DECEMBER 2008

geomagnetic survey, and artistic clay figurines that the clay used in their manufacture. Hence via their are completely different from earlier and contempora- deposition in the earth at Nyoo, it can be suggested neous sites in the Chad Basin are indications of cul- that they were returned to the medium from which tural change that might have led to an initial form of they came. This might in turn be linked to the former social complexity. existence of beliefs surrounding earth cults. This dis- cussion seeks to explore the possible place of pots within earth cults; through the material from which 14.7 Didier N’Dah, Universitée de they are made and potentially also through the proc- Abomey-Calavi, Benin / Université de esses utilised in their manufacture, but also through Ouagadougou, Burkina Faso. Les sites their acting as mnemonical objects invoking relation- du Late Stone Age de la Pendjari (nord- ships with the earth. Finally, the possible interpretive ouest de la République du implications of the Nyoo material and of ideas sur- Benin). rounding pots and earth cults will be briefly consid- ered in relation to archaeological contexts outside La localisation de plusieurs sites du Late Stone sub-Saharan Africa with reference to aspects of Age tout au long de la Pendjari, au cours des ‘structured deposition’, specifically Grooved Ware recherché menées par l’équipe mixte benino- pits from the later Neolithic of the British Isles. allemande et les premiers travaux permettent d’avoir quelques nouveaux elements d’appreciation de cette periode de la préhistoire. Les travaux récents que 14.9 Benjamin W. Kankpeyeng and nous venons de mener sur l’un des sites permettent Samuel Nkumbaan, University of Ghana, de disposer de données archéologiques ainsi que de Ghana. Ancient shrines? Koma stone datations radiocarbons croisées. Le but de la présente circle mound sites revisited in northern communication est de partir de ces nouveaux ele- Ghana: a preliminary report. ments pour contribuer au debat sur la question du Late Stone Age en Afrique de l’Ouest. The discovery This paper reports preliminary archaeological of several Late Stone Age sites along the Pendjari survey and excavations in the archaeological region during research undertaken by the joint Benin-Ger- within the basins of the Sisili and Kulpawn rivers in man team, and first investigations provided some new northern Ghana in June 2006 and January 2007. Pre- background information on this period of prehistory. vious research had described the stone mounds con- Recent work at one of the sites yielded archaeologi- taining terracotta as burial mounds. The current re- cal data as well as radiocarbon dates. Based on these search provides new perspectives and based on the new elements, it is the goal of this paper to contrib- material arrangements within and around the mounds, ute to the debate on the question of the Late Stone the types of artifacts and their distribution, the na- Age in West Africa. ture and burial of incomplete human remains, and residue on figurines and complete pots, it is possible to propose the hypothesis that the mounds were 14.8 Timothy A. InsolI, University of shrines. Ethnographic analogues of some West Afri- Manchester, Great Britain. Pots and can ethnolinguistic groups provide insights into the earth cults. The context and materiality likely multiple functions of the ancient shrines. of archaeological ceramics amongst the Tallensi of northern Ghana and their 14.10 Lucas Petit, Frankfurt, Germany. interpretive implications. Crime scene investigations at Oursi hu- This paper seeks to consider what factors might beero, Burkina Faso. underlie the deposition of a large spread of pottery Probably the only good thing about the event recorded during excavations in a shrine, the Nyoo of violent destruction in history is when it turns out shrine, in the Tongo Hills of northern Ghana. Primary to be, many centuries later, a most interesting find for in understanding what this pottery might represent archaeologists. The burnt structures of Oursi hu- would appear to be concepts of materiality associ- beero in northern Burkina Faso give us a good idea ated with the pots as products of the earth, i.e. through

116 NYAME AKUMA No. 70 DECEMBER 2008

of how a rural African village may have looked like in 14.12 Olalekan Akinade, National the Middle Ages. The destructive fire had ‘frozen’ Commission for Museums and Monu- th the situation of the 11 century AD, such as beauti- ments, Nigeria. The differentiation in fully finished architectural elements, storage jars filled with organic remains, ground stones with color traces, space, location and time of Nok charred wooden furniture, a rolled-up rope, a slave terracotta objects, Nigeria. chain, basketry and, surprisingly, also the skeletal The paper considers the differences in the na- remains of some murdered inhabitants. Numerous ture and existence of conventional Nok terracotta scientists have recently been involved in trying to objects inside and outside the Nok Culture area. A reanimate this unique picture of sub-Saharan village- report on recent fieldwork in regions outside the Nok life of which some details were presented at previous Culture area will be presented. A summary of find- SAfA conferences. In this paper Oursi hu-beero will ings is given and seemingly similar attributes of the be viewed through the glasses of a crime scene in- objects under discussion will be highlighted. The vestigator. author calls for rethinking some generally held views By closely studying the victims, the causes about Nok Culture phenomena and the spatial spread and manners of their deaths as it is undertaken in of that culture. forensic anthropology and by studying the archaeo- logical and historical contexts, we will present new 14.13 Musa O. Hambolu, National information about the last hours of the inhabitants of Oursi hu-beero and the surrounding village. What Commission for Museums and Monu- has really happened at the crime scene? Who were ments, Nigeria. Constructing a relation- the suspects and what were their motives? ship between northwestern Nigerian terracotta sculptures and those 14.11 Oumarou Amadou Idé, Université of Nok. Abdou Moumouni Niamey, Niger. La Widespread in the northwestern part of Nigeria moyenne vallée de la Mekrou (Niger are terracotta sites that have been ravaged by illegal S.O.): elements d’une occupation diggers. The sculptures are known to the art market humaine ancienne. world as Katsina/Sokoto terracottas. Recent efforts at making the best out of a bad situation (through La Mékrou, province archéologique peu connue scientific excavations) have produced some modest jusqu’alors, s’avere un haut-lieu de la préhistoire en results. While a date of 3,337 ± 39 BP obtained at one région sahelienne. C’est un affluent de la rive droite of the sites (Tsunkwui) led to a tentative belief that du fleuve Niger d’une largeur moyenne d’environ 25 we might be dealing with the antecedents of the more m qui délimite la frontière entre le Niger et le Benin, sophisticated and more developed Nok terracotta ou elle prend sa source. La moyenne vallée de la sculptures, a second date of 1555 ± 38 BP from an- Mékrou, milieu nature1 au paysage ensavane, est other site (Tsauni Lamba) has dampened the initial privilegiée en regard de son environnement, car elle a assumptions. Still, while more excavations are cer- du jouer depuis des temps immemoriaux un role tainly called for, we can begin to compare the two attractif pour les nombreuses populations seemingly disparate traditions. préhistoriques qui s’y sont succedées. Les travaux Therefore in this paper I intend to (1) present meneès dans la moyenne vallée de la Mékrou font data from Tsunkwui and Tsauni Lamba; (2) compare emerger la richesse d’une region sahelienne, dont the cultural manifestations; and (3) compare them l’occupation humaine ancienne est restée longtemps conjointly with data from, and interpretations on, Nok inconnue. Toutes les périodes sont representées, en terracotta sites. The total impression one gets from une longue sequence qui paraît ininterrompue de l’ the northwestern Nigerian sites is that of basic simi- Acheuléen au subactuel, en passant par le larities with Nok in terms of archaeological content Paléolithique recent, le Néolithique et l’apparition and context of finds. More investigations should d’une métallurgie du fer. throw further light on the purpose of the terracottas, the societies that produced them and their relation-

117 NYAME AKUMA No. 70 DECEMBER 2008

ship to other centers of artistic productions during this part of the Sahara. Although the ecological con- the early Iron Age of Nigeria. ditions were better during the Holocene humid phase (c. 8,500 to 5,000 cal BC) a sedentary way of life was improbable in the Djara region. Due to high evapora- 14.14 Giuseppina Mutri, Universita tion rates surface water was only seasonally avail- degli Studi di Roma “La Sapienza”, Italy. able. Thus, during times of rainfall Djara formed a Natural resources as cultural features in temporary ‘refuge’ in the midst of an unpredictable the settlement system of Late desert. However, during the dry seasons of the year Pleistocene foragers in the Jebel Garb, the dwellers retreated to regions offering permanent Libya. water. This paper presents the settlement system of upper Later Stone Age groups living during the Late 14.16 Heiko Riemer, Universitat zu Pleistocene in the Jebel Garb (Libya), correlating with Köln, Germany. Filling a gap in Saharan the geomorphologic and palaeoclimatic characteris- prehistory: the pottery chronology of tics of the region with the techniques used by the Uweinat. local hunter-gatherers to acquire lithic raw material. The natural environment where past populations Since Ahmed Hassan rediscovered Jebel lived is not merely a ‘natural’, but a ‘culturally per- Uweinat in 1923, it has become known for its immense ceivable’ context. These hunter-gatherers chose the rock art heritage. However, there is as yet only very best areas to settle, weighing up the presence of water scanty information on other artifacts, site contexts supplies against the proximity of habitats for animals and settings, as well as on the internal chronology of that were probably concentrated on the banks of the Uweinat. To fill a gap in the research program on wadis. The presence of water possibly led to the di- Holocene settlement history and climatic change be- versification of the economic activities and allowed tween the Nile Valley and the Central Sahara, the 2005 to make genuine’ cultural choices’, providing a greater ACACIA expedition of Cologne University carried and wider choice of food. The availability of raw ma- out a large-scale survey in the western (Libyan) sec- terials for manufacturing stone arteiacts was also tor of Uweinat. More than half of the 144 sites re- important for the upper LSA communities in Jebel corded yielded pottery that allows approximate dat- Garb and was a determining factor in their settlement ing of assemblages by typo-chronological compari- system. son with other regions in Egypt, Libya, Sudan, and Chad. The results indicate a chronological sequence with three phases lasting from c. 6600 to 3000 cal BC, 14.15 Karin Kindermann, Universitat zu as well as pottery roughly affiliated to later periods. Köln,Germany. Seasonal cycles and Moreover, the distribution of sites and their geo-topo- settlement patterns of Holocene hunter- graphical settings yield first insights for understand- gatherers at Djara, eastern Sahara. ing the settlement potential and development of Uweinat. From 1998 to 2002 several interdisciplinary ex- peditions to the Egyptian Limestone Plateau in the eastern Sahara were undertaken within the rramework 14.17 Michael Brass, University ofthe Cologne University Collaborative Research College London, Great Britain. Investi- Centre 389 ACACIA. A large settlement area labelled gating long-term changes in Saharan Djara was located at a distance in excess of 100 km pastoral social organization, 3rd to 1st from permanent water sources. Within this area sur- millennium BC. vey work was carried out at more than 240 predomi- nantly mid-Holocene sites (c. 6,400 to 4,500 cal BC). My doctoral research develops research pre- Detailed examinations of the sites, their distribution sented at SAfA 2006 (Calgary) which employed rock and positions in the landscape and the characteristic art, tumuli and skeletal remains from the central and geomorphological features allowed to reconstruct eastern Sahara to evaluate and model the develop- Holocene hunter-gatherers subsistence strategies in ment of social complexity in early Saharan pastoral

118 NYAME AKUMA No. 70 DECEMBER 2008

societies through examining pastoral settlement pat- artwork largely showing animals which were of sig- terns, monument/mortuary distributions and grave nificance to them. While the ecosphere declined, ar- assemblages, particularly considering the presence tistic production seemed to ignore the alarming de- and point of origin of valued items/prestige goods. velopment and evolved an aesthetics which initially An approximate date of 4000 BC was proposed for foregrounded animals in the function for which they the first archaeological signals of transient elite lead- had been domesticated whereas later they were cel- ers within the pastoral societies. More specifically, I ebrated for their looks, symbolic capital, and their am reexamining the early 20th century site reports and potential to express status. Even representations of materials from the c. 8000-3000 BC mortuary site of humans appear to adapt to this principle. But be- Jebel Moya in the south-central Sudan. The early yond such inclusive patterns, the landscape seems 20th century Wellcome excavations at Jebel Moya to have been divided into many small partitions, each were undertaken to a relatively good and detailed connected to a specific identity. In the youngest pe- standard for the time and an amazing 2,793 graves riods the prevailing aesthetic attitude is replaced with were excavated and recorded. Today, the materials a yet different symbolism that emphasizes the human are housed in the Duckworth Laboratory (Cambridge), role in using the landscape in a nomadic lifestyle the British Museum and the Petrie Museum (Lon- which, in itself, is entirely based on and communi- don), and the Pitt Rivers Museum (Oxford). The ma- cated through a specific animal, i.e. the camel. In the terials include grave goods and human remains. The wake of this development artistic sophistication loses site has been recently re-seriated by Gerharz (1994) in importance and eventually representational art with some further work on the pottery by Isabella partly becomes replaced by script and signs. Caneva (published in Antiquity for 1990). However, neither has considered the social aspects of the 1 2 graves in terms of the individual burial assemblages 14.19 Donatella Usai , A. Di Matteo , P. 2 l nor the distribution of the graves themselves. The Iacumin and S. Salvatori : Emerging grave cards are kept with the collection in the complexity: a view from the Late Cavendish Laboratory to which I have secured ac- Palaeolithic, Mesolithic, Neolithic and cess. Structural and spatial analyses of grave goods Post-Meroitic cemetery of Al-Khiday 2, are being undertaken alongside physical re-examina- Central Sudan. (1. Istituto Italiano per tion of the composition of certain classes of grave goods. Furthermore, I will also attempt to link the l’Africa e l’Oriente (IsIAO), Italy; 2. published physical anthropology to the grave cards. Universita degli Studi di Parma). This brief presentation will give the results of my The cultural evolution of human groups inhab- research to date and open it up for discussion iting the Central Sudanese Nile valley since the Late amongst a wider audience of Africanists. Pleistocene can be still considered a debatable mat- ter. The poor preservation of the archaeological de- 14.18 Tilman Lenssen-Erz, Universitat posits, due to a plethora of post-depositional natural zu Köln, Germany. The aesthetics of and human disturbances has made the studies on this subject to slide on a sort of ‘bog’. The recent aridification: the evolution of herder rock discovery of a well-preserved village and its associ- art in NE Chad. ated cemetery, by the IsIAO project working at El The prehistoric inhabitants of the Sahara re- Salha (Central Sudan), may open us new frontiers. gion articulated their agency under the conditions of The cemetery, moreover, seems to be an authentic ongoing aridification in part through aesthetic sym- mine of information as the presence of human re- bolic behavior which became petrified, as it were, in mains dating from the Late Palaeolithic to the Post- ubiquitous rock art. When pastoralists started to set- Meroitic period, with phases of the Mesolithic and tle the Ennedi Highlands, climatic deterioration was the Neolithic, represents a unique opportunity to already under way, with declining tendency. How- deploy a range of specialized studies, from bio-ar- ever, despite an environment that degraded from their chaeology to DNA analysis. This paper will briefly accustomed way of life and required adaptation to present the archaeological results accompanied by a new modes of livelihood, people kept on producing preliminary assessment of isotopic and environmen- tal data.

119 NYAME AKUMA No. 70 DECEMBER 2008

14.20 Friederike Jesse, Karl Peter 15 Early Human Behavior and Technol- Wendt, Franziska Bartz, Thomas Frank, ogy in the Middle Stone Age of Eastern Fenna Godhoff, Robin Peters and and Southern Africa. Session Chair: Bernhard Buhs, Universität zu Köln, Pierre-Jean Texier. Germany. Shades of the past: GIS- based spatial analysis of prehistoric 15.1 Pamela R. Willoughby, University surface sites in the Lower WadiBowar of Alberta, Canada. The Middle and (northern Sudan). Later Stone Age of Iringa, southern Settlement sites in arid areas such as the Sa- Tanzania. hara often present themselves as large surface scat- The Iringa region lies in the southern high- ters of artifacts which at first sight do not seem to be lands of Tanzania. It is best known for its Acheulean very specific. Features structuring a site (e.g. hearths sites such as Isimila Korongo, first studied by F. Clark or postholes) are either missing or poorly preserved. Howell in the 1950s. But it also contains numerous Often the simple mapping of finds and features al- large granite koppies with associated rockshelters. lows neither for statements concerning internal set- In the summer of 2006, Pamela Willoughby, Pastory tlement structures nor for the spatial and chronologi- Bushozi and Katie Biittner collected artifacts from cal delimitation of settlement areas. In this paper we the surface of three rockshelters - Mlambalasi, tackle these problems using a GIS-based approach. Magubike and Kitelewasi. Test excavations were also Our study area is the Lower Wadi Howar in Northern carried out at Mlambalasi and Magubike. Mlambalasi Sudan where under the aegis of the ACACIA project contains a record of the historic period, the Iron Age, numerous large settlement sites have been discov- a Holocene Later Stone Age (LSA), LSA human buri- ered and partly examined. For the present study five als, a Pleistocene LSA and the Middle Stone Age surface sites have been chosen, mostly dating to the (MSA). The first test pit at Magubike contains the 4th and 3rd millennium BC. There, small-scale excava- Iron Age, the LSA, and a LSA to MSA transitional tions combined with a detailed survey using a total industry or a true MSA. A second test pit contains station have been carried out. Different methods have the Iron Age above 1.6 meters of MSA deposits. been applied to detect the inherent structures of the Seven fossilized human teeth were recovered from sites. Besides the elaboration of maps showing the these MSA deposits. Faunal remains (bones and distribution patterns of the artifacts, approaches shells) were recovered from all sites. This paper re- known from GIS-based analysis on large scales such views the cultural history of Mlambalasi and as Thiessen-Polygons, Largest Empty Circle with Magubike in the light of current models of the ori- Kriging, density isolines and Nearest Neighbors gins of anatomically and behaviourally modern hu- Analysis have been tested. The results will be pre- mans. sented in this paper and displayed in the poster ses- sion. 15.2 Katie Biittner, University of Alberta, Canada. Raw material variability in MSA lithic assemblages from Iringa region, Tanzania. Stone tools have a critical role to play in our understanding of the behaviour of early humans. In particular, the types of raw materials that are present in stone tool assemblages, and the sources from which they are acquired, provide information relating to decision-making processes, planning, organization of technology, and group mobility. The characteri- zation of lithic artifact assemblages from two rockshelter sites, Magubike and Mlambalasi, in

120 NYAME AKUMA No. 70 DECEMBER 2008

southern Tanzania is currently ongoing in order to plex. In the meantime, alternative dating methods with evaluate inter- and intra-assemblage variability. Pre- a larger time range, such as OSL and ESR dating, liminary macroscopic analyses demonstrate a con- promise to bridge this gap. For this reason, the old siderable range in the raw material types used. trench was re-opened in 2007 and half a square metre was excavated. The improved excavation method is expected to allow a better chronological classifica- 15.3 Pastory M. Bushozi, University of tion of the lithic assemblages. In collaboration with Alberta, Canada. Middle Stone Age Zenobia Jacobs, Bert Roberts and Rainer Grün, sedi- technology and hunting behavior in ment samples for OSL dating and tooth samples for Tanzania. ESR dating were collected. If the measurements prove successful, Apollo 11 might become one of the best- This paper describes the preliminary results dated MSA sites in southern Africa. on the possible functions of lithic points from four Middle Stone Age (MSA) rockshelters in Tanzania: Mumba, Nasera, Magubike and Mlambalasi. It de- 15.5 Jayne Wilkins, University of scribes ways in which stone points were used and Toronto, Canada, Luca Pollarolo and the technological differences between spearheads Kathleen Kuman, University of the and arrowheads. The method used in this study was Witwatersrand, South Africa. Prepared first developed in North America. Later on it was used in the Levant and for some MSA sites in sub-Saha- core technology at Kudu Koppie and the ran Africa. Goodwin and Van Riet Lowe recognized modern human behavior debate. the presence of scrapers and points as the defining The aim of this paper is to describe the pre- character when they initially defined the MSA in 1929. pared core reduction strategies employed at Kudu Scrapers and points seem to have been deliberate Koppie, a stratified terminal Earlier Stone Age/ target forms for many toolmakers during the MSA Sangoan and MSA archaeological site located in the period, but the ways in which these artifacts have Limpopo region of northern South Africa, and relate been utilized is not well understood. It is important to lithic reduction to the variables of time and raw mate- understand the technological ability and food acqui- rial. A technological analysis of the prepared cores sition strategies of the earliest anatomically modern and end-products of Kudu Koppie suggests that both humans who evolved in sub-Saharan Africa during the Sangoan and MSA toolmakers employed the the MSA. Preliminary results from this study sug- Levallois Volumetric Concept, but often exploited a gest that MSA points were hafted and used to form nodule’s natural convexities and form. The MSA projectile armatures. toolmakers used a greater variety of prepared core methods and more intensely exploited cryptocrystal- 15.4 Ralf Vogelsang, Universität zu Köln, line nodules, the scarcity of which may have resulted Germany. New excavations and dating in a more ‘formalized’ application of the Levallois Volumetric Concept. These observations are consid- of the Middle Stone Age layers at Apollo ered in light of their cognitive and economic implica- 11 (Namibia). tions, and within the context of the behavioral mo- The rockshelter ‘Apollo 11’, situated in the dernity debate. southwestern part of Namibia, exhibits one of the most important MSA stratigraphies in southern Af- rica. This is not only due to finds of painted slabs dated to c. 27,000 BP, but also to an extraordinarily comprehensive cultural sequence covering all major MSA phases. Unfortunately, at the time of analysis, numerical ages obtained by using radiocarbon dat- ing were only available for the youngest MSA com-

121 NYAME AKUMA No. 70 DECEMBER 2008

15.6 Guillaume Porraz1, Pierre-Jean 15.7 Pierre-Jean Texier1, Guillaume Texier2, Aude Coudenneau3, John Porraz2, Chantal Tribolo3 & Jean- Parkington4 and Jean-Philippe Rigaud2. Philippe Rigaud1. Technology and tech- An overview of the Howiesons Poort niques involved in the shaping of the complex at Diepkloof rockshelter (West- Stillbay biJacial points after the analysis ern Cape, South Africa): implications of of the more characteristic remains from the truncated and backed pieces (1. Eales Cave, Hollow rockshelter and CEPAM, CNRS (Centre National de la Diepkloof rockshelter lithic assemblages Recherche Scientifique), France; 2. (Western Cape, South Africa) (1. Institut de Préhistoire et Geologie du l’Institut de Préhistoire et Géologie du Quaternaire, CNRS, Université de Quaternaire, CNRS – Université de Bordeaux I, France; 3. LAMPEA, Bordeaux, France; 2CEPAM, CNRS Université Aix-Marseille I, France; 4. (Centre National de la Recherche University of Cape Town, South Africa. Scientifique), France; 3. Institut de Truncated and backed pieces, often referred to Recherche sur les Archéomateriaux - as geometrics, segments, trapezes, or crescents, are Centre de Recherche en Physique considered one of the main typological and func- Appliquée à l’ Archéologie, CNRS - tional innovations that occurred among prehistoric Université de Bordeaux, France. hunter-gatherer societies. Their prevalence during the Later Stone Age has often been used as a parallel The nature as well as the status of the lithic to establish comparisons with older industries com- waste resulting from a bifacial shaping is opposite to prising such geometric pieces. One of the most ex- the nature and the status of the waste resulting from pressive complexes is the Howiesons Poort, once a debitage. But when producing the blanks to be considered as a transitional phase between the MSA shaped those two antagonistic stone knapping proc- and the LSA. However, even if the originality and esses are sometimes successively combined in the independence of this southern African Middle Stone same bifacial chaine operatoire. The bifacial points Age complex is now fully accepted, its nature and of three Stillbay lithic assemblages from the Western variability is still debated. Truncated and backed tools Cape Province, as well as the more characteristic by- of the Howiesons Poort constitute the ‘fossiles products resulting from their shaping, are taken into directeurs’ ofthis complex. They syncretize most of consideration. The main stages of the identified the debates that lead the scientific literature. The chaines operatoires and their variability according to standardization of these tools, whether they are the raw materials implemented are documented and microlithic or not, their association with blade pro- debated as well as the identification criteria of the duction, their confection on non-local fine-grained techniques involved for the completion of those raw materials as well as their function (associated bifacial points. The excavation under way at Diepkloof with hunting technology) are some presuppositions rockshelter (DRS) already brought to light, on a four remaining to be fully demonstrated. The excavation meter deep section, the existence of a Stillbay com- of the MSA site of Diepkloof rockshelter has high- plex subsequent to a MSA still undifferentiated but lighted the presence of several layers of occupations prior to a Howiesons Poort complex. 22 TL dates have belonging to the Howiesons Poort, dating to 55-65 already been calculated (see also paper by Tribolo et ka. Our study focuses on truncated and backed al. Session 9) on those two complexes. The techno- pieces, as one avenue to interrogate broader techno- logical analysis under way already allows us to evalu- logical and economic features. The tendency of sty- ate the segmentation of the bifacial chaine operatoire listic and functional attributes of these tools, often at DRS and moreover to rely on the part played by being key behavioral interpretations for this the sites in this specific knapping activity. complex,will also be discussed.

122 NYAME AKUMA No. 70 DECEMBER 2008

15.8 Karen Loise van Niekerk, Univer- 15.9 Moleboheng Mohapi, University of sity of Cape Town, South Africa. A the Witwatersrand, South Africa /Na- preliminary report on the fish remains tional University of Lesotho. A new from the Middle Stone Age sites of angle on Middle Stone Age hunting Blombos Cave and Klasies River with technology in South Africa. reference to the taphonomy offish bones Archaeologists have always been interested from Later Stone Age assemblages in the in hunting strategies of people living in the Stone southern Cape, South Africa. Age. Early interpretations of weaponry were specu- lative but in the last ten years technological studies There are currently only two well stratified have transformed lithic interpretations in Europe and Middle Stone Age sites in southern Atrica that have America. This paper presents results of a compara- preserved marine fish remains, namely Blombos Cave tive technological study of Middle Stone Age (MSA) and Klasies River, both located along the southern points from two sites in KwaZulu-Natal, South Af- Cape coast. The question of whether these fish re- rica, namely: Sibudu Cave and Umhlatuzana mains accumulated through human activity is impor- rockshelter. The points have ages between about 75 tant as fishing would indicate an awareness, and to 33 ka. The study examines changing hunting strat- broadening of, the diet base at a time period where egies through time in the MSA because points have other evidence for increasingly modern human been considered parts of hunting weaponry; it has behavior is becoming apparent. In order to deter- been suggested they were used as spears or darts. mine whether these remains were accumulated Points at the bottom of the sequence at Sibudu through human agency, it is necessary to develop a (Stillbay) are likely to have been multi-functional tools better understanding of the taphonomic processes while those in the younger MSA phases (post- that influence the preservation of fish bone. The fish Howiesons Poort, late MSA and final MSA) are likely assemblages from these sites are small relative to to have been tips of handheld spears. All points at those from Later Stone Age sites, but both contain Umhlatuzana are likely to have been tips of handheld bones of large specimens that are unlikely to have spears. been brought in by birds or other small predators. Due to the greater antiquity of the MSA assemblages, it is assumed that what remains is not fully repre- 15.10 Andrew W. Kandel, Heidelberger sentative due to natural processes of attrition over Akademie der Wissenschaften, Germany thousands of years. Therefore the better preserved and N. J. Conard, Eberhard Karls fish remains from several Later Stone Age sites were analyzed to gauge the survivability of specific ele- Universität Tübingen, Germany. The ments per species and under varying conditions. The significance of coastal adaptations along material from the LSA layers of Blombos Cave, open the southern and western coasts of South Garcia State Forest sites, Nelson Bay Cave and Africa. Hoffman’s Cave has been analyzed for comparison. This data can then be used to better interpret the The early use of coastal resources has been presence and absence of species and elements in the well documented at many archaeological sites along less well preserved MSA assemblages. Evidence for the southern and western coasts of South Africa. fishing during the MSA would support other research The spectrum of marine resources that contributed that suggests that the origin of modern human to prehistoric subsistence includes molluscs, crusta- behavior is of much greater antiquity than previously ceans and fish, as well as marine mammals and birds. assumed, and that much of this evidence is currently The intensification of coastal adaptations has been seen in the area of the southern Cape, South Africa. noted at many sites and signifies periods when hu- man population densities increased. The ability of prehistoric gatherers and hunters to adapt to this unique environment is a likely cause for their expan- sion within and out of Africa. In addition to present- ing a brief summary of coastal adaptations in South

123 NYAME AKUMA No. 70 DECEMBER 2008

Africa, this paper provides a case study trom several 16.2 Aneila Nhamo, University of Zim- open-air sites in the Geelbek Dunes. Located 100 km babwe, Zimbabwe. The ‘crocodile-men’ north of Cape Town on the shores of Langebaan of Harare: diversity in Zimbabwean rock Lagoon, this ecological setting between sea and land provides a backdrop for examining coastal adapta- art themes. tions and their significance. It is believed that rock art sites in Zimbabwe might exceed 10,000 and most of these are generally believed to have been executed by hunting and gath- ering communities who populated the sub-region 16 Subsistence, Settlement, and Sym- before the coming of Bantu communities around 2000 bolism in the Later Stone Age of South- years ago. The most popular current interpretation ern Africa. Session Chair: Judith Sealy. of the art is based on ethnographic material from present day hunter-gatherers in the sub-region es- th 16.1 Seke Katsamudanga, University of pecially those from the Kalahari region and the 19 century San communities from Northern Cape Prov- Zimbabwe, Zimbabwe. Researching the ince of South Africa. This approach has afforded re- ancestral searchers a chance to examine the meanings rather landscape of Manyikaland, southeastern than the aesthetics of the rock art but it also led to Zimbabwe - a spatial investigation. overgeneralizations of both the meaning and charac- ter of the art in the sub-region. Although most hunter- This paper presents the results of the archaeo- gatherer groups have similar core traits in their cul- logical research in Zimunya communal lands, south- ture, they also have other distinct values, beliefs and eastern Zimbabwe and the investigation into the spa- issues that they addressed through their art forms. tial behaviour of the various cultural periods repre- This paper aims to explore the motif variability in sented. UsingGeographic Information Systems (GIS) Zimbabwean hunter-gatherer rock art by analyzing and other technologies relevant for analysing spatial the occurrence of a particular theme that has been data, the distribution of the archaeological sites in dubbed ‘crocodile-men’. This motif is found only in the area shows a significant avoidance of the pres- and around the city of Harare. I hope to show that ently wetter areas of the Vumba Mountains and other although southern African rock art is similar it is not upland parts of the area during the Stone Age and a the same. I also endeavor to explore the thinking that shift towards these areas in later times. The paper is these differences in motifs might allude to different an attempt to answer questions on whether that dif- social issues being addressed by the art. ference in settlement preference, among other spatial patterns during the different cultural periods, was a result of climatic and environmental variations or 16.3 Justine Wintjes, University of the whether it was a result of technological and cogni- Witwatersrand, South Africa. Restoring tive differences. The conclusion that can be derived Good Hope: rock art in the age of digital from the results is that climate and the physical envi- reproduction. ronment had a role in the settlement behavior in the research area, and probably in the whole of the east- This presentation is a contribution to the long ern highlands of Zimbabwe. The material evidence controversy about the restoration of works of art - in also makes a significant contribution to the pertinent this case rock painting. We apply digital imaging tech- issues relating to the known sequences in the ar- niques to the historical pictorial record relating to the chaeological cultures of southern Africa in general paintings of Good Hope Shelter, a badly vandalized and Zimbabwe in particular. and deteriorated rock art site in the Sani Pass area of the southern Drakensberg in KwaZulu-Natal, South Africa. We demonstrate ways in which the pixel im- age is qualitatively different from pre-digital tech- niques of recording and replication. Using its spe- cific malleability we propose an innovative technique of digital restoration, bypassing many of the prob-

124 NYAME AKUMA No. 70 DECEMBER 2008

lems of physical restoration. We also argue that the that terrestrial C4 grasses (and animals grazing on idea of an ‘original’ rock art image can be dynami- them) were very minor items of diet. Sheep were cally re-imag(in)ed in the digital era, allowing the re- herded from around 2000 BP, but whatever their role turn of the images to their context, their original place in peoples’ diets, there is no significant shift in the in the world. Instead of creating a removed and simu- isotope ratios of human skeletons in the first millen- lated world of virtual reality, digital imaging can bring nium AD. During the second millennium AD, how- us closer to the materiality and physical presence of ever, people began to eat significantly more C4-based rock art, its granularity and impressionism, and its foods, probably in the form of animal products (meat, essentially dynamic and unfinished nature. milk) from animals grazing on C4 grasses. In one or two cases for which we have information on burial style, the body was interred in a seated, flexed posi- 16.4 Peter Mitchell, University of tion, and the grave capped with stones. Thus, al- Oxford, Great Britain: The canine con- though living sites remain elusive, there is evidence nection: dogs and southern African of new economic (dietary) and cultural practices af- hunter-gatherers. ter 1000 AD, with domestic stock - probably cattle becoming more central to peoples’ diets at this time. Southern African hunter-gatherers were among This evidence supports the hypothesis that the his- the last in the world to make the acquaintance of the torically documented Khoekhoe way of life devel- domestic dog, most likely around 2000 years ago. oped gradually over the last two millennia. It also This paper begins by surveying the evidence for the highlights the question of the origin of the cattle, presence of dogs in hunter-gatherer contexts in given several centuries of contact with Iron Age farm- southern Africa. It then takes advantage of the re- ers to the east. gion’s rich ethnographic and historical sources to consider what the consequences of this acquaint- ance may have been. The ways in which southern 16.6 Lesley Harrington, University of African hunter-gatherers used the dog in their sub- Toronto, Canada. Physical activity sistence strategies, the social role of the animal and patterns among forager children: evi- the symbolic importance accorded it are placed within dence from measures of bone strength. a broader comparative context that draws on exam- ples from elsewhere in the world. The architectural properties of long bones are developed, in large part, as an adaptive response to individual levels of physical activity. Patterns of bone 16.5 Judith Sealy, University of Cape strength observed in the upper and lower limbs of Town, South Africa: The emergence of adults are laid down during childhood when bony pastoralism in southernmost Africa: new tissue is most responsive. Through the cooperation evidence from stable isotopes. of four curatorial institutions in South Africa, the skel- etons of approximately 60 southern African Later Seventeenth-century Dutch records report Stone Age juveniles were studied. Measures of long Cape Khoekhoe people bringing huge herds of sheep bone (diaphyseal) strength were acquired through a and cattle to trade with early European settlers, but combination of nondestructive methods including archaeological evidence of these communities is hard direct measurement, bi-planar radiography and cast- to find. Possible explanations include (1) low archaeo- ing. Bending and torsional strength measures for the logical visibility due to the mobile lifestyle required upper and lower limbs were calculated using a beam to pasture large herds of domesticated animals; (2) model for the diaphysis. These data illustrate the large herds (especially of cattle) may have accumu- developmental timing of strength properties reported lated only recently, around the time of European con- for southern and western Cape Later Stone Age tact, and thus be difficult to detect in the archaeo- adults that have been linked to regional and sex-based logical record. Stable carbon and nitrogen isotope subsistence behaviors. Comparative data from pre- analyses of human skeletons from the southern Cape historic agriculturalist and modern sedentary chil- coast show that, prior to 2000 BP, hunter-gatherers dren demonstrate how subsistence behaviors can be ate varying mixes of marine and terrestrial foods, but reconstructed through analysis of long bone archi- tecture.

125 NYAME AKUMA No. 70 DECEMBER 2008

16.7 Susan Pfeiffer and Lesley marine sources typically makes samples appear sub- Harrington, University of Toronto, stantially older, so this adjustment is non-trivial. Cali- Canada. Indicators of bration of mixed carbon samples can be done by cal- culating the proportion of marine carbon (% Marine) childhood health among foragers and in the diet. This study uses two examples of co-in- agriculturists of southern Africa. terred human burials from coastal South Africa to Every human society must assure the survival compare methods for identifying % Marine in human of its children. Both the roles those children play and bone collagen. One method identifies the expected the environment in which they find themselves can range and endpoints of 100 % terrestrial and 100 % influence survivorship. From the skeletal remains of marine diets, based on stable isotope analysis of the those children who fail to reach adulthood we can local flora and fauna. The other method uses iso- ascertain some aspects of their experience. Growth in topes to identify % Marine, but sets the endpoints linear dimensions, development of tooth crowns and based on the observed isotope values of skeletons. skeletal indicators of non-specific stressors to de- The method using observed endpoints generates velopment contribute to this understanding. Age at more plausible results. When the % Marine values death must be determined from tooth crown devel- are used to calibrate published, uncalibrated radio- opment, which is less environmentally labile than carbon dates of 122 Holocene period adult skeletons, bone growth. Comparing a large sample from Later shifts in the dates range from -622 to +904 years. The Stone Age (LSA) burial sites and a small sample from magnitude of the adjustment is dependent on the % Iron Age (IA) kraal sites, different growth trajecto- Marine value, while the direction is based on time ries can be described from birth to maturity. Rather (pre- and post- 4000 BP). than analyzing absolute linear dimensions, the analy- sis is approached in terms of proportional attainment 16.9 C. Garth Sampson, Texas State of the adult size of the population, since LSA and IA University, USA. Late Holocene faunal adults are very different in average statures. The cross-sectional growth of the LSA children is con- diversity in the upper Seacow River sistent with that of healthy children whereas that of valley, South Africa. the IA children is not. Indicators of non-specific Diversity in the faunal remains from a dozen stressors (anaemia, growth arrest lines, enamel hy- Later Stone Age rockshelters in the semi-arid upper poplasia) contribute to the interpretative framework. Karoo region of central South Africa is reviewed. It appears that causes of death were frequently acute Occupied since about BC/AD 0 by ancestral Karoo among foragers, but were often linked to chronic Bushmen, they accumulated a very wide range of health problems among agriculturists. It is therefore faunal remains including not only the full range of especially important that dental development is the local game animals and attendant carnivores and scav- basis for age assessment of juvenile material from engers, but also micromammals, tortoises, snakes, Iron Age sites. lizards, birds, amphibians, fish, crabs, and molluscs - some 80 species in all. Proportional composition be- 16.8 Genevieve Dewar and Susan tween shelters is very uneven, reflecting local catch- Pfeiffer, University of Toronto, Canada. ment characteristics within the upper Seacow River valley. Perhaps most interesting is the ease with which Calibration of radiocarbon dates from shelters in the path of migratory Springbok may be human skeletons that reflect marine food identified, even from quite modest samples. By c. AD intake: a comparison of methods. 1000 the upper valley was invaded by prehistoric herders who appear to have been accommodated by Accurate calibration of radiocarbon dates into the resident population, who were in turn left in peace calendrical years must account for conditions of the to pursue their ‘eat everything that moves’ subsist- organism in life, including the incorporation of car- ence strategy. The herders themselves either failed bon from marine sources. A variable ratio of marine or did not attempt to dislodge the resident hunter- and terrestrial protein makes neither of the usual ap- foragers from their rockshelters, preferring instead to proaches to calibration fully appropriate. Carbon from camp in the open with their sheep and cattle. Conse-

126 NYAME AKUMA No. 70 DECEMBER 2008

quently very few domestic livestock entered the 17 Researching the Iron Age in Southern faunal record, in spite of the fact that some Africa. Session Chair: Tom Huffmann. rockshelters were surrounded by herder stock enclo- sures with stone walling. In a very short time hunter- foragers were incorporating small herds into their own 17.1 Wim Moritz Biemond, University of highly diversified economy, and building kraals of South Africa, South Africa. Reinterpret- their own, sometimes outside the shelters. But this ing the origin and spread of the Toutswe was not uniformly done and the boundary between chiefdom in Botswana. invading herders and the local ‘hunters-with-sheep’ remained porous, with full-time hunters also moving One of the great polities of southern Africa at about freely between the two zones. But by the time the beginning of the second millennium AD, the the first Dutch stock-farmers invaded the upper val- Toutswe chiefdom, occupied eastern Botswana from ley in the AD 1770s they evidently encountered no the fringes of the Kalahari in the west to the Limpopo stockowners, and their northward advance was suc- River in the east. The chiefdom rose to power around cessfully blocked for about a decade by a well-or- AD 1000, a process which led to the establishment of ganized hunter-forager resistance. The uppermost three major capitals, Toutswemogale, Bosutswe and levels of nearly all the rockshelters reflect the exter- Sung, at around AD 1200. A period of decline fol- mination of the larger game with firearms, and the lowed, culminating in total abandonment by AD 1290. rapid increase in livestock remains as the shelter in- By using a stylistic ceramic analysis method, I re- habitants, now with guns, became enveloped in the interpret existing ideas on the diverse cultural origin complexities of the Dutch frontier economy in which of the Toutswe people. It is suggested that their ori- local Bushmen became employed as shepherds and gin can be traced back to Bantu-speaking people liv- settled around the new farmsteads. ing in the southern Congo around AD 400. They moved south into Zambia and crossed the Zambezi River to spread into eastern Botswana and the north- 16.10 Maria M. Van der Ryst, Univer- ern Limpopo Province (RSA) as Taukome and Zhizo sity of South Africa, South Africa. people by AD 900. The intrusion of sK2 people into Olieboomspoort shelter: home is where the Limpopo Valley led to a gradual movement of the hearth is. Zhizo/Leokwe people into eastern Botswana, thereby giving rise to the Toutswe chiefdom. Further research Data recovered from the last 2000 years of has extended the boundaries of the Toutswe chiefdom hunter-gatherer occupation at the Olieboomspoort to the south, with the Limpopo River serving as a shelter in the Limpopo Province, South Africa, are permanent water source for their cattle herds. used to demonstrate how formal spaces were differ- entially structured over time by multi-band clusters and small hearth groups to meet particular social and 17.2 Morongwa N. Mosothwane, Univer- economic requirements. The two main pulses of in- sity of the Witwatersrand, South Africa. tensification and changes in organisation at approxi- Isotopic evidence of change in Toutswe mately 2000 BP and again at 1500 BP correspond to environment, east central Botswana. the movement of herders and African farmers into the lowlands of the Waterberg. The basic spatial pat- The Toutswe area, in east-central Botswana, is tern of hunter-gatherer open camps and cave/shelter known to have sustained large herds of cattle and sites manifested at Olieboomspoort in an area- small stock approximately 1000 years ago. However, focussed communal space contiguous to the talus this area is currently characterized by low, seasonal and private hearth-focused activities with sleeping and unreliable rainfall as well as poor soils. In order hollows against the wall. Differential uses of space to determine past environmental conditions under through time, and the spatial distributions of the dif- which large herds were maintained, carbon isotope ferent classes of material remains and waste, are dem- ratios of animals from EIA contexts at Bosutswe and onstrated by using a model of unconstrained cluster Toutswemogala were compared to those of present- analysis. day animals from the same sites. Bone collagen °l3C o results of archaeological cattle are around -6 % /oo

127 NYAME AKUMA No. 70 DECEMBER 2008

o compared to -12 /oo recorded on present-day cattle. 17.4 Sarah M. Mothulatshipi, Univer- Sheep and goats from archaeological contexts have sity of Botswana, Botswana. Landscape o °l3C values around -10% /oo whereas their present- dynamics: a case study of the confluence day counterparts have more negative values averag- ing -19% o. Less negative °13C values of both spe- zone of the Shashe and Limpopo Rivers, cies during the EIA are associated with a C4 (grasses) eastern Botswana. dominant component in diet. Present-day animals, This paper presents the archaeological contri- on the other hand, have °l3C values indicating a bution of the confluence zone of the Shashe-Limpopo strong C3 (trees, bushes and shrubs) component in Basin (SLB), a prehistoric landscape with an extraor- their overall diet. The abundance of C4 grasses dur- dinarily dynamic environment. The SLB landscape ing the EIA is a reflection of adequate rainfall, possi- has attracted research attention for decades, and bly longer rainy seasons as well as good herd man- sites such as Mapungubwe have indeed contributed agement strategies. At present, there is scarcity of considerably to our understanding of socio-cultural C4 grasses due to draught and overgrazing and hence and economic changes in Southern Africa over the animals have to depend largely on C3 mophane trees past millennium. However, research has remained for food. Thus, environmental conditions under heavily skewed and sites explicitly targeted for in- which large herds of cattle and small stock were main- vestigation were those deemed important for their tained about 1000 years ago in east central Botswana ability to contribute towards our understanding of were most likely more conducive than they are at the development of social complexity. This study present. demonstrates that the development of complex so- cial formations represents settlement structures that 17.3 Edwin N. Wilmsen, University of epitomize interaction of both long- and short-term Texas, USA, David Killick and Dana cultural and economic processes and that the organi- zation of such structures is randomly distributed Drake Rosenstein, University of Arizona, throughout the landscape. Using remote sensing tech- USA. The social geography of pottery in niques, archaeological sites on the eastern Botswana Botswana as reconstructed by optical side of the SLB have been located, and these attest petrography. to the long-term attraction of the basin for human settlement. This paper shows that this confluence An optical petrography research program was zone remained unexplored because of its initiated in order to determine the extent to which geomorphological setting and the otherwise poor pots, represented by shards excavated at various sites visibility of archaeological sites that could parallel in in Botswana, moved from place to place and what size and status neighboring sites across borders. could be deduced from this about interactions of Further analysis using GIS spatial and geochemical peoples who made and used them. We have analyzed methods on the sites located suggests a significant 267 shards from 23 EIA sites, 7 shards from 3 historic influence by the geomorphologic units on the type sites of known calendar date, and 10 shards from of activities undertaken which encouraged inhabit- pots made in 2006 by local potters. Clays from 74 ants to employ various management strategies to make sources in all parts of Botswana, including those from this apparently hostile environment habitable. which the modern pots were made, were also Through landscape analysis and archaeological ex- analyzed. This allows us to associate with a high cavations, this study has revealed that the fluctuat- degree of confidence the majority of shards in our ing environmental conditions made human habita- sample with clays from specific geological environ- tion of the floodplain problematic and restricted set- ments. Not only pots but also clays circulated among tlement and social organization to its periphery largely sites, as is true among potters today. The petro- on high ground and hill summits, whilst different parts graphic study is integrated with parallel studies of of the floodplain terrain were exploited as water the circulation of glass beads (from the Indian Ocean sources, cultivation and grazing resources. trade) and specularite mined in the Tsodilo Hills to trace pathways of internal exchange and the relation of these pathways to wider interregional networks.

128 NYAME AKUMA No. 70 DECEMBER 2008

17.5 Justin du Piesanie, University of entering into mutually beneficial relationships. For the Witwatersrand, South Africa. Under- example, during the K2 period (AD 1000-1220), the standing social complexity in the Middle proximity of certain hunter-gatherer sites to farmer settlements suggests close relationships between the Iron Age of South Africa: the case of K2 two groups. These relationships included providing and Leokwe. specialized services, such as helping farmers control Calabrese (2005) clearly identified and defined rain. Current research shows that hunter-gatherers two distinct ceramics styles in the Shashe-Limpopo and their places formed patry od the K2 farmers’ ritual basin. Due to this important research, the first landscape. In addition, hunter-gatherer material cul- etlmically distinct groups can be seen archaeologi- ture occurring on K2 period rain-control hills and in cally in sub-Saharan Africa. With the identification shelters on the slopes of these hills differs from that of Leokwe and K2 ceramic styles co-occurring, comes of more classic ‘residential’ material culture foud at the question of how did these groups interact, and other hunter-gatherer sites in the SLCA. Taken to- what was the relative status of these groups in rela- gether, these factors suggest that hunter-gatherers tion to one another through time. Calabrese argues played a role in SLCA farmer rain control. that at least some Leokwe groups maintained a rela- tively higher or equal status on initial contact with 17.7 M. H. (Alex) Schoeman, University the K2 group. His basis for this claim is through the of Pretoria: Rain and smoke: the archae- identification of what he termed ‘Elite Symbolic Ob- jects’. Using GIS, site distributions have been ology of rain-control places and things in analyzed to determine the locale of sites within the the Shashe-Limpopo Confluence Area. landscape, on the escarpment, floodplain or vlei, and Rain-control is materialized ideology. Similar to their position in relation to primary and secondary real rain, it is fluid and - contrary to claims of eternal- resources. The hopes are to assess whether there ness rain-control beliefs are context-specific and thus were forms of control in terms of access to resources, continually re-shaped and re-interpreted. In turn, rain- and what implication this would have in an analysis control beliefs and symbols can influence the socie- of the relative status of K2 and Leokwe groups. Ad- ties that shaped them. During the occupation of the ditionally, reassessment of the validity of the ‘elite Shashe-Limpopo Confluence Area (SLCA) by Leop- symbolic objects’ is addressed with relation to the ard’s Kopje people (AD 1000-1300), rain-control was reexamination of Castle Rock in 2007. shaped though complex interactions between farm- ers, hunter-gatherers, the landscape and material cul- 17.6 Bronwen Van Doornum, Natal ture. Interaction also influenced the selection of rain- Museum, South Africa. Hunter-gatherer- control places. The choice of places - hills - related to two factors. The first was an imagining of appropri- farmer interaction in the Shashe- ate rain-control places, including the association of Limpopo confluence area, South Africa: ‘nature’ with the plateau, where the rain-control hills the K2 period (AD 1000-1220). are located as well as the link between hills, pools/ rock-tanks, caves/shelters and rain-control. The sec- Hunter-gatherers have occupied the Shashe- ond factor was an association of hunter-gatherers Limpopo confluence area (SLCA) for at least 11,000 with the hills and plateau. Through ritual interaction years. During the last 2000 years, the arrival of vari- with the sites knowledge of rain became entangled in ous groups offarmers, including the Zhizo and Leop- the topographic features used in rain-control. A spe- ard’s Kopje people, lead to complex and changing cific material culture, botanical and faunal signature relationships with the indigenous hunter-gatherers. mark the spaces used during rain-control. In this pa- Differences in farmer social structure and identity per I explore the choices, places and things of SLCA played a role in how they interacted with hunter- rain-control. gatherers, and vice versa. Hunter-gatherers appear to have had a variety of responses to the presence of farmers, ranging from distancing themselves and re- treating, to settling in close proximity to farmers and

129 NYAME AKUMA No. 70 DECEMBER 2008

17.8 Tom Huffmann, University of the Iron Age, it is absolutely essential to employ an Witwatersrand, South Africa. archaeometric technique with better resolution in this Mapungubwe and Great Zimbabwe: the period. With luminescence, an absolute dating method using quartz grains from clay, mud and midden origin and spread of social complexity in features, chronometric results with good resolution southern Africa. and high precision can be obtained for Late Iron Age Social complexity in southern Africa first de- sites. For the recent past, luminescence dating is sig- veloped in the Shashe-Limpopo Basin. As is well nificantly more accurate than is possible with radio- known, rank-based society at K2 developed into class carbon. With an explicit, well-planned sampling strat- distinction at Mapungubwe. The transfer of this new egy, luminescence can be used as an anchor for ideology to Great Zimbabwe, however, has received archaeomagnetic dating, a relative dating technique. less attention. New research on rainmaking practices Both luminescence and archaeomagnetism require suggests that a Mapungubwe dynasty introduced particular considerations for use in the southern class structures at Great Zimbabwe. Poor climatic African Late Iron Age, however, and this paper conditions at the end of the 13th century undermined focuses on the use of these techniques for relatively sacred leaders at Mapungubwe itself, and while vul- young, shallowly buried archaeological contexts. nerable, the elite at Great Zimbabwe took over the The construction of an archaeomagnetic master curve important gold and ivory trade. Among other things, requires the location of in-situ, heat-exposed archaeo- the new elite used the now-famous Zimbabwe logical features that can be measured for birdstones to establish their legitimacy. archaeomagnetic direction and also dated by an ab- solute dating method. Hearths, door slides, daga walls and furnaces, all of which are constructed with sedi- 17.9 Dana Drake Rosenstein, Univer- ment datable by luminescence, are types of features sity of Arizona, USA. Considerations for sufficiently fired to obtain archaeomagnetic direc- absolute dating of Late Iron Age sites in tions. Extrapolation of records for magnetic declina- southern Africa: comparing radiocarbon, tion and inclination at Cape Town beginning in AD 1595 suggests that the drift of magnetic field in the luminescence and archaeomagnetism. region and time of interest would be sufficiently rapid Radiocarbon dating is the chronometric tech- to yield high-precision archaeomagnetic dates over nique most frequently utilized by archaeologists a short period. Archaeomagnetism is potentially the worldwide, however, this method is inadequate for best method for relative, and eventually absolute, sites occupied throughout the later Late Iron Age in dating of sites in this temporal and geographic con- southern Africa. Because of anomalies in atmospheric text. This research will be an important contribution production of radiocarbon in the time range 1650- to our field, because it will become possible to obtain 1950 cal CE, there are acute De Vries effects, or ‘wig- meaningful archaeometric dates for Late Iron Age gles’, in the radiocarbon calibration curve over this sites. Several of the most important urban centers of period. Any radiocarbon age from this time calibrates precolonial Africa date to second millennium AD time to a calendar date spanning two or more centuries. ranges that were subject to anomalies in radiocarbon Although Bayesian radiocarbon calibration, dates production. Once the archaeomagnetic master curve from oral-historical records and stratigraphic infor- is complete, archaeomagnetism can replace lumines- mation from ceramic seriation can help to refine these cence as the preferred chronometric method for the broad radiocarbon calendar age ranges, in order to second half of the Late Iron Age. understand the sequence of settlement in the Late

130