26 November 2013

Shaping the southern African ‘Iron Age’ archaeological landscape

Shaping the southern African
‘Iron Age’ archaeological landscape:
Indigenous African ritual, technology, and knowledge

A conceptual summary and project description
Robert Thornton
Anthropology, University of the Witwatersrand
Johannesburg, South Africa
November 2013
Abstract: Consideration of the role of African indigenous cultural, ritual, and symbolic thought and practices is largely absent from contemporary Iron Age archaeology and history.  This project seeks to fill this gap through a reinterpretation of the archaeological landscape and material culture in the light of southern African traditional knowledge systems.  A similar shift in the way rock art was interpreted allowed us to see rock art as a window into deep spiritual practices of San shamans rather than as simple pictures of economic activities such as hunting.  The change of perspective that this research introduces and seeks to validate will allow a profound re-visioning of South African history, material culture and archaeology for this period. Current models and histories emphasize cattle and kingship; the model to be developed in this discussion focuses instead on a social order built around guilds of ritual and technological specialists such as blacksmiths, miners, glass-bead makers, potters, healers and shamans in a landscape of sacred sites and secret knowledge.  Aspects of history are preserved in contemporary sangoma’s knowledge, but also in the material culture and landscape.  The research is transdisciplinary, utilizing high-level scientific technology (e.g. electron & light microscopy), experimental re-creation of ancient technologies of glass and metals, together with humanistic and social scientific methods of anthropology, archaeology and history.  It brings to bear new, recently developing perspectives (e.g. on landscapes, ancient mining, & early fabrication techniques) and new research technologies (GPS, XRF, GoogleEarth).

I seek to develop ideas that have arisen in my long-term engagement with bungoma (‘traditional healing’ or ‘indigenous knowledge systems’/IKS), and to develop data, argument, and theory more fully.  My preliminary research had led me to a major conceptualization of anthropological, historical and archaeological interpretations of the pre-colonial history of the region comprising Limpopo, Mpumalanga and Gauteng during the southern African ‘Iron Age’ from approximately 500-1600 CE.  This project will allow me to develop preliminary findings into a tight case for an alternative model.
   Second, it provides a significantly new understanding of bungoma/’traditional healing’ as a specialised technical knowledge system (or ‘science’), organised in secret guild-like social structures, that is partially preserved yet today amongst contemporary sangoma’s practices, material culture (‘regalia’), and knowledge systems.  I have already made considerable contributions to knowledge in this field (Thornton 2000; 2003; 2005; 2008; 2009; 2010; 2012; 1998).
The central argument, then, is that exclusive guilds or syndicates of specialists practiced these technologies, and that this specialized knowledge is historical and culturally continuous with some contemporary sangoma practices and knowledge. 
This project began with the question: ‘can a history of bungoma (South African ‘traditional healing’) be written, given the lack of textual sources?’ It is clear that knowledge, and thus historical evidence, but not history itself, is indeed passed down from teacher to student (Thornton 2009).  Bungoma knowledge has a history, therefore, like any product of human thought and practice, but this history is not known. To recover this history is to understand the role of bungoma in South African history, and to understand its place in the cultural landscape as well as in current and past cultural practices. This is therefore also an intellectual history of African thought in the context of a broader South African history over the previous two millennia. 
Throughout, the existence of strong historical and cultural continuities, rather than ruptures, is proposed.
Previous research shows clearly that the main features of material culture in the southern African archaeological landscape consists of: (1) large numbers (in excess of 100,000) of stone-walled structures of monumental scale, often circles or complexes of multiple circles and other linear features; (2) the presence of glass beads at most sites dating from 600 CE to the present, with high concentrations at a few sites such as Mapungubwe; (3) a profusion of ceramics and small metal objects made of iron, bronze, copper, brass and gold across a wide range of sites; (4) extensive evidence of high-temperature technologies including metal smelting, melting, casting and forging, and production of ceramics.  The landscape itself shows evidence of extensive mining and metal extraction technologies from pre-colonial times to contemporary large-scale, small-scale and artisanal mines.  None of this has been studied in an integrated way.
A key problem is the lack of textual accounts of these industries, and almost no oral history, memory, or identification with these sites among any contemporary southern African populations.  Since the knowledge of these technologies was the property of secret guilds of a few specialised technicians, this knowledge of pyrotechnology and mining was lost at the time that metal and glass objects and materials from European and Indian Ocean sources swamped local production, certainly by 200-300 years ago, and possibly a century or two earlier.  Suppression of sangomas from the nineteenth century by missionaries and African Christians, and by the state, resulted in further loss of knowledge.  It is now preserved largely in the material culture, and landscapes. 
Much of the southern African material culture from the previous three centuries is now only available in museums, especially European museums in London, Paris, Prague, Berlin, Leiden and others. SA scholars rarely access these.  These collections preserve 18th - 19th C material culture, much of it belonging to ‘witchdoctors’, ‘healers’ and 'shamans’. This is effectively new material or new phenomena to be investigated.
I intend to examine the archaeological landscape through the lens of bungoma (‘traditional healing’), that is, through an anthropological- archaeological- historical model built on southern African traditional knowledge systems that I have studied intensively over a period of 12 years.  An ‘archaeological landscape’ is a past cultural landscape.  Landscapes, like languages or other cultural products, preserve historical traces that can be interpreted.  Peter Johansen defines a cultural landscapes (in the archaeological context) as  “spatial and temporal fields of action in which material and conceptual contexts are constructed and negotiated through the processual articulation of social action, structure and the physical environment” (Johansen 2004; citing Smith 2003).  It is not possible to recover the full complex of meanings associated with the cultures of this period, but by assuming that historical continuities with ethnographic realities and landscapes exist, it is possible to recover significant structures of their practice and conceptual systems.
This exploratory project lies in the borderlands between anthropology, archaeology and the sciences (life, environmental, and physical sciences, geography, etc.), and goes beyond the horizon of contemporary knowledge of both academic anthropology and of bungoma because (1) there are no historical texts in IKS on which to base this enquiry, (2) for sangomas, all knowledge comes from ancestors (and therefore does not preserve what might be called historical knowledge), and (3) no history of southern African indigenous philosophy and healing yet exists.  Similarly, almost nothing is known of early African fabrication techniques in glass, metals or other materials, and almost nothing of mining, landscape and trade (except that it happened).
This approach complements and extends the current models by adding the dimension of ritual, healing, and the sacred.  Interpretation of the archaeology and historical landscape has so far been focused on agricultural-pastoral economies of ‘Bantu-speaking’ peoples with political structures characterized as early states, chiefdoms, and kingdoms during the southern African ‘Iron Age’.  The major role of ritual, ‘healing’ (bungoma) and the sacred in African cultures has been largely neglected (with the exception of rock art, attributed to San peoples), or treated as an adjunct to political order.  By contrast, the model I am developing interprets major archaeological sites (such as Mapungubwe, the ‘stone circles’, Thulamela and others) and their associated material cultures as elements of ritual practices, healing, and sacred sites in this landscape.  Many aspects of metal-working, metallurgy, mining, bead-making (especially glass bead production, but also using many other materials) and other elements of material culture can be better understood in terms of ritual practice and their symbolic significance.
The project is important because it helps to establish links between early southern African material culture with the rest of Africa and the Indian Ocean civilizations, because it provides a richer interpretive framework, and because it will enrich heritage management and tourism development. 
It is clear that the principal users of amulets, glass beads, and probably other early metals and high-temperature technologies were ritual practitioners, that is, sangomas, or ‘shamans’ and ‘healers’. They were also, probably, the makers of these objects, and therefore masters of specialized technologies.  This is certainly true in all other historical and ethnographic cases, so it is likely to be true of southern African too.  The link between secret and sacred ritual and early technologies is well attested globally, but has never been applied to southern African materials. 
Bungoma is an African science of its environment and utilizes natural resources to accomplish ritual and healing goals.  Close examination of its products in historical-archaeological context, using current advanced technologies as well as interpretive techniques of the humanities and social sciences, will help to unravel its history and significance.  New theoretical approaches will be developed and new methodologies deployed.  Breakthroughs in archaeometallurgy, chemistry and physics of glass and metal, new analytical techniques (XRF, OLM, SEM, EDS, others) and accurate geo-referencing and imaging systems (digital photography, GPS, GoogleEarth, specialized software, etc.) make the proposed research possible, and do-able, as never before.  It is cutting edge, integrative, and innovative.
For instance, my preliminary investigation of the physical structure of glass beads suggests that many of the earliest beads were produced in the region for local use but also for trade in Africa-wide and Indian Ocean networks.  This investigation is different from the investigation of the chemical composition and trace-element and REE analysis that has been done, for instance, by Davison(1972)), Saitowitz (Saitowitz and Reid 1996) and Robertshaw and Wood et al. (2000; 2002; 2012; 2012), all of which attempts to show the external origin of all glass beads.  So far, this has been unsuccessful in matching southern African beads with beads from anywhere else in the world.
Experimental reconstructions with local materials show how glass beads could have been locally manufactured, and suggests that in fact it could be locally manufactured. Source of plant ash used as flux (Salsola kali (L.) & S. soda (L.)) is plentiful in the vicinity of bead production sites, and high quality silica is also plentiful, as are grinding sites where quartz was processed.  These sites are also associated with metal production, too, especially gold and iron.  I have experimentally produced ‘virgin' glass from raw local materials, using an open charcoal fire.  This glass has colours and characteristics of the glass in first millennium beads. I have made beads that resemble very closely the physical structure and appearance of beads from the first millennium archaeological sites.  Materials and technologies were well within the range of capabilities that first millennium southern Africa societies already possessed but this has not been recognized in any study to date.  Ethnographic and archaeological evidence from other places in which beads were made, and examination of internal structures such as included bubbles, inclusions of ash and crystals, heat fractures, and other microscopic detail in the glass gives evidence of manufacturing processes involved.
The dominant historical narrative claims that all beads were imported.  I argue that the region’s trade with the rest of Africa was a two-way trade that exchanged beads for other beads, that is manufactured goods for other goods.   Other small ritually significant objects were also involved, especially metal, but also organic materials (‘muti’), shells, etc., in an economy of ritual objects.  This economic exchange of ritual objects probably went hand in hand with more general trade, but was probably the predominant focus of trade in this region as it was in most other regions in earlier times. The model I propose shows the role of independent African manufacturing capability in the later first millennium southern African cultures, and shifts our understanding of early pre-colonial trade networks considerably.  
Other anomalies exist. Despite an interpretive model (the ‘Central cattle pattern’ or CCP) that emphasises cattle pastoralism and agriculture, there is a surprising lack of evidence for cattle (‘bovines’) at most sites. Again, standard histories of the region emphasise warfare among ethnic groups, the rise of kingdoms, and agriculture, even though most artefacts are small and of little use for warfare or agriculture.
Another major problem for historical and archaeological interpretation has been the function and significance of the numerous (more than 100,000) and extensive monumental ‘stone circles’ of dry stone ‘walling’ across parts of the southern African landscape.  These are composed of local stone stacked without mortar into linear or circular features.  Surprisingly, there is very little other cultural material associated with these features (apart from the features themselves). There is a diversity of interpretation of what these may have been (Sadr 2005; Sadr 2013; Sadr, et al. 2013; Sadr and Rodier 2012), and extremely intense debate among scholars and a wider public of ‘amateur archaeologists’.  
This research offers potential insight into these problems.  Under this interpretation, large well-known sites such as Mapungubwe can be interpreted as regional sacred sites that functioned in regional ritual practices of pilgrimage involving sacrifice, feasts, and fasting.  This is true of many sacred sites today in southern Africa, and is more consistent with the archaeological material than other interpretations. The stone structures, for instance, are tentatively interpreted as temporary sacred sites built to provide ritual protection.  Ritual and processes that would require protection include the full range of ‘ritual’ activities (circumcision, healing, trance-dance,), as well as ‘manufacturing’ processes such as metal smithing (iron, gold, copper), bead making, and preparation of medicines, amulets and other paraphernalia.  The stone structures occur largely on ultramafic rocks and soils, that is, in places that are especially rich in metals, especially iron and gold.
In other words, the structures offer identity, geographic position, and ritual protection for a range of secular and sacred activities. In many ways, this initiative parallels the ground-breaking work of David Lewis-Williams who was able to chart radically new research direction in the field of rock art through interpreting the images through the lens of ‘shamanic’ practices and beliefs of early southern African populations ( For instance, among many others: Lewis-Williams 1996; Lewis-Williams and Pearce 2012; Lewis-Williams 1981; Lewis-Williams 1990; Lewis-Williams and Challis 2011; Lewis-Williams and Pearce 2005).  
There is currently considerable interest in recovering lost landscapes using mixed archaeological, technological and anthropological methods.  Herr describes the efforts to recover the sacred landscape of the Western Apache that was lost through forced removals, expropriation of native lands, and genocide that uses mixed methodologies.
When the fragile archaeological remains are considered with ecological information, historical documents, ethnographic reports, and the accounts provided by members from descendant communities, a robust historic Apache landscape can still be found (Herr 2013:679).
Jan Jansen, in the most recent issue of the journal History in Africa, discusses uses of new technologies in conservation of historical heritage.
Historians long have recognized the often ‘vulnerable’ nature of African historical sources, including the deterioration of manuscripts, the destruction of archives in conflict zones, the loss of recorded interview to decay, to name just three.  … [Today] new opportunities and challenges [exist] especially due to the introduction of new technologies and media and their roles in the collection, preservation, and distribution of historical sources.(Jansen, et al. 2013)
This is especially true of new GPS technologies, extremely portable iPad-type computing devices, GoogleEarth geo-imaging and sophisticated mapping and geo-location technologies.  These make investigation of the landscape possible in ways that have never before been possible. Chirikure and Pikirayi, professors of archaeology at University of Cape Town and University of Pretoria, make this point strongly with respect to studies of Great Zimbabwe archaeology.
Since 1980, there has never been an integrated archaeological research programme on Great Zimbabwe, only isolated and often fragmented approaches … on stone architecture, … on soapstone birds and … on metalwork. This fragmented approach … frustrates attempts to develop a coherent history of … the site as revealed through artefact studies. (Chirikure and Pikirayi 2008)
The proposed research here addresses this issue with respect a different area, but with similar past ‘fragmented’ studies. 
In sum, I have found that by interpreting a large part of the archaeological evidence in terms of ritual and religious systems of bungoma, rather than in the economic-evolutionist paradigms that are more usual, it is possible to see the southern African archaeological landscape in an entirely new light. 

References cited

[Titles have been shortened to save space, but should be fully identifiable. This is a only a small selection of the material available in this field.]
Chirikure, S, and I Pikirayi.  2008  Inside and outside the dry stone walls. Antiquity 82:976-993.
Davison, C C. 1972  Glass beads in African archaeology. PhD, University of California.
Herr, S A.  2013  In search of lost landscapes. Am. Antiquity 78(4):679-701.
Jansen, Jan, et al. 2013  In search of Africans' Histories. History in Africa 40(2013):1-3.
Johansen, P G  2004  Landscape, monumental architecture, and ritual. J Anthr Archaeology 23:209-330.
Lewis-Williams, D, and D G Pearce  2012  The southern San and the trance dance. Antiquity 86(333):696-706.
---. 2005  Inside the neolithic mind. Thames & Hudson.
Lewis-Williams, D.  1996  Discovering African rock art. CapeTown:David Philip.
---. 1981  Believing and seeing.  New York: Academic Press.
---. 1990  Discovering Southern African rock art. Cape Town: D. Phillip.
Lewis-Williams, D, and S Challis  2011  Deciphering ancient minds. London: Thames & Hudson.
Sadr, Karim.   2005  Hunter-gatherers and herders of the Kalahari during the late Holocene. In Desert Peoples: Archaeological perspectives. P. Veth, M. Smith, and P. Hiscock, eds. Pp. 206-221. Oxford: Blackwell Publishing.
 ---.  2013  The archaeology of herding in southernmost Africa. In The Oxford Handbook of African Archaeology. P. Mitchell and P. Lane, eds. Pp. 641-651. Oxford: Oxford University Press.
Sadr, Karim, J Gribble, and G Euston-Brown.   2013  Archaeological survey on the Vredenburg peninsula. In The Archaeology of the West Coast of South Africa. A. Jerandino, A. Malan, and D. Braun, eds. Pp. 50-67. Cambrdge Monographs in African Archaeology, Vol. 84. Oxford: Archaeopress.
Sadr, K, and X Rodier.   2012  Google Earth, GIS and stone-walled structures in southern Gauteng, South Africa. J Archaeological Science 39(2012):1034-1042.
Saitowitz, S J., and D L. Reid.  1996  Glass bead trade from Islamic Egypt to South Africa c. AD 900-1250. S A J  Science 92(2):101.
Smith, A T.   2003  The Political Landscape. Berkeley, CA: U of California Press.
Thornton, R J.  1998  Toward a theory of persons and things. African Anthropology / Anthropologie Africaine 4(1):14-16.
---. 2000  The Landspace, Land and Landscapes in Contemporary South Africa. In Skalnik, P (ed.) Sociocultural Anthropology at the Turn of the Century. Prague
---.  2003  Traditional healers and bio-medical practice: Prospects and barriers to co-operation. Adler Museum bulletin 29(2):8-16.
---.  2005  Four principles of South African political culture at the local level. Anthropology Southern Africa 28(1):22-30.
 ---. 2008  Unimagined community : sex, networks, and AIDS in Uganda and South Africa. Berkeley: University of California Press.
---. 2009  The transmission of knowledge in South African traditional healing. Africa 79(01):17-34.
---.  2010  The market for healing and the elasticity of belief: Medical pluralism in Mpumalanga, South Africa. Markets of Well-Being: Navigating Health and Healing in Africa 9:144.
 ---. 2012  Magical Empiricism and the ‘exposed being’ in Public Health and Traditional Healing. In Workshop: “The (Un)healthy Body in Southern Africa”.  . Wits Reproductive Health and HIV Institute, Hillbrow, Johannesburg, South Africa: unpublished.
Wood, M.  2000  Making Connections: Relationships between International Trade and Glass Beads from the Shashe-Limpopo Area. Goodwin Series 8.  South African Archaeological Society:78-90.
---. 2002  The glass beads of Kaole. In Southern Africa and the Swahili World. F. Chami and G. Pwiti, eds. Pp. 50-65. Dar es Salaam: Dar es Salaam University Press.
---. 2012  Interconnections: glass beads and trade in southern and eastern Africa and the Indian Ocean, 7th to 16th centuries AD. Uppsala: Department of Archaeology and Ancient History.
Wood, M, L Dussubieux, and P Robertshaw.  2012  The glass of Chibuene, Mozambique: New insights into early Indian Ocean trade. S A Archaeo Bull. 67(195):59-74

Making glass--experimental archaeology for southern African materials

My first home made glass.

Doesn’t look like much at the moment, but this is a photo of a cross section of the first glass that I made from quartz and soda carbonate flux, approximating the basic glass recipe.  It was made in the back yard, literally, on a charcoal fire with a forced draft.  It proves that glass can indeed be made quite easily from locally, even universally available materials, and that it is not particularly hard to make glass. It has always been said in the historical literature, and still is, that glass was so difficult to manufacture initially from raw materials that it must have been made only in a very limited number of places, and that the raw glass, once made, was traded or shipped all over the world, or all over trade networks within whatever limits.  This seems not to be true.  Certainly, one has to know how glass CAN be made, and HOW it can be made. It requires a craftsman who is skilled with high temperature technologies such as metal smelting, melting, casting, or working, but apart from that the materials are universally available.  

  •   Especially important is the fact that the quantities of inputs--the recipe--seems to be relatively unimportant.  The glass forms as the fluxed silicon dioxide gradually melts and seems to form eutectic substance as the sodium ions and other ions reach saturation points in the silicon dioxide solution, and excess materials either never dissolve into the substance at all, and remain unchanged, or crystallise out of the glass solution.  
  • )b carbon dioxide form throughout the glass initially.  The way these eeliminated, it appears, is by allowing the first formation of glass (what I made, anas in the picture above) to cool, regrinding it, and melting the eutectic mixture, now with the volitile component already removed, to melt again. This give a relatively bubble-free glass.  Thus, large numbers of bubbles in the glass probably indicate that the glass is made directly from raw materials and has not been reground.  Typically, the second melt would also include cullet, or ‘scrap’ glass from broken vessels in the melt.  this assists in the melting and provides a better glass with fewer inclusions and bubbles.