The development of iron technology in precolonial western Uganda

Abstract

The local production of iron was an important technology in eastern Africa up until the earlier twentieth century, when the use and reuse of imported iron overtook vernacular smelting industries and cemented their decline. Prior to this, the utilisation of local ores had produced iron for agricultural implements, household tools and weapons, serving the needs of many generations of farmers and herders across the region. The smelters of western Uganda enjoyed a particularly esteemed reputation in recent history, especially among their neighbours in Buganda. Prior to archaeometallurgical research undertaken in 2007, little was known about the technologies upon which this reputation was fostered. This paper presents an overview of the results of six months of fieldwork in Uganda and subsequent archaeometallurgical analysis, which together revealed the complexities of smelting in Mwenge, western Uganda, between the fourteenth and twentieth centuries. Technological reconstructions indicate that some iron producers in the region were supplementing the iron ores in their smelts with an additional manganese-rich material, imparting a tangible effect on the process and outcomes of these smelting episodes, hypothetically increasing the metal yield and improving operating parameters. Although such harnessing of beneficial manganese-rich minerals was an unexpected and unusual finding, it correlates with existing ethnoarchaeological and ethnohistorical data. Through these avenues, and building upon social approaches to iron technologies, it has been possible to explore this variation in the technological record of western Uganda in the context of the known history of the region.

La production du fer fut une technologie locale importante en Afrique orientale jusqu'au début du vingtième siècle, auquel moment l'utilisation et la réutilisation du fer importé dépassa les traditions de fonte locales et concrétisa leur déclin. Avant cette date, les minerais locaux avaient constitué la source du fer pour les outils agricoles, les outils ménagers et les armes, desservant les besoins de multiples générations d'agriculteurs et de pasteurs à travers la région. Les fondeurs de l‘Ouganda occidental jouissaient d'une réputation particulièrement éminente dans l'histoire récente, en particulier auprès de leurs voisins du Buganda. Avant des recherches archéométallurgiques entreprises en 2007, on savait vraiment peu des technologies sur lesquelles se basait cette réputation. Cet article présente les résultats de six mois de recherches de terrain en Ouganda et des analyses archéométallurgiques qui leur firent suite. Combinés, ces deux éléments révèlent la complexité de la fonte à Mwenge, en Ouganda occidental, entre les quatorzième et vingtième siècles. Les reconstructions technologiques indiquent que certains producteurs de la region supplémentaient les minerais de fer utilisés dans leurs fontes avec du matériel riche en manganèse, ce qui avait un effet tangible sur les procédés et les issues de ces épisodes de fonte et qui augmentait, on le suppose, la production de métal et améliorait les parameters d'opération. La découverte de l'utilisation de tels minéraux riches en manganèse représente certes une trouvaille inattendue et inhabituelle, mais elle recoupe les données ethnoarchéologiques et ethnohistoriques. De par ces avenues, et guidée par les enseignements des approches sociales aux technologies du fer, il a été possible d'explorer cette variation dans l'archive technologique de l'Ouganda occidental, et ce dans le contexte de l'histoire connue de la région.

Keywords:

• Uganda
• Bunyoro
• iron production
• archaeometallurgy
• Late Iron Age

Acknowledgements

The Arts and Humanities Research Council funded the PhD on which this paper is based. Additional financial assistance was received from the UCL Graduate School, the University of London Central Research Fund and the UCL Institute of Archaeology. Radiocarbon dates were funded by an AHRC/NERC ORADS grant. I should like to acknowledge Rose Nkaale, Dismas Ongwen and colleagues at the Uganda Museum in Kampala, and staff and students at the British Institute in Eastern Africa for their support in the field. Thanks must also go to Charlotte Karungi, Elijah Kisembo, Terry Childs, Pete Robertshaw, John Sutton, Andrew Reid and Marcos Martinón-Torres, as well as to two anonymous referees for their insightful comments on an earlier draft.

Notes

1. Similar research was carried out in Masindi to the north, and Rakai to the southeast (Figure 1), but these results are presented elsewhere (Iles 2011).

2. See Doyle (2006b) for an extensive discussion of the effect of Bunyoro's recent political history on the construction of its historical narrative.

3. Kyakaturi, 625±26 BP (1290–1398 cal. AD, OxA-20936); Mirongo, 553±27 BP (1314–1430 cal. AD, OxA-20933); Rugombe, 479±26 BP (1410–1450 cal. AD, OxA-20931) (OxCal 4.1; IntCal09; Bronk Ramsey 2009; Reimer et al. 2009).

4. A crude mass balance calculated from the chemical compositions of 11 slag blocks and a feasible ore sample at Kyakaturi estimates that for every 100 units of slag produced, 100 units of iron were also produced. This supposes that given the slag composition (50wt% iron oxide, 30wt% silica (primarily from the silicate flux) and 20wt% ‘other’) and ore composition (90wt% iron oxide and 10wt% ‘other’), two measures of ore are required to match the proportion of ‘other’ left in the slag (if contributions of ‘other’ from ash and technical ceramics are ignored). This gives 180 units of iron oxide, 50 of which are lost to the slag. The remaining 130 units of iron oxide convert to approximately 100 units of iron metal, per each 100 units of slag. The average weight of a complete slag block was 80 kg. If 17 smelts are presumed (i.e. 1360 kg slag), this gives an estimation of 1360 kg iron produced; if 50 smelts are presumed, together they would have produced around 4000 kg iron. This is an overestimate, but provides a rough guide to output.

5. The Lunyoro word mirongo is the name of one of three types of tree that Roscoe (1923: 218) notes specifically as being used to prepare charcoal for the Nyoro iron smelting technologies that he documented around Hoima.

6. Kirongo, 148±25 BP (1667–1949 cal. AD, OxA-20938); Kisamura, 102±25 BP (1684–1929 cal. AD, OxA-20932); Rukomero Furnace 1, 148±25 BP (1667–1949 cal. AD, OxA-20934); Rukomero Furnace 2, 170±28 BP (1661–1954 cal. AD, OxA-20935) (OxCal 4.1; IntCal09; Bronk Ramsey 2009; Reimer et al. 2009).

7. Markets seem to have existed for raw materials as well as for bloomery iron and finished products, at least during later periods. Childs’ and Robertshaw's informants specified markets that sold iron ore to smelters.