This article examines copper production in the Nyali Basin of the Republic of Congo from the mid-15th to mid-17th centuries AD. It assesses the microstructure and composition of slag and technical ceramics from sites associated with two different regions, pottery traditions. The two sites are characterized by the use of fire-resistant household pottery as crucibles for copper smelting.
introduce
Copper production in the Nyali Basin in the southern Republic of the Congo during the 15th-17th centuries AD (below).
Since it was first mined in Central Africa in the middle of the first millennium AD, copper, not gold, has been the most valuable metal for most of Central African history, revered for its color, luster and sonic properties.
Copper was used in status creation projects and showed a close relationship with social/political prestige, evident from the early 2000s in the graves of the elite of the Upemba depression, where copper also played an important role as a casting ingot in long-distance trade. In 19th-century Central Africa, similar value to unalloyed copper was given to imported European brass.
The Portuguese were active along the coast from the late 15th century, particularly within their southern colonies of Angola and the Kingdom of Congo, while Dutch influence grew rapidly after their arrival in 1597, thanks to trade with Loango in the northern coastal region.
Until the middle of the 17th century, when the demand for slaves seemed to overshadow commerce, as was the case in Portugal Angola in the 16th century.
Miners in Niari mined mineralization formed between sandstone and limestone and in faults inside limestone, particularly veins filled with malachite.
Pottery found in copper deposits during this period can be typically divided into two types: the Moubiri type and the Kindangakanzi type. Moubiri-type pottery is mainly found at sites near Mindouli in the east, and Kindangakanzi type pottery near Boko-Songho in the west.
This article describes copper production associated with Moubiri-type pottery around Mindouli and copper production associated with Kindangakanzi type pottery near Boko-Songho. It seeks to understand whether and how miners, smelters and potters in the two regions are technically and socially interrelated.
We considered the similarities and differences of Chaînes Opératoires, evaluating material and action options at each step as potential evidence for knowledge sharing. An important element is the common practice of reusing decorated domestic pottery as crucibles, the Moubiri type near Mindouli and the Kindangakanzi type near Boko-Songho.
Materials and methods
All analyses were carried out in the Archaeological Sciences Laboratory of the Department of Archaeology at the University of Cambridge. Of approximately 10.5 kg of metallurgical waste at both sites, 85 samples of slag, ore, crucible, air vents, furnace walls and household pottery were selected for analysis.
These are initially screened by macroscopic observation and pXRF, and then a portion of the sample is prepared as a polished block for light microscopy and SEM–EDS.
The sample is cut and mounted in epoxy, cured overnight, and then ground and polished to 1 μm. The polished block is first examined with planar and cross-polarized light on the Keyence VHX-6000 digital microscope to assess macroscopic size and inclusion distribution, and to locate and correlate areas for SEM analysis.
In addition to on-site analysis for slag phase identification and qualitative identification of impurities in metal particles, the overall chemical composition is calculated by averaging more than 500 × readings from typical ceramic and slag areas, avoiding macropores/inclusions. The spectra are visually evaluated, quantified as stoichiometric oxides where appropriate, and normalized to 100% weight.
While SEM-EDS quantification is limited by the porous nature of ceramics/slag and the presentation of data in single oxide form, its ability to collect a combination of specific background composition, microstructure, and chemical data is a key reconstruction reaction parameter. This does not affect ontology chemistry.
outcome
Technical ceramics:
Niari's technical ceramics include domestic pottery for use as crucibles as well as special vents and furnaces. The trace element pattern shows a relatively condensed group from around Mindouli, including Moubiri-type pottery from Kingoyi, and a second group from around Boko-Songho, including Kindangakanzi and its pottery style.
It is not currently possible to distinguish subgroups, and a comprehensive study of Niari pottery production is beyond the scope of the present.
These groups of elements correspond to different fabric types: Moubiri type pottery is rich in quartz, Kindangakanzi type pottery is rich in talc, chemically distinguished by the percentage of alumina with magnesium oxide.
Both fabrics are highly refractory and have no significant expansion, and the interface between the ceramic layer and the slag layer is relatively clear. A typical refractory chart of alumina vs. alkali used to assess ceramic refractory distorts the Niari material because the high magnesium oxide content of Kindangakanzi type pottery overcomes the offset of other bases.
In fact, the magnesium oxide content in Kindangakanzi crucibles makes it truly fire resistant.
The SEM-EDS data discussed below demonstrate high values for alumina and magnesium oxide for Moubiri type and Kindangakanzi type pottery, respectively, suggesting that the relative qualitative patterns identified in broader pXRF screening may be related to the refractoriness of these different types.
Air vents/furnaces:
The vents at both sites are formed around reeds, and organic imprints are visible on the inner surface. They are not fired before use, turn into ceramics by heat from metallurgical reactions, and often have a large amount of slag. The Tuyères in Kindangakanzi is slightly larger and more conical than the Kingoyi.
These are typical diameters for air vents operated using bellows instead of blowpipes, for which the opening will be narrower (about 0.5-1.0 cm), or for natural ventilation it will be wider (> 5 cm) (Rehder 1994). Wooden blowers for copper production were reported in the region in the 19th century.
The air vent MKU3b15_2 made of refractory quartz-based fabric with low expansion and high structural integrity. Other Mindouli tuyères, which were examined by pXRF clusters, had slightly more Zr than Moubiri-type pottery, while MKU3b15_2 were located outside the two groups.
This indicates a relatively discrete source of local clay developed for the vent, related but distinct from those used to make Moubiri-type pottery, and a second source of the vent MKU3b15_2.
conclusion
The chaînes opératoires at each site show that Kingoyi and Kindangakanzi share metallurgical principles: carbonate ores are smelted using crucibles in reused domestic pottery, without the use of fluxes. These commonalities are based on shared knowledge/learning and processes, but occur in spaces that are geographically and socially distant from each other.
In other words, sharing a space, such as a copper mine, encourages the community to share ideas and technologies, such as crucible smelting, thus forming clusters of practice around metallurgy.
In the context of copper metallurgy of the 15th-17th centuries, the understanding of the refractory of Kindangakanzi type pottery and its suitability for use in metallurgy seems to be based on an understanding of the feel of its soap. For example, in post-medieval Europe, refractory crucibles were traded according to color and texture, and even imitated refractory crucibles by replicating their sensory qualities.
Another possibility is that local potters and smelters in the 15th to 17th centuries realized that domestic pottery made of widely used and historically mined local clay had proper fire resistance, thus avoiding the need to make specialized crucibles during this period.
Technical data from evidence of copper production in Kingoyi near Mindouli and Kindangakanzi near Boko-Songho contributed to the metallurgical reconstruction of copper production in the Nyali Basin in the 15th-17th centuries. Analysis of crucibles, vents and furnaces revealed material and technological choices within the chaînes opératoires of both sites and allowed the identification of metallurgical practice constellations formed by discrete Moubiri and Kindangakanzi potting communities.
However, there are still many aspects of copper production in Niari in the 15th-17th centuries that remain to be explored.
Bibliography:
[1] Identification of slag at archaeological sites, provisional paper. London Institute of Archaeology
[2] Bandama F, Moffett AJ, Chirikure S (2017) Typology and technical properties of metallurgical crucibles in the Great Zimbabwe (1000–1700 AD) Remnants Collection. J Archaeol Sci Rep 12:646–657。
[3] Reconstruction of the first copper smelting process in Europe between 4 and 3 millennium BC: where did oxygen come from?
[4] Ceramics decorated with woven patterns: archaeological identifiers for the Kingdom of Kongo? In: Brinkman I, Bostoen K (eds) Kingdom of the Congo: Origins, Dynamics and World Culture of African Polity. Cambridge University Press, Cambridge, pp. 165-196.
[5] Maret P (1999) Symbolic Power and Symbolism of Power Through Time: Exploring Luba's Past.
[6] The nearest agricultural community and country in and around the Congo Basin. Published in: Mitchell P, Lane P (eds) Oxford Handbook of African Archaeology. Oxford University Press, Oxford, pp. 875-886.